Revised Glyphosate Issue Paper:
Evaluation of Carcinogenic Potential
EPA's Office of Pesticide Programs
December 12, 2017
DP Barcode: D444689
TXR#: 0057688
-------
Table of Contents
List of Acronyms 7
List of Tables 10
1.0 Introduction 12
1.1 Background 12
1.2 Evaluation of Carcinogenic Potential 12
1.3 Overview of "Framework for Incorporating Human Epidemiologic & Incident Data
in Health Risk Assessment" 14
1.4 Summary of the Exposure Profile in the United States 15
1.5 Organization of this Document 19
2.0 Systematic Review & Data Collection 19
2.1 Data Collection: Methods & Sources 20
2.1.1 Open Literature Search 20
2.1.2 Studies Submitted to the Agency 21
2.2 Evaluation of Relevant Studies 22
3.0 Data Evaluation of Epidemiology 23
3.1 Introduction 23
3.2 Considerations for Study Quality Evaluation and Scope of Assessment 23
3.2.1 Study Designs 24
3.2.1.1 Analytical Studies 26
Page 2 of216
-------
3.2.1.2 Descriptive Studies 27
3.2.2 Exposure Measures 28
3.2.3 Outcome Measures 28
3.2.4 Confounding 28
3.2.5 Statistical Analyses 29
3.2.6 Risk of Bias 29
3.3 Review of Quality Results 30
3.3.1 "High" Quality Group 31
3.3.2 "Moderate" Quality Group 32
3.3.3 "Low" Quality Group 32
3.4 Assessment of Epidemiological Studies for Relevance to Analysis 44
3.5 Summary of Relevant Epidemiological Studies 45
3.5.1 Solid Tumor Cancer Studies 45
3.5.2 Non-Solid Tumor Cancer Studies 53
3.6 Discussion 63
4.0 Data Evaluation of Animal Carcinogenicity Studies 69
4.1 Introduction 69
4.2 Consideration of Study Quality for Animal Carcinogenicity Studies 69
4.3 Assessment of Animal Carcinogenicity Studies 71
Page 3 of216
-------
4.4 Summary of Animal Carcinogenicity Studies 74
4.5 Rat Carcinogenicity Studies with Glyphosate 74
4.5.1 Lankas, 1981 (MRID 00093879) 74
4.5.2 Stout and Ruecker, 1990 (MRID 41643801) 75
4.5.3 Atkinson et al, 1993a (MRID 49631701) 79
4.5.4 Brammer, 2001 (MRID 49704601) 80
4.5.5 Pavkov and Wyand 1987 (MRIDs 40214007, 41209905, 41209907) 80
4.5.6 Suresh, 1996 (MRID 49987401) 81
4.5.7 Enemoto, 1997 (MRID 50017103-50017105) 81
4.5.8 Wood et al, 2009a (MRID 49957404) 81
4.5.9 Summary of Rat Data 82
4.6 Mouse Carcinogenicity Studies with Glyphosate 85
4.6.1 Reyna and Gordon, 1973 (MRID 00061113) 85
4.6.2 Knezevich and Hogan, 1983 (MRID 00130406) 85
4.6.3 Atkinson, 1993b (MRID 49631702) 87
4.6.4 Wood et al, 2009b (MRID 49957402) 88
4.6.5 Sugimoto, 1997 (MRID 50017108 - 50017109) 89
4.6.6 Pavkov and Turnier, 1987 (MRIDs 40214006, 41209907) 90
4.6.7 Summary of Mouse Data 90
Page 4 of216
-------
4.7 Absorption, Distribution, Metabolism, Excretion (ADME) 93
4.8 Discussion 94
5.0 Data Evaluation of Genetic Toxicity 98
5.1 Introduction 98
5.2 Scope of the Assessment Considerations for Study Quality Evaluation 99
5.3 Tests for Gene Mutations for Glyphosate Technical 100
5.3.1 Bacterial Mutagenicity Assays 100
5.3.2 In vitro Tests for Gene Mutations in Mammalian Cells 106
5.4 In vitro Tests for Chromosomal Abnormalities 108
5.4.1 In vitro Mammalian Chromosomal Aberration Test 108
5.4.2 In vitro Mammalian Micronucleus Test 109
5.5 In Vivo Genetic Toxicology Tests 114
5.5.1 In Vivo Assays for Chromosomal Abnormalities 114
5.5.1.1 Mammalian Bone Marrow Chromosomal Aberration Assays 114
5.5.1.2 Rodent Dominant Lethal Test 114
5.5.1.3 In Vivo Mammalian Erythrocyte Micronucleus Assays 115
5.6 Additional Genotoxicity Assays Evaluating Primary DNA Damage 122
5.7 Summary and Discussion 129
6.0 Data Integration & Weight-of-Evidence Analysis Across Multiple Lines of Evidence
132
Page 5 of216
-------
6.1 Background 132
6.2 Dose-Response and Temporal Concordance 132
6.3 Strength, Consistency, and Specificity 133
6.4 Biological Plausibility and Coherence 135
6.5 Uncertainty 136
6.6 Evaluation of Cancer Classification per the 2005 EPA Guidelines for Carcinogen
Risk Assessment 138
6.6.1 Introduction 138
6.6.2 Discussion of Evidence to Support Cancer Classification Descriptors 141
6.7 Proposed Conclusions Regarding the Carcinogenic Potential of Glyphosate 143
7.0 Collaborative Research Plan for Glyphosate and Glyphosate Formulations 145
8.0 References 147
Page 6 of216
-------
List of Acronyms
ADME: Absorption, Distribution, Metabolism, and Excretion
AHS: Agricultural Health Study
AOP: Adverse Outcome Pathway
AMPA: Aminomethylphosphonic Acid
BrdU: Bromodeoxyuridine
CA: Chromosomal Aberration
CARC: Cancer Assessment Review Committee
CBPI: Cytokinesis Block Proliferation Index
CHL: Chinese Hamster Lung
CHO: Chinese Hamster Ovary
CPRC: Carcinogenicity Peer Review Committee
EFSA: European Food Safety Authority
EPSPS: 5-enolpyruvylshikimate-3-phosphate synthase
FAO: Food and Agriculture Organization
FIFRA: Federal Insecticide, Fungicide, and Rodenticide Act
FISH: Fluorescence in situ Hybridization
GC-MS: Gas Chromatography-Mass Spectrometry
HL: Hodgkin Lymphoma
HPLC: High-Performance Liquid Chromatography
HPRT: Hypoxanthine-Guanine Phosphoribosyl Transferase
IARC: International Agency for Research on Cancer
JMPR: Joint FAOWHO Meeting on Pesticide Residues
MGUS: Monoclonal Gammopathy of Undetermined Significance
MN: Micronuclei
MOA: Mode of Action
MPCE: Micronucleated Polychromatic Erythrocytes
MRID: Master Record Identifier (a unique number assigned to each study submitted to EPA)
MTD: Maximum Tolerated Dose
NB: Nuclear Bud
NCR: National Research Council
NHL: Non-Hodgkin Lymphoma
NPB: Nucleoplasmic Bridges
NTP: National Toxicology Program
OCSPP: Office of Chemical Safety and Pollution Prevention
OECD: Organization for Economic Cooperation and Development
OPP: Office of Pesticides Program
PCE: Polychromatic Erythrocytes
PK: Pharmacokinetic
PPE: Personal Protective Equipment
PWG: Pathology Work Group
RED: Registration Eligibility Decision
ROS: Reactive Oxygen Species
SAP: Scientific Advisory Panel
SCE: Sister Chromatid Exchange
Page 7 of216
-------
SCGE: Single Cell Gel Electrophoresis
TAC: Total Antioxidant Capacity
TK: Thymidine Kinase
UDS: Unscheduled DNA Synthesis
USGS: United States Geological Survey
UV: Ultraviolet
WHO: World Health Organization
XPRT: Xanthine-Guanine Phosphoribosyl Transferase
Page 8 of216
-------
List of Figures
Figure 1.1. Source to outcome pathway (adapted from NRC, 2007)
Figure 1.2. Glyphosate agricultural usage (pounds applied annually) from 1987- 2014. Boxes
indicate years when glyphosate-resistant crops were introduced. Source: Proprietary Market
Research Data (1987 - 2014)
Figure 1.3. Map of estimated agricultural use for glyphosate in 1994 from USGS
Figure 1.4. Map of estimated agricultural use for glyphosate in 2014 from USGS
Figure 3.1. Study evaluation process for epidemiological studies
Figure 3.2. Forest plot of effect estimates (denoted as ES for effect sizes) and associated 95%
confidence intervals (CI) for non-Hodgkin lymphoma (NHL)
Figure 7.1. Results of HepG2 exposures following 24 hour incubation using different glyphosate
formulations
Page 9 of216
-------
List of Tables
Table 3.1 Epidemiology Study Quality Considerations
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and
Overall Ranking
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Table 3.4. Summary of Findings: Non-Solid Tumor Cancer Studies
Table 4.1. Testicular Interstitial Cell Tumors in Male Sprague-Dawley Rats (Lankas, 1981)
Cochran-Armitage Trend Test & Fisher's Exact Test Results
Table 4.2. Pancreatic Islet Cell Tumors in Male Sprague-Dawley Rats (Stout and Ruecker,
1990) Cochran-Armitage Trend Test & Fisher's Exact Test Results
Table 4.3. Historical Control Data — Pancreatic Islet Cell Adenomas in Male Sprague-Dawley
Rats (MRID No. 41728701)
Table 4.4. Hepatocellular Tumors in Male Sprague-Dawley Rats (Stout and Ruecker, 1990)
Cochran-Armitage Trend Test & Fisher's Exact Test Results
Table 4.5. Historical Control Data - Hepatocellular Tumors in Male Sprague- Dawley Rats
(MRID No. 41728701).
Table 4.6. Thyroid C-Cell Tumors in Male Sprague-Dawley Rats (Stout and Ruecker, 1990)
Cochran-Armitage Trend Test & Fisher's Exact Test Results
Table 4.7. Thyroid C-Cell Tumors in Female Sprague Dawley Rats Cochran-Armitage Trend
Test & Fisher's Exact Test Results (Stout and Ruecker, 1990)
Table 4.8. Historical Control Data - Thyroid C-Cell Tumors in Female Sprague-Dawley Rats
(MRID No. 41728701).
Table 4.9. Thyroid Non-Neoplastic Lesions (Stout and Ruecker, 1990)
Table 4.10. Liver Adenomas in Male Wistar Rats (Brammer, 2001) Cochran-Armitage Trend
Test and Fisher's Exact Test Results
Table 4.11. Mammary Gland Tumor Incidences in Female Rats (Wood et al., 2009a)
Fisher's Exact Test and Cochran-Armitage Trend Test Results
Table 4.12. Summary of Rat Carcinogenicity Studies
Table 4.13. Renal Tubular Cell Tumors in Male CD-I Mice (Knezevich and Hogan, 1983)
Cochran-Armitage Trend Test & Fisher's Exact Test Results
Table 4.14. Historical Control Data- Kidney tumors in CD-I Mice (TXR #0007252).
Page 10 of 216
-------
Table 4.15. Kidney Histopathological Alterations in Male CD-I Mice (Knezevich and Hogan,
1983)
Table 4.16. Hemangiosarcomas in Male CD-I Mice (Atkinson, 1993b) Cochran-Armitage Trend
Test and Fisher's Exact Test Results
Table 4.17. Lung Tumors in Male CD-I Mice (Wood et al., 2009b) Fisher's Exact Test and
Cochran-Armitage Trend Test Results
Table 4.18. Malignant Lymphomas in Male CD-I Mice (Wood et al., 2009b) Fisher's Exact
Test and Cochran-Armitage Trend Test Results
Table 4.19. Hemangioma Incidences (Sugimoto, 1997) Fisher's Exact Test and Cochran-
Armitage Trend Test Results
Table 4.20. Summary of Mouse Carcinogenicity Studies
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical
Table 5.2. In vitro Mammalian Gene Mutation Assays: Glyphosate Technical
Table 5.3. In vitro Tests for Chromosome Aberrations in Mammalian Cells- Glyphosate
Technical
Table 5.4. In vitro Tests for Micronuclei Induction in Mammalian Cells- Glyphosate Technical
Table 5.5. In Vivo Tests for Chromosomal Aberrations in Mammals- Glyphosate Technical
Table 5.6. In Vivo Tests for Micronuclei Induction in Mammals- Glyphosate Technical
Table 5.7 Assays for Detecting Primary DNA Damage- Glyphosate Technical
Page 11 of 216
-------
1.0 Introduction
1.1 Background
Glyphosate is a non-selective, phosphonomethyl amino acid herbicide registered to control
weeds in various agricultural and non-agricultural settings. The herbicide acts by inhibiting the
5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) enzyme, which is not present in
mammalian systems. Glyphosate was initially registered in 1974. Since then, several human
health analyses have been completed for glyphosate. In 1986, EPA issued the Glyphosate
Registration Standard which updated the agency's toxicity database for this compound. In 1993,
EPA issued the registration eligibility decision (RED) that indicated that glyphosate was eligible
for re-registration.
Currently, glyphosate is undergoing Registration Review1, a program where all registered
pesticides are reviewed at least every 15 years as mandated by the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA). The initial docket opening for glyphosate occurred in 2009 with
the publication of the human health scoping document and preliminary work plan2. As part of
this process, the hazard and exposure of glyphosate are reevaluated to determine its potential risk
to human and environmental health. Risks are assessed using current practices and policies to
ensure pesticide products can still be used safely. Registration Review also allows the agency to
incorporate new science. For human health risk assessment, both non-cancer and cancer effects
are evaluated for glyphosate and its metabolites, aminomethylphosphonic acid (AMPA) and N-
acetyl-glyphosate; however, this document will focus on the cancer effects only. EPA expects to
complete its complete human health risk assessment in 2017 that will include an assessment of
risk from anticipated exposures resulting from registered uses of glyphosate in residential and
occupational settings.
1.2 Evaluation of Carcinogenic Potential
Since its registration, the carcinogenic potential of glyphosate has been evaluated by EPA several
times. In 1985, the initial peer review of glyphosate was conducted in accordance with the
Proposed Guidelines for Carcinogen Risk Assessment. The agency classified glyphosate as a
Group C chemical (Possible Human Carcinogen), based on the presence of kidney tumors in
male mice. In 1986, the agency requested that the FIFRA Scientific Advisory Panel (SAP)
evaluate the carcinogenic potential of glyphosate. The panel determined that the data on renal
tumors in male mice were equivocal (only an increase in adenomas was observed and the
increase did not reach statistical significance). As a result, the panel recommended a Group D
chemical classification (Not Classifiable as to Human Carcinogenicity) for glyphosate and
advised the agency to issue a data call-in notice for further studies in rats and/or mice to clarify
the unresolved questions (FIFRA SAP Report, 1986)3.
1 Additional information on the Registration Review process can be found at: https://www.epa.gov/pesticide-
reevaluation/registration-review-process
2 Documents of the Registration Review can be found in the public docket at: EPA-HQ-OPP-2009-0361, accessible
at www.regulations.gov,,
3 Review available at: http://www.epa.gov/pesticides/chem_search/cleared_reviews/csr_PC-103601_24-Feb-
86_209.pdf
Page 12 of 216
-------
With the submission of two rat carcinogenicity studies following this data call-in, a second peer
review was conducted in 1991 by the agency's Carcinogenicity Peer Review Committee (CPRC)
to incorporate the new data. In accordance with the agency's 1986 Draft Guidelines for
Carcinogen Risk Assessment, the CPRC classified glyphosate as a Group E Chemical:
"Evidence of Non-Carcinogenicity for Humans" based upon lack of evidence for carcinogenicity
in mice and rats and the lack of concern for mutagenicity (TXR# 0008897).
Most recently, in September 2015, a third review was done by the Cancer Assessment Review
Committee (CARC). Relevant glyphosate data available to EPA at that time for glyphosate were
reevaluated, including studies submitted by the registrant and studies published in the open
literature. The agency performed this evaluation in support of Registration Review in
accordance with the 2005 Guidelines for Carcinogen Risk Assessment, classified glyphosate as
"Not Likely to be Carcinogenic to Humans" (CARC, 2015; TXR #0057299).
In recent years, several international agencies have evaluated the carcinogenic potential of
glyphosate. In March 2015, the International Agency for Research on Cancer (IARC), a
subdivision of the World Health Organization (WHO), determined that glyphosate was a
probable carcinogen (group 2A) (IARC, 2015). Later, in November 2015, the European Food
Safety Authority (EFSA) determined that glyphosate was unlikely to pose a carcinogenic hazard
to humans (EFSA, 2015). In May 2016, the Joint Food and Agriculture Organization
(FAO)/WHO Meeting on Pesticide Residues (JMPR), another subdivision of the WHO,
concluded that glyphosate was unlikely to pose a carcinogenic risk to humans from exposure
through the diet (JMPR, 2016). Some individual countries in Europe (e.g., France, Sweden)
have considered banning glyphosate uses based on the IARC decision, while other countries
(e.g., Japan, Canada, Australia, New Zealand) have continued to support their conclusion that
glyphosate is unlikely to pose a carcinogenic risk to humans.
The recent peer review performed by CARC served as an initial analysis to update the data
evaluation for glyphosate at that time. Based on an evaluation of the studies included in the
recent analyses by IARC, JMPR, and EFSA, the agency then became aware of additional
relevant studies not available to EPA. As a result, EPA also requested information from
registrants about studies that existed, but had never been submitted to the agency. The current
evaluation incorporates these additional studies. In addition, the agency conducted a systematic
review of the open literature and toxicological databases for glyphosate by using a "Framework
for Incorporating Human Epidemiologic & Incident Data in Health Risk Assessment". As such,
the current evaluation also provides a more thorough evaluation than the 2015 CARC review.
In December 2016, the FIFRA SAP was convened to evaluate the agency's Issue Paper
regarding the human carcinogenic potential of glyphosate. The panel's report was published in
March 2017 and the current document incorporates revisions based on the panel's report (G.
Akerman; 12-DEC-2017; TXR#0057689). Additionally, information from a recently published
analysis of glyphosate use and cancer incidence in the Agricultural Health Study (AHS) cohort
(Andreotti et al., 2017) with a longer follow-up than the previously published data (De Roos et
al., 2005) has been considered in this evaluation.
Page 13 of 216
-------
1.3 Overview of "Framework for Incorporating Human Epidemiologic & Incident Data
in Health Risk Assessment"
In 2010, the Office of Pesticide Programs (OPP) developed a draft "Framework for Incorporating
Human Epidemiologic & Incident Data in Health Risk Assessment" which provides the
foundation for evaluating multiple lines of scientific evidence in the context of understanding of
the mode of action (MOA)/adverse outcome pathway (AOP) (U.S. EPA, 2010). The draft
framework, which includes two key components, problem formulation and use of the MOA/AOP
pathway frameworks, was reviewed favorably by the FIFRA SAP in 2010 (FIFRA SAP, 2010).
In 2016, a final version of the framework was published4, which incorporated improvements
based on recommendations from the SAP, public comments, and the experience gained since
2010 conducting assessments on several pesticides for which epidemiological data were
available. Recently, EPA has applied this framework to the evaluation of atrazine and
chlorpyrifos5.
OPP's framework is consistent with updates to the World Health Organization/International
Programme on Chemical Safety MO A/human relevance framework, which highlights the
importance of problem formulation and the need to integrate information at different levels of
biological organization (Meek et al., 2014). Consistent with recommendations by the National
Research Council (NRC) in its 2009 report on Science and Decisions, OPP's framework
describes the importance of using problem formulation at the beginning of a complex scientific
analysis. The problem formulation stage starts with planning dialogue with risk managers to
identify goals for the analysis and possible risk management strategies. This initial dialogue
provides the regulatory context for the scientific analysis and helps define the scope of such an
analysis. The problem formulation stage also involves consideration of the available information
regarding the pesticide use/usage, toxicological effects of concern, and exposure pathways and
duration along with key gaps in data or scientific information. Specific to glyphosate, the
scoping document prepared for Registration Review (J. Langsdale et al., 2009) along with the
review conducted by the CARC (CARC, 2015) represent the problem formulation analyses for
the weight-of-evidence evaluation for carcinogenic potential. A summary of the US exposure
profile is provided in Section 1.4 to provide context for interpreting the various lines of evidence.
One of the key components of the agency's framework is the use of the MO A framework/ AOP
concept (Figure 1.1) as a tool for organizing and integrating information from different sources
to inform the causal nature of links observed in both experimental and observational studies.
Specifically, the modified Bradford Hill Criteria (Hill, 1965) are used to evaluate strength,
consistency, dose response, temporal concordance and biological plausibility of multiple lines of
evidence in a weight-of-evidence analysis.
4 https://www3.epa.gov/pesticides/EPA-HQ-OPP-2008-0316-DRAFT-0075.pdf
5 Chlorpyrifos Revised Human Health Risk Assessment for Registration Review; 29-DEC-2014; D424485.
U.S. EPA 2010 SAP Background White Paper - Re-evaluation of Human Health Effects of Atrazine: Review of
Experimental Animal and In Vitro Studies and Drinking Water Monitoring Frequency. EPA-HQ-OPP-2010-0125.
U.S. EPA 2011 SAP Issue Paper-Re-evaluation of Human Health Effects of Atrazine: Review of Cancer
Epidemiology, Non-cancer Experimental Animal and In Vitro Studies and Drinking Water Monitoring Frequency.
EPA-HQ-OPP-2011-0399.
Page 14 of 216
-------
Human Incidents Epidemiology
Figure 1.1. Source to outcome pathway (adapted from NRC, 2007).
1.4 Summary of the Exposure Profile in the United States
All pesticide products provide critical information about how to safely and legally handle and
use pesticide products. Pesticide labels are legally enforceable and all carry the statement "it is a
violation of Federal law to use this product in a manner inconsistent with its labeling." In other
words, the label is law. As a result, a key function of the pesticide product label is to manage the
potential risk from pesticides.
Labeled uses of glyphosate include over 100 terrestrial food crops as well as other non-
agricultural sites, such as greenhouses, aquatic areas, and residential areas. It is also registered
for use on glyphosate-resistant (transgenic) crop varieties such as corn, soybean, canola, cotton,
sugar beets, and wheat. Registered tolerances in the United States include residues of the parent
compound glyphosate and iV-acetyl-glyphosate, a metabolite found in/on glyphosate-tolerant
crops6.
Dietary (food and water) exposures are anticipated from applications to crops. Since there are
registered uses of glyphosate that may be used in residential settings, residential handlers may be
exposed to glyphosate during applications. Exposures may also occur from entering non-
occupational areas that have been previously treated with glyphosate. Occupational/commercial
workers may be exposed to glyphosate while handling the pesticide prior to application (mixing
and/or loading), during application, or when entering treated sites. The agency considers all of
the anticipated exposure pathways as part of their evaluation for human health.
Oral exposure is considered the primary route of concern for glyphosate. Oral absorption has
been shown to be relatively low for glyphosate (-30% of administered doses) with negligible
accumulation in tissues and rapid excretion (primarily unchanged parent) via the urine. Due to
its low vapor pressure, inhalation exposure to glyphosate is expected to be minimal. Dermal
penetration has also been shown to be relatively low for human skin (<1%) indicating dermal
exposure will only contribute slightly to a systemic biological dose. Furthermore, in route-
6 All currently registered tolerances for residues of glyphosate can be found in the Code of Federal Regulations (40
CFR §180.364).
Page 15 of 216
-------
specific inhalation and dermal toxicity studies, no adverse effects were observed. This all
suggests that there is low potential for a sustainable biological dose following glyphosate
exposure.
In residential/non-occupational settings, children 1-2 years old are considered the most highly
exposed subpopulation with oral exposures from dietary (food and water) ingestion and
incidental oral ingestion (e.g., hand-to-mouth activities) in treated areas. There is also potential
for dermal exposures in previously treated areas. Using OPP's standard exposure assessment
methodologies which are based on peer-reviewed and validated exposure data and models7, a
high-end estimate of combined exposure for children 1-2 years old is 0.47 mg/kg/day (see
Appendix E).
At the time of initial registration (1974), total use of glyphosate in the United States was
approximately 1.4 million pounds (Benbrook, 2016). In 1995, total use of glyphosate increased
to approximately 40 million pounds with agriculture accounting for 70% of use. With the
introduction of transgenic crop varieties in the United States circa 1996, (such as soybean,
cotton, and corn) use of glyphosate increased dramatically (Green and Owen, 2011), and in 2000
the total use of glyphosate in the United States was approximately 98.5 million pounds. By
2014, total annual use of glyphosate was approximately 280-290 million pounds (based on
Benbrook, 2016 and industry proprietary data accessible to EPA) with agriculture accounting for
90% of use. Although glyphosate use has continuously increased up to 2012, the stabilization of
glyphosate usage in recent years is due to the increase in a number of glyphosate-resistant weed
species, starting with rigid ryegrass identified in California in 1998 and currently totaling 16
different weed species in the United States as of March 2016. Figure 1.2 below provides a visual
representation of the increased agricultural use of glyphosate in the United States using
proprietary market research data from 1987-2014.
The increased use of glyphosate may be partly attributed to an increase in the number of farmers
using glyphosate; however, it is more likely that individuals already using glyphosate increased
their use and subsequent exposure. With the introduction of transgenic crop varieties, glyphosate
use shifted from pre-emergent to a combination of pre- and post-emergent applications.
Additionally, application rates increased in some instances and more applications were allowed
per year (2-3 times/year). Maps from the United States Geological Survey (USGS) displaying
glyphosate use in the United States indicate that although use has drastically increased since
1994, areas treated with glyphosate for agricultural purposes appear to be approximately the
same over time (Figures 1.3-1.4). The introduction of transgenic crops in some cases led to a
shift in crops grown on individual farms, such that more acreage within the farm would be
dedicated to growing the glyphosate-tolerant crops replacing other crops. In addition, during the
2000s there was also an increase in growing corn for ethanol production, which could also have
resulted in increased acreage dedicated glyphosate-tolerant corn.
7 Available: http://www2.epa.gov/pesticide-science-and-assessing-pesticide-risks/standard-operating-procedures-
residential-pesticide
Page 16 of 216
-------
300,000,000
Pounds
Al
Corn
Cotton
Alfalfa
and
Sugar
Soybean and
Canola
r-~ooaio*HOv-i88.06
I I No estimated use
Figure 1.3. Map of estimated agricultural use for glyphosate in 1994 from USGS
(http://water.usgs.gov/nawqa/pnsp/usage/maps/show_map.php ?year=1994&map=GLYPHOSATE&hilo=H)
Page 17 of 216
-------
Estimated Agricultural Use for Glyphosate , 2014 (Preliminary)
EPest-Hiah
Estimated use on
agricultural land, in
pounds per square mile
I l< 4.52
~ 4.52- 21.12
^¦21.13 - 88.06
M|> 88.06
I I No estimated use
Figure 1.4. Map of estimated agricultural use for glyphosate in 2014 from USGS
(http://water.usgs.gov/nawqa/pnsp/usage/maps/show_map.php ?year=2014&map=GLYPHOSATE&hilo=H)
The potential exposure to occupational handlers is dependent on the formulation, specific task
(mixer, loader, and/or applicator), rate of application, and acreage treated. Using HED's
standard occupational exposure assessment methodologies which are based on peer-reviewed
and validated exposure data and models8, mixer/loaders result in the highest potential exposure
estimates. Assuming no personal protective equipment (PPE), exposure estimates for
mixer/loaders range from 0.03-7 mg/kg/'day using the maximum application rate for high acreage
agricultural crops (6 lb ai/acre)9. For applicators, exposure would be lower with estimates
ranging from 0.02-0.03 mg/kg/day using the same application rate and acreage.
The maximum potential exposures from currently registered uses of glyphosate in residential and
occupational settings in the United States are used in the current evaluation to aid in the
determination of whether findings in laboratory studies are relevant for human health risk
assessment. In Sections 4.0 and 5.0, descriptions are provided for animal carcinogenicity and
genotoxicity studies, respectively. Results from these studies, particularly those administering
high doses, are put into context with the human exposure potential in the United States.
8 Available: https://www.epa.gov/pesticide-science-and-assessing-pesticide-risks/occupational-pesticide-handler-
exposure-data
9 Based on use information provided by the Joint Glyphosate Task Force for the following end-use products: EPA
Registration Nos.: 100-1182, 228-713,"524-343, 524-475, 524-537, 524-549, 524-579, 4787-23. and 62719-556.
Page 18 of 216
-------
1.5 Organization of this Document
In this analysis of the human carcinogenic potential of the active ingredient glyphosate, the
agency has performed a comprehensive analysis of available data from submitted guideline
studies and the open literature. This includes epidemiological, animal carcinogenicity, and
genotoxicity studies. Consistent with the framework described in Section 1.3, the agency has
evaluated these multiple lines of evidence and conducted a weight-of-evidence analysis.
Although there are studies available on glyphosate-based pesticide formulations, the agency is
soliciting advice from the FIFRA SAP on this evaluation of human carcinogenic potential for the
active ingredient glyphosate only at this time. The remainder of this document is organized by
the following:
• Section 2.0 Systematic Review & Data Collection Methods provides a description of
methods used to compile all relevant studies used in the current evaluation.
• Section 3.0 Data Evaluation of Epidemiology describes the available epidemiological
studies, evaluates relevant studies for study quality, and discusses reported effect
estimates.
• Section 4.0 Data Evaluation of Animal Carcinogenicity Studies provides a description
and evaluation of the available animal carcinogenicity studies for glyphosate.
• Section 5.0 Data Evaluation of Genetic Toxicity summarizes and discusses the various
genotoxicity assays that have been tested with glyphosate.
• Section 6.0 Data Integration & Weight of Evidence Analysis Across Multiple Lines of
Evidence integrates available data discussed in Sections 3.0-5.0 to consider concepts,
such as strength, consistency, dose response, temporal concordance and biological
plausibility in a weight-of-evidence analysis. This section also provides discussion of the
data in the context of cancer descriptors provided in the 2005 Guidelines for Carcinogen
Risk Assessment.
• Section 7.0 Collaborative Research Plan for Glyphosate and Glyphosate Formulations
provides a discussion of planned research that is intended to evaluate the role of
glyphosate in product formulations and the differences in formulation toxicity.
2.0 Systematic Review & Data Collection
In recent years, the National Academy of Sciences National Research Council (NRC) has
encouraged the agency to move towards systematic review processes to enhance the transparency
of scientific literature reviews that support chemical-specific risk assessments to inform
regulatory decision making (NRC, 2011). The NRC defines systematic review as "a scientific
investigation that focuses on a specific question and uses explicit, pre-specified scientific
methods to identify, select, assess, and summarize the findings of similar but separate studies"
(NRC, 2014). Consistent with NRC's recommendations, EPA's Office of Chemical Safety and
Pollution Prevention (OCSPP) is currently developing systematic review policies and
procedures. In short, OCSPP employs "fit for purpose" systematic reviews that rely on standard
methods for collecting, evaluating, and integrating the scientific data supporting the agency's
decisions. The concept of fit for purpose implies that a particular activity or method is suitable
for its intended use. Inherent in this definition is the concept that one size does not fit all
Page 19 of 216
-------
situations and thus flexibility is allowed. However, it is notable that with flexibility comes the
importance of transparency of documented processes; including the importance of transparency
and clarity in approaches to data collection, evaluation, and integration. These are described
throughout the document with data collection in Sections 2.1.1-2.1.2, evaluation in Sections 3-5,
and integration in Section 6.
As a result, more recent evaluations are starting to reflect this progression in the agency's
process. Similar to the framework for incorporating human epidemiologic and incident data,
systematic review begins with a problem formulation to determine the scope and purpose of the
search. Studies are considered based on their relevance to answer specific questions and those
studies deemed relevant are then further considered for their usefulness in risk assessment.
The agency strives to use high-quality studies when evaluating the hazard potential of pesticidal
chemicals and considers a broad set of data during this process. This includes registrant
generated studies required under FIFRA, as well as peer-reviewed scientific journals and other
sources, such as other governments and academia. A wide range of potential adverse effects are
assessed using acute, subchronic, chronic, and route-specific studies; predominately from studies
with laboratory animals, in addition to epidemiologic and human incident data. All studies are
thoroughly reviewed to ensure appropriate conduct and methodologies are utilized, and that
sufficient data and details are provided. In this way, hazards are identified and potential risks
characterized to ensure that decisions are informed by the best science available.
2.1 Data Collection: Methods & Sources
Data were collected by searching the open literature (Section 2.1.1) and other publicly available
sources (e.g., recent internal reviews, evaluations by other organizations) (Section 2.1.2).
Internal databases were also searched for submitted studies conducted according to Organization
for Economic Cooperation and Development (OECD) test guidelines, OCSPP harmonized test
guidelines, and other pesticide test guidelines (OPP guidelines) (Section 2.1.2).
It should be noted that glyphosate is primarily manufactured as various salts with cations, such as
isopropylamine, ammonium, or sodium. These salts are derivatives of the active substance
glyphosate and increase the solubility of technical-grade glyphosate acid in water. All of these
forms were considered for the current evaluation.
2.1.1 Open Literature Search
As part of the evaluation of the human carcinogenic potential of glyphosate, the literature review
described here uses concepts consistent with fit for purpose systematic review, such as detailed
tracking of search terms and which literature have been included or excluded. The primary goal
of the literature search was to identify relevant and appropriate open literature studies that had
the potential to inform the agency on the human carcinogenic potential of glyphosate. Therefore,
non-mammalian studies were not considered, and several terms were used in the search string in
an attempt to exclude non-mammalian studies.
Page 20 of 216
-------
To obtain literature studies, OPP worked with EPA librarians to conduct searches in PubMed,
Web of Science, and Science Direct. A search was conducted on May 6, 2016 in PubMed and
Web of Science using the following search string to yield 141 and 225 results, respectively:
((glyphosate OR "1071-83-6" OR roundup OR "N-(Phosphonomethyl)glycine") AND
(aneuploid* OR chromosom* OR clastogenic* OR "DNA damag*" OR "DNA adduct*" OR
genome* ORgenotoxic* ORmicronucle* OR cancer* OR carcinogen* OR oncogenic* OR
mutagen* OR cytotoxic* OR tumor* OR tumour* ORmalignanc* OR neoplasm* OR *oma))
NOT (fish* OR frog* OR tadpole* OR insect* OReco* OR amphibian* ORreptil* OR
invertebrate* OR fly OR flies OR aquatic OR bird* OR aqueous OR water OR yeast* OR worm*
OR earthworm* OR bacteria* OR lichen OR resist* OR "herbicide resist")
Due to differences with using Science Direct, the search string was slightly changed. This search
was also conducted on May 6, 2016 and yielded 459 results:
((glyphosate OR "1071-83-6" OR roundup OR "N-(Phosphonomethyl)glycine") AND
(aneuploid* OR chromosom* OR clastogenic* OR (DNA pre/2 (damag* OR adduct*)) OR
genome* ORgenotoxic* ORmicronucle* OR cancer* OR carcinogen* OR oncogenic* OR
mutagen* OR cytotoxic* OR tumor* OR tumour* ORmalignanc* OR neoplasm* OR *oma))
AND NOT (eco* OR fish* OR frog* OR tadpole* OR invertebrate* OR bird* OR insect* OR fly
OR flies OR amphibian* OR reptil* OR yeast* OR aquatic OR aqueous OR water OR worm*
OR earthworm* OR bacteria* OR lichen OR resist* OR "herbicide resist")
After cross-referencing the results obtained from the three open literature searches for duplicates,
a total of 735 individual articles were obtained (Appendix A) and one additional study (Alvarez -
Moya et al., 2014) not identified in the search was added to this list for a total of 736 individual
articles. Three staff members independently evaluated all of the studies and came to consensus
on which studies would be considered relevant to the issue of concern (i.e., human carcinogenic
potential of glyphosate). Many of the articles were not considered to be within the scope of the
search or not considered relevant in general (657 articles). Additionally, 27 articles were not
appropriate due to the type of article (i.e., correspondence, abstract only, not available in English,
retraction). Of the 52 relevant articles, 42 were used in the current evaluation (31 genotoxicity, 9
epidemiological, and 2 animal carcinogenicity). Three articles also reported on the potential of
glyphosate and its metabolites to be developed into therapeutic drugs for cancer treatment. The
remaining 7 articles evaluated effects on glyphosate or glyphosate formulations on cellular
processes, mostly focusing on epidermal cells, and were not considered informative for the
current evaluation.
2.1.2 Studies Submitted to the Agency
For all pesticides, there are toxicology data requirements that must be submitted to the agency
for registration. These studies, defined under the 40 CFR Part 158 Toxicology Data
Requirements, provide information on a wide range of adverse health outcomes, routes of
exposure, exposure durations, species, and lifestages. They typically follow OECD, OCSPP, or
OPP accepted protocols and guidelines, which ease comparisons across studies and chemicals.
Page 21 of 216
-------
The toxicological databases for glyphosate10 were reviewed and all relevant animal,
genotoxicity, and metabolism studies were collected for consideration.
Several resources were used to ensure all relevant studies were included in the current
evaluation. The list of studies obtained from the toxicological database and the open literature
search were cross-referenced with recent internal reviews (CARC, 2015; S. Recore el al., 2014).
This list was also cross-referenced with review articles from the open literature [Chang and
Delzell (2016), Greim et al. (2015), Kier and Kirkland (2013), Kier (2015), Mink etal. (2012),
Schinasi and Leon (2014), and Williams etal. (2000)]11. EPA requested studies from registrants
that were not previously available to the EPA. As a result, numerous studies were subsequently
submitted to the agency. Study reports for one animal carcinogenicity study and 17 genotoxicity
studies were not available to the agency and have been noted in the relevant sections below. For
these studies, data and study summaries provided in Greim et al. (2015) and Kier and Kirkland
(2013) were relied upon for the current evaluation.
2.2 Evaluation of Relevant Studies
Studies submitted to the agency are evaluating based on OECD, OCSPP, or OPP test guideline
requirements to determine whether studies are acceptable for use in risk assessment. In the
current evaluation, animal carcinogenicity, genotoxicity, and metabolism studies located in the
internal databases with access to full study reports were evaluated in this manner. Those
classified as unacceptable were noted and subsequently excluded from the current evaluation.
In order to evaluate open literature studies, criteria described in the OPP guidance for
considering and using open literature toxicity studies to support human health risk assessment
was utilized (U.S. EPA, 2012). This guidance assists OPP scientists in their judgement of the
scientific quality of open literature publications. More specifically, the document discusses how
to screen open literature studies for journal articles/publications that are relevant to risk
assessment, how to review potentially useful journal articles/publications and categorize them as
to their usefulness in risk assessment, and how the studies may be used in the risk assessment.
As with submitted studies, those deemed unacceptable were noted and subsequently excluded
from the current evaluation.
10 Glyphosate pesticide chemical (PC) codes: 103601, 103603, 103604, 103605, 103607, 103608, 103613, 128501,
and 417300.
11 All review articles, except Schinasi and Leon (2014), were funded and/or linked to Monsanto Co. or other
registrants.
Page 22 of 216
-------
3.0 Data Evaluation of Epidemiology
3.1 Introduction
Epidemiological studies are valuable for risk assessment since they may provide direct evidence
on whether human exposure to a chemical may cause cancer. Studies of high quality and
adequate statistical power are preferable and remove the need to account for extrapolation from
animals to humans or extrapolation from high to low doses. Epidemiological studies can also be
integrated with experimental evidence when determining or clarifying the carcinogenic potential
of a chemical for risk assessment. The key considerations in evaluating epidemiologic studies
are study design, exposure assessment, outcome assessment, confounding control, statistical
analyses, and risk of other bias.
OPP routinely evaluates the available epidemiological literature. As part of Registration Review
of glyphosate, an evaluation was initially conducted in 2014 (S. Recore et al., 2014) and
subsequently another evaluation was performed in 2015 (CARC, 2015). The 2015 evaluation
began with the epidemiological studies previously identified in the 2014 evaluation and included
three additional studies that were not included in the 2014 evaluation. These studies were
identified in review articles, included in the evaluation by IARC (2015), or were published since
the 2014 OPP evaluation. Both the 2014 and 2015 OPP evaluations considered the design and
overall quality of the epidemiological studies; however, formal study quality evaluations and
rankings were not conducted. In the current review, all of the studies in the 2015 report, as well
as additional epidemiological articles identified from a comprehensive search and cross-
referencing with available resources as described under Section 2.0, were considered in the
current evaluation. The following sections provide a description of how epidemiological studies
were evaluated for study quality and subsequent overall rankings, a summary of relevant studies,
and a discussion of the overall results.
3.2 Considerations for Study Quality Evaluation and Scope of Assessment
This section summarizes how specific study characteristics were factored into the determination
of a study's overall quality category. It should be noted that these study quality considerations
are specific to the issue of concern (i.e., carcinogenic potential of glyphosate). These
considerations are considered 'fit-for-purpose' under this context and could differ in another
regulatory or scientific context. Although the basic concepts apply broadly, the study quality
considerations are tailored specifically to studies investigating the association between
glyphosate exposure and cancer outcomes. As with all research studies, the design elements of
an epidemiological study have potential impacts on study quality and relevance to the research
question under investigation. Each study was, therefore, judged to be of high, moderate, or low
quality in each of the following six domains affecting study quality: study design, exposure
assessment, outcome assessment, confounder control, statistical analysis, and susceptibility to
bias (See Section 3.2.1 and Table 3.1 for general considerations under each domain). A similar
approach was recently used by OPP for the evaluation of epidemiological studies for
organophosphate pesticides (A. Lowit et al., 2015).
Page 23 of 216
-------
Primary literature and associated meta-analyses evaluating the association between glyphosate
exposure and a cancer outcome were the focus of this analysis. Reviews were only used to
identify individual studies that should be considered for study evaluation. Commentaries,
correspondence, and letters to the editor without original data were excluded. Of the relevant
studies identified, studies with the most complete analyses utilizing the greatest number of cases
and controls (e.g., pooled case-control studies) were evaluated for ranking (see Appendix B for
visual representation of these studies). If studies did not collect exposure information on
glyphosate from individual subjects, did not assess an outcome (e.g., biomonitoring studies),
and/or did not provide a quantitative measure of an association between glyphosate and a cancer
outcome, then these studies were assigned a low quality ranking and were not further evaluated
in detail (see Figure 3.1). A similar process was used by JMPR for their identification of
epidemiological studies for evaluating the carcinogenic potential of glyphosate and two other
pesticides (JMPR, 2016).
Artu les u)i^< led from litet,»tu;e scan h, nt^d in leview p.ifjfcrs, tami/or induced in urt isions
Yes
Not included m study quality ranking ewkution
NO
Yes
Ho
Yes
No
Yes
No
AsMj'nu? ksv ijj.iSiJv rnnkif»' urrhrj
im4 ii.y jii i\if\ Mnk,r > FiHkr
t-vaii itior^ ir» at rrt t uruku tod
:fH ludfri sh btusiv uuahly
evaluating nutt-a revaluation talk-
Assigned low quality ranking. Further
evaluation fn detail not conducted
Does the study assess a cancer outcome?
is the study a review article, commentary, correspondence, or tetter to the editor without original data?
»•> it the study with the most rumpled analyses utihrmgthe fue.ntubt number ot i .Tses and controls?
Does the study collect exposure information on glyphosate from individual subjects?
Does the study ev«>hiam glypbos,3t!' exposure jnd a urn^r nuh »nne individually for 3 potential qiuiit»lJtiv>»
m.MSUn- oJ mIiom •'
Detailed evaluation conducted tu aisi^n j^tj^Sif vr Forking
Figure 3.1. Study evaluation process for epidemiological studies.
3.2.1 Study Designs
In judging an individual study's contribution to the strength of evidence in the epidemiologic
literature base, the following general hierarchy of observational study designs was considered
(from most to least preferred): prospective cohort study (including nested case-control studies),
case-control study, and cross-sectional study. It is important to note, however, that this hierarchy
of study designs reflects the potential for the collection of high quality information (related to
exposure, outcome, confounders, and effect modifiers) and potential for efficient and valid
estimation of the true association. Thus, in deliberating on quality, care has been taken to
Page 24 of 216
-------
consider the circumstances and particulars of each individual study to consider whether the study
was well conducted independent of the type of study design.
The study designs used in the epidemiological literature reviewed were analytical and descriptive
studies. Cohort and case-control study designs are analytical studies used to evaluate relative
incidence of health and disease outcomes by exposure status. Cross-sectional and ecological
studies are generally considered descriptive or hypothesis-generating study designs; however,
they can also be used to test hypotheses regarding prevalence of health outcomes and, under
certain conditions, incidence as well.
Table 3.1. Epidemiological Study Quality Considerations3.
Parameter
High Score
Moderate Score
Low Score
Study Design
Cohort
Case-control
Cross-sectional/Ecological
Exposure Assessment
Questionnaire and/or
interview answered by
subjects for chemical-
specific exposure
Questionnaire and/or
interview for chemical-
specific exposure answered
by subjects or proxy
individuals
Low-quality questionnaire
and/or interview; information
collected for groups of
chemicals rather than
chemical-specific; no
chemical-specific exposure
information collected;
ever/never use of pesticides
in general evaluated
Outcome Assessment
State or National registries,
physicians, and/or special
surveillance programs with
cases verified by
histopathological evaluation
for tumors; appropriate
consideration of prevalent vs.
incident cases; analysis by
valid method specific for
biomarkers
State or National registries,
physicians, and/or special
surveillance programs
without histopathological
verification for tumors;
analysis by assays that are
less specific for biomarkers
of interest
No outcome evaluated;
unclear/no consideration for
whether prevalent or incident
cases are appropriate;
biomarker methods not
validated
Good control for important
confounders related to
cancer, standard
confounders, and known
confounders for glyphosate
and cancer outcomes (e.g.,
exposure to multiple
pesticides) through study
design or analytic control
with well measured co-
exposures (i.e., cumulative
exposure)
Moderately good control
for confounders related to
Confounder Control
cancer; standard variables
accounted for and; attempt
to control for known
confounders via a less
efficient measure of co-
exposure (e.g., ever/never
use)
No adjustments for
confounders
Statistical Analyses
Appropriate to study
question and design,
supported by relatively
adequate sample size,
maximal use of data,
reported well
Acceptable methods,
lower/questionable study
power or sample size
Minimal attention to
statistical analyses, sample
size evidently low,
comparison not performed or
described clearly
Page 25 of 216
-------
Table 3.1. Epidemiological Study Quality Considerations".
Parameter
High Score
Moderate Score
Low Score
Risk of (Other) Bias
Major sources of other
potential biases not likely
present, present but analyzed,
unlikely to influence
magnitude and direction of
effect estimate, no/low
potential of selection bias
Other sources of bias
present, acknowledged but
not addressed in study,
may influence magnitude
but not direction of
estimate, evidence of
potential selection bias
with low impact on effect
estimate
Major study biases present,
unacknowledged or
unaddressed in study, cannot
exclude other explanation for
study findings, evidence of
selection bias with high
potential to impact effect
estimate
a Overall study quality ranking based on comprehensive assessment across the parameters.
3.2.1.1 Analytical Studies
(1) Cohort Study
In a typical cohort study, such as the AHS, individuals are classified according to exposure status
(i.e., presence, absence, or magnitude of exposure) and then followed over time to quantify and
compare the development (i.e., incidence) of the health outcome of interest by exposure group.
Conceptually, the non-exposed comparison group in a cohort study provides an estimate of the
incidence of the outcome among the exposed, had they, counter-to-fact, not been exposed. Apart
from chance variations, a valid cohort study comparing exposed individuals to non-exposed
individuals provides an estimate of the relative risk (or rate) of the disease associated with
exposure. Ideally, the exposed and non-exposed groups are exchangeable, in the sense that
switching the exposed to non-exposed, and non-exposed to exposed would yield the same
measure of association (e.g., relative risk). If this were the case then, apart from chance, a cohort
study would yield a measure of association equivalent to that produced in a corresponding
(intervention) study where exposure status was randomly assigned.
The chief advantage of the cohort study design is that it affords the investigator the opportunity
to avoid and/or adjust for potential biases (i.e., selection bias, information bias, and
confounding); however, these biases may also be avoided in other well-designed study designs,
such as a case-control study. Cohort studies also allow for discernment of the chronological
relationship between exposure and outcome, and can be particularly efficient for studying
uncommon exposures. The primary disadvantage of the cohort study design is logistical
inefficiency with respect to the necessary time, expense, and other resources needed to conduct
them. Cohort studies are particularly inefficient for evaluating associations with rare outcomes
and diseases with long induction or latency periods. Case-control studies that are nested within a
cohort study (nested case-control studies) share the attributes of the cohort study and may be
more efficient. However, when follow-up throughout the study period is incomplete, the
potential for selection bias is increased, especially if follow-up rates are related to exposure
status.
Two sub-categories of cohort studies - prospective and retrospective - are often applied to
distinguish between studies in which the health outcome has occurred (retrospective study), or
has not occurred (prospective study) at the time the investigators initiate the study. This
distinction is important primarily as it pertains to the potential differences in the quality (e.g.,
Page 26 of 216
-------
completeness, accuracy, and precision) of information that can be ascertained by the
investigators, and also as it relates to potential sources of bias. Although not always true, the
prospective study design is considered the preferable of the two, as investigators can potentially
have more choices in determining how exposure, outcome, and covariate information is
collected. In a retrospective study conducted to evaluate the same hypothesis, by contrast, the
investigators would have to rely on exposure information based on self-reporting or historical
records. Such reporting is subject to (human) errors in recall, however when such errors are
uncorrected with disease state, there can be a bias towards the null due to random exposure
measurement error (information bias) and only when such errors are correlated with the disease
state can there be bias away from the null.
(2) Case-Control Study
In a typical case-control study, individuals are classified according to their outcome status (i.e.,
cases who have developed the outcome of interest, and controls who represent the population
from which the cases arise). The relative odds of exposure are then compared between cases and
controls. The primary advantage of case-control studies is that they are logistically efficient
relative to cohort studies, often being conducted at a fraction of the cost and in a fraction of the
time as a corresponding cohort study. Case-control studies can be used to examine associations
between multiple exposures and a given health outcome. They are particularly efficient for
evaluating rare outcomes, but are inefficient for studying uncommon exposures. An important
point to evaluate in each case-control study is the potential for selection bias, which arises if the
exposure distribution among the control subjects is not representative of the exposure
distribution among the population that gave rise to the cases. When participation rates between
cases and controls are low or distinctly imbalanced, the potential for selection bias is increased,
especially if participation rates are related to exposure status. Case-control studies that rely on
self-reported exposure measures are also potentially susceptible to information bias which could
result in bias towards the null or away from the null.
3.2.1.2 Descriptive Studies
Cross-sectional studies are used to evaluate associations between exposure and outcome
prevalence in a population at a single point in (or period of) time. The primary advantage of a
cross-sectional study is logistical efficiency. They are relatively quick and inexpensive to
conduct, as a long period of follow-up is not required, and exposure and outcome assessments
occur simultaneously. Cross-sectional studies have three primary potential disadvantages: 1)
potential difficulty in discerning the temporal relationships (i.e., whether the exposure precedes
the outcome); 2) estimating outcome prevalence rather than incidence of the outcome; and 3) the
possible overrepresentation of cases of the outcome with long duration relative to the average in
the population, and often with a better prognosis.
Ecological studies are used to evaluate associations between exposures and outcomes using
population-level rather than individual-level data. The primary advantages of ecological studies
are related to logistical efficiency, as they often rely on pre-existing data sources and require no
individual-level exposure, outcome, or covariate assessments. The primary weakness of the
ecologic study is the potential for confounding and resultant inappropriate extrapolation of
associations observed on the aggregate-level to associations on an individual level. The
Page 27 of 216
-------
discrepancy that associations observed at the population level are not observed at the individual
level is referred to as the ecological fallacy. Semi-ecological studies are less susceptible to the
ecological fallacy due to incorporation of individual-level data on outcomes and/or confounders.
The quality of these studies depends on the ability of the group exposure data to represent
individual exposure and the research question of interest.
3.2.2 Exposure Measures
As described in Section 3.2 and Figure 3.1, studies assigned a low quality ranking based on an
initial evaluation were not further evaluated in detail. In all of the studies included in the
analysis that were reviewed and ranked for study quality, exposure information was collected
from subjects and/or proxy individuals via questionnaires and/or interviews. These exposure
assessments typically include questions to determine the amount of direct pesticide use or to
collect information on behaviors and conditions associated with pesticide use (e.g., occupation,
tasks). This type of reporting likely misclassifies actual pesticide exposure. If conducted as part
of a prospective exposure assessment, these errors are likely to be non-differential with respect to
the outcome(s) of interest. In a retrospective assessment, the subject or proxy has knowledge of
the outcome; therefore, these errors may be differential or non-differential. Studies that
exclusively used subjects rather than including proxy individuals were considered more reliable
and given a higher weight given that the subjects would have a more accurate recollection of
their own exposure.
3.2.3 Outcome Measures
All of the studies evaluated in detail, except one, utilized state or national cancer registries,
physicians, and/or special surveillance programs to determine outcome status (i.e., subjects with
or without a cancer of interest). In several studies, the cases were also verified by
histopathological evaluation. Overall, outcome measures were relatively consistent across
studies and these assessments are likely to have minimal errors. The remaining study evaluated
in detail (Koureas et al., 2014) assessed oxidative DNA damage rather than a type of cancer. For
this evaluation, the oxidation by-product 8-hydroxydeoxyguanosine (8-OHdG) was measured by
enzyme immunoassay. This type of assay generally exhibits low specificity. More sensitive
quantitative methods are available to analyze genomic DNA for 8-OHdG by high-performance
liquid chromatography (HPLC) with electrochemical detection, gas chromatography-mass
spectrometry (GC-MS), and HPLC tandem mass spectrometry. Consideration of incident or
prevalent cases should also be carried out. By using only incident cases, there is greater
confidence that exposures occurred prior to the development of the outcomes. Inclusion of
prevalent cases can lead to an over-representation of cases with a long course of disease.
3.2.4 Confounding
The degree to which confounders were controlled varied across studies. Some studies adjusted
for particular medical variables, while others did not. Some standard variables, such as age,
geographical location, and sex, were either adjusted for analytically or by matching in case-
control studies. Several studies collected information on potential confounders; however, not all
of these variables were evaluated or results of the evaluation were not reported. The direction
and magnitude for confounders are, in general, difficult to determine because they are dependent
Page 28 of 216
-------
upon the relationship of each confounding factor with glyphosate and the cancer under
investigation. Several studies considered the potential for confounding from co-exposure to
other pesticides; however, only a few reported effect estimates between glyphosate exposure and
cancer risk adjusted for the use of other pesticides. Given most people in the epidemiological
studies who use pesticides occupationally will be exposed to multiple pesticides and, in some
instances, those other pesticides were observed to be risk factors for the same cancer, this is a
particularly important concern to address in either the study design or in the statistical analyses.
Across numerous studies, co-exposures to other pesticides was found to be positively correlated
with exposure to glyphosate and exposure to those other pesticides appear to increase the risk of
some cancers. As a result, the direction of confounding would be to inflate any true effect of
glyphosate in the absence of statistical control. This underlines the importance of adjusting for
co-exposures to other pesticides.
For NHL, other potential confounders, such as exposure to diesel exhaust fumes, solvents,
ultraviolet radiation, livestock, and viruses, have been identified. Some of these are more
plausible than others. For example, occupational exposure to diesel exhaust fumes (e.g.,
McDuffie et al., 2002; Karunanayake et al. 2008; Baris et al. 2001; Maizlish et al. 1998) and
solvents (Wang et al., 2009; Kato et al., 2005; Olsson and Brandt, 1988) are considered likely to
increase the risk of NHL. Agricultural workers are exposed to diesel fumes when using
agricultural vehicles when applying pesticides, such as glyphosate, and when using heavy
equipment during mixing, loading, and/or applying pesticides. Agricultural workers are also
exposed to solvents. Solvents are often used in pesticide products to aid the delivery of the
active ingredient and enhance efficacy. Solvents are also used for cleaning and
maintenance/repair of agricultural equipment used for mixing, loading, and/or applying
pesticides. With an association between exposure and outcome of interest, it is reasonable to
consider diesel exhaust fumes and solvents as probable confounders; however, neither of these
factors were accounted for in any of the studies evaluated in detail. There is also evidence that
ultraviolet (UV) radiation may increase the risk of NHL (Karipidis et al., 2007; Zhang et al.,
2007). As a result, there is a support that UV radiation is also a potential confounder given the
extended amount of time agricultural workers spend outside performing activities, including
those associated with pesticide use. Lastly, contact with farm and other animals has been
investigated as a suspected risk factor for hematopoietic and lymphoid tumors (McDuffie et al.,
2002). Hypothesized mechanisms to explain this association include viral transmissions, chronic
antigenic stimulation, and exposure to endotoxins, fungi, and mycotoxins. None of the
aforementioned potential confounders were accounted for in the studies evaluated in detail.
3.2.5 Statistical Analyses
Statistical analyses that were appropriate to the study question and study design, supported by
adequate sample size, maximized the use of available data, and were well characterized in the
report were weighted most highly. Acceptable statistical methods, questionable study power or
sample size, and analytical choices that resulted in the loss of information were given moderate
weight. Reports with only minimal attention paid to the conduct and reporting of the statistical
analyses were given the lowest weight.
3.2.6 Risk of Bias
Page 29 of 216
-------
The internal validity of the studies reviewed was judged by noting the design strategies and
analytic methods used in each study to constrain or eliminate selection bias, information bias,
and confounding. Selection bias can occur when the sampling of the population by the
investigator yields a study population that is not representative of the exposure and outcome
distributions in the population sampled. Put simply, selection bias occurs if selection of the
study sample yields a different estimate of the measure of association than that which would
have been obtained had the entire target population been evaluated. Although there are
numerous sources of selection bias, there are several mechanisms that may have induced
selection bias in the studies reviewed: low participation rates of eligible individuals due to non-
responsiveness or refusal (self-selection bias); loss to follow-up (i.e., failure to retain all study
participants initially enrolled in the study); and, in a case-control study, control selection bias
arising because the exposure distribution in the control sample does not represent the exposure
distribution of the study base (i.e., the population that gave rise to the cases or more formally, the
person-time experience of that population).
Information bias (also referred to as observation bias) arises when study participants are
incorrectly categorized with respect to their exposure or outcome status, or when errors arise in
the measurement of exposure or outcome, in the case of continuously distributed measures.
Epidemiologists often distinguish between two mechanisms or types of misclassification - those
that are non-differential (or random) and those that are differential (non-random). Non-
differential misclassification of exposure (or non-differential exposure measurement error)
occurs when the probability or magnitude of error in the classification or measurement of
exposure is independent of the outcome status of the study participants. Non-differential
exposure measurement error typically results in a bias towards the null which may obscure any
true effect of the exposure of interest. Similarly, non-differential misclassification of outcome
(or outcome measurement error) occurs when the probability or magnitude of error in the
assignment of outcome status or level is independent of exposure status. Non-differential
outcome measurement error typically does not cause bias but does decrease the precision of
effect estimates and therein inflates the width of confidence intervals. In contrast, differential
exposure misclassification (or measurement error) occurs when the error in the exposure
assignment is not independent of the outcome status. The mechanisms that cause non-
differential misclassification in the currently reviewed literature include random errors in
exposure recall from subjects or proxy respondents. The mechanisms that could induce
differential misclassification include recall bias and interviewer/observer bias. Note that
mismeasurement of confounders can result in residual confounding of the association of interest,
even when adjustment for that confounder has been conducted in the analysis.
Studies in which major sources of potential biases were not likely to be present, studies in which
potential sources of bias were present, but effectively addressed and analyzed to maximize the
study validity, and studies in which sources of bias were unlikely to influence the magnitude and
direction of the effect estimate were given more weight than studies where sources of bias may
be present, but not addressed in the study.
3.3 Review of Quality Results
Each study was judged to be of high, moderate, or low quality in each of the six domains
affecting study quality, as discussed above and in Table 3.1. The results of the quality
Page 30 of 216
-------
assessment are presented separately for each group below. The quality rankings presented are
specific to the current evaluation of the carcinogenic potential of glyphosate. As noted above
and in Table 3.2, several studies were not included in the ranking evaluation because they did not
represent the most complete analysis. Rather, the subjects were included in a larger analysis
(e.g., pooled case-control study) to produce a greater number of cases and controls (see
Appendix B for visual representation of these studies). For example, Cantor et al. (1992) was
not individually evaluated for ranking because the data from this study were pooled with data
from other studies in De Roos et al. (2003), which was included.
3.3.1 "High" Quality Group
Three studies were given a high quality ranking: De Roos et al. (2005), Eriksson et al. (2008),
and Koutros et al. (2013).
De Roos et al. (2005) was a prospective cohort study that evaluated associations between various
pesticide exposures, including glyphosate, and cancer incidence for numerous solid and non-
solid tumors in the AHS. The aim of the AHS is to evaluate the role of agricultural exposures in
the development of cancer and other diseases in the farming community. AHS recruited 52,934
licensed private pesticide applicators along with 32,345 of their spouses between 1993 and 1997.
In the first two phases of the study, the cohort also included 4,916 commercial pesticide
applicators from Iowa. As a prospective analysis of the AHS cohort, information was obtained
from exposed subjects at enrollment and no proxies were necessary. Exposure was evaluated as
ever/never use, cumulative lifetime exposure, and intensity-weighted cumulative exposure. Due
to the study design, the potential for many biases were reduced. Additionally, the study adjusted
and/or considered numerous factors, including use of other pesticides. Study participants
provided detailed pesticide exposure information prior to enrollment in the study and this
information has been incorporated into the study evaluation by determining tertile cut points and
calculating effect estimates by comparing to the lowest tertile. Additional evaluations with
quartiles and quintiles were performed for cancers with elevated effect estimates in the study and
for NHL. As noted earlier in this document, an analysis of the AHS cohort was recently
published (Andreotti etal., 2017) and the findings were considered as part of this evaluation.
Eriksson et al. (2008) was a population-based case-control study that recruited a consecutive
series of incident cases of NHL in several regions of Sweden from physicians treating lymphoma
within specified health service areas. Cases were verified pathologically and matched to
randomly selected controls from the national population registry by age, sex and health service
area. Exposure information was collected from exposed individuals (i.e., no use of proxy
respondents) using a comprehensive questionnaire including a total work history with in depth
questions about exposures to pesticides, solvents, and other chemicals. Interviewers were
blinded to case/control status. The study only reported minimal demographic information on
subjects (age and sex) and a table with subject characteristics (e.g., smoking status, alcohol
intake, physical activity, education) that could potentially be used to adjust effect estimates was
not provided. Glyphosate exposure was reported in 29 cases and 18 controls during the study
period. Multivariate analyses were adjusted for co-exposure to different agents, including
MCPA, "2,4,5-Y and/or 2,4-D", mercurial seed dressing, arsenic, creosote, and tar. An analysis
for a potential exposure-response relationship was also conducted; however, it was not clear
Page 31 of 216
-------
whether this analysis adjusted for co-exposure to other pesticides based on the statistical methods
description. The number of cases and controls were also not reported for this analysis.
Koutros et al. (2013) was a prospective cohort study within the AHS that evaluated the
association between pesticide use and prostate cancer. Exposure information was collected from
exposed subjects (no proxies necessary) through the enrollment questionnaires, as well as in a
follow-up questionnaire administered 5 years after enrollment. This study evaluated the
association between glyphosate and prostate cancer diagnoses from enrollment (1993-1997)
through 2007 resulting in a longer follow-up time than many of the other case-control studies
that utilized AHS subjects. The study used lifetime cumulative exposure and intensity-weighted
cumulative exposure metrics. Analyses were also conducted using untagged exposure and 15-
year lagged exposure, which excluded the most recent 15 years of exposure for both exposure
metrics. Although the effect estimate reported for glyphosate in this study was not adjusted for
co-exposure to other pesticides, additional analyses were not considered necessary since there
was no association observed.
3.3.2 "Moderate" Quality Group
Twenty-one case-control studies were assigned a moderate quality rating (Table 3.2). In general,
these studies share many study design characteristics. Exposure information was collected from
subjects and/or proxy individuals, the outcome measurement(s) utilized state/national registries
and surveillance programs, appropriate statistical analyses were performed, some covariates but
maybe not all relevant covariates were evaluated and/or considered, and risks of bias were
minimized to some extent. Sample sizes varied across studies. Case-control studies
investigating solid tumors included study populations in the United States and Canada. For non-
solid tumors, study populations were located in the United States, Canada, Sweden, France,
Germany, Italy, Ireland, Spain, and the Czech Republic. Although several nested case-control
studies shared most of the characteristics of the AHS cohort study, these studies were primarily
given a moderate quality ranking since co-exposure to other pesticides was not accounted for in
the analyses.
3.3.3 "Low" Quality Group
Seven case-control and 27 cross-sectional/ecological studies were assigned a low quality
ranking. All of these studies, except one case-control study (Cocco et al., 2013) and one
descriptive study (Koureas et al., 2014), were not subjected to a detailed evaluation because they
did not report a quantitative measure of an association between glyphosate exposure and a cancer
outcome, did not collect information on glyphosate exposure from all subjects, and/or did not
evaluate risk to a cancer outcome (Appendix D). In many instances, effect estimates were
reported only for total pesticide exposure. Additionally, exposure was assumed and glyphosate-
specific exposure information was not collected. In other studies, the aim of the study was to
assess exposure methods for epidemiological studies and/or to evaluate the impact of exposure
misclassification; therefore, there was no evaluation of a cancer outcome.
It should be noted that some of the studies assigned a low quality ranking in the current
evaluation were included in the recent evaluation by IARC. There were a number of descriptive
studies that evaluated the genotoxicity in human populations; however, these studies did not
Page 32 of 216
-------
meet the criteria for inclusion in the ranking as described in Section 3.2 and Figure 3.1. In most
instances, these studies reported effect estimates for total pesticide exposure and/or assumed
glyphosate exposure without collecting glyphosate-specific exposure information. For case-
control studies, Cocco et al. (2013), Dennis et al. (2010) and Ruder et al. (2004) were included in
the 2015 IARC evaluation, but were not considered informative in the current evaluation.
Detailed evaluations were not performed in the current evaluation for Dennis et al. (2010) and
Ruder et al. (2004) because a quantitative measure of an association between glyphosate and a
cancer outcome was not reported. Cocco et al. (2013) received a detailed evaluation and was
assigned a low quality ranking. This case-control study, which evaluated lymphoma risk across
six European countries, was not considered informative due to a combination of numerous
limitations in the study. The sample size of the study was low with only four cases and two
controls exposed to glyphosate. Control ascertainment was not consistent across countries, with
a mix of hospital- and population-based controls used. The overall participation rate for
population-based controls was found to be much lower than the overall participation rates of the
cases or hospital-based controls. Lastly, the study was limited to ever/never use of glyphosate
and did not adjust for confounders, in particular co-exposure to other pesticides. Although this
study was included in the IARC evaluation, IARC also stated that the study had very limited
power to assess the effects of glyphosate on risk of NHL.
The other study subjected to a detailed evaluation and assigned a low quality ranking was
Koureas et al. (2014). This cross-sectional study evaluated the association between glyphosate
exposure and oxidative DNA damage in 80 Greek pesticide sprayers. Although the study
reported a non-statistically significant effect estimate for glyphosate, it is limited in its ability to
contribute to the overall evaluation of the carcinogenic potential of glyphosate. The effect
estimate was not adjusted for any standard covariates or potential confounders, including co-
exposure to other pesticides. The sample size of the study was questionable. There were 80
subjects, but the number exposed to glyphosate was not reported. The outcome is measured
using an immunoassay that is less specific for measuring the biomarker of interest than other
available analytical methods. Lastly, the study evaluates primary DNA damage, but does not
measure the consequence of genetic damage. An increase in oxidative DNA damage may lead to
cell death or initiate DNA repair rather than lead to a mutation.
Due to the limitations in the studies assigned a low quality ranking, they do not provide reliable
information to evaluate associations between glyphosate exposure and cancer outcomes.
Therefore, the remaining sections of this document do not further discuss these studies except to
note when a study is included in meta-analyses.
Page 33 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
Alavanja et al. (2003)
This study was not included in the study quality ranking because the data were used in the updated analysis by Koutros et al. (2013).
Andreotti et al. (2009)
Nested Case-
control
Questionnaire answered
by subjects at study
enrollment followed by
take-home questionnaire;
examined exposure for
glyphosate as ever/never,
and intensity-weighted
cumulative exposure
days; spouses either self-
administered
questionnaire (81%) or
telephone interview
(19%)
State cancer registries
without histopathological
verification; exclusion of
subjects with prevalent
cancer at enrollment;
follow-up ~ 9 years
Adjusted for age,
smoking, and diabetes
for both exposure
metrics as well as
applicator type
forever/never exposure
metric
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
logistic regression
to obtain OR and
95% CI
Exposure
misclassification
particularly for spouses,
low response rate to take-
home questionnaire
(40%o) but unclear if
affected cases and
controls differently,
insufficient power for
pesticide exposure
interactions
Moderate
Bandef al. (2011)
Population-based
case-control
Males only
Self-administered
questionnaire answered
by subjects or proxies for
deceased subjects
requesting work history
and demographic
information; use of a job
exposure matrix to
estimate exposure to
pesticides
Cancer registry with
histopathological
verification; excluded
farmers that worked all
outside of British
Columbia; included
prostate cancer cases
prior to the PSA era
Adjustment for alcohol
consumption, cigarette
years, education level,
pipe years, and
respondent type.
Marital status and
ethnicity not
significant
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Conditional logistic
regression to obtain
ORs and 95% CIs
Recall bias, use of proxy
for deceased, exposure
misclassification,
participation rates cited
from another study, use
of cancer patients as
controls (excluding lung
and unknown cancer)
Moderate
Brown etal. (1990)
Pooled population-
based case-control
Males only
In-person interviews
using standardized
questionnaire with
subjects or proxies for
deceased/incapacitated;
supplementary
questionnaire
administered by
telephone for Iowa
subjects to obtain more
State cancer registry
(Iowa) and special
surveillance network
including hospitals and
pathology laboratories
(Minnesota); cases
ascertained
retrospectively and
prospectively (2 years
after start of study);
Adjusted for vital
status, age, state, ever
used tobacco daily,
close relative with
lymphopoietic cancer,
nonfarming job related
to risk of leukemia in
the study, exposure to
substances related to
risk in this study
Unconditional
logistic models to
obtain OR and 95%
CI; questionable
sample size (15
cases)
Recall bias; exposure
misclassification, use of
proxy respondents
Moderate
Page 34 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
detailed information
from those indicating
pesticide use
-26% of cases deceased
or too ill when identified
and ~15% deceased or
too ill at time of
interview;
histopathological
verification by
pathologists
(benzene, napthalene,
hair dyes)
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Brown etal. (1993)
Population-based
case-control
Males only
In person interviews with
standardized
questionnaire to obtain
detailed information on
farm activities and use of
pesticides from subjects
or proxies
State cancer registry
(Iowa) ascertained
retrospectively and
prospectively (2 years
after start of study);
-26% of cases deceased
or too ill when identified
and -15% deceased or
too ill at time of
interview;
histopathological
verification by
pathologists
Adjusted for vital
status and age;
smoking and education
evaluated and not
found to be significant
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Logistic models to
obtain OR and 95%
CI; questionable
sample size (11
cases)
Recall bias; exposure
misclassification, use of
proxy respondents
Moderate
Cantor etal. (1992)
This study was not included in the study quality ranking because the data were used in the pooled analysis conducted by De Roos etal. (2003).
Carreon et al. (2005)
This study was not included in the study quality ranking because the data were used in the pooled analysis conducted by Yiin et al. (2012).
Cocco etal. (2013)
European multi-
center case-control
Hospital-based and
population-based
(mixed for 2
countries, only
hospital-based for
the rest)
Trained interviewers
conducted in person
interviews using
structured questionnaire
answered by subjects;
those identified as
agricultural worker on
questionnaire given
subsequent questions
about pesticide use,
crops, etc.
Surveillance centers,
20% of slides from each
center reviewed by
pathologist
Adjustment for age,
sex, education, and
center.
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
logistic regression
to obtain ORs and
95% CIs; Low
sample size (4
cases, 2 controls)
Recall bias, selection
bias (low response rate
for population-based
controls and differed
from cases), exposure
misclassification, mix of
hospital- and population-
based controls,
Low
De Roos et al. (2003)
Population-based
case-control
Males only
Pooled analysis of
Interviews with subjects
or proxy for deceased
subjects. Different
interview techniques
across states. One study
collected information on
State cancer registries
(one state chose a
random sample, other
states chose all cases),
surveillance programs,
and hospitals without
Adjustment for age,
study site, and other
pesticides.
First degree relative
with haematopoietic
Logistic regression
and hierarchical
regression to obtain
ORs and 95% CIs
Recall bias, exposure
misclassification,, use of
proxy for deceased,,
varying quality of
questionnaire/ interview
techniques across studies
Moderate
Page 35 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
Cantor et al., 1992;
Hoar etal., 1986;
Zahmefa/., 1990
pesticide use and then
followed-up with
questions on selected
specific pesticides,
another study had a
direct question about a
selected list of specific
pesticides, and the last
study used an open ended
question without
prompting for specific
pesticides
histopathological
verification
cancer, education, and
smoking not found to
be important
confounders.
No adjustment for
other potential
confounders (e.g.,
solvents, diesel fumes,
UV radiation)
De Roos et al. (2005)
Prospective cohort
(licensed pesticide
applicators)
Questionnaire answered
by subjects at enrollment
and with subsequent
take-home questionnaire;
examined exposure as
ever/never, cumulative
lifetime days, and
intensity-weighted
cumulative exposure
days
State cancer registries
without histopathological
verification; follow-up
~7 years
Adjustment for state of
residence, age,
education, smoking
history, alcohol
consumption, family
history of cancer, use
of other common
pesticides
No adjustment for
other potential
confounders (e.g.,
solvents, diesel fumes,
UV radiation)
Poisson regression
to obtain RRs and
95% CIs
Major sources of
potential biases unlikely,
potential exposure
misclassification due to
any changes in exposure
since enrollment, follow-
up period may be limited
High
Engel etal. (2005)
Nested case-
control
Females only
Take-home questionnaire
from spouses of enrolled
applicators used to obtain
farm exposures, general
health information, and
reproductive health
history; Information
obtained from applicators
used as measure of
possible indirect
exposure to spouses
State cancer registries
identifying malignant
breast cancer; ~5 years
average follow-up time
Adjusted for age, race
and state.
Evaluated BMt, age at
menarche, parity, age
at first birth,
menopausal status, age
at menopause, family
history of breast
cancer, physical
activity, smoking,
alcohol consumption,
fruit and vegetable
consumption and
education but none
Poisson regression
to obtain RRs and
95% CIs
Exposure
misclassification,
exposure to other
pesticides (however no
association observed),
lack of information on
length of marriage could
result in overestimating
exposure based on
husband
Moderate
Page 36 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
found to be significant
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Eriksson et al. (2008)
Population-based
case-control
Questionnaire answered
by subjects; follow-up by
phone if incomplete
answers; excluded
exposures that occurred
during the same calendar
year and year before
diagnosis (cases) or
enrollment (controls);
minimal demographic
information reported
Physicians treating
lymphoma within
specified health service
areas and verified by
pathologists
Adjustment for age,
sex, year of
diagnosis/enrollment,
as well as exposure to
other pesticides in
multivariate analyses.
Not stated what
adjustments were
made for other
pesticides in latency
analyses.
No adjustment other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
logistic regression
and multivariate
analyses to obtain
ORs and 95% CIs;
not clear how
multivariate was
performed;
questionable sample
size (29 cases, 18
controls); also
included analysis of
<10 vs. >10 years
exposure
Recall bias, exposure
misclassification, lack of
subject demographics/
characteristics (e.g.,
smoking, alcohol
consumption, race, etc)
High
Flower et al. (2004)
Nested case-
control
Questionnaire answered
by applicators at
enrollment; spouses
enrolled through a
questionnaire brought
home by applicator;
females (applicators and
spouses) were asked to
complete a questionnaire
on female and family
health that collected
information on children
born during or after 1975
State cancer registry to
identify childhood cancer
cases (diagnosed from
birth through 19 yrs of
age) for children of
parents enrolled; hybrid
prospective/retrospective
ascertainment; excluded
female applicators
Child's age at parent's
enrollment was
included in model;
parental age at child's
birth, child's sex,
child's birth weight,
history of parental
smoking, paternal
history of cancer, and
maternal history of
miscarriage were
evaluated but not
found to be significant
and not included in
model
No adjustment for co-
Logistic regression
to obtain OR and
95% CI; calculated
standardized
incidence ratios to
compare observed
number of
childhood cancer
cases identified to
the expected
number;
low/questionable
sample size (6
parental cases, 13
maternal cases)
Exposure
misclassification, lack of
timing data to determine
if exposure occurred
prior to conception or
during pregnancy,
exposure to other
pesticides (however no
association observed and
lack of power for
adjustment)
Moderate
Page 37 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Hardell and Eriksson
(1999)
This study was not included in the study quality ranking because the data were used in the pooled analysis conducted by Hardell et al. (2002).
Hardell et al. (2002)
Population-based
case-control
Males only
Pooled analysis of
Hardell and
Eriksson 1999 and
Nordstrom etal.,
1998
Questionnaire answered
by subjects or proxy for
deceased subjects to
obtain complete working
history and exposure to
different chemicals;
follow-up with interview
for clarification
Registries with
histopathological
verification
Adjustment for age,
vital status, and county
(by matching).
Exposure to other
pesticides in
multivariate analysis.
No adjustment for
other potential
confounders (e.g.,
solvents, diesel fumes,
UV radiation)
Conditional logistic
regression to obtain
OR and 95% CI
(univariate and
multivariate
analyses).
Questionable
sample size (8
cases/8 controls)
Recall bias, exposure
misclassification, use of
proxy for deceased
Moderate
Hohenadel et al. (2011)
This study was not included in the study quality ranking because a more complete analysis was conducted by McDuffie et al. (2001).
Kachuri et al. (2013)
(extended analysis of
Pahwa etal. 2012)
Population-based
case-control
Males only
Questionnaire answered
by subjects or proxies;
pesticide use collected
via detailed telephone
interview on all
participants with 10+
hours of pesticide use
during lifetime and 15%
random sample of those
who did not; exposure
based on lifetime
exposure to glyphosate
Cancer registries or
hospitals in 6 Canadian
provinces with
histopathological
verification for 36.55%
of samples
Adjustment for age,
province, selected
medical conditions,
family history of
cancer, use of proxy
respondent, smoking
status
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
logistic regression
to obtain OR and
95% CI; trends
examined using
multiple logistic
regression
Recall bias, exposure
misclassification, control
selection based on three
different sources
depending on province of
residence, low
participation rates among
controls, use of proxy
respondents
Moderate
Karunanayake etal. (2012)
Population-based
case-control
Males only
Questionnaire answered
by subjects; pesticide use
collected via detailed
telephone interview on
all participants with 10+
hours of pesticide use
during lifetime and 15%
Cancer registries or
hospital in 6 Canadian
provinces with
histopathological
verification for 49% of
samples; difficulty
recruiting control
Adjusted for age,
province of residence,
and significant
medical history
variables
No adjustment for co-
Conditional logistic
regression to obtain
OR and 95% CI
Recall bias, exposure
misclassification, control
selection based on three
different sources
depending on province of
residence, low
participation rates among
Moderate
Page 38 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
random sample of those
who did not; exposure
based on lifetime
exposure to glyphosate
participants for older age
groups
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
controls, unable to
evaluate Epstein-barr
virus exposure
Koureas etal. (2014)
Cross-sectional
Questionnaire answered
by pesticide sprayers
Genomic DNA extracted
from peripheral blood
samples and oxidation
by-product 8-
hydroxydeoxyguanosine
(8-OHdG) was
determined by enzyme
immunoassay; more
specific methods (HPLC,
GC-MS) are available for
measurement
No adjustments. In
univariate,
occupational exposure,
sex and alcohol
consumption were
statistically significant
while DAP
concentrations and
smoking were not.
For univariate, chi-
square test used to
obtain RR and 95%
CI; 8-OHdG levels
transformed into
binary variables
(categorized as high
and low using the
75th percentile cut-
off); unknown
number of exposed
and unexposed
cases (questionable
sample size possible
given total number
of subjects is only
80)
Recall bias, did not
control for risk factors
identified as statistically
significant for univariate
analysis, does not
measure the consequence
of genetic damage
Low
Koutros et al. (2013)
Prospective cohort
Males only
Questionnaire answered
by subjects at study
enrollment; examined
exposure as cumulative
lifetime days and
intensity-weighted
cumulative exposure
days
State cancer registries
with histopathological
verification; total and
aggressive prostate
cancers evaluated
Adjustment for age,
state, race, smoking,
fruit servings, family
history of prostate
cancer, and leisure
time physical activity
in the winter.
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Poisson regression
to obtain RRs and
95% CIs; also
included unlagged
vs. lagged analysis
Exposure
misclassification
High
Landgren et al. (2009)
Nested case-
controP
Males only
Questionnaire answered
by subjects at enrollment
in AHS cohort and
subsequent take-home
questionnaire to collect
Venous blood collected
from antecubital vein and
analyzed for MGUS;
same method as used for
controls group in
Adjusted for age and
education level
Association with other
pesticides examined
Logistic regression
models to obtain
OR and 95% CI
comparing to
population-based
Exposure
misclassification, control
group not from
geographical area (used
control group with
Moderate
Page 39 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
information on 50
pesticides; occupational
expoures, medical
histories, and lifestyle
factors updated with 5-
year follow-up interview;
subjects with prior
history of
lymphoproliferative
malignancy excluded
Minnesota
and not found to be
significant so no
adjustment performed
No adjustment for
other potential
confounders (e.g.,
solvents, diesel fumes,
UV radiation)
screening study in
Olmsted County,
Minnesota;
questionable sample
size (27 cases; 11
controls)
similar demographics
from Minnesota)
Lee et al. (2004a)
This study was not included in the study quality ranking because the data were used in the pooled analysis conducted by De Roos etal. (2003).
Lee et al. (2004b)
Population-based
case-control
White males and
females only
Subjects or proxies were
interviewed by
telephone; those
living/working on a farm
asked for detailed history
of pesticide use and
farming information
State cancer registry or
review of discharge
diagnosis and pathology
records at 14 hospitals;
only newly diagnosed
cases with confirmed
adenocarcinoma of
stomach or esophagus
retained; controls
randomly selected from a
prior study conducted in
geographical area
Adjusted for age and
sex; evaluated BMI,
smoking, alcohol
consumption,
educational level,
family history of
stomach or esophageal
cancer, respondent
type, dietary intake of
particular vitamins and
minerals, protein, and
carbohydrates
(included in model if
changed value of OR
by more than 10%)
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
logistic regression
to obtain OR and
95% CI;
questionable sample
size (12 cases for
stomach; 12 cases
for esophagus)
Recall bias, exposure
misclassification, use of
proxy respondents,
control selection
Moderate
Lee etal. (2005)
Population-based
case-control
Questionnaire and/or
interview with subject or
proxy individuals to
collect information on
use of specific pesticides;
telephone follow-up for
unclear responses
Referral by hospitals or
through state cancer
registries with
histopathological
verification; controls
selected from a previous
study
Adjusted for age and
respondent type;
evaluated history of
head injury, marital
status, education level,
alcohol consumption,
medical history of
diabetes mellitus,
Unconditional
logistic regression
to obtain OR and
95% CI
Recall bias, exposure
misclassification, large
number of proxy
respondents, control
selection (historical
control group from
another cancer
evaluation, differences in
Moderate
Page 40 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
dietary intake of a- and
P-carotene, and dietary
fiber (included in
model if changed value
of OR by more than
10%)
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
exposure time period
evaluated, needed to add
younger controls,
exposure information
collected for different
time periods for cases vs.
controls)
Lee et al. (2007)
Nested case-
control
Questionnaire answered
by subjects at enrollment
in AHS cohort and
subsequent take-home
questionnaire to collect
information on 50
pesticides
State cancer registries
without histopathological
verification; follow-up ~
7 years
Adjustment for age,
smoking, state, total
days of pesticide
application
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Unconditional
multivariate logistic
regression to obtain
OR and 95% CI
Exposure
misclassification,
limited data on dietary
factors, NSAID drug use
and family cancer history
Moderate
McDuffie et al., 2001
Population based
case-control
Males only
Questionnaire answered
by subjects; pesticide use
collected via detailed
telephone interview on
all participants with 10+
hours of pesticide use
during lifetime and 15%
random sample of those
who did not; exposure
based on lifetime
exposure to glyphosate
Cancer registries or
hospital in 6 Canadian
provinces with
histopathological
verification for 84% of
samples; ascertainment
of cases stopped in each
province once target
numbers were reached
Adjustment for age,
province, and
significant medical
variables (including
history of cancer in
study participants and
family history).
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Conditional logistic
regression to obtain
OR and 95% CI
Recall bias, exposure
misclassification, control
selection based on three
different sources
depending on province of
residence, relatively low
participation rates
Moderate
Nordstrom et al., 1998
This study was not included in the study quality ranking because the data were used in the pooled analysis conducted by Hardell et al. (2002).
Orsi etal., 2009
Hospital-based
case-control
Data collection in 2
stages: 1) self-
Hospital catchment area
with histopathological/
Adjustment for age,
center, and
Unconditional
logistic regression
Recall bias, exposure
misclassification,
Moderate
Page 41 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
Males only
(occupationally
exposed)
administered
questionnaire on
socioeconomic
characteristics, family
medical history, and
lifelong residential and
occupational histories
and more specific
information for each job
held for at least 6
months, and 2) face-to-
face interview with
trained staff (blinded)
using standardized
questionnaire
cytological verification
Controls were hospital
based with no prior
history of lymphoid
neoplasms, excluding
patients with cancer or a
disease directly related to
occupation, smoking or
alcohol abuse (but
history of any of these
did not prevent selection
as a control)
socioeconomic
category. Education
and housing not found
to impact results. pju
immunization,
previous history of
mononucleosis, skin
type, smoking, and
drinking did not
change results.
Evaluated particular
crops and animal
husbandry as well.
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
to obtain OR and
95% CI.
Questionable
sample size (12
cases/24 controls)
hospital-based controls
Pahwa etal. (2011)
Population-based
case-control
Males only
Questionnaire answered
by subjects; pesticide use
collected via detailed
telephone interview on
all participants with 10+
hours of pesticide use
during lifetime and 15%
random sample of those
who did not; exposure
based on lifetime
exposure to glyphosate
Cancer registries or
hospitals in 6 Canadian
provinces with
histopathological
verification for 30% of
samples
Adjustment for age
group, province of
residence, and
statistically significant
medical history
variables
No adjustment for co-
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Conditional logistic
regression to obtain
OR and 95% CI;
trends examined
using multiple
logistic regression
Recall bias, exposure
misclassification, control
selection based on three
different sources
depending on province of
residence, low
participation rates among
controls
Moderate
Pahwaefa/. (2012)
Population-based
case-control
Males only
Questionnaire answered
by subjects; pesticide use
collected via detailed
telephone interview on
all participants with 10+
hours of pesticide use
during lifetime and 15%
random sample of those
Cancer registries or
hospitals in 6 Canadian
provinces with
histopathological
verification for 36.5% of
samples
Adjustment for age
group, province of
residence, and
statistically significant
medical history
variables
No adjustment for co-
Conditional logistic
regression to obtain
OR and 95% CI;
trends examined
using multiple
logistic regression
Recall bias, exposure
misclassification, control
selection based on three
different sources
depending on province of
residence, low
participation rates among
controls
Moderate
Page 42 of 216
-------
Table 3.2. Summary of Study Design Elements Impacting Study Quality Assignment and Overall Ranking.
Journal Article
Study Design
Exposure Assessment
Outcome Assessment
Confounder Control
Statistical Analyses
Risk of (Other) Bias
Overall
Ranking
who did not; exposure
based on lifetime
exposure to glyphosate
exposure to other
pesticides or other
potential confounders
(e.g., solvents, diesel
fumes, UV radiation)
Yiin et al. (2012)
Population-based
case-control
Pooled analysis of
men with women
analyzed in
Carreon et al.
(2005)
Questionnaire and/or
interview for chemical-
specific exposure
answered by subjects or
proxy individuals
Cases referred by
physicians or through
state cancer registries
with histopathological
verification; controls
matched within state, but
not county of residence
Adjustment for age,
education, sex, and,
sex, and farm
pesticide exposure
(yes/no)
No adjustment for
other potential
confounders (e.g.,
solvents, diesel fumes,
UV radiation)
Unconditional
logistic regression
to obtain ORs and
95% CIs
Acknowledge other
sources of bias. Recall
bias, exposure
misclassification, control
selection (low number of
deceased controls
obtained)
Moderate
a Mixed methods used in the Landgren et al (2009) study, with cross-sectional study design used to calculate prevalence rates comparing the AHS to a reference population MN.
Pesticide risk estimates (including glyphosate) calculated using nested case-control approach, comparing AHS exposed/unexposed (ever/never) study participants.
Page 43 of 216
-------
3.4 Assessment of Epidemiological Studies for Relevance to Analysis
Using the criteria summarized in Section 3.2, a total of 63 individual literature studies were
identified in the literature review and were judged as high, moderate, or low quality. The data
from 7 of these studies were used in pooled analyses by other studies; therefore, they were not
subjected to detailed evaluation. Overall, 3 studies, 19 studies, and 34 studies were assigned
high, moderate, or low rankings, respectively. All of the high and moderate quality studies were
considered relevant to the current evaluation. Additionally, the findings of a recently published
analysis of the AHS cohort (Andreotti el al., 2017) have been considered in this evaluation, when
appropriate.
The majority of the studies were case-control studies evaluating a wide-range of cancers in the
United States and Canada. There were several case-control studies from Canada that utilized the
same study population (Kachuri el al., 2013; Karunanayake el al, 2012; McDuffie el al, 2001;
Pahwa et al., 2011; Pahwa el al., 2012). In a similar fashion, numerous studies in the United
States were nested case-control studies, where the AHS cohort served as the source population
for selecting cases and controls (Andreotti et al., 2009; Engel et al., 2005; Flower et al., 2004;
Landgren et al., 2009; Lee et al., 2007). In these studies, a subset of the AHS cohort was
selected based on their outcome status for a particular cancer and exposure information was used
from the AHS enrollment questionnaire and/or during follow-up interviews. Nested case-control
studies allow for testing of hypotheses not anticipated when the cohort was initially assembled.
In the AHS prospective cohort studies (De Roos etal, 2005; Koutros etal, 2013; Andreotti et
al., 2017), exposure and demographic information were also obtained from the questionnaires at
enrollment; however, subjects were enrolled prior to developing cancer outcomes of interest.
Subjects were then followed from enrollment to a subsequent time point to determine if subjects
developed cancer outcomes of interest. As such, all available subjects in the cohort are included
in the evaluation of whether there was an association between a risk factor (e.g., glyphosate
exposure) and outcome.
The moderate studies included a varying degree of control for confounding and biases across
studies. As moderate studies, they encompass a combination of strengths and limitations. In
particular, important factors that impacted the quality assessment for these studies included
whether there was adjustment for known confounders, identification of control selection issues,
sample size issues, and length of follow-up. As noted previously, most people in these
epidemiological studies used pesticides occupationally and were exposed to multiple pesticides
over their working lifetime. Therefore, exposure to other pesticides is a particularly important
factor to adjust for and studies that made this adjustment were given more weight than those that
did not. Similarly, control selection issues were noted in a few studies and were given less
weighting than those without control selection issues. The issues ranged from concerns using
hospital-based controls, using different population sources to ascertain controls within the same
study, and appropriateness of using controls ascertained for another research question.
Numerous studies were limited by small sample sizes, which results in large confidence intervals
and reduces the reliability of the results to demonstrate a true association. Studies demonstrating
low or questionable sample size were therefore given less weighting. Lastly, the length of
follow-up time varied across studies.
Page 44 of 216
-------
3.5 Summary of Relevant Epidemiological Studies
A summary of the relevant studies evaluating the association between glyphosate exposure and
cancer are discussed below. Results of the studies reporting data on glyphosate exposure and
solid tumors (non-lymphohematopoietic) at various anatomical sites are presented in Table 3.3.
Results of the studies reporting data on glyphosate exposure and non-solid tumors
(lymphohematopoietic) are presented in Table 3.4. For study details, see Table 3.2 above and
Appendix C.
3.5.1 Solid Tumor Cancer Studies
(1) Cancer at Multiple Sites from the AHS Cohort
De Roos et a/., (2005) evaluated associations between glyphosate exposure and cancer incidence
of all cancers combined in the AHS cohort study and did not find an association [ever/never use
relative risk ratio (RR) =1.0 with 95% confidence interval (CI) of 0.90-1.2) when adjusting for
age, demographic and lifestyle factors, and exposure to other pesticides]. In addition, De Roos et
al., 2005 evaluated cancer at specific anatomical sites. Along with several nested case-control
studies, no statistical evidence of an association with glyphosate was observed at any specific
anatomical site (Table 3.3). Specifically, AHS researchers reported no evidence of an
association between glyphosate use and cancers of the oral cavity (De Roos et al., 2005), colon
(De Roos et al., 2005; Lee et al., 2007), rectum (De Roos et al., 2005; Lee et al., 2007), lung (De
Roos et al., 2005), kidney (De Roos et al., 2005), bladder (De Roos et al., 2005), pancreas (De
Roos et al., 2005; Andreotti etal., 2009), breast (Engel etal., 2005), prostate (De Roos et al.,
2005; Koutros et al., 2013) or melanoma (De Roos et al., 2005). The adjusted RR or odds ratio
(OR) and 95% CI for these studies are provided in Table 3.3.
Findings from the recently published analysis of the AHS cohort (Andreotti et al., 2017) with a
longer follow-up period than De Roos et al. (2005) also did not find associations between
glyphosate exposure and incidence of all cancers based on intensity-weighted lifetime days of
glyphosate use. Furthemore, there was no evidence of an association between glyphosate use
and cancers of the oral cavity, colon, rectum, pancreas, lung, melanoma, prostate, testes, bladder,
or kidney. Although there was evidence of a significant positive association in one quartile only
relative to intensity-weighted lifetime days of glyphosate exposure for pancreatic and lung
cancer, there was no evidence of a significant positive association in any other quartile for either
cancer type and the exposure-response trends were not statistically significant. As a result, these
isolated findings were not considered suggestive of an association.
(2) Prostate Cancer
In a Canadian population-based study (Band etal., 2011), researchers reported non-statistically
significant elevated odds of prostate cancer in relation to glyphosate use (OR=1.36; 95%
CI=0.83-2.25). There was no adjustment made for exposure to other pesticides. This study
included prostate cancer cases from 1983-1990, prior to the prostate-specific antigen (PSA) era.
Consequently, the study included more advanced tumors before diagnosis. The AHS related
studies (De Roos et al., 2005; Koutros et al., 2013; Andreotti et al., 2017), reflect PSA-era cases
Page 45 of 216
-------
(i.e., cases which are typically identified at an earlier stage in the progression of the disease) and
also did not identify an association with prostate cancer.
(3) Brain (Glioma) Cancer
Lee et al. (2005) investigated the association between brain cancer with farming and agricultural
pesticide use. Matching for age, sex, vital status, and region, study authors reported a non-
significant elevated odds of glioma (OR=1.5; 95% CI=0.7-3.1) in relation to glyphosate use by
male farmers; however, the results were significantly different between those who self-reported
pesticide use (OR=0.4; 95% CI=0.1-1.6), and for those for whom a proxy respondent was used
(OR=3.1; 95%) CI=1.2-8.2), indicating recall bias was a potential factor in this study.
Furthermore, there was no adjustment for co-exposure to other pesticides and issues noted with
control selection.
A population-based case-control study evaluated the risk of brain cancer, specifically, glioma
risk, among men and women participating in the Upper Midwest Health Study (Yiin et al.,
2012). Using a quantitative measure of pesticide exposure (in contrast to an ever-use metric),
Yiin et al. (2012) observed no statistical evidence of an association with glyphosate with effect
estimates roughly equal to the null value following adjustment for age, education, sex, and use of
other pesticides (home and garden use: OR=0.98; 95% CI=0.67-1.43; non-farm jobs: OR=0.83;
95% 0=0.39-1.73).
(4) Stomach and Esophageal Cancer
In a population-based case-control study in eastern Nebraska, Lee et al. (2004b) investigated
pesticide use and stomach and esophageal adenocarcinomas. There was no association observed
between glyphosate exposure and either stomach cancer (OR=0.8; 95% 0=0.4-1.5) or
esophageal cancer (OR=0.7; 95% 0=0.3-1.4) after adjustment for age and sex. No adjustment
was made for exposure to other pesticides.
(5) Soft Tissue Sarcoma
A Canadian case-control study (Pahwa et al., 2011) examined exposure to pesticides and soft
tissue sarcoma and found no relation with the use of glyphosate after adjustment for age,
province of residence, and medical history variables (OR=0.90; 95% CI= 0.58-1.40); however,
control selection issues were noted, including low response rate and selection from three
different sources depending on the province of residence.
(6) Total Childhood Cancer
Flower et al. (2004), a nested case-control study in the AHS cohort, examined the relation
between parental pesticide use and all pediatric cancers reported to state registries among
children of AHS participants and did not observe a significant association with maternal use
exposure to glyphosate (OR=0.61; 95% CI= 0.32-1.16) or paternal (prenatal) exposure to
glyphosate (OR=0.84; 95% CI= 0.35-2.54). The models adjusted for the child's age at the time
of parents' enrollment. There was no adjustment for exposure to other pesticides.
Page 46 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
All Cancers Combined
Ever/never
1.0(0.9-1.2)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.0 (0.9-1.1)
1.0 (0.9-1.1)
Age, demographic and lifestyle factors, and
other pesticidesb
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.9(0.8-1.0)
0.9(0.8-1.1)
Age, demographic and lifestyle factors, and
other pesticidesb
Lung
Ever/never
0.9(0.6-1.3)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
0.9(0.5-1.5)
0.7(0.4-1.2)
Age, demographic and lifestyle factors, and
other pesticidesb
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.1 (0.7-1.9)
0.6 (0.3-1.0)
Age, demographic and lifestyle factors, and
other pesticides'3
Oral Cavity
Ever/never
1.0(0.5-1.8)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
0.8(0.4-1.7)
0.8(0.4-1.7)
Age, demographic and lifestyle factors, and
other pesticidesb
Page 47 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.1 (0.5-2.5)
1.0 (0.5-2.3)
Age, demographic and lifestyle factors, and
other pesticides'3
Kidney
Ever/never
1.6(0.7-3.8)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
0.6 (0.3-1.4)
0.7(0.3-1.6)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.3(0.1-0.7)
0.5(0.2-1.0)
Age, demographic and lifestyle factors, and
other pesticidesb
Bladder
Ever/never
1.5(0.7-3.2)
Age, demographic and lifestyle factors, and
other pesticides'3
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.0 (0.5-1.9)
1.2 (0.6-2.2)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.5(0.2-1.3)
0.8(0.3-1.8)
Age, demographic and lifestyle factors, and
other pesticidesb
Page 48 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Melanoma
Ever/never
1.6 (0.8-3.0)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.2 (0.7-2.3)
0.9(0.5-1.8)
Age, demographic and lifestyle factors, and
other pesticidesb
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.6 (0.3-1.1)
0.7(0.3-1.2)
Age, demographic and lifestyle factors, and
other pesticidesb
Colon
Ever/never
1.4 (0.8-2.2)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.4 (0.9-2.4)
0.9(0.4-1.7)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.8(0.5-1.5)
1.4 (0.8-2.5)
Age, demographic and lifestyle factors, and
other pesticidesb
Lee et al. (2007)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
1.0(0.7-1.5)
Age, smoking, state, total days of pesticide
application
Rectum
Ever/never
1.3(0.7-2.3)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.3(0.7-2.5)
1.1 (0.6-2.3)
Age, demographic and lifestyle factors, and
other pesticides'3
Page 49 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.0 (0.5-2.0)
0.9(0.5-1.9)
Age, demographic and lifestyle factors, and
other pesticidesb
Lee et al. (2007)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
1.6(0.9-2.9)
Age, smoking, state, total days of pesticide
application
Colorectal
Lee et al. (2007)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
1.2 (0.9-1.6)
Age, smoking, state, total days of pesticide
application
Pancreas
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
0.7(0.3-2.0)
Age, demographic and lifestyle factors, and
other pesticidesb
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.6 (0.6-4.1)
1.3(0.5-3.6)
Age, demographic and lifestyle factors, and
other pesticidesb
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
2.5(1.0-6.3)
0.5(0.1-1.9)
Age, demographic and lifestyle factors, and
other pesticidesb
Andreotti et al. (2009)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
1.1 (0.6-1.7)
Age group, cigarette smoking, diabetes, and
applicator type
Intensity-Weighted Exposure Days
(by control median):
<184
>185
1.4 (0.9-3.8)
0.5(0.2-1.3)
Age group, cigarette smoking, and diabetes
Prostate
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
1.1 (0.9-1.3)
Age, demographic and lifestyle factors, and
other pesticides'3
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
0.9(0.7-1.1)
1.1 (0.9-1.3)
Age, demographic and lifestyle factors, and
other pesticidesb
Page 50 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.0 (0.8-1.2)
1.1 (0.9-1.3)
Age, demographic and lifestyle factors, and
other pesticidesb
Koutros etal. (2013)c
Prospective cohort
USA: Iowa and
North Carolina
Intensity-Weighted Cumulative Exposure
Days (by quartile):
Qi
Q2
Q3
Q4
Total prostate cancer:
0.91 (0.79-1.06)
0.96 (0.83-1.12)
1.01 (0.87-1.17)
0.99 (0.86-1.15)
Age, state, race, smoking, fruit servings,
family history of prostate cancer, and
leisure time physical activity in the winter
Intensity-Weighted Cumulative Exposure
Days (by quartile):
Qi
Q2
Q3
Q4
Aggressive prostate cancer:
0.93 (0.74-1.16)
0.91 (0.73-1.13)
1.01 (0.82-1.25)
0.94 (0.75-1.18)
Age, state, race, smoking, fruit servings,
family history of prostate cancer, and
leisure time physical activity in the winter
Bandef al. (2011)
Case-Control
Canada: British
Columbia
Ever/never
1.36 (0.83-2.25)
Alcohol consumption, cigarette years,
education level, pipe years, and respondent
type
Esophagus
Lee et al. (2004b)
Case-Control
USA: Nebraska
Ever/never
0.7(0.3-1.4)
Age and sex
Stomach
Lee et al. (2004b)
Case-Control
USA: Nebraska
Ever/never
0.8(0.4-1.5)
Age and sex
Breast
Engel etal. (2005)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
Wives who apply
pesticides:
0.9(0.7-1.1)
Wives who never used
pesticides:
1.3(0.8-1.9)
Age, race, and state of residence
Soft Tissue Sarcoma
Pahwaefa/. (2011)
Case-Control
Canada
Ever/never
0.90 (0.58-1.40)
Age group, province of residence, and
statistically significant medical history
variables
Page 51 of 216
-------
Table 3.3. Summary of Findings: Solid Tumor Cancer Studies
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Brain (glioma)
Overall:
1.5(0.7-3.1)
Leeetal. (2005)
Case-Control
USA: Nebraska
Ever/never
Self-reported:
0.4 (0.1-1.6)
Proxy respondents:
3.1 (1.2-8.2)
Age for overall analysis; age and
respondent type for other analyses
Yiin et al. (2012)
Case-Control
USA: Iowa,
Michigan,
Minnesota, and
Wisconsin
Ever/never
House/garden use:
0.98 (0.67-1.43)
Non-farm jobs:
0.83 (0.39-1.73)
Age, education, sex, and use of other
pesticides
Total Childhood
Flower et al. (2004)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
Maternal use:
0.61 (0.32-1.16)
Paternal use:
0.84 (0.35-2.34)
Child's age at enrollment
a Some studies report multiple quantitative risk measurements. This table reports the most highly adjusted quantitative measurements.
b De Roos et al. (2005) excluded subjects missing covariate data for demographic and lifestyle factors and exposure to other pesticides; therefore, the number of subjects included
in each analysis varies.
c Effect estimates for glyphosate reported in the supplemental web material for Koutros et al. (2013).
Page 52 of 216
-------
3.5.2 Non-Solid Tumor Cancer Studies
(1) Leukemia
De Roos et al. (2005) reported no association between leukemia and glyphosate-exposed
(ever/never used) pesticide applicators in the AHS cohort. For applicators with the full data set
(54,315), the RR was 1.1 (95% CI=0.6-2.4) with only adjustment for age. In the fully adjusted
model, the RR was similar (RR=1.0; 95% CI=0.5-1.9). The number of participants included in
the adjusted analysis was lower (n=40,716) due to the exclusion of subjects with missing
covariate data. Effect estimates using cumulative lifetime exposure and intensity-weighted
cumulative exposure were also found to be non-statistically significant and did not demonstrate a
trend with increasing exposure. In the recently published analysis of the AHS cohort with a
longer follow-up period (Andreotti et al., 2017), there was no association reported between
chronic lymphocytic leukemia/small lymphocytic lymphoma and chronic myeloid leukemia. For
acute myeloid leukemia, an elevated but non-statistically significant association was reported in
only one quartile relative to glyphosate exposure; however, there was a low number of observed
cases in each of the quartiles and the overall trend was not significant. There are no other studies
available evaluating acute myeloid leukemia. Given the limitations of the acute myeloid
leukemia analysis, the agency will continue to follow the literature regarding the association
between glyphosate exposure and risk of acute myeloid leukemia.
In a population-based case-control study in Iowa and Minnesota, Brown et al. (1990) did not
observe an association with the ever-use of glyphosate (OR=0.9; 95% CI=0.5-1.6). A limitation
in the study was the low number of cases exposed to glyphosate (n=15). Adjustments were made
for several covariates, including vital status, age, tobacco use, family history of lymphopoietic
cancer, high risk occupations, and high risk exposures; however, no adjustment was made for
exposure to other pesticides.
Chang and Delzell (2016) conducted a meta-analysis exploring glyphosate exposure and
leukemia using 3 studies (De Roos et al., 2005; Brown et al., 1990; and Kaufman et al., 2009).
I2 values were reported, which represented the percentage of the total variance explained by
study heterogeneity and measure inconsistency in results. Larger I2 values indicate greater
inconsistency. A meta-risk ratio of 1.0 (95% CI=0.6-1.5) was obtained with an I2 value of 0.0%,
indicating consistency across the data sets. It should be noted that this analysis included data
from Kaufman et al. (2009), which is not considered in the current evaluation because it was
assigned a low quality ranking because a quantitative measure of an association between
glyphosate and a cancer outcome was not reported for that study.
(2) Multiple Myeloma
In a follow-up analysis of the study population from Iowa and Minnesota used in Brown et al.
(1990), Brown et al. (1993) investigated whether pesticide use was related to multiple myeloma.
Among men in Iowa, the authors observed a non-statistically significant elevated association
with glyphosate use (OR=1.7; 95% CI=0.8-3.6; 11 exposed cases); however, no adjustment was
made for exposure to other pesticides. The authors cautioned that while the study may lend
Page 53 of 216
-------
support to the role of pesticides in general, the study limitations preclude use of the evidence as a
definitive finding for any one compound.
De Roos et al. (2005) reported a suggestive association between multiple myeloma and
glyphosate-exposed pesticide applicators based on 32 multiple myeloma cases observed in the
AHS cohort. For applicators with the full data set, the RR was 1.1 (95% CI=0.5-2.4) with only
adjustment for age. In the fully adjusted model excluding subjects with missing covariate data,
there was a non-statistically significant elevated risk following adjustment for age, demographic
and lifestyle factors, and exposure to other pesticides (RR=2.6; 95% CI=0.7-9.4). The authors
postulated that the increased myeloma risk could be due to bias resulting from a selection of
subjects in adjusted analyses that differed from subjects included in unadjusted analyses or may
be due to a confounder or effect modifier that is prevalent among the subgroup and has not been
accounted for in the analyses. When exposure data were also stratified by tertiles with the lowest
tertile of exposure as the referent category, trend analyses were not statistically significant. Non-
statistically significant elevated RRs of 1.9 (95% CI: 0.6-6.3) and 2.1 (95% CI: 0.6-7.0) were
estimated for the highest tertile of both cumulative and intensity-weighted exposure days,
respectively. The study authors did note that small sample size precluded precise estimation
(n=19 for adjusted analyses). When using never exposed as the referent category, the trend
analysis was again non-statistically significant, but the RRs ranged from 2.3 (95% CI: 0.6-8.9) to
4.4 (95%) CI: 1.0-20.2) from the lowest tertile to the highest tertile, respectively. When stratified
by quartiles, a statistically significant trend is achieved and the RR increased to 6.6 (95% CI:
1.4-30.6); however, the authors noted that the cases were sparsely distributed for these analyses.
In the recently published analysis of the AHS cohort with a longer follow-up period and 88
exposed cases (Andreotti et al., 2017), there was no association observed between glyphosate
exposure and multiple myeloma.
Sorahan (2015)12 re-analyzed the AHS data reported by De Roos et al. (2005) to examine the
reason for the disparate findings in relation to the use of a full data set versus the restricted data
set. Using Poisson regression, risk ratios were calculated without excluding subjects with
missing covariate data. When adjusted for age and sex, the RR for ever-use of glyphosate was
1.12 (95%) CI of 0.5-2.49). Additional adjustment for lifestyle factors and use of other pesticides
did not have a large impact (RR=1.24; 95%> CI=0.52-2.94). The authors concluded that the
disparate findings in De Roos et al. (2005) could be attributed to the use of a restricted dataset
that was unrepresentative.
Landgren et al. (2009), within the AHS study population, also investigated the association
between pesticide use and prevalence of monoclonal gammopathy of undetermined significance
(MGUS). MGUS is considered a pre-clinical marker of multiple myeloma progression. The
authors did not observe an association with glyphosate use and MGUS using subjects from the
AHS cohort (OR=0.50; 95%> CI=0.20-1.0). No adjustment was made for exposure to other
pesticides.
In a population-based case-control study (Pahwa et al., 2012) among men in six Canadian
provinces, a non-statistically significant elevated odds of multiple myeloma was reported in
relation to glyphosate use (OR=1.22; 95%> CI = 0.77-1.93), based upon 32 glyphosate exposed
12 Funded by Monsanto
Page 54 of 216
-------
multiple myeloma cases and 133 controls. There was no adjustment for exposure to other
pesticides. In an extended analysis of these data, Kachuri etal. (2013), using the same Canadian
study population, further explored multiple myeloma in relation to days per year that glyphosate
was used. Adjustment for exposure to other pesticides was also not performed in this study. For
ever-use, there was a slight non-statistically significant increased odds ratio (OR=1.19; 95%
CI=0.76-1.87). For light users (>0 and <2 days/year), there was no association (OR=0.72; 95%
CI = 0.39-1.32; 15 exposed cases); whereas, for heavy users (>2 days/ year), there was a non-
statistically significant increased odds ratio (OR=2.04; 95% 0=0.98-4.23; 12 exposed cases).
Similar results were obtained when proxy respondents were excluded from the analysis. The low
number of cases and controls exposed to glyphosate, particularly when exposed subjects were
divided into light and heavy users, was a limitation of the study. It would be expected that effect
estimates would be reduced if adjustment for co-exposure to other pesticides had been
performed.
In a hospital-based case-control study conducted by Orsi etal. (2009) in France, 56 multiple
myleoma cases and 313 age- and sex-matched controls were identified. A non-statistically
significant elevated risk was observed (OR=2.4; 95% 0=0.8-7.3; 5 exposed cases and 18
exposed controls). The wide CI range can primarily be attributed to the low number of exposed
cases, which reduces the reliability of the results to demonstrate a true association. Additionally,
the study did not adjust for exposure to multiple pesticides.
Chang and Delzell (2016) conducted a meta-analysis exploring glyphosate exposure and multiple
myeloma using data from the 6 studies described above (Brown et al., 1993; De Roos et al.,
2005; Sorahan, 2015; Pahwae^a/., 2012; Kachuri etal., 2013; Orsi et al., 2009). Meta-risk
ratios were obtained using data from each of the 4 independent study populations, such that if a
study population was already represented in the analysis by one study, then the same population
analyzed by another study would not be included (e.g., Sorahan, 2015 and De Roos et al., 2005
could not be used simultaneously in a meta-analysis). The combined meta-risk ratio based on
data from prioritized studies (Brown etal., 1993; Kachuri etal., 2013; Orsi etal., 2009; and
Sorahan, 2015) was 1.4 (95% 0=1.0-1.9) using random-effects and fixed-effects models and the
I2 value = 0.0% indicating consistency across data sets. There was relatively no impact on the
meta-risk ratio and associated 95% CI when secondary analyses were conducted using
alternative estimates for a study population (e.g., substituting the data from Sorahan, 2015 for De
Roos et al., 2005).
(3) Hodgkin Lymphoma
In a Canadian case-control study, Karunanayake etal., (2012) evaluated Hodgkin lymphoma
(HL) and observed no association with glyphosate exposure following adjustment for age,
province of residence, and medical history variables (OR=0.99; 95% 0=0.62-1.56; 38 cases).
No adjustment was made for exposure to other pesticides.
Page 55 of 216
-------
In a hospital-based case-control study conducted by Orsi et al. (2009) in France, authors
identified 87 HL cases and 265 age-and sex-matched controls. There was a non-statistically
significant elevated odds ratio observed (OR=1.7; 95% CI=0.6-5.0; 6 exposed cases and 15
exposed controls). The wide CI range can primarily be attributed to the low number of exposed
cases. Also, as noted earlier, this study did not adjust for exposure to multiple pesticides.
Chang and Delzell (2016) conducted a meta-analysis exploring glyphosate exposure and HL
using data from both of these studies. A meta-risk ratio of 1.1 (95% CI=0.7-1.6) was obtained
with a I2 value of 0.0%, indicating consistency across the data sets.
HL was also evaluated in the recently published analysis of the AHS cohort (Andreotti el al.,
2017) and no association was observed with glyphosate use; however, the number of cases
available for this analysis was limited.
(4) Non-Hodgkin Lymphoma
NHL has about 60 subtypes classified by the WHO, which may have etiological differences
(Morton et al., 2014). There are analyses available for particular subtypes of NHL; however,
these are particularly limited by the small sample sizes. As a result, this evaluation only presents
results for total NHL with the exception of the recently published analysis of the AHS cohort
(Andreotti et al., 2017) where sample sizes were not limited for all subtypes.
There were six studies available that investigated the association between glyphosate exposure
and NHL, which was the most for any type of cancer. As discussed in Section 3.4, these studies
encompass a combination of strengths and limitations. These studies are therefore discussed in
more detail in this section as compared to discussions of other cancer types in order to highlight
the strengths and identify the limitations for each study.
De Roos et al. (2005) was the only prospective cohort study available; therefore, subjects were
enrolled prior to developing cancer outcomes. Disease status was determined through state
cancer registries. Exposure information was obtained from a large number of licensed pesticide
applicators and no proxies were used. Exposure was evaluated as ever/never use, cumulative
lifetime exposure, and intensity-weighted cumulative exposure. Due to the study design, the
potential for many biases were reduced. Additionally, the study adjusted and/or considered
numerous factors, including use of other pesticides. Median follow-up time was approximately 7
years.; however, as discussed in Section 3.3.1, study participants provided exposure information
prior to enrollment and this information was incorporated into the cumulative lifetime and
intensity-weighted cumulative exposure metrics. As a result, the amount of time exposed was
longer than just the follow-up time since enrollment. For applicators with the full data set, the
RR for ever/never use was 1.2 (95% CI=0.7-1.9; 92 cases) with only adjustment for age. In the
fully adjusted model excluding subjects with missing covariate data, the RR was similar
following adjustment for age, demographic and lifestyle factors, and exposure to other pesticides
(RR=1.1; 95%) 0=0.7-1.9). Effect estimates obtained using cumulative lifetime exposure and
intensity-weighted cumulative exposure were below 1 (RR = 0.6-0.9 when comparing to the
lowest tertile). The recently published analysis of the AHS cohort with a longer follow-up
Page 56 of 216
-------
period of approximately 17.5 years (Andreotti et al., 2017) also reported no association between
glyphosate exposure and NHL overall or any of its subtypes.
De Roos et al. (2003) used pooled data from three case-controls studies evaluating NHL in white
males from Nebraska, Kansas, and in Iowa and Minnesota (Cantor et al., 1992; Hoar et al., 1986;
Zahm et al., 1990; Appendix B). Exposure information was obtained from exposed individuals
or their next of kin (i.e., proxy respondents) if the subjects were dead or incapacitated; however,
techniques varied across the three studies. There is potential for selection bias due to exclusion
of observations with missing covariate data, but only if the lack of the covariate data was
associated with glyphosate exposure. The effect estimates for the association between
glyphosate exposure and NHL was significant (OR=2.1; 95% CI=1.1-4.0) in the logistic
regression analyses adjusting for co-exposure to other pesticides. However, utilizing alternative
hierarchical regression techniques to adjust for co-exposure to other pesticide exposures, the
odds ratio was still elevated, but the increase was not statistically significant (OR=1.6;
95% C 1=0.90-2.8).
Eriksson et al. (2008) is a Swedish case-control study that used detailed exposure information
from exposed individuals (i.e., no use of proxy respondents), but only minimal demographic
information was provided on subjects (age and sex) and a table with subject characteristics (e.g.,
smoking status, alcohol intake, physical activity, education) was not provided. Cases were
identified through physicians and verified histopathologically. Glyphosate exposure, which was
reported in 29 cases and 18 controls between 1999 and 2003, produced a statistically significant
increased OR in the univariate analysis (OR=2.02; 95% CI= 1.10—3.71); however, in the
multivariate analysis adjustments were conducted for co-exposure to different agents including
MCPA, "2,4,5-Y and/or 2,4-D", mercurial seed dressing, arsenic, creosote, and tar and the OR
reduced to 1.51 (95% CI=0.77-2.94) and was not statistically significant. Additional analyses
were conducted to investigate the impact of various exposure times. When exposure was for
more than 10 cumulative days (the median number of days among exposed controls), the OR was
2.36 (95%) CI=1.04-5.37; 17 exposed cases) and for exposure less than 10 cumulative days, the
OR was 1.69 (95% 0=0.7-4.07; 12 exposed cases). By dividing the exposed cases and controls
using this exposure metric, wider CIs were observed due to smaller sample sizes, which reduces
the reliability of the results to demonstrate a true association. Additionally, these analyses did
not account for co-exposure to other pesticides. Similarly, wider CIs were also observed when
exposed cases and controls were divided by a longer exposure metric. ORs of 1.11 (95%
0=0.24-5.08) and 2.26 (95% 0=1.16-4.40) were obtained for 1-10 years and >10 years,
respectively. It was not clear whether this analysis adjusted for co-exposure to other pesticides
based on the statistical methods description and the subjects for each exposure group were not
reported. This finding, while limited to a single study, suggests that cohort studies without
sufficient follow-up time or other case-control studies which did not stratify by time since first
exposure may be less sensitive in detecting risk.
Hardell et al. (2002) used pooled data from two case-control studies in Sweden (Hardell and
Eriksson, 1999; Nordstrom etal., 1998; Appendix B) that examined hairy cell leukemia, a
subtype of NHL, and NHL (not including hairy cell leukemia). Exposure information was
collected from individuals or proxy respondents based on a working history with specific
questions on exposures to different chemicals. Cases were identified from regional cancer
Page 57 of 216
-------
registries and verified histopathologically. In the univariate analysis, risk of NHL associated
with glyphosate exposure was found to be significantly increased (OR=3.04; 95% CI=1.08-
8.52), but when study site, vital status, and co-exposure to other pesticides were considered in
the multivariate analysis, the OR noticeably attenuated and was found to be non-statistically
significant (OR=1.85; 95% CI=0.55-6.20). The wide range of the CI resulting from the small
sample size (only 8 glyphosate-exposed cases and 8 glyphosate-controls).
McDuffie etal. (2001) is a multicenter population-based study among men of six Canadian
provinces. This case-control study utilized a well-conducted exposure assessment and cases
were ascertained from cancer registries or hospitals in six provinces with histopathological
verification for 84% of the samples. There are concerns with control selection. There was low
control participation (48%) and different sources were used for selecting controls depending on
the province of residence. Effect estimates were obtained using a considerable number of
exposed cases and controls (51 cases and 133 controls); however, the study did not assess co-
exposure to other pesticides. There was a non-statistically significant increased risk of NHL
from glyphosate exposure when adjusting for age and province (OR=1.26; 95% CI=0.87-1.80)
and when adjusting for age, province and medical variables (OR=1.20; 95% CI=0.83-1.74).
Medical variables found to be statistically significant included history of measles, mumps,
previous cancer, skin-prick allergy tests, allergy desensitization shots, and a positive family
history of cancer in a first-degree relative. It would be expected that effect estimates would
attenuate if adjustment for co-exposure to other pesticides had been performed. Additional
analyses were conducted to investigate differences in exposure time. When exposure was for
more than 2 days/year, the OR was 2.12 (95% CI=1.20-3.73; 23 exposed cases and 36 exposed
controls) compared to unexposed subjects and for exposure more than 0 and < 2 days/year, the
OR was 1.00 (95%) CI=0.63-1.57; 28 exposed cases and 97 exposed controls) compared to
unexposed subjects.
Orsi et al. (2009) is a French hospital-based case-control study that obtained exposure
information from subjects (no proxies used) using a detailed questionnaire with lifelong
residential and occupational histories followed by a discussion with a trained interviewer who
was blinded to case status. No issues regarding exposure or outcome assessment were identified;
however, there is potential for selection bias given the study utilized hospital-based controls
(primarily from orthopedic and rhematological departments) that may not be representative of
the general population that gave rise to the cases. The study evaluated several potential
confounders; however, it did not assess co-exposure to other pesticides. There was no
association observed between NHL and glyphosate use (OR=1.0; 95% CI=0.5-2.2; 12 exposed
cases and 24 exposed controls). The low number of cases and controls exposed to glyphosate
and lack of adjustment for exposure to multiple pesticides were limitations of the study.
Schinasi and Leon (2014) conducted a meta-analysis exploring occupational glyphosate exposure
and NHL using data from six of the above mentioned studies (McDuffie et al., 2001; Hardell et
al., 2002; De Roos et al., 2003; De Roos et al., 2005; Eriksson et al., 2008; and Orsi et al.,
2009). Since the authors identified a variety of sources of heterogeneity between publications,
they decided a priori to calculate meta-risk ratio estimates and 95% CIs using random effect
models, allowing between study heterogeneity to contribute to the variance. I2 values were
reported as a measure of inconsistency in results. For glyphosate, the meta-risk ratio was 1.5
Page 58 of 216
-------
with a 95% CI of 1.1-2.0 and the I2 value was 32.7% indicating relatively low levels of
heterogeneity among these studies. This study combined multiple smaller studies that on their
own had limitations, including small sample sizes.
The 2015 IARC evaluation noted that fully adjusted effect estimates in two of the Swedish
studies (Hardell et al., 2002 and Eriksson et al., 2008) were not used in the analysis conducted
by Schinasi and Leon (2014). Consequently, the IARC Working Group conducted a
reexamination of the results of these studies (IARC 2015). For an association between
glyphosate exposure and NHL, the IARC estimated a meta-risk ratio of 1.3 (95% CI=1.03-1.65,
I2=0%; p=0.589 for heterogeneity).
Chang and Delzell (2016) conducted their own meta-analysis exploring glyphosate exposure and
NHL using six independent studies (De Roos et al., 2003; De Roos et al., 2005; Eriksson et al.,
2008; Hardell et al., 2002; McDuffie et al., 2001; and Orsi et al., 2009). A meta-risk ratio of 1.3
(95%) CI=1.0-1.6) was obtained with an I2 value of 0.0%>. In a secondary analysis, the De Roos et
al. (2003) OR using hierarchical regression was replaced by the logistic regression OR. This
change had no impact on the meta-risk ratio and associated confidence interval (meta-risk
ratio=1.3; 95% CI=1.0-1.6). In another secondary analysis, the OR from McDuffie et al. (2001)
was replaced by the OR from Hohenadel et al. (2011), which evaluated the same study
population (minus four previously misclassified NHL cases). This analysis also yielded similar
results (meta-risk ratio=1.3; 95% 0=1.0-1.7). A final analysis was performed with the
replacements for both secondary analyses [i.e., logistic regression OR from De Roos et al. (2003)
and OR from Hohenadel et al. (2011)]. The results were relatively the same as the other meta-
analyses (meta-risk ratio=1.4; 95% 0=1.0-1.8). Chang and Delzell (2016) also tested for
publication bias using Egger's linear regression approach to evaluating funnel plot asymmetry,
and found no significant asymmetry indicating little evidence of publication bias; however, given
the small sample size (n=6), this analysis would lack power and the results are not considered
meaningful.
Page 59 of 216
-------
Table 3.4. Summary of Findings: Non-Solid Tumor Cancer Studies.
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Leukemia
De Roos etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
1.0 (0.5-1.9)
Age, demographic and lifestyle factors, and
other pesticidesb
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.9(0.8-4.5)
1.0 (0.4-2.9)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.9(0.8-4.7)
0.7(0.2-2.1)
Age, demographic and lifestyle factors, and
other pesticides'3
Brown etal. (1990)
Case-Control
USA: Iowa and
Minnesota
Ever/never
0.9(0.5-1.6)
Vital status, age, tobacco use, family history
of lymphopoietic cancer, high occupations,
and high risk exposures
Multiple Myeloma
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
2.6 (0.7-9.4)
Age, demographic and lifestyle factors, and
other pesticides'5
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
1.1 (0.4-3.5)
1.9(0.6-6.3)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
1.2 (0.4-3.8)
2.1 (0.6-7.0)
Age, demographic and lifestyle factors, and
other pesticides'5
Brown etal. (1993)
Case-Control
USA: Iowa
Ever/never
1.7(0.8-3.6)
Age and vital status
Kachuri etal. (2013)
(extended analysis of
Pahwa2012)
Case-Control
Canada
Ever/never
1.19(0.76-1.87)
Age, province of residence, smoking status,
selected medical conditions, family history
of cancer, and use of a proxy respondent
Days per year of use:
0 to <2 days/year
>2 days/year
0.72(0.39-1.32)
2.04 (0.98-4.23)
Age, province of residence, smoking status,
selected medical conditions, family history
of cancer, and use of a proxy respondent
Pahwaetal. (2012)
Case-Control
Canada
Ever/never
1.22 (0.77-1.93)
Age group, province of residence, and
statistically significant medical history
variables
Page 60 of 216
-------
Table 3.4. Summary of Findings: Non-Solid Tumor Cancer Studies.
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Orsi et al. (2009)
Case-Control
France
Ever/never
2.4 (0.8-7.3)
Age, centre, and socioeconomic category
Sorahan (2015)
Reanalysis of De Roos
etal. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
1.12(0.5-2.49)
Age and sex
1.24 (0.52-2.94)
Age sex, lifestyle factors, and other
pesticides
Monoclonal Gammopathy of Undetermined Significance (MGUS)
Landgren et al. (2009)
Nested Case-Control
USA: Iowa and
North Carolina
Ever/never
0.5(0.2-1.0)
Age and education
Hodgkin Lymphoma (HL)
Karunanayake et al.
(2012)
Case-Control
Canada
Ever/never
0.99 (0.62-1.56)
Age group, province of residence, and
statistically significant medical history
variables
Orsi et al. (2009)
Case-Control
France
Ever/never
1.7(0.6-5.0)
Age, centre, and socioeconomic category
Non-Hodgkin Lymphoma (NHL)
De Roos et al. (2005)
Prospective Cohort
USA: Iowa and
North Carolina
Ever/never
1.1 (0.7-1.9)
Age, demographic and lifestyle factors, and
other pesticidesb
Cumulative Exposure Days
(by tertile cut points):
1-20
21-56
57-2,678
1.0
0.7(0.4-1.4)
0.9(0.5-1.6)
Age, demographic and lifestyle factors, and
other pesticides'3
Intensity-Weighted Cumulative Exposure
Days
(by tertile cut points):
0.1-79.5
79.6-337.1
337.2-18,241
1.0
0.6 (0.3-1.1)
0.8(0.5-1.4)
Age, demographic and lifestyle factors, and
other pesticidesb
De Roos etal. (2003)
Case-Control
USA: Iowa,
Nebraska,
Minnesota, and
Kansas
Ever/never
1.6(0.9-2.8)
Age, study site, and use of other pesticides
Eriksson etal. (2008)
Case-Control
Sweden
Ever/never
Multivariate:
1.51 (0.77-2.94)
Age, sex, year of diagnosis or enrollment,
and exposure to other pesticides
Days per year of use:
<10 days
>10 days
1.69 (0.70-4.07)
2.36(1.04-5.37)
Age, sex, and year of diagnosis or
enrollment
Years of use:
1-10 years
>10 years
1.11 (0.24-5.08)
2.26(1.16-4.40)
Unknown
Page 61 of 216
-------
Table 3.4. Summary of Findings: Non-Solid Tumor Cancer Studies.
Study
Study Design
Study Location
Exposure Metric
Adjusted Effect Estimate:
RR or OR (95% CI)a
Covariate Adjustments in Analyses
Hardell et al. (2002)
Case-Control
Sweden
Ever/never
Multivariate:
1.85 (0.55-6.20)
Study, study area, vital status, and exposure
to other pesticides
McDuffie et al. (2001)
Case-Control
Canada
Ever/never
1.20 (0.83-1.74)
Age, province of residence, and statistically
significant medical variables
Days per year of use:
>0 and < 2 days
>2 days
1.00 (0.63-1.57)
2.12(1.20 -3.73)
Age and province of residence
Orsi et al. (2009)
Case-Control
France
Ever/never
1.0(0.5-2.2)
Age, centre, and socioeconomic category
a Some studies report multiple quantitative risk measurements. This table reports the most highly adjusted quantitative measurements.
b De Roos et al. (2005) excluded subjects missing covariate data for demographic and lifestyle factors and exposure to other pesticides; therefore, the number of subjects included
in each analysis varies.
Page 62 of 216
-------
3.6 Discussion
A total of 63 individual studies were identified in the systematic review. The data from 7 of
these studies were used in pooled analyses by other studies; therefore, they were not subjected to
detailed evaluation. Overall, 3 studies, 19 studies, and 34 studies were assigned high, moderate,
or low rankings, respectively. All of the high and moderate quality studies were considered
informative with regard to the carcinogenic potential of glyphosate. Additionally, the recently
published analysis of the AHS cohort (Andreotti el al., 2017) was also considered as part of this
evaluation.
There was no evidence of an association between glyphosate exposure and solid tumors,
leukemia, or HL. These conclusions are consistent with those recently conducted by IARC,
EFSA, and JMPR who also concluded there is no evidence of an association for these tumors at
this time. The data should be considered limited though with only one or two studies available
for almost all of the cancer types investigated. The remainder of this discussion focuses on
multiple myeloma and NHL. Study elements for the available studies and their potential to
impact effect estimates are examined; however, the discussion is applicable in most cases to all
of the epidemiological studies used in this evaluation.
Multiple Myeloma
Four studies were available evaluating the association between glyphosate exposure and risk of
multiple myeloma in the initial evaluation presented to the SAP in December 2016 (Brown et al.,
1993; De Roos et al., 2005; Orsi et al., 2009; Pahwa et al., 2012). Since that time, a recent
analysis of the AHS cohort has been published (Andreotti et al., 2017), which included
evaluation of multiple myeloma. One reanalysis (Sorahan, 2015) and one extended analysis
(Kachuri et al., 2013) were also included in the evaluation. The effect estimates for ever/never
use ranged from 1.19 to 2.6 although none were found to be statistically significant. Only one
study (De Roos etal., 2005) adjusted for co-exposures to other pesticides; therefore, potential
confounding was not addressed in the other studies. There was an indication of a possible
exposure-response relationship; however, this was the only study that evaluated the exposure-
response relationship for multiple myeloma. Reanalysis of the full dataset by Sorahan (2015)
raised concerns about whether the restricted dataset used for these analyses was representative of
the whole cohort. Furthermore, in the recent analysis of the AHS cohort (Andreotti et al., 2017)
with a longer follow-up period and almost 5 times more exposed cases, there was no evidence of
an association between glyphosate exposure and risk of multiple myeloma. There was a single
study of MGUS, a precursor to multiple myeloma, which showed decreased risk with exposure
to glyphosate; however, the study did not adjust for exposure to other pesticides. Overall, the
available evidence does not link glyphosate exposure to multiple myeloma.
NHL
Six studies were available evaluating the association between glyphosate exposure and risk of
NHL in the initial evaluation presented to the SAP in December 2016. Since that time, a recent
analysis of the AHS cohort has been published (Andreotti et al., 2017), which included
evaluation of NHL. Effect estimates for ever/never use ranged from 1.0-1.85 in adjusted
Page 63 of 216
-------
analyses with none reaching statistical significance (Figure 3.2). Two of these studies did not
adjust for co-exposures to other pesticides (McDuffie el al., 2001; Orsi el al., 2009). Many of
the evaluated studies were limited by small sample sizes, which resulted in large confidence
intervals and reduced the reliability of the results to demonstrate a true association. Meta-
analyses were performed by IARC (2015) and Chang and Delzell (2016) using these results for
the ever/never use metric. Both analyses reported similar meta-risk ratios ranging from 1.3-1.5,
depending on the effect estimates and studies included in the analyses. Any of the meta-analysis
estimates that were statistically significant were all borderline with the lower limit of the 95% CI
just slightly over 1. For example, the lower 95% confidence limit reported by IARC (2015) was
1.03 and the lower 95% confidence limit displayed in Figure 3.2 generated by the agency is 1.01.
It should also be noted that publication bias may play a role in this evaluation given there is a
tendency to only publish positive results and potential concerns regarding glyphosate have only
been raised in recent years.
With respect to meta-analyses, caution should be taken when interpreting results. Meta-analyses
are a systematic way to combine data from several studies to estimate a summary effect.
Analyses were performed with 6 studies, which many would consider small for performing meta-
analyses. Rarely will meta-analyses synthesize data from studies with identical study designs
and methods. In the meta-analyses performed by IARC (2015) and Chang and Delzell (2016),
inclusion was primarily based on whether a study addressed the broader question regarding the
association between glyphosate exposure and risk of NHL. For meaningful results, careful
consideration of whether studies are similar and should be combined in the analysis.
Furthermore, the bias and confounding issues inherent for each individual study are carried over
into the meta-analyses. Across the NHL studies, study characteristics varied, such as overall
study design (i.e., cohort and case-control), source population, proxy respondent use, covariate
adjustments, and confounding control. Even if these differences are not detected statistically, the
meta-analysis estimate should be considered in the context of the data that are used to generate it.
Using cumulative lifetime and intensity-weighted cumulative exposure metrics, all effect
estimates were less than 1 (OR = 0.6-0.9 when comparing to the lowest tertile) in the AHS
cohort study (De Roos et al., 2005). Similar results were obtained in the recent analysis of the
AHS cohort (Andreotti el al., 2017). Two case-control studies (Eriksson el a!., 2008; McDuffie
et al., 2001) evaluated the association of glyphosate exposure and NHL stratifying exposure by
days per year of use. These studies obtained effect estimates greater than 1, which conflicted
with the results in the prospective cohort study; however, these estimates from the case-control
studies do not appear to be adjusted for co-exposures to other pesticides. By dividing the total
number of exposed cases and controls by these exposure metrics in Eriksson et al. (2008), wider
confidence intervals were observed due to small sample sizes, which reduces the reliability of the
results to demonstrate a true association. Furthermore, as mentioned previously (and will be
discussed further below), there was clearly strong potential for confounding from exposure to
other pesticides. In each instance where a study adjusted for co-exposure to other pesticides, the
adjusted effect estimate decreased in magnitude, including other analyses performed in one of
these case-control studies. Consequently, lack of adjustment for co-exposure to other pesticides
in these analyses could partially explain the conflicting results between the cohort and case-
control studies.
Page 64 of 216
-------
Sttiiy
E5(35%C!)
%
Weight
De Rcci it at (200i!
Be Rods tt si i200J,>
£ nisson et al iICPS;
Harder »' sL iiCC2,
McOulie et al COD 5}
Qra-iela" CSCSj
Overs? rt-squarsa « 0 o*«. p » a.;
V
I 60 !0 90 2 801
1 10(0.70. i 90)
1 51.0 77, 2,94;
I 85 (0.55, 8 20)
1,20(0 83,1.74)
1 90 (0 SO, 2.201
1 27(1.0'!, 1.53)
16.22
20.95
U.fri
3,58
38,12
9.52
fCO 00
BOTE. Weights are tow random effects snaiysv
-r—1 - -j-- - -j-
"> 3 5 10
Figure 3.2. Forest plot of effect estimates (denoted as ES for effect sizes) and associated 95% confidence
intervals (CI) for non-Hodgkin lymphoma (NHL).
The possible effect of confounding factors, which are related to both the exposure of interest and
the risk of disease, may make it difficult to interpret the results. Control for confounding varied
considerably across studies (Table 3.2). Studies primarily adjusted for standard variables, such
as age, gender, and residency location. Co-exposure to other pesticides was considered for
several of the NHL studies for ever/never use (De Roos et al., 2003; De Roos et al., 2005;
Eriksson etal., 2008; Hardell el al., 2002); however, analyses of exposure-response and latency
effects did not appear to adjust for these co-exposures. The recent analysis by Andreotti et al.
(2017) also adjusted for co-exposure to other pesticides.
There is clearly a strong potential for confounding by co-exposures to other pesticides since
many are highly correlated and have been reported to be risk factors for NHL. In the studies that
did report a quantitative measure adjusted for the use of other pesticides, the risk was always
found to be closer to the null than the risk calculated prior to this adjustment. For examples,
Eriksson etal. (2008) reported unadjusted and adjusted effect estimates of 2.02 (95% CI: 1.10-
3.71) and 1.51 (95% CI: 0.77-2.94), respectively. Comparing the magnitude of those effect sizes
on the natural log scale, the unadjusted effect was P=0.70 (95% CI: 0.10, 1.31) while the
adjusted effect was P=0.41 (95% CI: -0.26, 1.08), suggesting a difference compatible with a
degree of confounding by those herbicide co-exposures which appeared to have inflated the
unadjusted effect upwards by 70% on the natural log scale (or by 46% on the OR scale). This
demonstrates the profound effect this adjustment has on effect estimates and the concern for
residual confounding by other pesticides that cause NHL themselves. As discussed in Section
3.2.4, other potential confounders have also been identified. With an association between
glyphosate exposure and the outcome of interest, occupational exposure to diesel exhaust fumes,
solvents, livestock and other farm animals, and UV radiation are highly likely confounders in the
Page 65 of 216
-------
NHL studies; however, none of the studies accounted for these potential confounders. These
confounders and/or other unknown factors could explain the increased risk of NHL among
farmers, particularly since increased risk of NHL to farmers has been previously documented and
existed prior to the introduction of glyphosate.
Recall bias and missing data are also limitations in most of the studies. In epidemiologic studies,
the quality of the exposure assessment is a major concern since the validity of the evaluations
depends in large part on the ability to correctly quantify and classify an individual's exposure.
Variation in the quality of exposure assessment, study design and methods, as well as available
information concerning potential confounding variables could also explain discrepancies in study
findings. During their lifetime, farmers are typically exposed to multiple pesticides and often
several may be used together posing a challenge for identifying specific risk factors. Moreover,
there is no direct information on pesticide exposure or absorbed dose because analyses are based
on self-reported pesticide use. The studies included in this epidemiology assessment relied
primarily on questionnaires and interviews to describe participants' past and/or current exposure
to glyphosate. Since the questionnaires are commonly used to account for exposure and capture
self-reporting, the results can be subject to misclassification and recall bias.
Furthermore, the use of proxy respondents has the potential to increase recall bias and thus may
increase exposure misclassification, especially for proxy respondents not directly involved in
farming operations that may be more prone to inaccurate responses than directly interviewed
subjects. In some of the NHL studies, the study participants were interviewed directly to assess
exposure (De Roos etal., 2005; Eriksson et al., 2008; McDuffie etal., 2001; Orsi et al., 2009),
making proxy respondent use a non-issue for these studies. In other studies, however, study
participants or proxy respondents were interviewed to assess exposure (Hardell et al., 2002, De
Roos et al., 2003). De Roos et al. (2003) did not find type of respondent to be statistically
significant, but Hardell et al. (2002) did not conduct analyses to evaluate the impact of proxy
use. In non-NHL studies, proxy analyses were conducted in a small subset (Kachuri etal., 2013;
Lee etal., 2004b; Lee etal., 2005; Yiin etal., 2012) and differences in effect estimates were
often observed. In a few studies, respondent type was used as an adjustment variable when
calculating effect estimates (Band et al., 2011; Kachuri et al., 2013; Lee et al., 2005). As with
all study design elements of case-control studies, one concern is whether or not the use of proxy
respondents had a differential impact on the cases and controls included in the study because any
differential impact may result in differential exposure misclassification. When use of proxy
respondents was comparable for cases and controls in the full study population, it could be
assumed that there is less concern for potential recall bias from the use of proxy respondents. In
Hardell etal., (2002), the percentage of cases and controls with proxy respondents was not fully
reported for cases and controls though and this adds a potential source of uncertainty for the
study. Moreover, when proxy respondents were used in a study, the percentages were usually
reported only for the full study population and were not reported for the specific cases and
controls exposed to glyphosate. This lack of information makes it difficult to assess the degree
to which recall bias may have occurred due to the use of proxy respondents.
Previously, some have argued that the follow-up period (median = 7 years) in De Roos et al.
(2005) is not sufficiently long to account for the latency of NHL (Portier et al., 2016); however,
an analysis of the AHS cohort was recently published (Andreotti etal., 2017) with an extended
Page 66 of 216
-------
follow-up of 17.5 years. This study reported no association between glyphosate exposure and all
lymphohematopoietic cancers, NHL, or any of its subtypes across exposure metrics. No
association was observed in untagged or lagged analyses, after adjustment for pesticides linked
to NHL in previous AHS analyses, and after exclusion of multiple myeloma from the NHL
grouping.
It was also noted that reference groups differed across studies. For example, some studies
(McDuffie et al., 2001; Hardell et al., 2002; and Eriksson et al., 2008) eliminated cases and
controls who had been exposed to certain classes of pesticides, which may have resulted in
selection bias and/or recall bias that may ultimately impact the effect estimates obtained in these
studies. In the dose-response analysis by De Roos et al. (2005), the lowest exposed tertile was
used as the reference group in an effort to reduce the potential for residual confounding by
unmeasured covariates due to lack of comparability observed between the never exposed group
and the higher exposed groups. Analyses were also performed using the unexposed group as the
reference. This study consistently found no evidence of an association between glyphosate
exposure and NHL using different exposure metrics and reference groups. Similarly, there was
no evidence of an association observed in the recent analysis of the AHS cohort (Andreotti et al.,
2017) with a longer follow-up period.
There are conflicting views on how to interpret the overall results for NHL. Some believe that
the data are indicative of a potential association between glyphosate exposure and risk of NHL.
This is primarily based on reported effect estimates across case-control studies and the associated
meta-analyses greater than 1. Additionally, the analysis conducted by Eriksson et al. (2008)
observed a slightly statistically significant increase for those with more than 10 years of exposure
prior to diagnosis. There were also two case-control studies that investigated the association of
glyphosate exposure and NHL by stratifying exposure by days per year of use that reported
effect estimates greater than 1 for groups with the highest exposure.
Conversely, others have viewed the effect estimates as relatively small in magnitude and
observed associations could be explained by chance and/or bias, particularly since studies have
reported farmers develop NHL at excess rates and this risk existed prior to the introduction of
glyphosate. All of the effect estimates for ever/never use were non-statistically significant.
Several studies reported effect estimates approximately equal to the null. The widest confidence
intervals were observed for the highest effect estimates indicating these effect estimate are less
reliable. Sample sizes were limited in several of these case-control studies. Meta-analyses were
based on studies with varying study characteristics. Given the limitations and concerns
discussed above for the individual studies included in this evaluation, chance and/or bias cannot
be excluded as an explanation for the relatively small increase observed in the meta-risk ratios.
Meanwhile, analyses performed by De Roos et al. (2005) and Andreotti el al. (2017) reported
effect estimates less than 1 for cumulative lifetime exposure and intensity-weighted cumulative
exposure and these extensive analyses did not detect any exposure-response relationship, which
conflicts with the two case-control studies that indicate potential for an exposure-response
relationship comparing two groups stratified by days per year of use. Although increased effect
estimates were observed in one case-control study (Eriksson et al., 2008) for subjects exposed
more than 10 years prior to diagnosis and in two case-control studies (McDuffie et al., 2001;
Eriksson et al., 2008) that stratified exposure by days per year of use, none of these analyses
Page 67 of 216
-------
appeared to adjust for exposures to other pesticides, which has been found to be particularly
important for these analyses and would be expected to attenuate these estimates towards the null.
Furthermore, none of the studies in this evaluation of glyphosate exposure and risk of NHL
accounted for other potential confounders, such as diesel exhaust fumes, solvents, animals, and
UV radiation.
Based on the weight-of-evidence, the agency cannot exclude chance and/or bias as an
explanation for observed associations in the database. Due to study limitations and contradictory
results across studies of at least equal quality, a conclusion regarding the association between
glyphosate exposure and risk of NHL cannot be determined based on the available data. The
agency will continue to monitor the literature for studies and any updates to the AHS will be
considered when available.
Page 68 of 216
-------
4.0 Data Evaluation of Animal Carcinogenicity Studies
4.1 Introduction
Cancer bioassays in animals have historically been the primary studies available to evaluate
cancer hazard in humans since, until recently, epidemiological evidence was limited. The results
of these bioassays, as well as results from screening assays for genotoxicity, are considered in a
weight-of-evidence approach to determine the potential of a chemical to induce cancer in
humans. Carcinogenicity studies in two rodent species are required for the registration of food
use pesticides or when the use of a pesticide is likely to result in repeated human exposure over a
considerable portion of the human lifespan (40 CFR Part 158.500). Rodent carcinogenicity
studies identified from the data collection phase of the systematic review were evaluated for
study quality and acceptable studies were evaluated in the context of the 2005 EPA Guidelines
for Carcinogen Risk Assessment as described in Sections 4.2 and 4.3 below, respectively. This
included studies using glyphosate salts, which dissociate quickly in aqueous environments to the
glyphosate acid and the corresponding cation. The cations would not be expected to impact the
toxicity results compared to studies where animals are treated with glyphosate acid alone.
4.2 Consideration of Study Quality for Animal Carcinogenicity Studies
The agency has published test guidelines on how to conduct carcinogenicity studies (OCSPP
870.4200) and combined chronic/carcinogenicity studies (OCSPP 870.4300) in rodents which
have been harmonized with OECD guidelines (Test Nos. 451 and 453). Test substances are
typically administered in animal carcinogenicity studies by the oral route for food use pesticides.
The studies are generally conducted in mice and rats with exposure durations of 18-24 months
for mice and 24 months for rats, which represent exposures of the majority of the expected
lifespan in these animals. Guideline carcinogenicity studies are designed to test three or more
doses in both sexes (with at least 50 animals/sex/dose) with adequate dose spacing to
characterize tumor dose-response relationships. Key considerations when evaluating
carcinogenicity studies for cancer hazard assessment include identification of target organs of
carcinogenicity, increased incidence of tumors or proportion of malignant neoplasms, and
reduction in the time to appearance of tumors relative to the concurrent control group (OECD
TG 451).
There are a number of criteria the agency uses when evaluating the technical adequacy of animal
carcinogenicity studies. A primary criterion is the determination of the adequacy of dosing. The
2005 EPA Guidelines for Carcinogen Risk Assessment recommends that the highest dose level
selected should elicit signs of toxicity without substantially altering the normal life span due to
effects other than tumors; or without inducing inappropriate toxicokinetics (e.g., overwhelming
absorption or detoxification mechanisms); however, the high dose need not exceed 1,000
mg/kg/day (i.e., limit dose) (OCSPP 870.4200; OCSPP 870.4300). Additional criteria to judge
the technical adequacy and acceptability of animal carcinogenicity studies are provided in the
test guidelines as well as other published sources (NTP, 1984; OSTP, 1985; Chhabra et al.,
1990). As stated in the 2005 EPA Guidelines for Carcinogen Risk Assessment, studies that are
judged to be wholly inadequate in protocol, conduct or results, should be discarded from
Page 69 of 216
-------
analysis. Studies the agency consider acceptable are further evaluated for potential tumor
effects.
Following study quality evaluation, a total of 8 chronic/carcinogenicity studies in the rat and 6
carcinogenicity studies in the mouse were considered acceptable for use in the current evaluation
for the active ingredient glyphosate and were subsequently evaluated in the context of the 2005
EPA Guidelines for Carcinogen Risk Assessment as described in Section 4.3. A number of
studies were judged to be inadequate in protocol, conduct or reporting and were not considered
in the analysis of glyphosate. These studies and the justification for not including them in the
analysis are listed below:
1. A two-year chronic oral toxicity study in Albino rats by Reyna (1974)13. The study
was considered inadequate to assess carcinogenicity due to insufficient reporting on
the histopathology findings in the control and treatment groups. Approximately 70
animals were unaccounted for across the study.
2. A two-year drinking water study in Wistar rats with a formulated product (13.6%
ammonium salt) by Chruscielska el al., (2000). In addition to deficiencies including
inadequate reporting of water consumption and body weight data, this study was
conducted with a glyphosate formulated product and not the active ingredient
glyphosate, which is the focus of this review. Glyphosate formulations contain
various components other than glyphosate and it has been hypothesized these
components are more toxic than glyphosate alone. The agency is collaborating with
NTP to systematically investigate the mechanism(s) of toxicity for glyphosate and
glyphosate formulations. This project is discussed in more detail in Section 7.0 of
this document.
3. An initiation-promotion study (George et al., 2010) in male Swiss mice that tested a
commercial formulation of glyphosate (41%) on the skin. Study deficiencies
included small number (20) of animals, tested only males, and lack of
histopathological examination.
4. A carcinogenicity study in Swiss albino mice (Kumar, 2001)14. This study was not
included due to the presence of a viral infection within the colony, which confounded
the interpretation of the study findings. Malignant lymphomas were reported in this
study in all dose groups. However, lymphomas are one of the most common types of
spontaneous neoplastic lesions in aging mice (Brayton et al., 2012). Murine
leukemia viruses (MuLVs) are also a common cause of lymphoma in many different
strains of mice (Ward, 2006). For example, Tadesse-Heath el a/. (2000) reported
50% lymphoma (mostly B-cell origin) incidence in a colony of Swiss mice infected
with MuLVs. Although the lymphoma incidences in Kumar (2001) were within or
near normal background variation, it is not clear whether or not the viral infection
may have contributed to the lymphoma incidence reported or the lower survival seen
at the high dose in this study.
13 MRID 00062507.
14 MRID 49987403. In Greim et al. (2015), the same study is cited as Feinchemie Schwebda (2001).
Page 70 of 216
-------
5. A two year feeding study in Sprague-Dawley rats (Excel, 1997) was not included.
The agency does not have access to this study to perform an independent assessment
of its conduct and; however, Greim et al. (2015) stated that the study "is considered
unreliable for carcinogenicity evaluation" and there were "several deviations from
the OECD Test Guideline 453".
4.3 Assessment of Animal Carcinogenicity Studies
The agency considers many factors when interpreting the results of carcinogenicity studies.
The 2005 EPA Guidelines for Carcinogen Risk Assessment are intended as a guidance only and
does not provide a checklist for determining whether tumor findings are related to treatment.
These guidelines emphasize the importance of weighing multiple lines of evidence in reaching
conclusions regarding human carcinogenic potential of chemicals. Evaluation of observed
tumor findings takes into consideration both biological and statistical significance. There are
several factors in the 2005 EPA Guidelines for Carcinogen Risk Assessment used in the weight-
of-evidence evaluation of individual studies. For this evaluation, the interpretation of the
evidence related to tumor findings is described below.
Dose Selection
Doses should be selected based on relevant toxicological information. Caution is taken in
administering an excessively high dose that would confound the interpretation of the results to
humans. As mentioned above, the 2005 EPA Guidelines for Carcinogen Risk Assessment
recommends that the highest dose level selected should elicit signs of toxicity without
substantially altering the normal life span due to effects other than tumors; or without inducing
inappropriate toxicokinetics (e.g., overwhelming absorption or detoxification mechanisms);
however, the high dose is not recommended to exceed 1,000 mg/kg/day (OCSPP 870.4200;
OCSPP 870.4300). Doses should provide relevant dose-response data for evaluating human
hazard for human health risk assessment. In the case of glyphosate, the low (oral) systemic
toxicity and limited pharmacokinetic (PK) data for this chemical make it difficult to define a
maximum tolerated dose (MTD) for the cancer bioassays. A large number of the
carcinogenicity studies conducted with glyphosate approach or exceed the limit dose. The 2005
EPA Guidelines for Carcinogen Risk Assessment state that "weighing of the evidence includes
addressing not only the likelihood of human carcinogenic effects of the agent but also the
conditions under which such effects may be expressed". As such, the agency puts less weight
on observations of increased incidence of tumors that only occur near or above the limit dose.
Statistical analyses to evaluate dose response and tumor incidences
The main aim of statistical evaluation is to determine whether exposure to the test agent is
associated with an increase in tumor development, rather than due to chance alone. Tumors
were selected for statistical analyses in the current evaluation if the study report identified
tumors as statistically significant and/or have been identified by the reviewer as potentially
biologically significant based on the presence of an increasing monotonic dose-response and/or
relative increases from concurrent controls. For toxicological studies submitted to the agency
for pesticide registration, including animal carcinogenicity studies, detailed reviews are
performed, which summarize study findings and identify effects, such as tumors, for evaluation.
Page 71 of 216
-------
Statistical analyses should be performed on each tumor type separately. The incidence of
benign and malignant lesions of the same cell type, usually within a single tissue or organ, are
considered separately, but may be combined when scientifically defensible (McConnell et al.,
1986). Trend tests and pairwise comparison tests are the recommended tests for determining
whether chance, rather than a treatment-related effect, is a plausible explanation for an apparent
increase in tumor incidence. The 2005 Guidelines for Carcinogen Risk Assessment states that:
"A trend test such as the Cochran-Armitage test (Snedecor and Cochran, 1967) asks whether the
results in all dose groups together increase as dose increases. A pairwise comparison test such
as the Fisher exact test (Fisher, 1950) asks whether an incidence in one dose group is increased
over that of the control group. By convention, for both tests a statically significant comparison
is one for whichp is less than 0.05 that the increased incidence is due to chance. Significance
in either kind of test is sufficient to reject the hypothesis that chance accounts for the result."
In the current evaluation, animals sacrificed for interim evaluations or died prior to the interim
sacrifices were not included in the statistical evaluations to avoid dilution of a potential
carcinogenic effect. Additionally, survival was evaluated across dose groups and no significant
mortality differences were observed in any of the studies. As a result, there was no need to
incorporate survival adjustments into the analyses (e.g., Peto prevalence test). The Cochran-
Armitage Test for Trend (Snedecor and Cochran, 1967; one-sided) was used for trend analysis.
For pairwise comparisons, the Fisher Exact Test (Fisher, 1950; one-sided) was used to
determine if incidences observed in treated groups were different from concurrent controls.
Furthermore, the 2005 EPA Guidelines for Carcinogen Risk Assessment state that
"considerations of multiple comparisons should also be taken into account". Multiple
comparison methods control the familywise error rate, such that the probability of Type I error
(incorrect rejection of the null hypothesis or "false positive") for the pairwise comparisons in
the family does not exceed the alpha level. In the current evaluation, the Benjamini-Hochberg
correction method was used to adjust for multiple comparisons (Benjamini and Hochberg,
1995).
For the current evaluation, statistical significance observed in either test is judged in the context
of all of the available evidence. Statistically significant responses may or may not be
biologically significant and vice versa (Hsu and Stedeford, 2010; EPA, 2005). If a trend was
found to be statistically significant, a closer examination of the tumor incidence was taken to
determine whether the data demonstrate a monotonic dose-response where an increase in tumor
incidence is expected with corresponding increase in dose. Therefore, statistically significant
results with fluctuating tumor incidence across doses are not weighed as heavily as those
displaying a monotonic dose-response. If a pair-wise comparison was found to be statistically
significant, a closer examination of the tumor incidence and other lines of evidence was taken
to determine whether the response was biologically significant. Factors considered in
determining the biological relevance of a response are discussed below.
All statistical analyses were reanalyzed for the purposes of this evaluation to ensure consistent
methods were applied (M. Perron; 12-DEC-2017; TXR#0057690).
Page 72 of 216
-------
Historical Control Data
As indicated in the 2005 EPA Guidelines for Carcinogen Risk Assessment (Section 2.2.2.1.3),
the standard for determining statistical significance of tumor incidence comes from a comparison
of tumors in dosed animals with those in concurrent control animals. Additional insight into the
statistical and/or biological significance of a response can come from the consideration of
historical control data (Tarone, 1982; Haseman, 1995; EPA, 2005). Historical control data can
add to the analysis, particularly by enabling identification of uncommon tumor types or high
spontaneous incidence of a tumor in a given animal strain. Generally speaking, statistically
significant increases in tumors should not be discounted simply because incidence rates in the
treated groups are within the range of historical controls or because incidence rates in the
concurrent controls are somewhat lower than average.
Historical control data are also useful to determine if concurrent control tumor incidences are
consistent with previously reported tumor rates (Haseman, 1995). Historical control data
available to the agency from the performing laboratory for the same species and strain for a study
were considered in the current evaluation. These data were primarily generated within 3 years
and in limited cases within 5 years of the study date. Given the large number of age-related
tumor outcomes in long-term rodent bioassays, and thus the large number of potential statistical
tests run, caution is taken when interpreting results that have marginal statistical significance or
in which incidence rates in concurrent controls are unusually low in comparison with historical
controls since there may be an artificial inflation of the differences between concurrent controls
and treated groups. Consequently, in the current evaluation, unusually low incidence in
concurrent controls was noted when applicable and considered as part of the weight-of-evidence
for the tumor findings. Identification of common or uncommon situations prompts further
thought about the meaning of the response in the current study in context with other observations
in animal studies and with other evidence about the carcinogenic potential of the agent.
Evidence of supporting preneoplastic lesions or related non-neoplastic lesions
Carcinogenicity rodent studies are designed to examine the production of tumors as well as
preneoplastic lesions and other indications of chronic toxicity that may provide evidence of
treatment-related effects and insights into the way the test agent produces tumors (EPA, 2005).
As such, the presence or lack of supporting preneoplastic or other related non-neoplastic
changes were noted in the current evaluation of each study and considered in the weight-of-
evidence to aid in the determination of biological significance since these lesions would not be
expected for age-related tumors in carcinogenicity with continuous treatment. In the current
evaluation, the agency investigated lesions in organs where tumors were observed and
demonstrated biological significance based on the presence of an increasing monotonic dose-
response and/or relative increases from concurrent controls.
Additional Considerations
Other observations can strengthen or lessen the significance of tumor findings in carcinogenicity
studies. Such factors include: uncommon tumor types; tumors at multiple sites; tumors in
multiple species, strains, or both sexes; progression of lesions from preneoplastic to benign to
malignant; reduced latency of neoplastic lesions (i.e., time to tumor); presence of metastases;
unusual magnitude of tumor response; and proportion of malignant tumors (EPA, 2005). The
Page 73 of 216
-------
agency considers all of the above factors when determining the significance of tumor findings in
animal carcinogenicity studies.
4.4 Summary of Animal Carcinogenicity Studies
A total of 8 chronic toxicity/carcinogenicity studies in the rat15 and 6 carcinogenicity studies in
the mouse were considered acceptable and evaluated in the weight-of-evidence analysis for
glyphosate. This includes all of the studies that were part of the 2015 CARC evaluation plus an
additional 4 studies identified from the systematic review. In the 2015 CARC evaluation, for
some of the studies considered, the CARC relied on summary data that was provided in the
supplement to the Greim et al. (2015) review article. Due to the ongoing data collection effort
and the acquiring of studies not previously submitted, the agency no longer needs to rely on the
Greim et al. (2015) review article for the study data generated in relevant studies, allowing for a
more complete and independent analysis. It should be noted that studies have been cited
differently in this evaluation as compared to Greim et al. (2015) so these alternative citations
have been noted for applicable studies.
The carcinogenicity studies conducted in the rat and mouse that were considered for the analysis
are discussed in Sections 4.5 and 4.6, respectively. In these sections, short study summaries are
presented which include information on the study design (including test material, strain of animal
used, and doses and route of administration) as well as study findings including effects on
survival, general toxicity observed, relevant non-neoplastic lesions, and the incidence and
characterization of any tumor findings. The characterization of the tumor response(s) is based on
the considerations previously discussed in Section 4.3 for interpreting the significance of tumor
findings in animal carcinogenicity studies. The rat and mouse carcinogenicity studies are all
summarized in Table 4.11 and Table 4.18, respectively.
4.5 Rat Carcinogenicity Studies with Glyphosate
4.5.1 Lankas, 1981 (MRID 00093879)16
In a chronic toxicity/carcinogenicity study, groups of Sprague-Dawley rats (50/sex/dose) were
fed diets containing glyphosate (98.7%, pure) at dietary doses of 0, 3/3, 10/11, and 31/34
mg/kg/day (M/F) for 26 months.
There were no treatment-related effects on survival at any dose level. The highest dose tested of
approximately 31 mg/kg/day was not considered a maximum tolerable dose to assess the
carcinogenic potential of glyphosate. Consequently, a second study (Stout and Ruecker, 1990)
was conducted at higher doses, which is summarized in the Section 4.5.3.
A statistically significant trend was reported for the testicular interstitial tumors; however, closer
examination of the tumor incidence indicates that the data do not demonstrate a monotonic dose
response with greater incidence observed at the low-dose as compared at the mid-dose. The
15 Note: the original draft of this Issue Paper included 9 studies in rats; however, one study (Burnett, 1979) was
removed since the study was conducted with a contaminant of glyphosate, not the active ingredient glyphosate.
16 In Greim et al. (2015), the same study is cited as Monsanto (1981).
Page 74 of 216
-------
incidence at the high dose was found to be statistically significant as compared to the concurrent
controls (raw and adjusted p-values).
Table 4.1. Testicular Interstitial Cell Tumors in Male Sprague-Dawley Rats (Lankas, 1981)
Cochran-Armitage Trend Test & Fisher's Exact Test Results.
0 mg/kg/day
3.05 mg/kg/day
10.3 mg/kg/day
31.49 mg/kg/day
Incidence
0/47a
3/49
1/47
6/50
(%)
(0)
(6)
(2)
(12)
Raw p-value =
0.011**
0.129
0.500
0.016*
Adjusted p-value =
0.032*
0.172
0.500
0.032*
Note: Trend test results denoted at control: * denotes significance at p=0.05; ** denotes significance at p=0.01.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were sacrificed
prior to study week 52 (interim sacrifice).
The study report provided historical control information for 5 studies of similar duration (24-29
months) run concurrently within 9 months of the termination of the study in the same laboratory.
The historical control range for this tumor type was 3.4%-6.7% (mean = 4.5%) when considering
all animals. When only considering animals that survived to terminal sacrifice, the historical
control range was 6.2%-27.3% (mean = 9.6%). These data indicate that the incidence of
testicular cell tumors in concurrent controls (0%) appears to be unusually low for this tumor
type. Furthermore, the incidence at all doses, including the high dose, was within the historical
control range when evaluating animals at terminal sacrifice. There were no supporting
preneoplastic or other related non-neoplastic changes observed.
4.5.2 Stout and Ruecker, 1990 (MRID 41643801)17
In a chronic toxicity/carcinogenicity study, groups of Sprague-Dawley rats (60/sex/dose) were
fed diets containing glyphosate (96.5%, pure) at dietary doses of 0, 89/113, 362/457 or 940/1183
mg/kg/day M/F) for 24 months. The highest dose tested in this study approaches or exceeds the
highest dose recommended in the test guidelines on how to conduct carcinogenicity studies
(OCSPP 870.4200 and OCSPP 870.4300). Tumor findings at these high doses are given less
weight.
There was no significant increase in mortality. Three types of tumors were evaluated in this
study: pancreatic cell adenomas, hepatocellular adenomas, and thyroid C-cell adenomas in
males. A discussion of each tumor type by organ is presented below:
1. Pancreas: Tumor incidences of pancreatic islet cell tumors in male rats are presented in
Tables 4.2. The incidence of pancreatic islet cell tumors lacked monotonic dose-
responses and trend analyses were not statistically significant. There was also no
statistical significance of the pairwise comparisons. Historical control data were
provided for 7 studies conducted in the same laboratory from 1983-1989 (within 1-4
years of when Stout and Ruecker (1990) was performed). The historical control range for
the adenomas was 1.8%-8.3% (mean = 5.3%). These data are presented in Table 4.3 and
indicate that the incidence of adenomas in concurrent controls (2%) was at the lower limit
17 In Greim etal. (2015), the same study is cited as Monsanto (1990).
Page 75 of 216
-------
of the historical range. There were no supporting preneoplastic or other related non-
neoplastic changes observed and no evidence of progression from adenomas to
carcinomas.
Table 4.2. Pancreatic Islet Cell Tumors in Male Sprague-Dawley Rats (Stout and Ruecker, 1990)
Cochran-Armitage Trend Test & Fisher's Exact Test Results.
Tumor Type
0 mg/kg/day
89 mg/kg/day
362 mg/kg/day
940 mg/kg/day
Adenoma
Incidence
l/43a
8/45
5/49
7/4 8b
(%)
(2)
(18)
(10)
(15)
Raw p-value =
0.176
0.018*
0.135
0.042*
Adjusted p-value =
0.176
0.071
0.176
0.083
Carcinoma
Incidence
1/43°
0/45
0/49
0/48
(%)
(2)
(0)
(0)
(0)
Raw p-value =
_d
_d
_d
_d
Adjusted p-value =
_d
_d
_d
_d
Combined
Incidence
2/43
8/45
5/49
7/48
(%)
(5)
(18)
(10)
(15)
Raw p-value =
0.242
0.052
0.275
0.108
Adjusted p-value =
0.275
0.209
0.275
0.215
Note: Trend test results denoted at control: * denotes significance at p=0.05.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were
sacrificed prior to study week 55 (interim sacrifice).
b. First adenoma in the study was observed at week 81 in the 940 mg/kg/day group.
c. First carcinoma in the study was observed at week 105 in the controls.
d. Trend p-value not reported since tumor incidence decreased with increasing dose.
Table 4.3. Historical Control Data — Pancreatic Islet Cell Adenomas in Male Sprague- Dawley Rats (MRID No. 41728701).
Study No.
1
2
3
4
5
6
7
Mean
Study Year
07/83
02/85
10/85
6/85
9/88
1/89
3/89
-
Tumor Incidence
2/68
5/59
4/69
1/57
5/60
3/60
3/59
-
Percentage (%)
2.9%
8.5%
5.8%
1.8%
8.3%
5.0%
5.1%
5.3%
2. Liver: Tumor incidences of liver tumors in male rats are presented in Tables 4.4. There
was a statistically significant dose trend for liver adenomas; however, the trend was not
statistically significant with an adjustment for multiple comparisons. Closer examination
of the incidence indicates a relatively flat response at the low- and mid-dose with only an
increase observed at the high-dose (940 mg/kg/day); however, the incidence of liver
adenomas at the high-dose was not statistically significant when compared to the
concurrent controls (raw or adjusted p-values). Carcinomas and combined
adenomas/carcinomas lacked statistical significance in trend and pairwise comparisons
(Table 4.4). Historical control data were provided for 7 studies conducted in the same
laboratory from 1983-1989 (within 1-4 years of when Stout and Ruecker (1990) was
performed). The historical control range was 1.4%-l8.3% (mean = 9.2%) for the
adenomas and 0%-6.7% for carcinomas (mean = 2.6%). These data are provided in
Table 4.5 and indicate that the observed incidences at all dose levels were within the
historical control ranges. Except for a single animal at the mid-dose late in the study (89
Page 76 of 216
-------
weeks), no hyperplasia, preneoplastic foci or other non-neoplastic lesions were observed.
Furthermore, there was no evidence of progression from adenomas to carcinomas.
Table 4.4. Hepatocellular Tumors in Male Sprague-Dawley Rats (Stout and Ruecker, 1990)
Cochran-Armitage Trend Test & Fisher's Exact Test Results
Tumor Type
0 mg/kg/day
89 mg/kg/day
362 mg/kg/day
940 mg/kg/day
Adenoma
Incidence
(%)
Raw p-value =
Adjusted p-value =
2/44a
2/45
3/49
7/4 8b
(5)
0.022*
0.089
(4)
0.700
0.700
(6)
0.551
0.700
(15)
0.101
0.202
Carcinoma
Incidence
(%)
Raw p-value =
3/44
2/45
1/49
2/48°
(7)
d
(4)
d
(2)
d
(4)
d
Adjusted p-value =
_d
_d
_d
_d
Combined
Incidence
5/44
4/45
4/49
9/48
(%)
(11)
(9)
(8)
(19)
Raw p-value =
0.078
0.769
0.808
0.245
Adjusted p-value =
0.312
0.808
0.808
0.489
Note: Trend test results denoted at control: * denotes significance at p=0.05.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were
sacrificed prior to study week 55 (interim sacrifice).
b. First adenoma in the study was observed at week 88 in the 940 mg/kg/day group.
c. First carcinoma in the study was observed at week 85 in the 940 mg/kg/day group.
d. Trend p-value not reported since tumor incidence decreased with increasing dose.
Table 4.5. Historical Control Data — Hepatocellular Tumors in Male Sprague- Dawley Rats (MRID No. 41728701).
Study No.
1
2
3
4
5
6
7
Mean
Study Year
07/83
02/85
10/85
6/85
9/88
1/89
3/89
-
Adenomas
Tumor Incidence
5/60
11/68
1/70
3/59
11/60
5/60
4/60
-
Percentage (%)
8.3%
16.2%
1.4%
5.1%
18.3%
8.3%
6.7%
9.2%
Carcinomas
Tumor Incidence
4/60
0/68
1/70
2/59
3/60
1/60
0/60
-
Percentage (%)
6.7%
0%
1.4%
3.4%
5%
1.7%
0%
2.6%
3. Thyroid: Tumor incidences of thyroid tumors in male and female rats are presented in
Tables 4.6 and 4.7, respectively. For males, no statistically significant trends were
observed for adenomas, carcinomas, or combined adenomas/carcinomas. For females, a
statistically significant trend was observed for adenomas and combined
adenomas/carcinomas; however, the trend was not statistically significant with
adjustment for multiple comparisons. There was no statistical significance in pairwise
analyses. Historical control data were provided for 7 studies conducted in the same
laboratory from 1983-1989 (within 1-4 years of when Stout and Ruecker (1990) was
performed). The historical control range was 3.3%-10% (mean = 6.1%) for the adenomas
and 0%-2.9% for carcinomas (mean = 0.9%). These data are provided in Table 4.8.
Page 77 of 216
-------
Non-neoplastic lesions (thyroid C-cell hyperplasia) were observed; however, there was a
lack of a monotonic dose-response for these histopathological findings and no dose-
related increase in severity to support tumor findings (Table 4.9). There was also no
evidence of progression from adenomas to carcinomas.
Table 4.6. Thyroid C-Cell Tumors in Male Sprague-Dawley Rats (Stout and Ruecker, 1990)
Cochran-Armitage
Trend Test & Fisher's Exact Test Results
Tumor Type
0 mg/kg/day
89 mg/kg/day
362 mg/kg/day
940 mg/kg/day
Adenoma
Incidence
2/54a>b
4/55
8/58
7/58
(%)
(4)
(7)
(14)
(12)
Raw p-value =
0.079
0.348
0.060
0.099
Adjusted p-value =
0.132
0.348
0.132
0.132
Carcinoma
Incidence
0/54
2/55°
0/58
1/58
(%)
(0)
(4)
(0)
(2)
Raw p-value =
0.457
0.252
1.000
0.518
Adjusted p-value =
0.518
0.518
1.000
0.518
Combined
Incidence
2/54
6/55
8/58
8/58
(%)
(4)
(11)
(14)
(14)
Raw p-value =
0.087
0.141
0.060
0.060
Adjusted p-value =
0.116
0.141
0.116
0.116
Note: Trend test results denoted at control.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were
sacrificed prior to study week 55 (interim sacrifice).
b. First adenoma in the study was observed at week 54 in the controls.
c. First carcinoma in the study was observed at week 93 in the 89 mg/kg/day group.
Table 4.7. Thyroid C-Cell Tumors in Female Sprague Dawley Rats
Cochran-Armitage Trend Test & Fisher's Exact Test Results (Stout and Ruecker, 1990).
Tumor Type
0 mg/kg/day
113 mg/kg/day
457 mg/kg/day
1183 mg/kg/day
Adenoma
Incidence
2I5T
2/60
6/5 9b
6/55
(%)
(4)
(3)
(10)
(11)
Raw p-value =
0.040*
0.710
0.147
0.124
Adjusted p-value =
0.159
0.710
0.196
0.196
Carcinoma
Incidence
0/57
0/60
1/59°
0/55
(%)
(0)
(0)
(2)
(0)
Raw p-value =
0.494
1.000
0.509
1.000
Adjusted p-value =
0.509
1.000
0.509
1.000
Adenoma/Carcinoma
Incidence
2/57
2/60
7/59
6/55
(%)
(4)
(3)
(12)
(11)
Raw p-value =
0.042*
0.710
0.090
0.124
Adjusted p-value =
0.166
0.710
0.166
0.166
Note: Trend test results denoted at control: * denotes significant at p=0.05.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were
sacrificed prior to study week 55 (interim sacrifice).
b. First adenoma in the study was observed at week 72 in the controls.
c. First carcinoma in the study was observed at week 93 in the 457 mg/kg/day group.
Page 78 of 216
-------
Table 4.8. Historical Control Data— Thyroid C-Cell Tumors in Female Sprague- Dawley Rats (MRID No. 41728701).
Study No.
1
2
3
4
5
6
7
Mean
Study Year
07/83
02/85
10/85
6/85
9/88
1/89
3/89
-
Adenomas
Tumor Incidence
2/60
3/69
7/70
3/59
5/59
5/60
2/60
-
Percentage (%)
3.3%
4.3%
10.0%
5.1%
8.5%
8.3%
3.3%
6.1%
Carcinomas
Tumor Incidence
1/60
2/69
0/70
1/59
0/59
0/60
0/60
-
Percentage (%)
1.7%
2.9%
0%
1.7%
0%
0%
0%
0.9%
Table 4.9. Thyroid Non-Neoplastic Lesions (Stout and Ruecker, 1990)
Males
Dose
0 mg/kg/day
89 mg/kg/day
362 mg/kg/day
940 mg/kg/day
Total Incidences of thyroid
C-cell hyperplasia and
severity scores
5/60
(8%)
Diffuse (moderate) - 1
Multi-focal (minimal) - 3
Focal (mild) - 1
1/60
(2%)
Focal (mild) - 1
6/60
(10%)
Focal (minimal) - 4
Multi-focal (minimal) - 1
Multi-Focal (mild) - 1
5/60
(8%)
Focal (minimal) - 2
Focal (mild) - 1
Multi-focal (mild) - 1
Multi-focal (moderate) -1
Females
0 mg/kg/day
113 mg/kg/day
457 mg/kg/day
1183 mg/kg/day
Thyroid C-cell hyperplasia
and severity scores
10/60
(17%)
Diffuse (moderate) - 1
Focal (mild) - 1
Focal (minimal) - 1
Focal (mild) - 1
Focal (moderate) - 1
Multi-focal (minimal) - 3
Multi-focal (moderate) - 1
Diffuse (moderate) - 1
5/60
(8%)
Focal (mild) - 3
Focal (minimal) - 1
Multi-focal (minimal) - 1
9/60
(15%)
Focal (minimal) - 4
Multi-focal (minimal) - 2
Multi-focal (mild) - 3
5/60
(8%)
Focal (mild) - 1
Focal (minimal) - 1
Multi-focal (mild) - 2
Diffuse (moderate) - 1
*Data taken from pages 1071-2114 of the study report.
4.5.3 Atkinson et al, 1993a (MRID 49631701)18
In a combined chronic toxicity/carcinogenicity study, glyphosate (98.9% pure) was administered
to 50 Sprague-Dawley rats/sex/dose in the diet at doses of 0, 11/12, 112/109, 320/347, and
1147/1134 mg/kg/day for 104 weeks (M/F) for 104 weeks. An additional 35 rats/sex/dose were
included for 1-year interim sacrifice.
18 Note: In Greim et al. (2015), the same study is cited as Cheminova (1993a).
Page 79 of 216
-------
No adverse effects on survival were seen in either sex across the doses tested. There were no
changes in histopathological findings observed. There were no treatment-related increases in
tumor incidences in the study.
4.5.4 Brammer, 2001 (MRID 49704601)19
In a combined chronic toxicity/carcinogenicity study, glyphosate acid (97.6% pure) was
administered to groups of Wistar rats in the diet. Groups of 52 rats/sex received diets containing
doses of 0, 121/145, 361/437 or 1214/1498 mg/kg/day for 24 months, in males/females,
respectively. The highest dose tested in this study exceeds the highest dose recommended in the
test guidelines on how to conduct carcinogenicity studies (OCSPP 870.4200 and OCSPP
870.4300).
A statistically significant higher survival (p=0.02) was observed in males at the highest dose
tested at the end of 104 weeks relative to concurrent controls, and a statistically significant trend
for improved survival was observed in treated males (p=0.03). The inter-current (early) deaths
were 37/52, 36/52, 35/52, and 26/52 for the control, low-, mid-, and high-dose groups,
respectively. The terminal deaths were 16/52, 17/52, 18/52, and 26/52 for the control, low-, mid-
and high-dose groups, respectively. There were no treatment-related non-neoplastic lesions in
any organs of either sex at any dose level tested. As shown in Table 4.10, a statistically
significant trend in the incidences of liver adenomas was observed in male rats; however, a
monotonic dose-response was not observed upon closer examination of the incidence data.
Tumor incidences appear to fluctuate with increases observed at the low- and high-dose and no
tumors observed in the control and mid-dose. Statistical significance with raw (unadjusted) p-
values was observed for the tumor incidence at the high-dose (1214 mg/kg/day) when compared
to concurrent controls; however, it was not statistically significant with an adjustment for
multiple comparisons (p= 0.055). The improved survival in the high-dose group may help
explain a modestly higher incidence of an age-related background tumor like liver adenomas and
this corresponds with the lack of associated lesions observed in the study.
Table 4.10. Liver Adenomas in Male Wistar Rats (Brammer, 2001)
Cochran-Armitage Trend Test and Fisher's Exact Test Results.
0 mg/kg/day
121 mg/kg/day
361 mg/kg/day
1214 mg/kg/day
Adenoma
Incidence
0/44a
2/48
0/48
5/49
(%)
(0)
(4)
(0)
(10)
Raw p-value =
0.010*
0.269
1.000
0.037*
Adjusted p-value =
0.029*
0.269
1.000
0.055
Note: Trend test results denoted at control: * denotes significance at p=0.05; ** denotes significance at p=0.01
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were
sacrificed prior to study week 52 (interim sacrifice).
4.5.5 Pavkov and Wyand 1987 (MRIDs 40214007, 41209905, 41209907)
Glyphosate trimesium salt (sulfosate, 56.2% pure) was tested in a 2-year chronic
feeding/carcinogenicity study in male and female Sprague-Dawley (Crl:CD[SD]BR) rats. Sixty
19 Note: In Greim etal. (2015), the same study is cited as Syngenta (2001).
Page 80 of 216
-------
animals/sex were tested in control group 1 (basal diet, no vehicle), 80/sex were tested in control
group 2 (basal diet plus propylene glycol at 1% w/w vehicle) and in the low and mid-dose
groups, and 90/sex were tested in the high dose group. The following dose levels were tested: 0,
4.2/5.4, 21.2/27 or 41.8/55.7 mg/kg/day in males and females respectively.
Treatment had no effect on survival. There were no changes in histopathological findings
observed. There were no treatment-related increases in tumor incidences in the study.
4.5.6 Suresh, 1996 (MRID 49987401)20
In a combined chronic toxicity/carcinogenicity study, glyphosate (96.0-96.8% pure) was
administered to groups of Wistar rats in the diet. Groups of 50 rats/sex/group received diets
containing 0, 6.3/8.6, 59.4/88.5, and 595.2/886 mg/kg/day glyphosate for 24 months in males and
females respectively. The highest dose tested in females in this study approaches the highest
dose recommended in the test guidelines on how to conduct carcinogenicity studies (OCSPP
870.4200 and OCSPP 870.4300).
No adverse effects on survival were observed in either sex across the doses tested. There were
no changes in histopathological findings observed. There were no treatment-related increases in
tumor incidence observed in the study.
4.5.7 Enemoto, 1997 (MRID 50017103-50017105)21
In a combined chronic toxicity and carcinogenicity study, groups of 50 Sprague-Dawley
rats/sex/group received daily dietary doses of 0, 104/115, 354/393 and 1127/1247 mg/kg
bw/day glyphosate for males and females, respectively. In addition, 10 rats/sex/group were
included for interim sacrifices at 26, 52, and 78 weeks. The highest dose tested in this study
exceeds the highest dose recommended in the test guidelines on how to conduct
carcinogenicity studies (OCSPP 870.4200 and OCSPP 870.4300).
There were no changes in mortality at any of the doses tested. There were no changes in
histopathological findings observed. There were no treatment-related increases in tumor
incidence observed in the study.
4.5.8 Wood et al, 2009a (MRID 49957404)22
In a combined chronic toxicity/carcinogenicity study, glyphosate (95.7% pure) was administered
to groups of Wistar rats in the diet. Groups of 51 rats/sex/group received diets containing 0, 95.0,
316.9, and 1229.7 mg/kg/day glyphosate for males and female, respectively. The highest dose
tested in this study exceeds the highest dose recommended in the test guidelines on how to
conduct carcinogenicity studies (OCSPP 870.4200 and OCSPP 870.4300).
20 Note: In Greim etal. (2015), the same study is cited as Feinchemie Schwebda (1996).
21 Note: In Greim etal. (2015), the same study is cited as ArystaLife Sciences (1997b).
22 Note: In Greim et al. (2015), the same study is cited as NuFarm (2009b).
Page 81 of 216
-------
No adverse effects on survival were seen in either sex across the doses tested. There were no
treatment-related preneoplastic or related non-neoplastic lesions in either sex at any dose level.
In female rats, mammary gland tumors were noted. Tumor incidences for mammary gland
adenomas, adenocarcinomas, and combined adenomas/adenocarcinomas in female mice are
presented in Table 4.11. Statistically significant trends were observed for the adenocarcinoma
and combined analyses; however, statistical significance was only seen for the combined
analyses following adjustment for multiple comparisons. There was no statistical significance
observed in pairwise comparisons.
Table 4.11 Mammary Gland Tumor Incidences in Female Rats (Wood et al., 2009a)
Fisher's Exact Test and Cochran-Armitage Trend Test Results
Tumor Type
0 mg/kg/day
95.0 mg/kg/day
316.9 mg/kg/day
1229.7 mg/kg/day
Adenoma
Incidence
0/48a
0/51
0/50
2/50
(%)
Raw p-value =
(0)
0.062
(0)
1.000
(0)
1.000
(4)
0.258
Adjusted p-value =
0.124
1.000
1.000
0.258
Adenocarcinoma
Incidence
2/48
3/51
1/50
6/50
(%)
Raw p-value =
(4)
0.043*
(6)
0.529
(2)
0.886
(12)
0.148
Adjusted p-value =
0.172
0.705
0.886
0.296
Combined Incidence
(%)
Raw p-value =
Adjusted p-value =
2/48
3/51
1/50
8/50
(4)
0.007**
0.028*
(6)
0.529
0.705
(2)
0.886
0.886
(16)
0.053
0.105
Note: Trend test results denoted at control: * denotes significance at p=0.05; ** denotes significant at p=0.01.
a. Number of tumor-bearing animals/Number of animals examined, excluding those that died or were sacrificed
prior to study week 52 (interim sacrifice).
4.5.9 Summary of Rat Data
In 4 of the 8 rat studies conducted with glyphosate, no tumors were identified for evaluation. Of
the remaining 4 rat studies, a statistically significant trend was observed for tumor incidences in
the testes, liver, or mammary gland following adjustment for multiple comparisons. A
statistically significant pairwise comparison was only observed following adjustment for
multiple comparisons for testicular tumors at the highest dose tested (31 mg/kg/day) in one
individual study. In some cases, the tumor incidence across doses did not demonstrate a
monotonic dose response. There was no evidence of corroborating pre-neoplastic or related non-
neoplastic lesions or evidence of tumor progression (progression from pre-neoplastic to
malignancy) to support biological significance of tumor findings. In a limited number of cases,
the agency considered historical control data to inform the relevance of a tumor increase, which
indicated concurrent controls were unusually low or the observed incidences were within the
historical control range in most instances.
Page 82 of 216
-------
Table 4.12. Summary of Rat Carcinogenicity Studies
Study
Dose Range
Pre-Neoplastic or Related
Non-Neoplastic Lesions
Tumors Incidences, Statistical Significance, and Related Comments
Lankas (1981)
Sprague-Dawley rats
98.7% Technical in diet
0, 3/3, 10/11, and 31/34 mg/kg/day [M/F]
None observed
Statistically significant trend observed for testicular interstitial cell tumors;
however, did not observe monotonic dose-response with higher incidence at
low-dose than mid-dose. Incidences were 0/47 in controls, 3/49 at low-dose,
1/47 at mid-dose, and 6/50 at high-dose. Increased incidence at high-dose
statistically significant, but unusually low control incidence (based on
terminal sacrifice historical control data in study report) inflated increase at
high-dose.
Pancreatic tumors lacked statistically significant trend. Tumor incidence for
pancreatic adenomas in males were 1/43 in controls, 8/45 at the low-dose,
5/49 at the mid-dose, and 7/48 at the high-dose. Concurrent control incidence
for this tumor type was at the lower bound of the historical control range for
performing laboratory. Negative trend observed for carcinomas. Combined
adenoma/carcinoma incidence similar except low-dose was 2/43. No
statistically significant pairwise comparisons, including the highest dose
tested which is approaching/exceeding 1,000 mg/kg/day.
Stout and Ruecker (1990)
Sprague-Dawley rats
96.5% Technical in diet
0, 89/113, 362/457 and 940/1183 mg/kg/day [M/F] for
24 months
None observed
No statistically significant trends or pairwise comparisons for hepatocellular
tumors following adjustment for multiple comparisons. Negative trend
observed for carcinomas. The highest dose tested approached/exceeded 1,000
mg/kg/day. All incidences within historical control range for performing
laboratory.
No statistically significant trend for thyroid C-cell tumors in males. For
females, statistically significant trend for combined adenomas/carcinomas
following adjustment for multiple comparisons. Incidences for combined
adenomas/carcinomas were 2/57 in controls, 2/60 at the low-dose, 7/59 at the
mid-dose, and 6/55 at the high-dose. No statistically significant pairwise
comparisons, including the highest dose tested which is
approaching/exceeding 1,000 mg/kg/day.
Atkinson et al. (1993a)
Sprague-Dawley rats
98.9%) Technical in diet
0, 11/12, 112/109, 320/347, and 1147/1134 mg/kg/day
for 104 weeks (M/F)
None observed
There were no tumors identified for evaluation, including the highest dose
tested which exceeded 1,000 mg/kg/day.
Page 83 of 216
-------
Table 4.12. Summary of Rat Carcinogenicity Studies
Study
Dose Range
Pre-Neoplastic or Related
Non-Neoplastic Lesions
Tumors Incidences, Statistical Significance, and Related Comments
Brammer (2001)
Wistar rats
97.6% Technical in diet
0, 121/145, 361/437 and 1214/1498 mg/kg/day [M/F]
None observed
Statistically significant trend in liver adenomas in males. Non-monotonic
dose-response with incidences at 0/44 in controls, 2/48 at the low-dose, 0/48
at the mid-dose, and 5/49 at the high-dose. No statistically significant
pairwise comparisons following adjustment for multiple comparisons,
including the highest dose tested which exceeded 1,000 mg/kg/day.
Pavkov and Wyand (1987)
Sprague-Dawley rats
56.2% Technical (Trimesium salt; Sulfosate)
0, 4.2/5.4, 21.2/27 and 41.8/55.7 mg/kg/day [M/F]
None observed
There were no tumors identified for evaluation.
Suresh (1996)
Wistar rats
96.0-96.8%) Technical in diet
0, 6.3/8.6, 59.4/88.5, and 595.2/886 mg/kg/day [M/F]
None observed
There were no tumors identified for evaluation, including the highest dose
tested which exceeded 1,000 mg/kg/day.
Enemoto (1997)
Sprague-Dawley rats
94.61-97.56%) Technical in diet
0, 104/115, 354/393 and 1127/1247 mg/kg/day [M/F]
None observed
There were no tumors identified for evaluation, including the highest dose
tested which exceeded 1,000 mg/kg/day.
Wood et ul. (2009a)
Wistar rats
95.7%o Technical in diet
0, 86/105, 285/349 or 1077/1382 mg/kg/day [M/F]
None observed
Statistically significant trends were observed for the combined mammary
gland adenoma/adenocarcinoma analyses following adjustment for multiple
comparisons. Incidences were 2/48 in controls, 3/51 at the low-dose, 1/50 at
the mid-dose, and 8/50 at the high-dose. No statistically significant pairwise
comparisons, including the highest dose tested which exceed 1,000
mg/kg/day.
Page 84 of 216
-------
4.6 Mouse Carcinogenicity Studies with Glyphosate
4.6.1 Reyna and Gordon, 1973 (MRID 00061113)
In an 18-month carcinogenicity study, groups of 50 Swiss white mice/sex/dose were fed
glyphosate at dietary levels of approximately 17 mg/kg/day and 50 mg/kg/day. There was no
effect on survival at any of the doses tested. There were no changes in histopathological findings
observed. There were no treatment-related increases in tumor incidence observed in the study.
Although only ten mice/sex/dose were examined for histopathological changes, there were no
statistically significant increases in tumors observed in the study; therefore, this deficiency
would not impact the overall conclusion regarding tumor findings.
4.6.2 Knezevich and Hogan, 1983 (MRID 00130406)23
Groups of 50 male and female CD-I mice received glyphosate (99.78%, pure) at dietary doses of
0, 161/195, 835/968, 4945/6069 mg/kg/day for males and females, respectively for 24 months.
The highest dose tested in this study far exceeds the highest dose recommended in the test
guidelines on how to conduct carcinogenicity studies (OCSPP 870.4200 and OCSPP 870.4300).
Furthermore, the mid-dose tested in this study was approaching 1,000 mg/kg/day. Tumor
findings at these high doses are given less weight. No effect on survival was observed. A low
incidence of renal tubule adenomas, which are considered rare, were noted in males. The
incidences of renal tubule adenomas following initial evaluation of the study were reported as
follows: 0/49 in the controls; 0/49 at the low-dose; 1/50 at the mid-dose; and 3/50 at the high
dose (TXR# 0004370). In 1985, the registrant directed a re-evaluation of the original renal
sections by a consulting pathologist. This re-evaluation identified a small renal tubule adenoma
in one control male mouse, which was not diagnosed as such in the original pathology report. In
1986, at the request of the agency, additional renal sections (3 sections/kidney/mouse spaced at
150 micron intervals) were evaluated in all control and all glyphosate-treated male mice in order
to determine if additional tumors were present. The additional pathological and statistical
evaluations concluded that the renal tumors in male mice were not compound-related.
Subsequently, the agency requested a Pathology Work Group (PWG) evaluate the kidney
sections. The PWG examined all sections of the kidney, including the additional renal sections,
and were blinded to treatment group. The renal tubular-cell lesions diagnosed by the PWG are
presented below in Table 4.13 with results from statistical analyses. The PWG noted that
because differentiation between tubular-cell adenoma and tubular-cell carcinoma is not always
clearly apparent and because both lesions are derived from the same cell type, it is appropriate to
combine the incidences from these two tumor types for purposes of evaluation and statistical
analysis. The PWG unanimously concluded that these lesions are not compound-related based on
the following considerations: 1) renal tubular cell tumors are spontaneous lesions for which there
is a paucity of historical control data for this mouse stock; 2) there was no statistical significance
in a pairwise comparison of treated groups with the concurrent controls and there was no
evidence of a statistically significant linear trend; 3) multiple renal tumors were not found in any
23 Note: In Greim etal. (2015), the same study is cited as Monsanto (1983).
Page 85 of 216
-------
animal; and 4) compound-related nephrotoxic lesions, including pre-neoplastic changes, were not
present in male mice in this study (TXR# 0005590).
Table 4.13. Renal Tubular Cell Tumors in Male CD-I Mice (Knezevich and Hogan, 1983)
Cochran-Armitage Trend Test & Fisher's Exact Test Results.
Tumor Type
0 mg/kg/day
161 mg/kg/day
835 mg/kg/day
4945 mg/kg/day
Adenoma Incidence
(%)
Raw p-value =
Adjusted p-value =
1/49
(2)
0.442
1.000
0/49
(0)
1.000
1.000
0/50
(0)
1.000
1.000
1/50
(2)
0.758
1.000
Carcinoma
Incidence
(%)
Raw p-value =
Adjusted p-value =
0/49
(0)
0.063
0.190
0/49
(0)
1.000
1.000
1/50
(2)
0.505
0.505
2/50
(4)
0.253
0.379
Combined
Incidence
(%)
Raw p-value =
Adjusted p-value =
1/49
(2)
0.065
0.259
0/49
(0)
1.000
1.000
1/50
(2)
0.758
1.000
3/50
(6)
0.316
0.633
Note: Trend test results denoted at control.
Historical control data from 14 studies conducted between 1977 and 1981 (within <1 to 3 years
of when Knezevich and Hogan (1983) was performed) at the performing laboratory (Table 4.14)
indicated that the mouse renal tubular adenomas ranged from 0 to 3.3% and the incidence in the
current study was within the historical control range (TXR# 0007252).
Table 4.14. Historical Control Data- Kidney tumors in CD-I Mice (TXR #0007252).
Study Period
6/78-
7/80
12/77-
4/80
12/77-
3/80
10/78-
4/81
11/78-
4/81
11/77-
4/80
10/77-
4/80
No. Examined
57
54
61
51
53
59
60
60
60
60
60
60
60
60
Tubular Adenoma
0
1
0
0
0
0
0
0
0
2
0
0
0
0
Histopathological examinations noted chronic interstitial nephritis and tubular epithelial changes
(basophilia and hypertrophy) in the kidneys of male rats in the study (Table 4.15). The increased
incidence of chronic interstitial nephritis in males lacked a dose-response. The incidence in
controls of bilateral interstitial nephritis was higher than low-dose group and approximately the
same as the mid-dose group. Unilateral chronic interstitial nephritis was only seen in 1 animal in
the low- and high-dose groups. Furthermore, chronic interstitial nephritis is not considered to be
a precursor lesion for tubular neoplasms. A monotonic dose-response was not observed for the
epithelial basophilia and hypertrophy, such that the incidence fluctuated with dose and the lowest
incidence was observed at the highest dose tested. There was no increase in supporting
preneoplastic or related non-neoplastic renal tubular lesions (e.g., tubular epithelial
necrosis/regeneration, hyperplasia) observed in male mice.
Page 86 of 216
-------
Table 4.15. Kidney Histopathological Alterations in Male CD-I Mice (Knezevich and Hogan, 1983)
Males
Dose
0 mg/kg/day
161 mg/kg/day
835 mg/kg/day
4945 mg/kg/day
Bilateral Chronic
Interstitial Nephritis
5/49
(10%)
1/49
(2%)
7/50
(14%)
11/50
(22%)
Unilateral Chronic
Interstitial Nephritis
0/49
(0%)
1/49
(2%)
0/49
(0%)
1/50
(2%)
Proximal Tubule
Epithelial Basophilia
and Hypertrophy
15/49
(31%)
10/49
(20%)
15/50
(30%)
7/50
(14%)
*Data taken from page 305 and 306, and the study pathology report; incidences were moderate diffuse
4.6.3 Atkinson, 1993b (MRID 49631702)24
In a carcinogenicity study, glyphosate (>97% pure) was administered to groups of 50 CD-I
mice/sex/dose in the diet for 104 weeks at doses of 0, 98/102, 297/298, 988/1000 mg/kg/day for
males and females, respectively. No interim sacrifices were performed.
There was no effect on survival in the study. There were no preneoplastic lesions or related non-
neoplastic lesions observed. As shown in Table 4.16, hemangiosarcomas were found in 4/45
(9%) of high-dose male mice (1000 mg/kg/day) compared to none in the concurrent controls or
other treated groups. Hemangiosarcomas are commonly observed in mice (generally more
common in males for CD-I strain) as both spontaneous and treatment-related tumors arising
from endothelial cells. As vascular tumors, they can occur at different sites, with liver and
spleen tending to be the most common sites in mice. In the high-dose mice with
hemangiosarcomas, one had the tumors present in the liver and spleen, one had the tumor present
in the liver only, one had the tumors present in the liver, spleen, and prostate, and one had the
tumor present in the spleen only. A statistically significant trend was observed. Closer
examination of the incidence indicates a relatively flat response at the low- and mid-dose with
only an increase observed at the high-dose; however, the incidence of hemangiosarcomas at the
high-dose was not statistically significant when compared to the concurrent controls.
Table 4.16. Hemangiosarcomas in Male CD-I Mice (Atkinson, 1993b)
Cochran-Armitage Trend Test and Fisher's Exact Test Results.
Dose (mg/kg/day)
0
100
300
1000
Hemangiosarcoma
Incidence
0/47a
0/46
0/50
4/45
(%)
(0)
(0)
(0)
(9)
Raw p-value =
0.003**
1.000
1.000
0.053
Adjusted p-value =
0.006**
1.000
1.000
0.053
Note: Trend test results denoted at control; * denotes significance at p=0.05; ** denotes significance at p=0.01
a= Number of tumor bearing animals/Number of animals examined, excluding those that died before week 52.
24 Note: In Greim etal. (2015), the same study is cited as Cheminova (1993b).
Page 87 of 216
-------
4.6.4 Wood et al, 2009b (MRID 49957402)25
In a feeding study, CD-I mice (50/sex/dose) received glyphosate (95.7%) for 80 weeks at dietary
dose levels of 0, 71.4/97.9, 234.2/299.5, or 810/1081.2 mg/kg/day for males and females,
respectively. The highest dose tested in this study approaches or exceeds the highest dose
recommended in the test guidelines on how to conduct carcinogenicity studies (OCSPP 870.4200
and OCSPP 870.4300).
There was no effect on survival in the study. In male mice at the high dose, there were increases
in the incidences of lung adenocarcinomas and malignant lymphomas. A discussion of each
tumor type is presented below:
1. Lung: Tumor incidence for lung adenomas, adenocarcinomas, and combined
adenomas/adenocarcinomas are presented in Table 4.17. A statistically significant trend
was only noted for the adenocarcinomas; however, the trend was not statistically
significant with adjustment for multiple comparisons. Closer examination of the tumor
incidence indicates the dose-response was relatively flat at the low- and mid-dose with
only an increase observed at the high-dose and the incidence of lung adenocarcinomas at
the high-dose (810 mg/kg/day) was not statistically significant when compared to the
concurrent controls. There were no treatment-related preneoplastic or related non-
neoplastic lesions observed.
2. Malignant lymphoma: Tumor incidence for malignant lymphoma are also presented in
Table 4.18. A statistically significant trend was observed and the incidence at the high-
dose (810 mg/kg/day) was statistically significantly elevated as compared to concurrent
controls with the raw (unadjusted) p-value; however, with an adjustment for multiple
comparisons, the increased incidence at the high-dose was not statistically significant (p=
0.059). Historical control data have been submitted (MRIDs 50464501and 50464601)
from the same testing laboratory for 10 studies of similar duration. These data were
generated within approximately 5 years of the Wood et al. (2009b) study. The historical
control range was 0%-32% (mean = 8.7%). All observed incidences of this tumor type
were within the historical control range.
25 Note: In Greim et al. (2015), the same study is cited as NuFarm (2009a).
Page 88 of 216
-------
Table 4.17. Lung Tumors in Male CD-I Mice (Wood et al., 2009b)
Fisher's Exact Test and Cochran-Armitage Trend Test Results.
Dose (mg/kg/day)
0
71.4
234.2
810
Lung Adenoma
Incidence
(%)
Raw p-value =
Adjusted p-value =
9/44
(20)
_b
_b
7/46
(15)
_b
_b
9/48
(19)
_b
_b
4/45
(9)
_b
_b
Lung
Adenocarcinoma
(%)
Raw p-value =
Adjusted p-value =
5/44a
(11)
0.026*
0.103
5/46
(11)
0.659
0.659
7/48
(15)
0.443
0.590
11/45
(24)
0.091
0.182
Lung Combined
Incidence
(%)
Raw p-value =
Adjusted p-value =
14/44
(32)
0.328
0.706
12/46
(26)
0.797
0.797
16/48
(33)
0.527
0.706
15/45
(33)
0.529
0.706
Note: Trend test results denoted at control: * denotes significance atp=0.05:** denotes significance atp=0.01
a= Number of tumor bearing animals/Number of animals examined, excluding those that died before week
52 (interim sacrifice).
b = Trend and pairwise p-values not reported since tumor incidence decreased with increasing dose.
Table 4.18. Malignant Lymphomas in Male CD-I Mice (Wood et al., 2009b)
Fisher's Exact Test and Cochran-Armitage Trend Test Results.
Dose (mg/kg/day)
0
71.4
234.2
810
Malignant
Lymphoma
Incidence
(%)
Raw p-value =
Adjusted p-value
0/44
(0)
0.006**
0.025*
1/46
(2)
0.511
0.511
2/48
(4)
0.269
0.359
5/45
(11)
0.029*
0.059
Note: Trend test results denoted at control: * denotes significance at p=0.05; ** denotes significance at
p=0.01
a= Number of tumor bearing animals/Number of animals examined, excluding those that died before week
52 (interim sacrifice).
4.6.5 Sugimoto, 1997 (MRID 50017108 - 50017109)26
In a carcinogenicity study, glyphosate (purity 97.56 and 94.61%; two lots) was administered
to groups of 50 male and 50 female Specific-Pathogen-Free (SPF) ICR (Crj: CD-I)
mice/dose in the diet at dose levels of 0, 165/153.2, 838.1/786.8, or 4348/4116 mg/kg/day
for males and females, respectively, for 18 months. The highest dose tested in this study far
exceeds the highest dose recommended in the test guidelines on how to conduct
carcinogenicity studies (OCSPP 870.4200 and OCSPP 870.4300). Furthermore, the mid-
dose tested in this study was approaching 1,000 mg/kg/day. Tumor findings at these high
doses are given less weight.
26Note: In Greim et al. (2015), the same study is cited as Arysta Life Sciences (1997b)
Page 89 of 216
-------
There were no treatment-related effects on mortality or survival. There were no changes in
histopathological findings observed.
Hemangiomas in female mice were found to occur at different sites. The tumor incidences are
presented in Table 4.19. A statistically significant trend was observed. Tumor incidence at the
high-dose, which was approximately 4 times the recommended high-dose in test guidelines
(4116 mg/kg/day), was statistically significant as compared to concurrent controls.
Table 4.19. Hemangioma Incidences (Sugimoto, 1997)
Fisher's Exact Test and Cochran-Armitage Trend Test Results
Tumor Type
0 mg/kg/day
153.2 mg/kg/day
786.8 mg/kg/day
4116 mg/kg/day
Hemangioma
Incidence
(%)
Raw p-value =
Adjusted p-value
0/48
(0)
0.002**
0.005**
0/47
(0)
1.000
1.000
2/45
(4)
0.231
0.231
5/45
(11)
0.024*
0.035*
Note: Trend test results denoted at control: * denotes significance at p=0.05; ** denotes significance at p=0.01.
a= Number of tumor bearing animals/Number of animals examined, excluding those that died before week
52 (interim sacrifice).
4.6.6 Pavkov and Turnier, 1987 (MRIDs 40214006, 41209907)
Glyphosate trimesium salt (sulfosate, 56.2% pure) was tested in a 2-year chronic
feeding/carcinogenicity study in male and female CD-I mice. Sixty animals/sex were tested in
control group 1 (basal diet, no vehicle), 80/sex were tested in control group 2 (basal diet plus
propylene glycol at 1% w/w vehicle) and in the low- and mid-dose groups, and 90/sex were
tested in the high-dose group. The following dose levels were tested: 0, 11.7/16, 118/159, and
991/1341 mg/kg/day for males and females, respectively.
No adverse effects on survival were seen in either sex across the doses tested. There were no
changes in histopathological findings observed. There were no treatment-related increases in
tumor incidence observed in the study.
4.6.7 Summary of Mouse Data
No tumors were identified for evaluation in 2 of the 6 mouse carcinogenicity studies. In the
remaining 4 mouse studies, 3 observed a statistically significant trend in tumor incidences
in the hemangiosarcomas, malignant lymphomas, or hemangiomas following adjustment for
multiple comparisons. In one individual study, a statistically significant pairwise
comparison was only observed following adjustment for multiple comparisons for
hemangiomas at the highest dose tested, which was more than 4X the limit dose. There was
no evidence of corroborating pre-neoplastic or related non-neoplastic lesions or evidence of
tumor progression (progression from pre-neoplastic to malignancy) to support biological
significance of tumor findings. In a limited number of cases, historical control data were
available which the observed tumor incidences were within the historical control range.
Page 90 of 216
-------
Table 4.20. Summary of Mouse Carcinogenicity Studies
Study
Dose Range
Pre-Neoplastic or Related
Non-Neoplastic Lesions
Tumors Incidences, Statistical Significance, and Related Comments
Reyna and Gordon (1973)
Swiss white mice
0, 17 or 50 mg/kg/day for 18 months
None observed
There were no tumors identified for evaluation.
Knezevich and Hogan (1983)
CD-I mice
99.78% Technical in diet
0, 161/195, 835/968, 4945/6069 mg/kg/day for
[M/F] for 24 months.
Chronic interstitial nephritis
lacked dose-response and not
considered relevant to renal
tumors. Tubular epithelial
changes in kidney were
approximately the same in
controls, low- and mid-doses
and then decreased at high-
dose.
No statistical significance in trend or pairwise comparisons, including the
mid- and high-doses which approached or exceeded 1,000 mg/kg/day.
Incidence of adenomas within historical control range for performing
laboratory.
Atkinson et al (1993b).
CD-I mice
97.5 - 100.2% Technical in diet
0, 98/102,297/298, 988/1000 mg/kg/day for 104
weeks (M/F)
None observed
Statistically significant trend for hemangiosarcomas that were only
observed in 4/45 (9%) high-dose male mice. Increased incidence was not
statistically significant from the concurrent controls at all doses, including
the highest dose tested which is approximately 1,000 mg/kg/day.
Wood et al (2009b)
CD-I mice
95.7%) Technical in diet
0, 71.4/97.9, 234.2/299.5, or 810/1081.2
mg/kg/day [M/F] for 80 weeks
None observed
No statistically significance in trend or pairwise comparisons following
adjustment for multiple comparisons. Negative trend observed for
adenomas.
Statistically significant trend for malignant lymphoma with incidences of
0/44 in controls, 1/46 at the low-dose, 2/48 at the mid-dose, and 5/45 at the
high-dose. No statistically significant pairwise results following
adjustment for multiple comparisons, including the highest dose tested
which was approaching 1,000 mg/kg/day. All observed incidences within
historical control range for performing laboratory.
Page 91 of 216
-------
Table 4.20. Summary of Mouse Carcinogenicity Studies
Study
Dose Range
Pre-Neoplastic or Related
Non-Neoplastic Lesions
Tumors Incidences, Statistical Significance, and Related Comments
Sugimoto (1997)
CD-I mice
94.61 - 97.56% Technical in diet
0, 165/153.2, 838.1/786.8, or 4348/4116
mg/kg/day [M/F] for 18 months
None observed
Statistically significant trend for hemangiomas in female mice following
adjustment for multiple comparisons with incidences of 0/48 in controls,
0/47 at the low-dose, 2/45 at the mid-dose, and 5/45 at the high-dose.
Increased incidence at high-dose statistically significant following
adjustment for multiple comparisons. Highest dose tested was more than
4X the limit dose.
Pavkov and Turnier (1987)
CD-I mice
56.2% Technical (Trimesium salt; Sulfosate)
0, 11.7/16, 118/159, and 991/1341 mg/kg/day
[M/F] for 24 months.
None observed
There were no tumors identified for evaluation, including the highest dose
tested which approached/exceeded 1,000 mg/kg/day.
Page 92 of 216
-------
4.7 Absorption, Distribution, Metabolism, Excretion (ADME)
The 2005 EPA Guidelines for Carcinogen Risk Assessment also permit analysis of other key
data that may provide valuable insights into the likelihood of human cancer risk from exposure
to a chemical, such as information regarding the absorption, distribution, metabolism, and
excretion (ADME) of a test chemical. EPA's Harmonized Test Guidelines for pesticides include
a series of studies for characterizing a chemical's metabolism and pharmacokinetics. As
described in the test guideline (OCSPP 870.7485), testing of the disposition of a test substance is
designed to obtain adequate information on its: absorption, distribution, biotransformation
(metabolism), and excretion, which can all collectively aid in understanding the chemical's
mechanism of toxicity. Basic pharmacokinetic/toxicokinetic parameters determined from these
studies can also provide information on the potential for accumulation of the test substance in
tissues and/or organs and the potential for induction of biotransformation as a result of exposure
to the test substance. These data can be used to assess the adequacy and relevance of the
extrapolation of animal toxicity data (particularly chronic toxicity and/or carcinogenicity data) to
estimate human risk.
Oral exposure is considered the primary route of concern for glyphosate. The maximum
absorption from the GI tract for glyphosate was estimated to be -30% with one study showing up
to 40% based upon radiolabel detected in the urine. In general, the amounts of glyphosate
detected in tissues were negligible indicating low tissue retention following dosing. Parent
glyphosate is the principal form excreted in urine and feces. The primary route of excretion
following oral administration of glyphosate is the feces, as verified by the intravenous dosing
and bile cannulation experiments. Within the dose ranges tested, elimination was essentially
complete by 24 hours indicating that glyphosate does not bioaccumulate.
Multiple studies examined the pharmacokinetics of a single dose of radiolabeled glyphosate
ranging from 5.6 - 400 mg/kg. Across these studies, time to reach peak plasma concentrations
(Tmax) appeared to increase with increasing dose; however, the reported range of Tmax (1-5.5
hours) suggests only a slight shift in absorption kinetics occurs despite large increases in dose.
In the one study that tested two doses (NTP, 1992), data graphically show that peak blood levels
were only roughly 3-fold with a 10-fold increase between the two doses. Reported area under
the curve (AUC) values indicated conflicting results regarding whether linear or non-linear
absorption kinetics was occurring at higher doses.
In general, EPA and OECD guideline ADME studies are designed for a different purpose and do
not provide the information needed to adequately determine whether linear kinetics is still
occurring at high doses of glyphosate. These studies are often limited to one or two doses and do
not include time course data. A well-conducted pharmacokinetic study testing multiple doses is
needed to conclusively make this determination.
Page 93 of 216
-------
4.8 Discussion
Glyphosate has been extensively tested in rodents to evaluate its carcinogenic potential. A total
of 14 rodent carcinogenicity studies were considered to be adequate for this analysis. Eight
studies were conducted in the rat and 6 studies were conducted in the mouse. When a potential
tumor signal was identified in a study, the agency considered several factors. Consistent with the
EPA's 2005 Guidelines for Carcinogen Risk Assessment, the agency evaluated the tumor
responses for both statistical and biological significance by considering factors such as historical
control data; rarity of tumor types; tumors at multiple sites; tumors in multiple species, strains, or
both sexes; progression of lesions from preneoplastic to benign to malignant; reduced latency of
neoplastic lesions (i.e., time to tumor); presence of metastases; unusual magnitude of tumor
response; proportion of malignant tumors; and dose-related increases. When these factors were
considered together, the agency made a determination of whether or not the observed tumor was
related to treatment with glyphosate. A weight of the evidence approach was used to determine
the carcinogenic potential of glyphosate in rodents.
In 4 of the 8 rat studies conducted with glyphosate, no tumors were identified for evaluation. Of
the remaining 4 rat studies, tumor incidences were evaluated in detail for testicular, pancreatic,
hepatocellular, thyroid C-cell, and mammary gland tumors. In 2 of the 6 mouse studies, no
tumors were identified for evaluation. In the remaining 4 mouse studies, tumor incidences were
evaluated in detail for hemangiosarcomas, malignant lymphoma, hemangiomas, lung, and kidney
tumors. Below are the weight of evidence evaluations for each tumor type.
Testicular Tumors
In Table 4.1, a statistically significant trend was observed for testicular interstitial cell tumors
(adjusted p-value = 0.032) and pairwise significance was observed at the highest dose tested of
31 mg/kg/day (adjusted p-value = 0.032) in male Sprague-Dawley rats (Lankas, 1981). Closer
examination of the tumor incidence indicated that the data do not demonstrate a monotonic dose
response with greater incidence observed at the low-dose as compared at the mid-dose. There
was a lack of preneoplastic or related non-neoplastic lesions to support a treatment-related effect.
It was also noted that the incidence of testicular cell tumors in concurrent controls (0%) appears
to be unusually low for this tumor type as compared to historical controls from the performing
laboratory. These data also indicated that the incidence at the highest dose tested was outside the
historical control range when all animals were considered, but within the terminal historical
control range for the performing laboratory. Testicular interstitial cell tumors are relatively
common in Sprague-Dawley rats and this tumor type is difficult to distinguish from simple
hyperplasia. Testicular tumors were not seen in the other 7 rat studies, many of which tested up
to or beyond the limit dose (1000 mg/kg/day). More specifically, of the 4 other studies
performed in Sprague-Dawley rats (Stout and Ruecker, 1990; Atkinson et al., 1993a; Pavkov and
Wyand, 1987; Enemoto, 1997), 3 were tested at doses 3OX higher or more than the highest dose
tested in Lankas (1981) and no testicular tumors were observed. Furthermore, there were no
testicular tumors observed in the 6 mouse bioassays.
Pancreatic Tumors
In Table 4.2, no statistically significant trends were observed for pancreatic islet cell tumors in
male Sprague-Dawley rats (Stout and Ruecker, 1990). Raw p-values were statistically
Page 94 of 216
-------
significant for adenomas at the low and high dose for pairwise comparisons; however, these were
not statistically significant following adjustment for multiple comparisons. Closer examination
of the tumor incidence indicated that the data do not demonstrate a monotonic dose response
with greater incidence observed at the low-dose as compared at the mid-dose and high-dose.
There was a lack of preneoplastic or related non-neoplastic lesions to support a treatment-related
effect. Historical control data from the performing laboratory for pancreatic adenomas indicated
that the incidence in concurrent controls was at the lower limit of the historical control range.
There was no evidence of progression to malignancy. Notably, carcinomas demonstrated a
negative trend with decreasing tumor incidence with increasing dose. Pancreatic tumors were
not observed in the other 7 rat studies, including 4 other studies performed in Sprague-Dawley
rats (Atkinson et al., 1993a; Pavkov and Wyand, 1987; Enemoto, 1997; Lankas, 1981).
Furthermore, pancreatic tumors were not observed in the 6 mouse bioassays.
Hepatocellular Tumors
Hepatocellular tumors were evaluated in 2 rat carcinogenicity studies (Stout and Ruecker, 1990;
Brammer, 2001). In Table 4.4, the raw p-value for trend was statistically significant for
adenomas in male Sprague-Dawley rats (Stout and Ruecker, 1990); however, it was not
statistically significant following adjustment for multiple comparisons. There were no
statistically significant pairwise comparisons. Historical control data from the performing
laboratory indicated that the incidence of hepatocellular tumors was within the historical control
range at all doses. Closer examination of the tumor incidence indicated that the data fluctuated
with no tumors observed in concurrent controls or the mid-dose and increases seen at the low-
dose and high-dose. In Table 4.10, a statistically significant trend was observed for adenomas in
male Wistar rats (adjusted p-value = 0.029) (Brammer, 2001). For pairwise comparisons, the
raw p-value was statistically significant at the highest dose tested (1214 mg/kg/day); however, it
was not statistically significant following adjustment for multiple comparisons. There was a lack
of preneoplastic or related non-neoplastic lesions to support a treatment-related effect in both
studies. There was no evidence of progression to malignancy. In particular, carcinomas
demonstrated a negative trend with decreasing tumor incidence with increasing dose.
Hepatocellular tumors were not observed in the other 6 rat studies, including 4 other studies in
Sprague-Dawley rats (Atkinson et al., 1993a; Pavkov and Wyand, 1987; Enemoto, 1997;
Lankas, 1981) and 2 other studies in Wistar rats (Suresh, 1996; Wood et al. 2009a) the same rat
strains. Furthermore, hepatocellular tumors were not observed in the 6 mouse bioassays.
Thyroid Tumors
In Table 4.6, there were no statistically significant trends or pairwise comparisons were observed
for thyroid C-cell tumors in male Sprague-Dawley rats (Stout and Ruecker, 1990). In Table 4.7,
a statistically significant trend for adenomas and combined adenomas/carcinomas was observed
with raw p-values in female Sprague-Dawley rats (Stout and Ruecker, 1990); however, the
trends were not statistically significant with adjustment for multiple comparisons. There were no
statistically significant pairwise comparisons observed at any dose. There was a lack of
preneoplastic or related non-neoplastic lesions to support a treatment-related effect. There was
no evidence of progression to malignancy. Thyroid tumors were not observed in the other 7 rat
studies, including 4 other studies performed in Sprague-Dawley rats (Atkinson et al., 1993a;
Pavkov and Wyand, 1987; Enemoto, 1997; Lankas, 1981). Furthermore, thyroid tumors were
not observed in the 6 mouse bioassays.
Page 95 of 216
-------
Mammary Gland Tumors
In Table 4.11, raw trend p-values were statistically significant for mammary gland
adenocarcinomas and combined adenomas/adenocarcinomas in female Wistar rats (Wood et al.,
2009a); however, only the combined p-value for trend remained statistically significant
following adjustment for multiple comparison (adjusted p-value = 0.028). There were no
statistically significant pairwise comparisons. There was a lack of preneoplastic or related non-
neoplastic lesions to support a treatment-related effect. Mammary gland tumors were not
observed in the other 7 rat studies, including 2 other studies performed in Wistar rats (Brammer,
2001; Suresh, 1996). Furthermore, mammary gland tumors were not observed in the 6 mouse
bioassays.
Kidney Tumors
In Table 4.13, there were no statistically significant trend observed for renal tubular cell tumors
in male CD-I mice (Knezevich and Hogan, 1983). This study tested up to almost 5000
mg/kg/day. Historical control data from the performing laboratory indicated that the incidence
of adenomas was within the historical control range. There was a lack of preneoplastic or related
non-neoplastic lesions to support a treatment-related effect. There was no evidence of
progression to malignancy. Kidney tumors were not observed in the other 5 mouse studies,
including 4 other studies performed in CD-I mice (Atkinson et al., 1993b; Wood et al., 2009b;
Sugimoto, 1997; Pavkov and Turnier, 1987). Furthermore, kidney tumors were not observed in
the 8 rat bioassays.
Hemangiosarcomas
In Table 4.16, a statistically significant trend was observed for hemangiosarcomas in male CD-I
mice (adjusted p-value = 0.006) (Atkinson et al., 1993b). There were no statistically significant
pairwise comparisons. There was a lack of preneoplastic or related non-neoplastic lesions to
support a treatment-related effect. There was no evidence of progression to malignancy.
Hemangiosarcomas are commonly observed in mice (generally more common in males for CD-I
strain) as both spontaneous and treatment-related tumors arising from endothelial cells.
Hemangiosarcomas were not observed in the other 5 mouse studies, including 4 other studies
performed in CD-I mice (Knezevich and Hogan, 1983; Wood et al., 2009b; Sugimoto, 1997;
Pavkov and Turnier, 1987). Furthermore, hemangiosarcomas were not observed in the 8 rat
bioassays.
Lung Tumors
In Table 4.17, the raw p-value for trend was observed for lung adenocarcinomas; however, the
trend was not statistically significant following adjustment for multiple comparisons in male CD-
1 mice (Wood etal., 2009b). There were no statistically significant pairwise comparisons.
Tumor incidences at all doses were within the historical control range for the performing
laboratory. There was a lack of preneoplastic or related non-neoplastic lesions to support a
treatment-related effect. Lung tumors were not observed in the other 5 mouse studies, including
4 other studies performed in CD-I mice (Knezevich and Hogan, 1983; Atkinson et al., 1993b;
Sugimoto, 1997; Pavkov and Turnier, 1987). Furthermore, lung tumors were not observed in the
8 rat bioassays.
Page 96 of 216
-------
Malignant Lymphoma
In Table 4.18, statistically significant trend was observed in male CD-I mice (adjusted p-value =
0.025) (Wood et al., 2009b). For pairwise comparisons, the raw p-value for the highest dose
tested was statistically significant; however, it was not statistically significant following
adjustment for multiple comparisons. Malignant lymphoma was not observed in the other 5
mouse studies, including 4 other studies performed in CD-I mice (Knezevich and Hogan, 1983;
Atkinson et al., 1993b; Sugimoto, 1997; Pavkov and Turnier, 1987). Furthermore, malignant
lymphoma was not observed in the 8 rat bioassays.
Hemangiomas
In Table 4.19, a statistically significant trend was observed in female CD-I mice (adjusted p-
value = 0.005) (Sugimoto, 1997). For pairwise comparisons, the incidence at the highest dose
tested was statistically significant (adjusted p-value = 0.035). The highest dose tested in this
study was more than 4X the limit dose. Hemangiomas were not observed in the other 5 mouse
studies, including 4 other studies performed in CD-I mice (Knezevich and Hogan, 1983;
Atkinson et al., 1993b; Wood et al., 2009b; Pavkov and Turnier, 1987). Furthermore,
hemangiomas were not observed in the 8 rat bioassays.
Based on the weight-of-evidence evaluations, the agency has concluded that none of the tumors
evaluated in individual rat and mouse carcinogenicity studies are treatment-related due to lack of
pairwise statistical significance, lack of a monotonic dose response, absence of preneoplastic or
related non-neoplastic lesions, no evidence of tumor progression, and/or historical control
information (when available). Tumors seen in individual rat and mouse studies were also not
reproduced in other studies, including those conducted in the same animal species and strain at
similar or higher doses.
Page 97 of 216
-------
5.0 Data Evaluation of Genetic Toxicity
5.1 Introduction
Genotoxicity is a broad term for any damage to the genetic material, whether the damage is
transient or permanent. Transient damage refers to unintended modifications to the structure of
DNA, which may or may not undergo successful repair. Permanent damage refers to heritable
changes in the DNA sequence, known as mutations. Types of mutations include: 1) changes in
single base pairs, partial, single or multiple genes, or chromosomes, 2) breaks in chromosomes
that result in transmissible deletion, duplication or rearrangement of chromosome segments, and
3) mitotic recombination (OECD, 2015). In somatic cells, DNA-reactive chemicals can cause
cancer if the mutations occur within regulatory genes that control cell growth, cell division and
differentiation, such as proto-oncogenes, tumor suppressor genes and/or DNA damage response
genes (OECD, 2015). Additionally, DNA damage may signal the cell to undergo apoptosis (cell
death) rather than cell division and, therefore, the damage is not "fixed" as a mutation and is not
passed along to daughter cells.
Evaluation of genotoxicity data entails a weight-of-evidence approach that includes
consideration of the various types of genetic damage that can occur. Since no single genotoxicity
assay evaluates the many types of genetic alterations that can be induced by a chemical, one
must employ a battery of genotoxicity tests to adequately cover all the genetic endpoints
important for regulatory decisions. EPA, like other regulatory agencies, considers genotoxicity
information as part of the weight of evidence when assessing the potential of a chemical to
induce cancer in humans. Under FIFRA, OPP requires genotoxicity tests of the technical grade
active ingredient for the registration of both food and non-food use pesticides. The current
genotoxicity test battery (40 CFR Part 158.500) for pesticide registration consists of:
1) Bacterial reverse mutation test (typically conducted in bacteria strains Salmonella
typhimurium and Escherichia coli),
2) in vitro mammalian (forward) gene mutation and in vitro mammalian chromosomal
aberration test, and
3) in vivo test for micronucleus induction (mammalian erythrocyte micronucleus test)
or in vivo chromosomal aberration test (mammalian bone marrow chromosomal
aberration test).
In cases where equivocal or inconsistent results are obtained for the same endpoint in different
test systems, additional testing may be required. Test Guidelines on how to conduct the
genotoxicity tests have been published by the agency and have been harmonized with the
Organization for Economic Cooperation and Development (OECD, 2015; Cimino 2006). These
guidelines identify specific test species, genetic endpoints, test conditions, exposure durations as
well information on how to report data and interpret the results. The test guidelines provide a
level of consistency and predictability for regulatory compliance and regulatory decision making.
Page 98 of 216
-------
5.2 Scope of the Assessment Considerations for Study Quality Evaluation
Previous genotoxicity assessments conducted as part of the CARC reviews for glyphosate in
1991 and 2015, considered only studies conducted with glyphosate technical and included only
studies that provided adequate characterization of the test material {i.e. purity information
provided). In the current analysis, a fit-for-purpose systematic review process was conducted to
identify relevant genotoxicity data from regulatory studies and published literature from open
sources (published and unpublished) for both glyphosate technical and glyphosate-based
formulations. Studies conducted with glyphosate formulations that were identified and
considered relevant for genotoxicity evaluation are summarized in table form in Appendix F. As
described in Section 7.0 of this document, glyphosate formulations are hypothesized to be more
toxic than glyphosate alone. The agency is collaborating with NTP to systematically investigate
the mechanism(s) of toxicity for glyphosate and glyphosate formulations. However, the focus of
this section is the genotoxic potential of glyphosate technical.
As described previously in Section 2.1.3, the list of studies identified in this process were also
cross-referenced with genotoxicity review articles for glyphosate from the open literature [Kier
and Kirkland (2013), and Williams et al. (2000)], as well as recent international evaluations of
glyphosate (IARC 2015, EFSA 2015, JMPR 2016). The current analysis also includes studies
conducted by other registrants that were not previously available to the agency. Sixteen studies
for glyphosate technical that were included in Kier and Kirkland (2013) were not available to the
agency; therefore, data and study summaries provided in the review articles were relied upon in
the current review and are identified in the data tables with a footnote. The Kier and Kirkland
(2013) article serves as the original publication for these studies and provided relevant
information on study design and conditions as well as summary data. The data set includes in
vitro and in vivo studies conducted in mammalian systems, with the exception of standard
bacterial test strains, which have a long history of detecting chemicals that are mutagenic in
humans. Studies conducted in non-mammalian species (e.g. worms, fish, reptiles, plants), were
excluded because they were considered to be not relevant for informing genotoxic risk in
humans. Several epidemiological studies that evaluated biomarkers for genotoxicity were not
included in this evaluation because these studies were assigned a low quality ranking as
described in Section 3.3.
When evaluating the quality of the published and unpublished data for inclusion in the analysis,
the agency considered the reporting quality (how well a study was reported), the study design
and how well the study was conducted. Critical elements in study design and interpretation for
genotoxicity tests are described in the various EPA and OECD test guidelines. Elements such as
test conditions (e.g. solubility, pH, osmolality, and cytotoxicity) and study design (e.g. number
of test organisms, doses selected, use of positive and negative controls; blinded evaluation) were
used to evaluate the quality of published and non-published studies. In cases where
inappropriate testing conditions or study design clearly had an impact on the outcome the study,
the study was excluded from the analysis. For example, early studies by Majeska (1982) were
excluded from the analysis since it was clearly demonstrated that altered pH by the test chemical
can result in false positive responses in several of in vitro genotoxicity tests (Majeska,
1985d,e,f). In other cases, particularly with the published literature studies, where test
conditions and/or study design differed from what is generally considered as acceptable
Page 99 of 216
-------
following in the EPA or OECD guidelines, the differences are noted, but the studies were not
excluded from analysis unless the condition made the study unreliable. Summaries of relevant
genotoxicity studies can be found in TXR# 0057499. Studies that were excluded from the
analysis are listed in Appendix G.
The studies evaluating the genetic toxicity of the active ingredient glyphosate are presented in
the following sections according to the type of genetic endpoints evaluated: mutations,
chromosomal aberrations and other assays evaluating DNA damage. In vitro and in vivo assays
are discussed separately according to the genetic endpoint. For the purpose of this analysis,
glyphosate and its salts are considered together when evaluating the genotoxic potential of the
active ingredient glyphosate.
5.3 Tests for Gene Mutations for Glyphosate Technical
5.3.1 Bacterial Mutagenicity Assays
Bacteria have traditionally been employed as a primary test organism for the detection of
chemical mutagens. The bacterial reverse mutation assay is routinely performed in the test
strains of Salmonella typhimurium and Escherichia coli. These test strains are mutant strains
that are deficient for the synthesis of an essential amino acid. The assay detects mutations that
revert the test strains back to wild type for amino acid synthesis and the revertants are identified
by their ability to grow in culture medium deficient of the specific amino acid(s). This
mutagenicity test identifies point mutations, which includes base substitutions and deletions and
insertions of up to a few base pairs (OECD 471). The tests are typically conducted in the
presence and absence of an exogenous source of metabolic activation (e.g., S9 microsomal
fraction of activated liver homogenates) to identify potential mutagenic metabolites.
Glyphosate has been extensively evaluated for its potential to induce mutations in bacteria. Most
of the studies considered consist of the full battery of bacterial strains {i.e. the recommend strains
in EPA and OECD Test Guidelines) and were evaluated at appropriate test concentrations (up to
cytotoxic or assay limit concentrations).
EPA identified 27 studies that tested glyphosate technical in bacterial mutagenicity assays by
means of the standard plate incorporation method or the pre-incubation modification of the
standard assay. Glyphosate was negative in the presence and absence of metabolic activation in
all the studies. The results of the bacterial reversion mutation assays evaluating glyphosate
technical are presented in Table 5.1
Page 100 of 216
-------
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical.
Test/
Endpoint
Test System
Concentrations
Purity
Results
Reference
Comments
Bacterial Reverse
Mutation
S. typhimurium
TA1535,
TA1537, TA98
and TA100 and
WP uvrA ± S9
156-5000 (ig/plate
95.68%
Negative ± S9
Akanuma (1995)
[MRID 50017102]
Bacterial Reverse
Mutation
S. typhimurium
TA535, TA1537,
TA98 and
TA100 and E.
coli WP2P and
WP2P uvrA ± S9
100-5000 (ig/plate in
DMSO
95.6%
glyphosate
acid
Negative ± S9
Callander (1996)
[MRID 44320617]
Bacterial Reverse
Mutation
S. typhimurium
TA 1535,
TA1537, TA98
and TA 100 and
E. coli WP2P
and WP2P uvrA
±S9
100-5000 (ig/plate in
water
60%
potassium
glyphosate salt
Negative ± S9
Callander (1999)1
Bacterial Reverse
Mutation
S. typhimurium
TA97a, TA98,
TA100 and
TA102, ± S9
25-2000 (ig in
aqueous solution
Not provided
Negative ± S9
Chruscielska et al. (2000)
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537
± S9
10-1000 ng/plate
98.4%
Negative ± S9
Flowers and Kier (1978)
[MRID 00078620]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102, TA1535,
TA1537 ± S9
31.6-3160 ng/plate
98.8%
Negative ± S9
Fliigge (2009a)1
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102, TA1535,
TA1537 ± S9
31.6-3160 ng/plate
96.4%
technical
Negative ± S9
Fliigge (2010b)1
Page 101 of 216
-------
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical.
Test/
Endpoint
Test System
Concentrations
Purity
Results
Reference
Comments
Bacterial Reverse
Mutation
S. typhimurium
TA1535,
TA1537, TA98
and TA100
310-5000 (ig/plate
(+S9); 160-2500
(ig/plate (-S9)
98.6%
Negative ± S9
Jensen (1991a)
[MRID 49961502]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102, TA1535,
TA1537 ± S9
1-1000 (ig/plate
98.05%
Negative ± S9
Miyaji (2008)1
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537,
TA1538 ± S9
5000 (ig/plate
Not reported
Negative ± S9
Moriya et al. (1983)
Bacterial Reverse
Mutation
S. typhimurium
TA1535, TA97,
TA98 and
TA100 ± S9
33-10,000 (ig/plate
99%
Negative ± S9
NTP (1992)
Hamster and rat S9
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535 and
TA97a ± S9
1-5000 ng/plate
61.27%
Glyphosate
isopropyl-
ainine salt
Negative ± S9
Ranzani (2000)1
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102, TA1535,
TA1537 ± S9
648-5000 ng/plate
98.01%
Negative ± S9
Ribeiro do Val (2007)
[MRID 50000903]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 andit.
Coli WP2 uvrA ±
S9
31.6-5000 ng/plate
96.0%
technical
Negative ± S9
Schreib (2010)1
Page 102 of 216
-------
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical.
Test/
Endpoint
Test System
Concentrations
Purity
Results
Reference
Comments
Bacterial Reverse
Mutation
S. typhimurium
TA1535,
TA1537,
TA1538, TA98,
TA100 and E.
coli WP2 her ±
S9
10-5000 (ig/plate
98.4%
Negative ± S9
Sliirasu et al. (1978)
[MRID 00078619]
Published in Li &
Long, 1988
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 and it.
coli WP uvrA ±
S9
3-5000 ng/plate
(plate-incorporation),
33-5000 ng/plate
(pre-incubation test)
95.1%
Negative ± S9
Sokolowski (2007a)
[MRID 49957406]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 and it.
coli WP uvrA ±
S9
3-5000 ng/plate
(plate-incorporation)
33 - 5000 ng/plate
(pre-incubation test)
97.7%
Negative ± S9
Sokolowski (2007b)
[MRID 49957407]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 andit.
coli WP uvrA ±
S9
3-5000 ng/plate
(plate-incorporation)
33-5000 ng/plate
(pre-incubation test)
95.0%
Negative ± S9
Sokolowski (2007c)
[MRID 49957408]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 andit.
coli WP uvrA ±
S9
3-5000 ng/plate
96.66%
technical
Negative ± S9
Sokolowski (2009a)1
Page 103 of 216
-------
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical.
Test/
Test System
Concentrations
Purity
Results
Reference
Comments
Endpoint
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 and it.
coli WP2 uvrA
pKM 101 and
WP2 pKM 101 ±
S9
3-5000 ng/plate
96.3%
glyphosate
acid
Negative ± S9
Sokolowski (2009b)
[MRID 49961801]
Bacterial Reverse
S. typhimurium
3-5000 ng/plate
97.16%
Negative ± S9
Sokolowski (2010)
Mutation
TA98, TA100,
TA1535,
TA1537 and it.
coli WP uvrA ±
S9
[MRID 50000902]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537,
TA1538 ± S9
1-1000 ng/plate
96.0%
Negative ± S9
Suresh (1993a)1
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 andit.
coli WP uvrA ±
S9
0-5000 ng/plate
95.3%
Negative ± S9
Thompson (1996)
[MRID 49957409]
Bacterial Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102, TA1535,
TA1537 ± S9
31.6-5000 ng/plate
98.2%
Negative ± S9
Wallner (2010)1
Bacterial Reverse
Mutation
S. typhimurium
TA98 and
TA100 ± S9
25 ng/plate
Not reported
Negative ± S9
Wildennan and Nazar
(1982)
Rat S9 and plant cell-
free homogenates were
used for metabolic
activation
Page 104 of 216
-------
Table 5.1. In vitro Test for Gene Mutations in Bacteria: Glyphosate Technical.
Test/
Test System
Concentrations
Purity
Results
Reference
Comments
Endpoint
Bacterial Reverse
Mutation
S. typhimurium
TA1535,
TA1537,
TA1538, TA98
and TA100 ± S9
0.12-10 mg/plate -S9
0.56-15 mg/plate +S9
90%
glyphosate
trimesium salt
Negative ± S9
Majeska et al. (1982a)
[MRID 00126612]
Bacterial Reverse
Mutation
S. typhimurium
TA1535,
TA1537, TA98
and TA100 ± S9
0.005-50 (iL/mL
55.6%
glyphosate
trimesium salt
Negative ± S9
Majeska (1985a)
[MRID 00155527]
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline, test material purity,
control chemicals and summary data tables.
Page 105 of 216
-------
5.3.2 In vitro Tests for Gene Mutations in Mammalian Cells
In vitro gene mutation studies in mammalian cells are conducted in cell lines with reporter genes
for forward mutations. The most common reporter genes are the endogenous thymidine kinase
(TK) gene, endogenous hypoxanthine-guanine phosphoribosyl transferase (HPRT) gene and the
xanthine-guanine phosphoribosyl transferase transgene (XPRT). Mutations that occur within
these reporter genes result in mutant cells that are resistant to the cytotoxic effect of the
pyrimidine analogue trifluorothymidine (for TK) or the purine analogue 6-thioguanine (for
HPRT and XPRT) (OPPTS 870.5330). Suitable cell lines for this assay include L5178Y mouse
lymphoma cells, Chinese hamster ovary (CHO) cells, hamster AS52 and V79 lung fibroblasts
and human TK6 lymphoblastoid cells. Similar to other in vitro assays, chemicals are tested both
in the presence and absence of S9 metabolic activation.
A total of four studies were conducted for (forward) mutations in mammalian cells (Table 5.3).
Three studies were conducted with a high purity concentration of glyphosate technical (>95.6%)
and the remaining study was performed with glyphosate trimesium salt. In four of the assays,
mouse lymphoma L5178Y TK+/" cells were the target organism and one was conducted in CHO
cells with the HPRT endpoint. Glyphosate technical and the glyphosate trimesium salt were
negative in the mouse lymphoma cell assays (Jensen, 1991b; Clay, 1996; Majesak, 1985b) when
tested up to the current guideline limit concentration and glyphosate was negative in CHO/HPRT
cells when tested up to cytotoxic concentrations (Li, 1983a).
Page 106 of 216
-------
Table 5.2. In vitro Mammalian Gene Mutation Assays: Glyphosate Technical.
Test/Endpoint
Test System
Concentrations/
Conditions
Test Material
Purity
Results
Reference
Comments
Gene Mutations in
Mammalian Cells
Mouse lymphoma
L5178YTK+'" cells ± S9
296-1000 ng/mL
95.6%
Negative
Clay (1996)1
Relative survival was
90% (-S9) and 57%
(+S9) at top
concentration
Gene Mutations in
Mammalian Cells
Mouse lymphoma
L5178Y TK+/" cells ±S9
520-4200 ng/mL
(+S9); 610-5000
Hg/mL (-S9)
98.6%
Negative
Jensen (1991b)
[MRID 49961504]
Reported no significant
reduction in cloning
efficiency at any
concentration.
Gene Mutations in
Mammalian Cells
Chinese hamster ovary
(CHO) cells, HPRT
locus ± S9
500-25000 |ig/mL
(+S9); 500-22500
Hg/mL (-S9)
98.7%
Negative
Li (1983a);
[MRID 00132681]
Tested S9 from 1-10%
Cytotoxic at 22.5 mg/mL
(-S9, and with 1,2 and
10% S9) and at 17.5
mg/ml (10% S9)
Gene Mutations in
Mammalian Cells
Mouse lymphoma
L5178Y TK+/" cells ± S9
1-5 iil/mL
55.6%
Glyphosate
trimesium salt
Negative
Majeska (1985b)
[MRID 00155530]
Negative with pH
adjusted
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline, test material purity,
control chemicals and summary data tables.
Page 107 of 216
-------
5.4 In vitro Tests for Chromosomal Abnormalities
Cytogenetic assays are tests that can detect chemicals that cause structural chromosomal damage
(clastogenicity) or affect the segregation of chromosomes during cell division and alter
chromosome number (aneuploidy). Generally, there are two types of in vitro cytogenetic assays
that identify chemicals inducing chromosomal abnormalities: chromosomal aberration assays
and micronucleus assays. Although chromosomal damage observed in these assays are not
considered heritable mutations, chemicals that can induce these types of chromosomal damage
can also induce transmissible mutations to daughter cells indicating their role in cancer (Yauk et
a/., 2015; OECD 2015). In addition, assays such as (fluorescence in situ hybridization (FISH))
can provide additional mechanistic information on the formation of chromosomal abnormalities.
It is important to note that factors such as cytotoxicity, solubility of the test substance, changes in
pH or osmolality play a significant role in the outcome of the assay. Like other in vitro assays,
compounds are generally tested in the presence or absence of S9 metabolic activation to
determine if metabolism affects the genotoxic activity of the parent compound and to determine
if potential genotoxic metabolites are formed.
5.4.1 In vitro Mammalian Chromosomal Aberration Test
Chromosomal aberration assays detect both structural chromosomal and numerical aberrations.
Structural chromosomal aberrations are of two types: chromatid and chromosome and include
breaks, deletions and rearrangements (OPPTS 870.5375, OECD 2015). Numerical chromosomal
aberrations generally result from the loss of an entire chromosome mostly due to damage in the
spindle fiber resulting in aneuploidy. The types of cells that are most commonly used in
chromosomal aberration assays include established cell lines such as Chinese hamster lung
(CHL) and CHO cells or primary cell cultures such as human or other mammalian peripheral
blood lymphocytes. In this assay, cells are typically sampled at a time equivalent to the length of
approximately 1.5 cell cycles from the start of treatment. Prior to harvesting, cells are treated
with Colcemid® or colchicine to arrest cells at the first metaphase stage of the cell cycle
following the beginning of exposure to the test article. Once harvested, the cells are stained and
metaphase cells are evaluated microscopically for various types of chromosome aberrations.
(OECD TG 473). Data should be presented in a way that indicates the percentage of affected
cells in the population of cells scored (e.g., % cells with aberrations or # aberrant cells/100 cells).
Gaps should not be included in the analysis; they are scored but gaps alone in the absence of any
additional chromosomal aberrations (e.g., a fragment or a ring chromosome) are not sufficient to
define a cell as aberrant.
Glyphosate technical was evaluated in eight chromosomal aberrations tests to determine its
potential to induce clastogenic effects in vitro. The findings are presented in Table 5.3. Six of
the eight studies were negative. The two positive studies were both from the same laboratory
where, Lioi et al. reported an increase in chromosomal aberrations at glyphosate concentrations
of 8,5|iM and above in bovine lymphocytes (Lioi et al., 1998b) and at all concentrations of
glyphosate tested (7-170 (jM) in human lymphocytes (Lioi et al., 1998a) following a 72-hour
exposure period. No chromosomal aberrations were observed as a result of exposure to
glyphosate in one study using CHO cells (Majeska, 1985c) and in two studies with CHL cells
Page 108 of 216
-------
(Matsumoto, 1995; and Wright, 1996). Sivikova and Dianovsky (2006) reported no statistically
significant increases in chromosomal aberrations in bovine lymphocytes treated with glyphosate
(62% pure) at concentrations up 1120 [xM following 24-hour exposure. (Sivikova and
Dianovsky, 2006). In studies conducted with human lymphocytes treated with glyphosate
(>95%) for 24-96 hours at concentrations, no increase in chromosomal aberrations were seen at
concentrations as high as 6000 [xM (Fox, 1998; and Manas et al., 2009).
5.4.2 In vitro Mammalian Micronucleus Test
The in vitro micronucleus test can detect the induction of micronuclei in the cytoplasm of cells in
the interphase stage of the cell cycle. Micronuclei form from acentric chromosome fragments
(i.e., chromosome fragments lacking a centromere) or when whole chromosomes are unable to
migrate to the cellular poles during anaphase prior to cell division. (OECD 487). Thus, the
micronucleus assay can detect both structural and numerical chromosomal changes. It should be
noted, however, that additional work is required to distinguish whether induced micronuclei have
arisen from a clastogenic versus an aneugenic mechanism, e.g., staining micronuclei to detect the
presence of kinetochore proteins. The assay is typically performed with cell lines or primary cell
cultures of human or rodent origin. The assay can be conducted with the addition of
cytochalasin B which inhibits cytokinesis resulting in the formation of binucleated cells. The
presence of binucleated cells, indicates that cells have undergone one round of mitosis, a
necessary prerequisite for micronucleus formation.
Six studies evaluated glyphosate technical for its potential to induce micronuclei in vitro (Table
5.4). Four of the six studies were positive and the remaining two studies were equivocal. In a
study by Koller et al. (2012), TR146 cells (derived from a human neck metastasis of buccal
epithelial origin) were treated for 20 minutes with up to 20 mg/L (-0.12 mM) glyphosate (95%),
the authors reported a statistically significant increase in binucleated cells with micronuclei at 15
(-0.09 mM) and 20 (-0.12 mM) mg/L, and also indicated significant apoptosis and necrosis at
20 mg/L. The short exposure period in this study was unusually short (20 minutes) and was
conducted in a tumor cell line that had not been well characterized in regards to its degree of
chromosomal instability and DNA damage and repair capacity. In another study, Roustan et al.
(2014) reported positive findings +S9 only in CHO cells treated with glyphosate (unknown
purity) at 10- 100 [xg/mL with little evidence of a dose response over that concentration range.
Two other studies evaluated glyphosate technical in human lymphocytes (Mladinic et al., 2009a,
2009b). These studies used an exposure protocol that is different from the OECD
recommendations for the in vitro micronucleus assay. OECD recommends that whole blood or
isolated lymphocytes are cultured in the presence of a mitogen (e.g. phytohemagglutinin; PHA)
prior to exposure of a test chemical in order to detect micronuclei formed via an aneugenic
mechanism. However, in these two studies, blood cells were exposed to glyphosate for 4 hours,
washed, and then treated with PHA to stimulate cell division. Both studies reported a statistically
significant increase in micronucleated cells at 580 [xg/mL (-3.4 mM), but not at lower
concentrations, following 4-hour exposures in the presence of S9. The frequency of
micronucleated cells (+S9) ranged from 11.3 to 28.7 in one study (Mladinic et al., 2009a) and
33.3 to 65.2 in the other study (Mladinic etal., 2009b) over the 1000-fold concentration range.
No statistically significant increases in micronucleated cells were seen in either study in the
absence of S9 activation. When cells were evaluated with vital stains, cells treated with 580
Page 109 of 216
-------
[j,g/mL showed a significant (p<0.05) increase in the percentage of cells undergoing apoptosis
and necrosis compared to the negative controls.
Piesova et al. (2004, 2005) conducted two in vitro micronucleus studies using glyphosate
technical (62%) up to 560 |iM in bovine lymphocytes. In the 2004 study, bovine lymphocytes
from two donors were treated for 24 or 48 hours without S9 metabolic activation, and for 2 hours
(with and without S9 activation) or 48 hours (-S9) in the 2005 study. Both studies yielded
similar results following 48-hour exposure to glyphosate. In both cases, the authors reported a
weak induction of micronuclei in one donor at 280 [jM and at 560 [xM in the second donor. The
induction was approximately 2-fold (p < 0.05), but with no clear dose response. No effects on
micronuclei induction were seen at the 2- or 24-hour time points; however, with these early time
points it is unlikely that one cell division has occurred during or after treatment.
Page 110 of 216
-------
Table 5.3. In vitro Tests for Chromosome Aberrations in Mammalian Cells- Glyphosate Technical
Test/Endpoint
Test System
Concentrations/
Conditions
Test Material
Purity
Results
Reference
Comments
In vitro
Chromosomal
Aberration
Chinese hamster ovary
(CHO) cells
4-10 nl/mL, ± S9
55.6%
Glyphosate
trimesium salt
Negative
Majeska (1985c)
[MRID 00155530]
pH adjusted (7.4-7.6)
In vitro
Chinese Hamster lung
±S9: 0, 250, 500, 1000
95.68%
Negative
Matsumoto (1995)
Decline in pH noted at
Chromosomal
Aberration
(CHL) cells
and 2000 |ig/mL: 24
and 48 h treatment -
S9; 6 h treatment ±S9
harvest 24 h
[MRID 50017106]
500 and 1000 |ig/mL.
In vitro
Chinese hamster lung
-S9: 24 & 48-hr
95.3%
Negative
Wright (1996)
Excessive decrease in
Chromosomal
Aberration
(CHL) cells
exposure: 0-1250
lig/mL:
+S9: 0-1250 ng/mL
[MRID 49957410]
pH >1250 |ig/mL
In vitro
Bovine lymphocytes
-S9 only: 0, 7, 85 and
>98%
Positive
Lioi et al. (1998b)
Chromosomal
170 yM:
(all cones.)
Aberration
72 h exposure
In vitro
Bovine lymphocytes
±S9: 0, 28, 56, 140,
62.0%
Negative
Sivikova and
Decreased MI and PI at
Chromosomal
280, 560 and 1120
Dianovsky (2006)
> 560 nM
Aberration
HM;
24 h exposure
In vitro
Human lymphocytes
±S9: 100-1250 ng/mL
95.6%
Negative
Fox (1998)
Excessive decrease in
Chromosomal
Aberration
cultures analyzed;
68 & 92 h
[MRID 49961803]
pH >1250 |ig/mL
In vitro
Chromosomal
Human lymphocytes
-S9 only: 0, 5.0,
8.5, 17.0 and 51.1 \M:
>98%
Positive
> 8.5 (iM
Lioi etal. (1998a)
No significant \ in MI
observed.
Aberration
72 h exposure
In vitro
Chromosomal
Human lymphocytes
-S9: 0, 200, 1200 and
6000 (iM; 48 h
96.0%
Negative
Manas et al. (2009)
No toxicity observed up
to 6000 ^M
Aberration
exposure
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline, test material purity,
control chemicals and summary data tables.
CA= chromosomal aberrations, MI= mitotic index, PI= proliferation index.
Page 111 of 216
-------
Table 5.4. In vitro Tests for Micronuclei Induction in Mammalian Cells- Glyphosate Technical
Test/
Test System
Concentrations/
Test Material
Results
Reference
Comments
Endpoint
Conditions
Purity
In vitro
TR146 cells
10, 15 and 20 mg/L;
95%
Positive
Koller el al.
Apoptosis and necrosis
Cytokinesis
Block
Micronucleus
(human-derived
buccal carcinoma
cell line)
20-minute exposure.
Statistically significant
(p<0.05) increase inMN
(2012)
reported at 20 mg/L
Also reported t in NB
Assay
(witli FISH
analysis)
at 15 and 20 mg/L.
andNPB
In vitro
CHO-K1 cells
5 - 100 ng/mL, ±S9
Not stated
Negative -S9
Roustan et al.,
No clear dose response
Cytokinesis
Block
Micronucleus
Positive +S9 at 10-100
Hg/mL
(2014)
Test
In vitro
Cytokinesis
Block
Micronucleus
Bovine lymphocytes
(2 donors)
0, 28, 56, 140, 280
and 560 |iM
24 & 48 h exposure
62%
24 h: Negative
48 h: Equivocal
Piesova, 2004
No dose-response No
significant decrease in
CBPI observed.
Test
t MN at 280 pM only
(donor A) f MN at 560
liM only (donor B)
In vitro
Cytokinesis
Block
Bovine lymphocytes
(2 donors)
0, 28, 56, 140, 280
and 560 |iM; 2 h
(±S9) and 48 h (-S9)
62%
2 h: Negative
48 h: Equivocal
Piesova, 2005
No dose-response; No
significant decrease in
CBPI observed.
Micronucleus
exposure
Metabolic activation
Test
t MN at 280 (iM only
(donor A) and at 560
(iM only (donor B)
had no effect on MN
formation after 2 h
exposure.
Page 112 of 216
-------
Table 5.4. In vitro Tests for Micronuclei Induction in Mammalian Cells- Glyphosate Technical
Test/
Endpoint
Test System
Concentrations/
Conditions
Test Material
Purity
Results
Reference
Comments
In vitro
Cytokinesis
Block
Micronucleus
Assay
(witli FISH
analysis)
Human lymphocytes
(treated with
cytochalasin B)
4h treatment ±S9;
0.5,2.91,3.50, 92.8
and 580 |ig/mL:
harvested 72 h
98.0%
Negative -S9
Positive +S9, f MN at
580 ng/mL, but not at
0.5-92.8 ng/mL
Also observed t in NB
at 580 ng/mL (±S9); |
NPB at 580 ng/mL
(+S9)
Mladinic et al.
(2009a)
Cells were exposed to
glyphosate and washed
prior to treatment with
PHA.
Authors did not report
being blind to
treatment.
In vitro
Cytokinesis
Block
Micronucleus
Assay
(witli FISH
analysis)
Human lymphocytes
(treated with
cytochalasin B)
4h treatment ±S9;
0.5,2.91, 3.50, 92.8
and 580 ng/mL
98%
Negative -S9
Positive +S9
t MN at 580 |ig/mL.
but not at 0.5 -92.8
Hg/mL
t apoptosis and necrosis
at 580 ng/mL (-S9);
t apoptosis at > 2.91
Hg/mL and necrosis at
580 ng/mL (+S9)
t in NB at 580 ng/mL
(±S9) and NPB at 580
Hg/mL (+S9)
Mladinic et al.
(2009b)
Cells were exposed to
glyphosate and washed
prior to treatment with
PHA.
Authors did not report
being blind to
treatment.
CBPI= cytokinesis block proliferation index, FISH= fluorescent in situ hybridization; MN= micronuclei; NB= nuclear buds; NPB= nucleoplasms bridges.
Page 113 of 216
-------
5.5 In Vivo Genetic Toxicology Tests
5.5.1 In Vivo Assays for Chromosomal Abnormalities
5.5.1.1 Mammalian Bone Marrow Chromosomal Aberration Assays
The in vivo mammalian bone marrow chromosomal assay detects the ability of a chemical to
cause structural chromosomal damage in cells in the bone marrow. The assay is typically
conducted in rodents (mouse or rat) and detects both chromosome-type and chromatid-type
aberrations. Chromatid-type aberrations are expressed when a single chromatid break occurs
and/or a reunion between chromatids, and chromosome-type aberrations result from damage
expressed in both sister chromatids (OPPTS 870.5385). In this test, animals are exposed
(typically via oral route or intraperitoneal injection) and sacrificed at sequential intervals. Prior
to sacrifice, animals are treated with a spindle inhibitor such as colchicine or Colcemid® to arrest
cells at metaphase. Chromosome preparations from the bone marrow are stained and scored for
chromosomal aberrations. (OPPTS 870.5385). Generally, the optimal time to detect
chromosomal aberrations in the bone marrow is 24 hours after treatment.
Three in vivo mammalian bone marrow chromosomal assays were conducted with glyphosate
technical for regulatory purposes and all were negative (Table 5.8). In the first study, Sprague
Dawley rats were administered glyphosate (98%) at 0 or 1000 mg/kg and the bone marrow was
sampled at 6, 12 or 24 hours after dosing. No significant increase in bone marrow chromosomal
aberrations were observed (Li, 1983b). In the second study, Swiss albino mice were treated
twice by oral gavage (24 hours apart) with 0 or 5000 mg/kg glyphosate technical (96.8%)
resulting in no significant increase in bone marrow chromosomal aberrations (Suresh, 1994). In a
third study conducted with glyphosate trimesium salt, no increase in chromosomal aberrations
were seen in the bone marrow of rats treated by oral gavage with up to 188 mg/kg (Majeska,
1982c).
5.5.1.2 Rodent Dominant Lethal Test
Dominant lethal mutations cause embryonic or fetal death. The induction of a dominant lethal
mutation after exposure to a chemical indicates that the test chemical has affected the germinal
tissue (sperm at some point in development, from stem cell to spermatocyte). Dominant lethal
effects are considered to result from chromosomal damage (structural or numerical), but may
also reflect gene mutations or systemic toxicity (OPPTS 870.5450, OECD 2016). In this test,
male rodents are treated with the test material and mated with (untreated) virgin females. The
female animals are sacrificed at an appropriate time and the uteri are examined to determine the
number of implants, and live and dead embryos. Two dominant lethal studies were identified.
One study was conducted in the rat (Suresh, 1992) where male rats were dosed by oral gavage
with glyphosate up to 5000 mg/kg. The other study (Rodney, 1980) was conducted in male mice
treated with up to 2000 mg/kg glyphosate (98.7%) by oral gavage. No significant increase in
dominant lethal mutations were observed in either study (Table 5.5).
Page 114 of 216
-------
5.5.1.3 In Vivo Mammalian Erythrocyte Micronucleus Assays
The mammalian micronucleus test is the most commonly conducted in vivo test to detect
clastogenic or aneugenic chemicals. The test identifies chemicals that induce micronuclei in
proerythrocytes (progenitor cells) by assessing micronucleus frequency in immature erythrocytes
(polychromatic erythrocytes, PCEs) sampled from the bone marrow or from the peripheral blood
(reticulocytes). This test is typically conducted in mice or rats. When bone marrow
erythroblasts develop into erythrocytes, the main nucleus is extruded following the final cell
division (erythrocytes are the only mammalian cell that does not contain a nucleus). Any
micronuclei formed after the final cell division may remain in the cytoplasm following extrusion
of the main nucleus. The visualization of micronuclei is facilitated by the lack of a nucleus in
these cells (OPPTS 870.5395, OECD 474). Micronuclei can originate from acentric
chromosomes, lagging chromosome fragments, or whole chromosomes; thus, micronuclei are
biomarkers of both altered chromosome structure or chromosome number. The assay is based on
an increase in the frequency of micronucleated erythrocytes in treated animals, in either
peripheral blood samples or bone marrow samples (OPPTS 870.5395). Additional mechanistic
information on the formation of chromosomal abnormalities can be obtained from the
incorporation of centromeric and telomeric fluorescent probes (FISH) assay. According to EPA
test guidelines, a single dose of the test substance may be used in this test if the dose is the
maximum tolerated dose (MTD), a dose that produces some indication of bone marrow
cytotoxicity (e.g., a reduction in the proportion of immature erythrocytes (PCEs) to total
erythrocytes by >50%) or a maximum limit dose of 5000 mg/kg. The routes of administration
for this test are typically oral or intraperitoneal injection and generally involve a single
administration.
Glyphosate technical has been extensively evaluated for micronuclei induction in in vivo studies.
Fourteen studies were conducted for regulatory purposes, four were identified from the open
literature, and one study was conducted by the U.S. National Toxicology Program (NTP). This
included nine studies with administration of glyphosate by the intraperitoneal (i.p.) route and 10
studies by the oral route. The findings are presented in Table 5.10. Of the nine i.p. studies,
seven (Costa, 2008; Chruscielskae^a/., 2000; Durward, 2006; Gava, 2000; Marques, 1999; Rank
et al., 1993 and Zaccaria, 1996) were negative. These studies tested doses as high as 2016
mg/kg (single and double administration) with sampling times at 24- and 48-hours post-dose.
Two positive findings were reported when glyphosate technical was administered by i.p.
Bolognesi et al. (1997) reported a significant increase in micronuclei in the bone marrow of male
Swiss CD mice 24 hours after i.p. treatment with 300 mg/kg glyphosate technical (99.9%). The
dose in this study was administered as V2 dose (150 mg/kg) injections 24 hours apart to 3 male
mice. Manas et al. (2009) evaluated glyphosate technical (96%) in BALB/c male and female
mice (5/sex/dose) administered 50, 100 or 200 mg/kg by two i.p. injections, 24 hours apart. The
results showed a significant increase in micronucleated erythrocytes at 200 mg/kg, but not at 50
or 100 mg/kg. It should be noted that doses that resulted in the positive responses in these two
studies were above the reported i.p. LD50 value (130 mg/kg) for glyphosate in mice (NTP 1992).
Glyphosate technical was also evaluated in nine micronucleus assays with administration by the
oral route in mice and one in the rat. Eight of the nine oral studies in the mouse were negative
for micronuclei induction. The single positive response was seen in female mice treated with
Page 115 of 216
-------
two 5000 mg/kg doses (limit dose), 24 hours apart with bone marrow sampling at 24-hours post-
dose (Suresh, 1993b). No increase was observed at lower doses (50 and 500 mg/kg) in females
or at any dose in males. The eight negative oral studies in mice included single dose
administrations of 5000 mg/kg and bone marrow analysis at 24, 48, and/or 72 hours (Jensen,
1991c; Fox and Mackay, 1996) and one or two administrations of glyphosate technical with top
doses between 30 and 2000 mg/kg (Honarvar, 2005; Honarvar, 2008; Jones, 1999; and Zoriki-
Hosmi, 2007). It should be noted that evaluations at 48 and 72 hours post dose may be too late to
detect chemically-induced micronucleated PCEs in the bone marrow as these cells may have
already migrated into the peripheral blood. No significant increase in micronucleated
erythrocytes were seen in male or female mice following 13-weeks of dietary (feed)
administration of glyphosate technical at doses up to 3393 mg/kg/day (NTP, 1992). In the single
study that evaluated micronuclei induction in rats, glyphosate technical did not induce significant
induce micronuclei in CD1 rats treated by oral gavage at doses up to 2000 mg/kg (Fliigge,
2009b). When glyphosate trimesium salt was evaluated, no increase in micronuclei induction
was seen in mice treated orally up to 1100 and 800 mg/kg in males and females, respectively
(Majeska, 1987).
Page 116 of 216
-------
Table 5.5. In Vivo Tests for Chromosomal Aberrations in Mammals- Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Chromosomal
Aberration Test
Sprague Dawley rats
(males and females)
Intraperitoneal
injection;
sampled at 6, 12
and 24 h after
treatment
0, 1000 mg/kg
(6/sex/dose/samp
ling time)
98%
Negative
Li (1983b)
[MRID 00132683]
No toxicity observed.
A separate study
using 14C-glyphosate
showed that
glyphosate reaches
BM 0.5 h after dosing
with 'A life
elimination at 7.6 h.
Peak BM value was
400 ppm,
corresponding to 2000
ppm plasma value.
Bone Marrow
Chromosomal
Aberration Test
Sprague Dawley rats
(males and females)
Vehicle: distilled
water
Oral gavage,
sampling after 6,
12, 24, 48 h and
5 d
0, 21, 63 and
188 mg/kg
58.5%
Glyphosate
trimesium
salt
Negative
Majeska (1982c)
[MRID 00132176]
Bone Marrow
Chromosomal
Aberration Test
Swiss Albino mice
(males and females)
Vehicle: peanut oil
Oral gavage
(2 treatments, 24
h apart);
sampling after 24
h (last treatment)
0, 5000 mg/kg
(5/sex/dose)
96.8%
Negative
Suresh (1994)
[MRID 49987408]
Significant (p<0.05)
decrease in bw of
females at high dose.
Rodent
CD-I mice
Oral gavage
0, 200, 800,
98.7%
Negative
Rodwell (1980)
Dominant
Lethal Test
Each dosed male
mated with 2
and 2000
mg/kg
[MRID 00046364]
females/week for 8
weeks
Rodent
Wistar rat
Oral gavage
0, 200, 100 and
96.8%
Negative
Suresh (1992)
Dominant
Lethal Test
Each dosed male
mated with 1
female/week for 10
weeks
5000 mg/kg
[MRID 49987404]
Page 117 of 216
-------
Table 5.6. In Vivo Tests for Micronuclei Induction in Mammals-
Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
Swiss CD1 mice
(males only)
Intraperitoneal
injection; 2
injections of half
the dosage of 300
mg/kg 24 h apart;
sampling at 6 and
24 h
0, 300 mg/kg
(3/dose)
99.9%
Positive
Stat
significant
increase in
MN at 24 h
Bolognesi et al.
(1997)
Material and methods
indicate 3
animals/dose;
however. Table 1 of
article indicates 4
animals were
evaluated.
Bone Marrow
Micronucleus
Test
Balb C mice
(males and females)
Vehicle: Saline
Intraperitoneal
Injection (two
injections, 24 h
apart); sampling
after 24 h (last
treatment)
0, 50, 100, and
200 mg/kg
(5/sex/dose)
96%
Positive
|MN at 200
mg/kg, but
not at 50 or
100 mg/kg
Manas et al.
(2009)
No significant signs
of toxicity observed.
Bone Marrow
Micronucleus
Test
C3H mice
(males only)
Vehicle: water
Intraperitoneal
Injection
(single treatment);
sampling after 24,
48 and 72 h
0, 300 mg/kg
Not
reported
Negative
Chruscielska et
al. (2000)
Bone Marrow
Micronucleus
Test
Swiss Albino mice
(males and females)
Vehicle: corn oil
Intraperitoneal
Injection
(2 treatments, 24
h apart); sampling
after 24 h (last
treatment)
0, 15.62, 31.25,
and 62.5 mg/kg
(5/sex/dose)
980 g/kg
Glyphosate
technical
Negative#
Costa (2008)1
OECD guideline 474
#Was not tested up to
limit dose and did not
demonstrate that
compound was tested
up to toxic dose. No
mention of BM
toxicity or clinical
signs.
Page 118 of 216
-------
Table 5.6. In Vivo Tests for Micronuclei Induction in Mammals-
Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
Crl:CD-lTM(ICR)
BR mice
(males only1)
Vehicle: PBS
Intraperitoneal
Injection
(single treatment);
sampling after 24
and 48 h (high
dose only)
0, 150, 300 and
600 mg/kg
(7/dose)
95.7%
Negative
Durward (2006)
[MRID 49957411]
Clinical signs
reported at > 150
mg/kg. Significant j
in %PCEs reported at
24 h in 600 mg/kg
group, fin MN PCEs
observed at 600
mg/kg (1.9± 0.7 vs.
1.0 ± 1.2 control;
p<0.05), at 24 h, but
not 48 h, within
historical control
range.
Bone Marrow
Micronucleus
Test
Swiss Albino mice
(males and females)
Vehicle: water
Intraperitoneal
Injection
(2 treatments, 24
h apart); sampling
after 24 h (last
treatment)
0, 1008, 2016,
and 3024 mg/kg
5/sex/dose
612.7 g/kg
(glyphosate
technical
Nufann)
Negative
Gava (2000)1
LD50 was 4032
mg/kg
Mortality observed in
1 animal at high dose
(only 4 m/f scored for
MPCEs).
No effect on
PCE/NCE.
Bone Marrow
Micronucleus
Test
Swiss Albino mice
(males and females)
Vehicle: water
Intraperitoneal
Injection
(2 treatments, 24
h apart); sampling
after 24 h (last
treatment)
0, 187.5,375
and 562.5 mg/kg
5/sex/dose
954.9 g/kg
(glyphosate
technical
Nufann)
Negative
Marques (1999)
[MRID 49957412]
LD50 was 750 mg/kg
No significant signs
of toxicity observed
in main study
Bone Marrow
Micronucleus
Test
NMRI-Bom mice
Intraperitoneal
Injection (single
treatment);
sampling after 24
h (all doses) and
48 h (150 and 200
mg/kg)
0, 150, and 200
mg/kg
(5/sex/dose)
glyphosate
isopropyla
mine (purity
not
specified)
Negative
Rankefo/. (1993)
Page 119 of 216
-------
Table 5.6. In Vivo Tests for Micronuclei Induction in Mammals-
Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
Swiss albino mice
(males and females)
Intraperitoneal
Injection
(2 treatments, 24
h apart); sampling
after 24 h (last
treatment)
0, 68, 137, and
206 mg/kg (
360 g/L
Negative
Zaccaria (1996)
[MRID 49961501]
Doses selected were
reported as
corresponding to 25,
50 and 75% LD50
Bone Marrow
Micronucleus
Test
CD-I mice
(males and females)
Vehicle: saline
Oral gavage
(single treatment);
sampling after 24
and 48 h
0, 5000 mg/kg
5/sex/dose
95.6%
Negative
Fox and Mackay
(1996)
[MRID 44320619]
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
NMRI mice
(males and females)
Vehicle: PEG 400
Oral gavage
(single treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000,
and 2000 mg/kg
5 sex/dose
97.73%
Negative
Honarvar (2005)1
OECD guideline 474
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
NMRI mice
(males only)
Vehicle: 0.5%
carboxymethyl-
cellulose
Oral gavage
(single treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000,
and 2000 mg/kg
(5/dose)
99.1%
Negative
Honarvar (2008)
[MRID 49961802]
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
NMRI mice
(males and females)
Vehicle: 0.5%
carboxymethyl-
cellulose
Oral gavage
(single treatment);
sampling after 24,
48 and 72h
0, 5000 mg/kg;
5/sex/dose
98.6%
Negative
Jensen (1991c)
[MRID 49961503]
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
CD-I mice
(males only1)
Vehicle: water
Oral gavage
single treatment);
sampling after 24
and 48 h
0, 2000 mg/kg
5/dose
59.3%
potassium
glyphosate
salt
Negative
Jones (1999)1
OECD guideline 474
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
Swiss albino mice;
(males and females)
Vehicle: peanut oil
Oral gavage
(2 treatments, 24
h apart); sampling
0, 50, 500, 5000
mg/kg
5/sex/dose
96.8%
glyphosate
acid
Positive in
females at
5000
Suresh (1993b)
[MRID 49987407]
No significant signs
of toxicity observed
Page 120 of 216
-------
Table 5.6. In Vivo Tests for Micronuclei Induction in Mammals-
Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
after 24 h (last
treatment)
mg/kg
only.
Negative in
males at all
doses
Bone Marrow
Micronucleus
Test
Swiss mice
(males only)
Vehicle: corn oil
Oral gavage
(2 treatments, 24
h apart); sampling
after 24 h (last
treatment)
0, 8, 15 and 30
mg/kg
(6/dose)
980.1 g/kg
Negative
Zoriki Hosomi
(2007)
[MRID 50000901]
OECD guideline 474
No significant signs
of toxicity observed
Bone Marrow
Micronucleus
Test
CD-I mice
(males and females)
Vehicle: distilled
water
Oral gavage.
Sampling 24, 48
and 72 h after
treatment
Males: 0, 700,
900 and 1100
mg/kg
Females: 0,
400, 600 and
800 mg/kg
55.3%
Glyphosate
trimesium
salt
Negative
Majeska (1987)
[MRID 40214004]
Bone Marrow
Micronucleus
Test
B6CF3 Mice
(males and females)
Oral (dietary).
MN assay
conducted
following 13-
week feed study.
0, 205/213,
410/421,
811/844,
1678/1690 and
3393/3393
mg/kg (m/f)
(10/sex/dose)
99%
Negative
NTP (1992)
Bone Marrow
Micronucleus
Test
CD Rats
(males and females)
Vehicle: 0.8%
hydro xypropylmethyl
cellulose
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 mg/kg
(5/sex/dose)
98.8%
Negative
Fliigge (2009b)1
OECD guideline 474
No significant signs
of toxicity observed
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline followed, test material
purity, control chemicals and summary data tables.
2Only males tested; report indicated that there was no difference between sexes seen in range finding study.
CA= chromosomal aberrations, MPCE= micronucleated polychromatic erythrocytes, NCE= normochromatic erythrocytes, PCE=polychromatic erythrocytes.
Page 121 of 216
-------
5.6 Additional Genotoxicity Assays Evaluating Primary DNA Damage
There are a number of genotoxicity assays that evaluate primary DNA damage, but do not
measure the consequence of the genetic damage {i.e., mutation or chromosomal damage). As
discussed in the Guidance Document on Revisions to OECD Genetic Toxicology Test
Guidelines (OECD 2015), the endpoints measured in primary DNA damage tests such as DNA
adducts, comet assay, or unscheduled DNA synthesis may lead to cell death or may initiate DNA
repair, rather than a mutation. These types of assays can, however, provide mechanistic
information when interpreting positive findings in other genotoxicity tests or when determining
whether a chemical is acting through a mutagenic mode of action. Additionally, indirect
mechanisms of DNA damage such as oxidative DNA damage can be detected by these test
systems. Oxidative damage results from oxidative stress, which occurs when there is a
disturbance in the balance between the production of reactive oxygen species (ROS) and
antioxidant defense systems. Normal cellular metabolism is a source of endogenous reactive
oxygen species that accounts for background levels of oxidative damage in normal cells. Some
types of oxidative damage are repairable while others lead to serious consequences in the cell.
(Cooke et al, 2003). The various assays evaluating primary DNA damage in glyphosate
technical are presented in Table 5.7. Details of the findings are discussed below.
Glyphosate technical is not electrophilic and is not considered to be DNA-reactive. In a study to
evaluate the potential for glyphosate to directly interact with DNA, Peluso et al. (1998) reported
that glyphosate technical did not form DNA adducts in mice when tested up to 270 mg/kg via i.p.
Bolognesi et al. (1997) reported an increase in the oxidative damage biomarker 8-
hydroxydeoxyguanosine (8-OHdG) in the liver 24 h after i.p. injection of 300 mg/kg in mice.
No increase in 8-OHdG was seen in the kidney with glyphosate technical. The dose in this study
was high (300 mg/kg) for an i.p. injection and within the i.p. LDso range (134- 545 mg/kg) that
has been reported elsewhere (WHO, 1994).
The comet assay, also known as single cell gel electrophoresis (SCGE), is a sensitive and rapid
method to detect DNA strand breaks in individual cells. In this assay, individual cells are
embedded in agarose. The cells are then lysed (which digests the cellular and nuclear
membranes) and the DNA is allowed to unwind under alkaline or neutral conditions. During
electrophoresis, chromatin (which is in a supercoiled state) that has undergone steric relaxation
due to DNA damage migrates away from the nucleoid (nucleus) toward the anode, yielding
images that resemble a comet. The intensity of the comet tail relative to the comet head reflects
the amount of DNA breakage (Tice et al., 2000; Collins et al., 2008). The comet assay can
detect single and double strand breaks resulting from direct interactions with DNA, alkali labile
sites, or transient DNA breaks resulting during DNA excision repair. These types of strand
breaks may be, (a) repaired with no persistent effect, (b) be lethal to the cell or (c) be fixed as a
mutation (OECD TG 489). DNA strand breaks in the comet assay can be measured by endpoints
such as percent tail DNA (also referred to as % tail intensity), tail length, and tail moment.
However, % tail DNA is the recommended metric for evaluating and interpreting results using
this assay (OECD TG 489).
Page 122 of 216
-------
The five studies that evaluated glyphosate technical using the comet assay are summarized in
Table 5.12. Two of the studies were conducted using tumor cell lines. Koller et al. (2012)
reported positive comet effects (increased tail intensity) in a human buccal carcinoma cell line
(TR146) following a 20-minute treatment with > 20 mg/L (-0.118 mM) glyphosate. Although no
evidence of cytotoxicity was reported in this study, the authors did report an increase in
apoptosis and necrosis at the same concentrations (> 20 mg/L) when the same cell line was tested
for in vitro micronuclei induction (discussed previously). In a study using Hep-2 cells
(presumably a HeLa cell derivative), Manas et al. (2009) reported a statistically significant
increase in mean tail length, and tail intensity at all concentrations (3.0-7.5 mM) tested. In a
comet study conducted on human lymphocytes, Alvarez-Moya et al. (2014) reported significant
increases in tail length only (but not % tail DNA) following treatment with glyphosate
concentrations of 0.7-700 [xM. Mladinic et al. (2009a) evaluated DNA damage in non-dividing
human lymphocytes (±S9) following treatment from 0.5 to 580 |ag/mL using the standard
alkaline comet method and a modified comet method that detects DNA damage due to oxidative
damage (human 8-hydroxyguanidine DNA-glycosylase, hOGGl comet method). In this study,
the authors reported statistically significant increases in tail intensity at 3.5 |ig/mL and higher in
the absence of S9, with significance only at 580 |ag/mL (-3.4 mM) in the presence of S9 using
the alkaline method. This concentration also resulted in increased apoptosis and necrosis as well
as an increase in plasma total antioxidant capacity (TAC) and changes in plasma lipid
peroxidation (thiobarbituric reactive substances, TBARs); however, only a dose-related increase
in tail length (not % tail DNA) was observed at 580 |ig/mL (+S9) using the hOGGl method.
When the Manas et al. (2013) evaluated blood and liver cells following a 14-day drinking water
study in mice treated with 40 and 400 mg/kg/day glyphosate, significant increases in tail
intensity, tail length and tail moment were reported were observed at both doses in both tissues
(except for DNA tail intensity in liver at 40 mg/kg); however, there were no substantial effects
on oxidative stress measurements suggesting that DNA damage reported may not be due to
oxidative damage.
The Unscheduled DNA Synthesis (UDS) test with mammalian liver cells in vitro identifies
substances that induce DNA repair after excision and removal of a segment of damaged DNA.
The test is typically conducted in liver cells, which have relatively few cells in the S-phase of the
cell cycle. The assay measures the incorporation of radiolabeled nucleotide [3H]-thymidine into
DNA during the repair process in non-S phase cells. (OPPTS 870.5555). Substances that produce
either a statistically significant dose-related increase or statistically significant and reproducible
increase in 3H-TdR incorporation in at least one test point are considered to be positive in this
test. A UDS study that evaluated glyphosate technical in rat primary hepatocytes was negative
(Williams, 1978). Glyphosate technical was also negative in a DNA repair test conducted in
bacteria (Rec-Atest) (Shirasu, 1978).
In an alkaline elution assay, which detects single strand DNA breaks, Bolognesi et al. (1997)
reported an increase in single strand breaks (i.e. increased DNA elution rate) in the liver and
kidney 4 hours after a single i.p. injection of 300 mg/kg. The elution rate returned to control
levels at 24 hours. Glyphosate technical was also negative in a DNA repair test conducted in
Bacillus subtilis H17 (rec+) and M45 (rec") bacterial Rec-A test (Shirasu, 1978).
Page 123 of 216
-------
Finally, the sister chromatid exchange (SCE) test is an assay that can measure the consequence
of primary DNA damage. The mechanism(s) of action for chemical induction of SCE is unclear.
The SCE assay detects the exchange of DNA between two sister chromatid arms within a single
chromosome. The assay can be performed in vitro or in vivo. Following exposure, cells/animals
are treated with bromodeoxyuridine (BrdU) to allow for the differentiation of the two sister
chromatids (harlequin staining) and prior to harvest are treated with a spindle inhibitor to
accumulate cells in metaphase. The chromosome preparations are then stained and analyzed for
SCEs (OPPTS 870.5900, 870.5915). The SCE studies that evaluated glyphosate technical are
also presented in Table 12. Positive SCE findings were reported in all four studies; two
evaluating bovine lymphocytes (Lioi, 1988b, Sivikova and Dianovksy, 2006) and two studies
evaluating human lymphocytes (Lioi, 1988a; Bolognesi et al., 1997). In all four studies the
induction did not demonstrate a clear dose response.
Additionally, although it is recognized that mechanisms other than genotoxicity may be involved
in cell transformation, glyphosate trimesium salt was evaluated in the Balb/3T cell
transformation assay (an in vitro tumor formation assay) and was negative up to 5.0 mg/ml
(Majeska, 1982b).
Page 124 of 216
-------
Table 5.7 Assays for Detecting Primary DNA Damage- Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses/
Concentrations
Test Material
Purity
Results
Reference
Comments
DNA Adducts
32P-postlabeling
Swiss CD 1 mice
(males and females)
Liver and kidney
evaluated
Intraperitoneal
injection; 24 h
exposure
0, 130 and 270
mg/kg
Not reported
Negative
Peluso et al.
(1998)
DNA oxidative
damage:
8-OHdG
formation
Swiss CD-I mice
(males)
liver and kidney
evaluated
Intraperitoneal
injection (single
dose); sampling
4 and 24 h after
injection
0, 300 mg/kg
(3/dose)
99.9%
Kidney:
negative
Liver:
positive
(24 h)
Bolognesi et
al. (1997)
Single-cell gel
TR146 cells
NA (in vitro)
-S9: 10-2000
95%
Positive
Koller et al.
Also measured multiple
electrophoresis
(SCGE) assays-
Comet assay
(human-derived
buccal epithelial cell
line).
mg/L;
20-minute
exposure.
Increased
DNA
migration
at >20
mg/L
(2012)
cellular integrity
parameters to assess
cytotoxicity. No clear
evidence of cytotoxicity
seen except for increase
in enzyme activity
(indicative of membrane
damage) in LDHe
(extracellular lactate
dehydrogenase) assay at
>80 mg/L.
No mention of
monitoring pH
Single-cell gel
Hep-2 cells
NA (in vitro)
0,3,4.5,6, 7.5,
96%
Positive
Manas et al.
The authors did not report
electrophoresis
(SCGE) assays-
Comet assay
9, 12 and 15 mM
Stat.
significant
increase in
mean tail
length and
tail
intensity at
all cones.
(2009)
a source for the Hep-2
cells. The agency
presumes that this is a
HeLa derived cervical
carcinoma cell line.
Page 125 of 216
-------
Table 5.7 Assays for Detecting Primary DNA Damage- Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses/
Concentrations
Test Material
Purity
Results
Reference
Comments
Single-cell gel
electrophoresis
(SCGE) assays-
Comet assay
Human
lymphocytes
NA (in vitro)
0, 0.7, 7, 70, 700
HM
96%
Positive at
all doses
(increase in
tail length
only)
Alvarez-Moya
et al., (2014)
Issues were identified
with this study resulting
in a low quality ranking.
These include: 1) blood
was washed with PBS
and then held at 4° C for
an indeterminate amount
of time before exposure
to glyphosate. (2) Cells
were treated for 20 hours
at room temperature.
(3) The same amount of
damage was reported
across 2 orders of
magnitude concentration.
Single-cell gel
electrophoresis
(SCGE) assays-
Comet assay
Human
lymphocytes; ±S9
Alkaline and hOOGl
Comet assays
performed
NA (in vitro)
0,0.5,2.91,3.5,
92.8 and 580
Hg/mL
98%
Positive
±S9
Mladinic et al.
(2009a)
The alkaline comet assav
-S9: | in mean tail length
at 580 iig/mL and t in tail
intensity at > 3.5 |ig/mL).
+S9: t DNA tail length
at >3.5 |ig/mL. Tail
intensity t only at 580
Hg/mL
hOOGl comet assav:
-S9 no effect on tail
length, ftail intensity only
at 3.50 |ig/mL
+S9: t tail length at 580
Hg/mL, no effect on tail
intensity compared to
controls at any conc.
Page 126 of 216
-------
Table 5.7 Assays for Detecting Primary DNA Damage- Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses/
Concentrations
Test Material
Purity
Results
Reference
Comments
Single-cell gel
Balb/C mice;
Drinking water
0, 40, and 400
96%
Positive
Manas et al.
Only minor effects seen
electrophoresis
(SCGE) assays-
Comet assay
with oxidative
evaluated blood and
liver
(14 days)
mg/kg
Blood and
liver at
both doses
(2013)
on oxidative stress
measurements (TBARs,
SOD, CAT)
stress measures
Sister Chromatid
Bovine lymphocytes
NA (in vitro)
-S9: 0, 17, 85
>98%
Positive
Lioi (1998b)
1.8-, 2.1-, 1.6-fold
Exchange (SCE)
(3 donors)
and 170 ^M; 72
h exposure
Significant
(p>0.05)
increase in
SC/cell at
all
concentrati
ons
increases, respectively
Sister Chromatid
Human lymphocytes
NA (in vitro)
-S9: 0, 5, 8.5, 17
>98%
Positive
Lioi (1998a)
1.9-.2.8-. and 2.6-fold
Exchange (SCE)
and 51 ^M; 72 h
exposure
Significant
(p>0.05)
increase in
SCE/cell at
> 8.5 nM
increase at 8.5, 17 and 51
liM. respectively
Sister Chromatid
Human lymphocytes
NA (in vitro)
-S9: 0,0.33, 1,3
99.9%
Positive
Bolognesi et
Very limited information
Exchange (SCE)
and 6 mg/mL;
72 h exposure
al. (1997)
was provided on the
methods used in this
paper. Authors report a
dose -dependent increase
in SCE frequency;
however, no statistical
analysis for dose response
was reported. Data
presented graphically
with no error bars.
Sister Chromatid
Human lymphocytes
NA (in vitro)
28, 56, 140, 280,
62%
Positive
Sivikova and
The increases in SCEs
Exchange (SCE)
560 and 1120
liM: 24 h
exposure ±S9
Dianovsky
(2006)
observed did not show a
clear concentration
related increase across a
40-fold increase in the
concentrations tested
Page 127 of 216
-------
Table 5.7 Assays for Detecting Primary DNA Damage- Glyphosate Technical.
Test/Endpoint
Test System
Route of
Administration
Doses/
Concentrations
Test Material
Purity
Results
Reference
Comments
Alkaline elution
assay- DNA
single strand
breaks
Swiss CD-I mice
(males)
liver and kidney
evaluated
Intraperitoneal
injection (single
dose); sampling
8 and 24 h after
injection
0, 300 mg/kg
(3/dose)
99.9%
Positive
(Increased
elution
rate) at 4
hours in
liver and
kidney
At 24 h
elution rate
returned to
control
levels
Bolognesi et
al. (1997)
Return to control values
may indicate DNA repair
or reflect rapid
elimination of compound
DNA Repair
Test
(Rec-A test)
B. subtilisHll (rec+)
and M45 (rec-)
NA (in vitro)
20-2000 (ig/disk
98.4%
Negative
Shirasu (1978)
[MRID
00078619]
Unscheduled
DNA synthesis
(DNA repair)
F-344 rat primary
hepatocytes
NA (in vitro)
0,0.0125,
0.0625,0.125,
0.6.5, 1.25, 12.5,
125 iig/mL
98%
Negative
Li and Long
(1988)
Cell
Transformation
Assay
BALB/3T cells
NA (in vitro)
0.313-5.0
mg/mL
90%
Glyphosate
trimesium salt
Negative
Majeska
(1982b)
[MRID
00126616]
h- hour; CAT= catalase, G6PD= glucose 6-phosphate dehydrogenase, NA= not applicable, hOOGl = TBARs= tliiobarbituric acid reactive substances, SOD=
superoxide dismutase
Page 128 of 216
-------
5.7
Summary and Discussion
The genotoxic potential of glyphosate has been extensively investigated using a variety of test
systems and genetic endpoints. This assessment focuses only on test systems that the agency
considered relevant for assessing genotoxic risks in humans. The totality of the genetic
toxicology information was evaluated using a weight of evidence approach to determine the
genotoxic potential of glyphosate. This involves the integration of in vitro and in vivo results as
well as an overall evaluation of the quality, consistency, reproducibility, magnitude of response,
dose-response relationship and relevance of the findings. In the weight of evidence analysis,
studies evaluating endpoints that measured gene mutations and chromosomal aberrations (i.e.
permanent DNA damage) were given more weight than endpoints reflecting DNA events that
may be transient or reversible such as primary DNA damage (e.g., comet assays). In vivo studies
in mammals were given the greatest weight and more weight was given to doses and routes of
administration that were considered relevant for evaluating genotoxic risk based on human
exposure to glyphosate. Also, since the molecular mechanisms underlying the observation of
SCEs are unclear and thus, the consequences of increased frequencies of SCEs are unclear, the
data from this test were given low weight in the overall analysis. A summary of the various lines
of evidence of considered in the weight of evidence evaluation for the genotoxic potential of the
active ingredient glyphosate is presented below.
Evidence of primary DNA damage
Glyphosate technical is not considered to be electrophilic and did not induce DNA adducts in the
liver or kidney at an i.p. dose of 270 mg/kg. However, evidence of DNA strand breaks was
reported in a number mammalian cell studies using the comet assay. Additionally, transient
increases in alkali labile sites in the liver and kidney of mice and an induction of 8-OHdG in
DNA were seen in the livers of mice following i.p. injections with 300 mg/kg glyphosate. These
effects were seen at high doses for the i.p. route in mice (LDso for mouse =130 mg/kg; NTP,
1992). However, due to technical limitations identified in a number of these studies (e.g. use of
cancer cell lines that have not been well-characterized, atypical exposure protocols and no
indication of blind to treatment), caution should be exercised in interpreting the results.
In vitro mutations
Glyphosate technical was negative in all 39 studies for mutagenicity in bacteria. In the four
studies that tested for gene mutations in mammalian cells in vitro, no increase in mutations were
observed.
In vitro chromosomal alterations
Mixed results were observed in studies evaluating in vitro chromosomal alterations with
glyphosate treatment. Three SCE studies reported positive findings (Lioi, 1998a, b; Bolognesi et
al., 1997) bovine and human lymphocytes. As stated previously, low weight is given to SCE
results in the overall analysis given the uncertainty regarding the consequence of increases in the
frequencies of SCEs. The SCE responses were weak and not concentration dependent. Eight of
the 10 studies measuring in vitro chromosomal aberrations were negative. The two positive
Page 129 of 216
-------
findings were reported by Lioi et al., one study was conducted with bovine lymphocytes and the
other with human lymphocytes. The authors reported positive findings in these studies at
concentrations much lower than four other studies that reported negative results using the same
cell types. Additionally, in both studies, Lioi et al. used an atypical exposure protocol of 72
hours which is very long for analyzing one round of mitosis. Furthermore, in both studies,
nearly the same level effect for aberration frequency and percent of cells with aberrations were
observed for the same concentrations of glyphosate and the two other chemicals tested in those
experiments.
Four of the six studies evaluating micronuclei induction in vitro were positive and two showed
equivocal results. Three of the positive responses required S9 activation, two conducted with
human lymphocytes and one conducted with CHO cells. The remaining positive micronucleus
study was conducted using a TR146 cells which is a tumor cell line derived from human buccal
mucosa. The authors state that this cell line had not been previously used for genotoxicity
testing. It is difficult to interpret any genotoxicity findings conducted in a tumor cell line that
has not been well-characterized regarding its DNA damage response and repair capacity, and its
degree of chromosomal instability.
Glyphosate was negative in all three L5178Y mouse lymphoma cell studies which may detect
chromosomal damage in addition to mutations.
Mammalian in vivo chromosomal alterations
All three in vivo mammalian studies evaluating chromosomal aberrations with glyphosate
technical were negative. Two studies were conducted in rats (i.p. and oral) and one was
conducted in mice (oral). In addition, glyphosate was also negative in a rodent dominant lethal
test. Glyphosate was negative in 15 of the 19 bone marrow micronucleus studies evaluated. In
two of the positive studies, glyphosate technical was administered by i.p. injection. In these
studies, the authors reported positive findings at doses of 200-300 mg/kg. Based on the available
ADME data for glyphosate, assuming 30% oral absorption, an oral dose of-700-1000 mg/kg
would be needed to achieve a dose of 200-300 mg/kg in the blood. Seven other i.p. studies in
mice reported no increase in micronuclei induction at doses up to 3000 mg/kg. The remaining
positive finding was reported in an oral gavage study in mice where an approximately 2-fold
increase in micronuclei were reported in females only at a dose of 5000 mg/kg, which is
considerably higher than the current guideline recommended limit dose of 2000 mg/kg. The
effect was not seen in the 7 other oral gavage studies in mice when glyphosate was tested at
similar doses. In addition, glyphosate was negative for micronuclei induction following a 13-
week dietary study with a dose up to approximately 3000 mg/kg/day. A negative finding was
also reported in the only study that evaluated in vivo micronuclei induction in the rat using doses
up to 2000 mg/kg.
In a meta-analytic review of micronuclei frequency across mammalian and non-mammalian
species (primarily fish, amphibians, reptiles and plants), Ghisi et al. (2016), not surprisingly,
reported that different responses were observed when comparing mammalian results to
phylogenetically distant non-mammalian species for micronuclei induction. Their analyses
included most, but not all, of the mammalian studies that the agency evaluated and determined to
Page 130 of 216
-------
be negative for micronuclei induction. The authors reported a statistically significant increase in
micronuclei by the i.p. route across the studies in the data set they considered; however, when
glyphosate was administered by the oral route (which is the most physiologically relevant route
for human exposure to glyphosate), no significant difference was observed.
Conclusion for Glyphosate
The overall weight of evidence indicates that there is no convincing evidence that glyphosate
induces mutations in vivo via the oral route. When administered by i.p. injection, the
micronucleus studies were predominantly negative. In the two cases where an increase in
micronuclei were reported via this route, the effects occurred above the reported i.p. LD50 for
mice and were not observed in other i.p. injection studies at similar or higher doses. While there
is limited evidence genotoxic for effects in some in vitro experiments, in vivo effects were given
more weight than in vitro effects particularly when the same genetic endpoint was measured,
which is consistent with current OECD guidance. The only positive findings reported in vivo
were seen at relatively high doses that are not relevant for human health risk assessment.
Page 131 of 216
-------
6.0 Data Integration & Weight-of-Evidence Analysis Across Multiple Lines of Evidence
6.1 Background
In 2010, OPP developed a draft "Framework for Incorporating Human Epidemiologic & Incident
Data in Health Risk Assessment" which provides the foundation for evaluating multiple lines of
scientific evidence (U.S. EPA, 2010). In 2016, a final version of the framework was published.
OPP's framework is consistent with updates to the World Health Organization/International
Programme on Chemical Safety MO A/human relevance framework, which highlights the
importance of problem formulation and the need to integrate information at different levels of
biological organization (Meek et al, 2014).
One of the key components of the agency's framework is the use of modified Bradford Hill
Criteria (Hill, 1965) like those described in the 2005 Guidelines for Carcinogen Risk
Assessment. These criteria are used to evaluate the experimental support considers such
concepts as strength, consistency, dose response, temporal concordance and biological
plausibility in a weight-of-evidence analysis.
6.2 Dose-Response and Temporal Concordance
Given the lack of consistent positive findings particularly at doses < 1000 mg/kg/day across the
lines of evidence, lack of mechanistic understanding, and lack of biological activity in
mammalian systems to the parent compound glyphosate, there are few data to assess key events
in the biological pathway and any associated temporal or dose concordance. Temporal
concordance can be assessed using the experimental animal studies and epidemiological studies
that evaluated exposure prior to outcomes. Similarly, dose concordance can be assessed using
findings of apical outcomes in experimental animal studies, as well as epidemiological studies
that utilize exposure metrics that are stratified by the number of exposure days.
A prospective cohort study is designed to collect exposure information prior to the development
of cancer. As such, exposure is known to occur before the outcome. In De Roos et al. (2005), a
prospective cohort study, no association was observed between glyphosate exposure and all
cancer outcomes evaluated in the AHS cohort. Although the median follow-up time following
recruitment into the cohort was approximately 7 years in De Roos et al. (2005), an updated
analysis of the AHS cohort has been recently published (Andreotti etal., 2017), which included
an extended follow-up period of 17.5 years and also did not report an association between
glyphosate exposure and all cancer outcomes evaluated.
Two case-control studies evaluating the risk of NHL (Eriksson etal., 2008 and McDuffie et al.,
2001) observed increased effect estimates in the highest exposure categories analyzed. Eriksson
et al. (2008) found a greater effect estimate for subjects with >10 days (based on the median days
of exposure among controls) and >10 years of exposure (for latency analysis) when compared to
subjects with <10 days and 1-10 years of exposure, respectively; however, this analysis did not
appear to adjust for co-exposures to other pesticides. By dividing the total number of exposed
cases and controls using these exposure metrics, wider confidence intervals were observed due to
smaller sample sizes, which reduces the reliability of the results to demonstrate a true
Page 132 of 216
-------
association. This may indicate that a longer follow-up time is needed to detect the risk for NHL;
however, given the latency analysis of NHL was limited to Eriksson et al. (2008) and lack of
NHL latency understanding in general, further studies are needed to determine the true latency of
NHL. McDuffie el al. (2001), stratifying based on the average number of days per year of
exposure, observed similar effect estimates in the lower exposure category (>0 and <2 days/year)
while a greater effect estimate was observed in the highest exposure category (>2 days/year).
The results from these two case-control studies conflict with the results observed in the cohort
study (De Roos et al., 2005; Andreotti et al., 2017), where no dose-response was seen across
three exposure categories (stratified by tertiles); however, the case-control studies did not adjust
for co-exposure to other pesticides. It is also difficult to make conclusions regarding dose-
response with only two exposure categories used for the analyses by Eriksson et al. (2008) and
McDuffie et al. (2001). It should also be noted that these analyses combine all NHL subtypes,
which may have etiological differences (Morton el al., 2014). Although some studies did
provide effect estimates for subtypes, as stated in Section 3.5.2, these were not considered in the
current evaluation due to the limited sample sizes. At this time, there are no data available to
evaluate dose-response for NHL subtypes.
With respect to animal carcinogenicity studies, key events in a MOA/AOP are evaluated to
confirm that they precede tumor appearance. This temporal concordance evaluation cannot be
conducted for glyphosate since a MOA/AOP has not been established. It was noted that no
preneoplastic or related non-neoplastic lesions were reported in any of the animal carcinogenicity
studies to support any observed tumors. Furthermore, genotoxicity assays did not support a
direct mutagenic MO A. While there is limited evidence of genotoxicity in some in vitro
endpoints, multiple in vivo studies do not support a genotoxic risk at relevant human exposure
levels.
6.3 Strength, Consistency, and Specificity
A large database is available for evaluating the carcinogenicity potential of glyphosate. Across
animal carcinogenicity and genotoxicity studies, results were consistent. For epidemiological
studies, only one or two studies were available for almost all cancers investigated. The largest
number of studies was available investigating NHL; however, there were conflicting results
across studies.
In epidemiological studies, there was no evidence of an association between glyphosate exposure
and solid tumors, leukemia, and HL. This conclusion is consistent with those recently conducted
by IARC, EFSA, and JMPR. Furthermore, the available studies do not link glyphosate exposure
to multiple myeloma.
At this time, a conclusion regarding the association between glyphosate exposure and risk of
NHL cannot be supported based on the available data due to conflicting results. Chance and/or
bias cannot be excluded as an explanation for observed associations. The magnitude of adjusted
risk estimates for ever/never use were relatively small ranging from 1.0 (no association) to 1.85
in adjusted analyses, with the widest confidence intervals observed for the highest effect
estimates indicating less reliability in these estimates. All of the ever/never estimates were not
statistically significant with several effect estimates approximately equal to the null. There were
Page 133 of 216
-------
various limitations identified in Section 3.6 for these studies that could impact calculated effect
estimates and explain the weak responses observed. Meta-risk ratios using these studies were
also of small magnitude ranging from 1.3-1.5. As discussed in Section 3.6, meta-analyses should
be interpreted with caution and are susceptible to the same limitations identified for individual
studies.
Although none of the effect estimates were below 1 using the ever/never exposure metric, risk
estimates were all below 1 (0.6-0.9) when using cumulative lifetime and intensity-weighted
cumulative exposure metrics in the prospective cohort study (De Roos et al., 2005; Andreotti et
al., 2017). As discussed in Section 6.2, two case-control studies that investigated an exposure-
response relationship conflicted with the extensive analyses conducted for the AHS cohort. This
may be due to differences in confounding control, differences associated with study design,
limitations from small sample sizes, and/or, as some may suggest, a potentially short follow-up
time in the cohort. It should also be noted that publication bias may play a role in this evaluation
given there is a tendency to only publish positive results and potential concerns regarding
glyphosate have only been raised in recent years.
A total of 14 (8 rat and 6 mouse) animal carcinogenicity studies with glyphosate, glyphosate
acid, or glyphosate salts were analyzed for the current evaluation. None of the tumors evaluated
were considered to be treatment-related based on weight-of-evidence evaluations. Although
statistically significant trends were observed following adjustment for multiple comparisons in a
limited number of studies, statistically significant pairwise comparisons were only observed in 2
studies indicating tumor incidences were generally similar to concurrent controls. Additionally,
none of the tumor results were reproduced in other studies, including those testing the same
animal strain with similar or higher dosing. Furthermore, the tumors lacked a monotonic dose-
response and/or corroborating preneoplastic or related non-neoplastic lesions.
Over 80 genotoxicity studies with the active ingredient glyphosate were analyzed for the current
evaluation. The overall weight-of-evidence indicates that there is no convincing evidence that
glyphosate is genotoxic in vivo via the oral route. When administered via i.p. injection the
studies were predominantly negative. In the two cases where an increase in micronuclei were
reported via this route, the effects were not observed in other i.p. injection studies at similar or
higher doses. Technical glyphosate was negative in all gene mutation studies. There was limited
evidence of positive findings in studies evaluating primary DNA damage; however, as discussed
in Section 5.6, the endpoints measured in these assays are less specific in regards to detecting
permanent DNA changes (mutations) and can be attributed to other factors, such as cytotoxicity
or cell culture conditions. Although some positive findings were reported for chromosomal
alterations in vitro, these findings were limited to a few studies and are not supported by the in
vivo studies that are the most relevant for human risk assessment.
Overall, there is remarkable consistency in the database for glyphosate across multiple lines of
evidence. For NHL, observed associations in epidemiological studies were non-statistically
significant and were of relatively small magnitude. Chance and/or bias cannot be excluded as an
explanation for the observed associations. For all other cancer types, there were no associations
found; however, only one or two studies were available for evaluation of most cancer types.
Across species, strain, and laboratory, none of the tumors evaluated were considered to be
Page 134 of 216
-------
treatment-related based on weight-of-evidence evaluations. Statistically significant tumor results
seen in individual studies were not reproduced in other studies, including those conducted using
the same strain at similar or higher doses. The genotoxicity studies demonstrate that glyphosate
is not directly mutagenic or genotoxic in vivo.
6.4 Biological Plausibility and Coherence
The Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005) include the following
guidance regarding the criteria of biological plausibility and coherence:
"evaluation of the biological plausibility of the associations observed in epidemiologic
studies reflects consideration of both exposure-relatedfactors and toxicological evidence
relevant to identification ofpotential modes of action (MOAs). Similarly, consideration of
the coherence of health effects associations reported in the epidemiologic literature
reflects broad consideration of information pertaining to the nature of the biological
markers evaluated in toxicologic and epidemiologic studies, [p.39]."
The genotoxicity studies demonstrate that glyphosate is not directly mutagenic or genotoxic in
vivo. Immunodeficiency is another plausible MOA associated with tumorigenesis (i.e., altered
immune surveillance). Glyphosate was negative in an immunotoxicity study in mice at doses up
to 1448 mg/kg/day (MRID 48934207). Additionally, the toxicology database for glyphosate
does not reveal any evidence of immunotoxicity. Overall, the available data regarding non-
cancer endpoints also do not provide any support for a carcinogenic process for glyphosate, and
have shown glyphosate has relatively low toxicity. Laboratory animals generally display non-
specific effects (e.g., clinical signs, reduced body weight) following glyphosate exposure at
relatively high-doses, and no preneoplastic or related non-neoplastic lesions were observed to
corroborate any of the observed tumors in the carcinogenicity studies.
As discussed in Section 4.2, metabolism studies demonstrate low oral absorption and rapid
excretion of glyphosate. The data are not sufficient to determine whether linear kinetics is
occurring at high doses where tumors were observed. In the carcinogenicity test guideline
(OCSPP 870.4200) and the 2005 Guidelines for Carcinogen Risk Assessment, inappropriate
toxicokinetics (e.g., overwhelming absorption or detoxification mechanisms) should be avoided.
A study evaluating the toxicokinetic profile of glyphosate using multiple doses is needed to
further investigate the pharmacokinetic properties between low- and high-dose levels.
Although many of the studies included in this document focus on the potential for glyphosate to
cause a cancer outcome, the agency is also aware of a limited number of studies in the open
literature that have shown glyphosate and its metabolite, AMP A, can inhibit proliferation and
promote apoptosis in cancer cells indicating the compounds have potential to be developed into
therapeutic drugs for cancer treatment (Li et al, 2013; Parajuli etal., 2015; Parajuli et al., 2016).
It is unknown if this is due to lack of additional studies that have investigated these compounds
for cancer treatment or if this may be due to publication bias. The bias towards only publishing
positive and/or novel results can hamper the ability to evaluate whether there are plausible
biological mechanisms for observed outcomes and/or sufficient information to support a cause-
and-effect interpretation of an association. Overall, this further supports the need for
Page 135 of 216
-------
mechanistic data to elucidate the true mammalian MOA/AOP for glyphosate. There is a distinct
lack of mechanistic understanding for the toxicity of glyphosate in mammals and the plant MOA
is not relevant for mammalian systems.
The agency does not consider any of the tumors observed in the animal carcinogenicity studies to
be treatment-related; however, some believe that the increased tumor incidences in various
studies at the highest doses tested are treatment-related. In almost all of these studies, the highest
dose tested was approximately equal to or greater than the limit dose (1000 mg/kg/day). It is
very unlikely for people to be exposed to such large doses of glyphosate via the oral route.
Glyphosate is registered for pre- and post-emergence application to a variety of fruit, vegetable,
and field crops, as well as desiccant applications to several commodities. The highest dietary
exposure value for any population subgroup in an unrefined chronic dietary analysis would be
0.23 mg/kg/day for children (1-2 years old). Since glyphosate also has residential uses,
including application to turf, there is also the potential for children at this age to be exposed via
incidental oral exposures (e.g., hand to mouth, object to mouth and soil ingestion) from playing
on treated lawns. The highest exposure for the incidental oral and dermal exposures would be
0.16 mg/kg/day (from hand-to-mouth behaviors for children) and 0.08 mg/kg/day, respectively.
Combining exposures from the dietary and residential exposures for children would, therefore,
result in an aggregate exposure of 0.47 mg/kg/day. These calculations use a number of
assumptions that have been extensively peer-reviewed27 and yet the potential oral exposure of
glyphosate for the most highly exposed residential population subgroup is more than 2,000 times
lower than the highest doses tested in the animal carcinogenicity studies (see Appendix E for
more details regarding these calculations). The maximum potential exposure calculated for
occupational handlers would be 7 mg/kg/day, which is still significantly lower than the highest
doses tested in the animal carcinogenicity studies. As a result, even though increased tumor
incidences were observed in some of the animal carcinogenicity studies, the possibility of being
exposed to these excessive dietary doses over time is considered implausible based on the
currently registered use pattern and not relevant to human health risk assessment.
6.5 Uncertainty
When evaluating a database, it is also important to assess the uncertainties associated with the
available data. When uncertainty is high there is less confidence in the exposure and effect
estimates and, therefore, informs the reliability of results. Understanding the sources of
uncertainty within a database can help characterize observed results and aid in developing new
research with reduced uncertainty.
In some instances, the agency did not have access to all of the data underlying the studies
analyzed for the current evaluation. This includes all of the epidemiological studies, 17
genotoxicity studies, and 1 animal carcinogenicity study. For these studies, the agency had to
rely upon information the study authors reported. Without the raw data, statistical analyses could
not be replicated or recalculated. On the other hand, studies that include full reports with
detailed methodology, analytically measured doses, and individual animal data may provide a
27 Using the 2012 Standard Operating Procedures for Residential Exposure Assessment. Available:
http://www2.epa.gov/pesticide-science-and-assessing-pesticide-risks/standard-operating-procedures-residential-
pesticide
Page 136 of 216
-------
higher level of confidence. It also allows the agency to perform its own evaluation of the data
using current practices and policies.
Several uncertainties have already been identified throughout this document. There are
numerous metabolism studies available for glyphosate; however, the data are not sufficient to
determine whether linear kinetics is occurring at high doses where tumors were observed in
animal carcinogenicity studies. In the carcinogenicity test guideline (OCSPP 870.4200) and the
2005 Guidelines for Carcinogen Risk Assessment, inappropriate toxicokinetics (e.g.,
overwhelming absorption or detoxification mechanisms) should be avoided. A study evaluating
the toxicokinetic profile of glyphosate using multiple doses is needed to further investigate the
pharmacokinetic properties between low- and high-dose levels.
With respect to the epidemiological data, the database is limited for each investigated cancer
with only one or two studies available. Although numerous studies were used in the evaluation
of NHL, the results were constrained by the limitations of the individual studies, such as small
sample size, missing data, and control selection issues. The quality of the exposure assessment is
a major concern since the validity of the overall study results depend in large part on the ability
of the study to correctly quantify and classify a subject's exposure. There was no direct
information on pesticide exposure or absorbed dose because the exposures were self-reported.
All of the studies conducted exposure assessments through questionnaires and interviews that are
susceptible to recall bias, which can result in exposure misclassification. The cohort study (De
Roos et al., 2005), which was given a high ranking, did not find an association between
glyphosate exposure and NHL; however, it has been noted that the median follow-up time for
this study was ~7 years. Recently, an updated analysis was published (Andreotti el al., 2017)
with an extended follow-up period of 17.5 years that addresses concerns regarding follow-up
time. This study reported no association between glyphosate exposure and all
lymphohematopoietic cancers, NHL, or any of its subtypes across exposure metrics. No
association was observed in untagged or lagged analyses, after adjustment for pesticides linked
to NHL in previous AHS analyses, and after exclusion of multiple myeloma from the NHL
grouping. Furthermore, with the increased use of glyphosate following the introduction of
glyphosate-tolerant crops in 1996, there is a need for more recent studies since a large number of
studies were conducted prior to 1996. As described in Section 1.1, the use pattern changed
following the introduction of transgenic crops, which may impact overall effect estimates.
Another consideration is that farmers and other applicators apply formulations, not the active
ingredient alone. It is possible that different formulations were used across and/or within the
different epidemiological studies. Formulations are end-use products that are sold as a mixture
of registered pesticidal active ingredients, such as glyphosate, and additional substances that
increase the effectiveness of a pesticidal product, which are often referred to as "inert
ingredients." For example, inert ingredients may act as a solvent to allow a pesticide active
ingredient to penetrate a plant's outer surface, may facilitate and accentuate the dispersion of the
product, or may extend the pesticide product's shelf-life28. Inert ingredients and the proportion
of these inert ingredients vary across formulations. It has been hypothesized that glyphosate
formulations may be more toxic than glyphosate alone. Glyphosate has been studied in a
multitude of studies and there are studies that have been conducted on numerous formulations
28 https://www.epa.gov/pesticide-registration/inert-ingredients-overview-and-guidance
Page 137 of 216
-------
that contain glyphosate; however, there are relatively few research projects that have attempted
to systematically compare glyphosate and the formulations in the same experimental design.
Furthermore, there are even less instances of studies comparing toxicity across formulations.
This is one aspect of the uncertainty in the database that the agency has been working to address
by developing a strategic research plan in collaboration with NTP (see Section 7.0).
It is recognized that these uncertainties exist for the current database; however, the available data
are adequate for evaluating the carcinogenic potential of glyphosate and determine the cancer
classification using the 2005 Guidelines for Carcinogen Risk Assessment. As discussed in
Section 6.3, there are a large number of studies available and the database is remarkably
consistent across these studies.
6.6 Evaluation of Cancer Classification per the 2005 EPA Guidelines for Carcinogen
Risk Assessment
6.6.1 Introduction
In the 2005 Guidelines for Carcinogen Risk Assessment, five classification descriptors are
provided:
• Carcinogenic to Humans
• Likely to be Carcinogenic to Humans
• Suggestive Evidence of Carcinogenic Potential
• Inadequate Information to Assess Carcinogenic Potential
• Not Likely to be Carcinogenic to Humans
Descriptors are assigned using all available data from the multiple lines of evidence. The
following text has been excerpted/summarized from the guidelines regarding these descriptors:
Choosing a descriptor is a matter of judgment and cannot be reduced to a formula. Each
descriptor may be applicable to a wide variety of potential data sets and weights of
evidence. The weight-of-evidence, including the selected descriptor, is presented as a
narrative laying out the complexity of information that is essential to understanding the
hazard and its dependence on the quality, quantity, and type(s) of data available. The
descriptors and narratives are intended to permit sufficient flexibility to accommodate
new scientific understanding and new testing methods. The descriptors represent points
along a continuum of evidence; consequently, there are gradations and borderline cases
that are clarified by the full weight-of-evidence narrative. Rather than focusing simply
on the descriptor, the entire range of information included in the weight-of-evidence
narrative should be considered.
The weight-of-evidence presented in Sections 6.2-6.5 and based on the available
epidemiological, animal carcinogenicity, and genotoxicity data for glyphosate was considered for
each classification descriptor. For each descriptor, the guidelines provide examples and/or
conditions for when the descriptor may be appropriate and the weight-of-evidence for glyphosate
is assessed to determine which descriptor is supported by the available data. As stated in the
2005 EPA Guidelines for Carcinogen Risk Assessment, "the entire range of information included
Page 138 of 216
-------
in the weight-of-evidence should be considered". Based on all of the available data, the weight-
of-evidence clearly do not support the descriptors "carcinogenic to humans" and "likely to be
carcinogenic to humans" at this time. According to the 2005 Cancer Guidelines, "carcinogenic
to humans" is appropriate "when there is convincing epidemiologic evidence of a causal association
between human exposure and cancer." Similarly, "likely to be carcinogenic to humans" descriptor
is appropriate "when the weight of the evidence is adequate to demonstrate carcinogenic potential to
humans but does not reach the weight of evidence for the descriptor."
In epidemiological studies, there was no evidence of an association between glyphosate exposure
and solid tumors, leukemia, or HL. Furthermore, the available studies do not link glyphosate
exposure to multiple myeloma. A conclusion regarding the association between glyphosate
exposure and risk of NHL cannot be determined based on the available data due to conflicting
results and various limitations identified in studies investigating NHL. In 6 of the 14 animal
carcinogenicity studies, no tumors were identified for evaluation. In the remaining 8 studies, the
agency has concluded that none of the tumors evaluated in individual rat and mouse
carcinogenicity studies are treatment-related due to lack of pairwise statistical significance, lack
of a monotonic dose response, absence of preneoplastic or related non-neoplastic lesions, no
evidence of tumor progression, and/or historical control information (when available). Tumors
seen in individual rat and mouse studies were also not reproduced in other studies, including
those conducted in the same animal species and strain at similar or higher doses. The tumor
incidence increases in these studies were seen at or exceeding 1,000 mg/kg/day, except the
testicular tumors in a single rat study, and these high doses would also not be considered relevant
for human health risk assessment. The mammalian MOA/AOP is unknown for glyphosate and
precursor events are unknown; however, the genotoxicity data were highly reproducible and
consistent with a clear demonstration that glyphosate does not have a mutagenic MOA.
The descriptor "inadequate information to assess carcinogenic potential" is used when available
data are judged inadequate for applying one of the other descriptors. Given the extensive size of
the glyphosate database, which includes a multitude of well-designed and well-conducted
studies, this classification descriptor is not supported. The epidemiological data at this time are
limited and study results appear to be inconsistent for some cancer types. However, it is
important to note that epidemiological studies are not available for most pesticides. Similarly,
for most pesticides, generally, only two animal bioassays are available. EPA routinely evaluates
human cancer potential using the small, more typical datasets. As such, for glyphosate, given the
significant amount of information across multiple lines of evidence, the agency believes the
database is sufficient to designate a cancer classification descriptor for glyphosate and that
"inadequate information to assess carcinogenic potential" is not appropriate.
The remaining two cancer classification descriptors ( "Suggestive Evidence of Carcinogenic
Potential" and "Not Likely to Be Carcinogenic to Humans") from the 2005 EPA Guidelines for
Carcinogen Risk Assessment are described in detail below. Subsequently, these descriptors are
discussed in the context of whether the available evidence do or do not support them.
"Suggestive Evidence of Carcinogenic Potential"
This descriptor is appropriate when a concern for potential carcinogenic effects in humans is
raised, but the data are judged not sufficient for a stronger conclusion. It covers a spectrum of
Page 139 of 216
-------
evidence associated with varying levels of concern for carcinogenicity. Depending on the extent
of the database, additional studies may or may not provide further insights.
Some examples of when this descriptor may be appropriate include the following:
• If a small, and possibly not statistically significant, increase in tumor incidence observed
in a single animal or human study that does not reach the weight-of-evidence for the
descriptor of "likely to be carcinogenic to humans." The study generally would not be
contradicted by other studies of equal quality in the same population group or
experimental system;
• If there is evidence of a positive response in a study whose power, design, or conduct
limits the ability to draw a confident conclusion (but does not make the study fatally
flawed), but where the carcinogenic potential is strengthened by other lines of evidence;
• If there is a small increase in a tumor with a high background rate in that sex and strain,
when there is some but insufficient evidence that the observed tumors may be due to
intrinsic factors that cause background tumors and not due to the agent being assessed
(when there is a high background rate of a specific tumor in animals of a particular sex
and strain, then there may be biological factors operating independently of the agent
being assessed that could be responsible for the development of the tumors). In this
case, the reasons for determining that the tumors are not due to the agent are explained;
or
• If there is a statistically significant increase at one dose only, but no significant response
at the other doses and no overall trend.
"Not Likely to Be Carcinogenic to Humans "
This descriptor is appropriate when the available data are considered robust for deciding that
there is no basis for human hazard concern. In some instances, there can be positive results in
experimental animals when there is strong, consistent evidence that each MOA in experimental
animals does not operate in humans. In other cases, there can be convincing evidence in both
humans and animals that the agent is not carcinogenic.
This descriptor would be appropriate if any of the following was observed:
• Animal evidence demonstrates lack of carcinogenic effects in both sexes in well-designed
and well-conducted studies in at least two appropriate animal species in the absence of
other animal or human data suggesting a potential for cancer effects, or
• Convincing and extensive experimental evidence showing that the only carcinogenic
effects observed in animals are not relevant to humans, or
• Convincing evidence that carcinogenic effects are not likely by a particular exposure
route, or
• Convincing evidence that carcinogenic effects are not likely below a defined dose range.
Page 140 of 216
-------
6.6.2 Discussion of Evidence to Support Cancer Classification Descriptors
As stated above, the available data and weight-of-evidence clearly do not support the descriptors
"carcinogenic to humans", "likely to be carcinogenic to humans", or "inadequate information to
assess carcinogenic potential". The following discusses the remaining cancer classification
descriptors and how the evidence does or does not support the descriptors.
It could be argued that the "suggestive evidence of carcinogenic potential" descriptor would be
appropriate. The evidence to support this includes:
• Non-statistically significant effect estimates greater than the null were reported for NHL
across studies and meta-analyses based on ever/never use ranged from 1.3-1.5.
• There was limited evidence of a possible exposure-response relationship between
glyphosate exposure and NHL in case-control studies.
• In several animal carcinogenicity studies, a statistically significant trend was observed.
In two studies, tumor incidences at the highest doses tested were statistically significant
as compared to concurrent controls.
• Positive responses were observed in a limited number of genotoxicity assays evaluating
chromosomal and primary DNA damage.
However, according to the 2005 EPA Guidelines for Carcinogen Risk Assessment, in order for
the above evidence to support the "suggestive evidence of carcinogenic potential" descriptor,
"the study generally would not be contradicted by other studies of equal quality in the same
population group or experimental system". Furthermore, the guidelines state that "rather than
focusing simply on the descriptor, the entire range of information included in the weight-of-
evidence narrative should be considered". For the epidemiological studies evaluating NHL,
several studies reported effect estimates approximately equal to the null. The widest confidence
intervals were observed for the highest effect estimates indicating these effect estimate are less
reliable. All of the effect ever/never estimates were non-statistically significant. There were
conflicting results in exposure-response assessments investigating glyphosate exposure and the
risk of NHL. Although two-case control studies (McDuffie et al., 2001; Eriksson et a!., 2008)
reported elevated effect estimates when analyzing for exposure-response relationships across two
exposure categories, extensive analyses in a study of equal or higher quality (De Roos et al.,
2005) for cumulative lifetime exposure and intensity-weighted cumulative exposure contradicted
these results reporting effect estimates less than null (ranging from 0.6-0.9) when analyzing
across tertiles and these analyses were further supported by the recent evaluation of the AHS
cohort by Andreotti et al. (2017). Furthermore, the two-case control studies did not account for
co-exposure to other pesticides, which would be expected to cause inflated effect estimates.
Various limitations that could impact the calculated effect estimate were identified for these
studies and discussed in Section 3.6. The effect estimates greater than the null were not
strengthened by other lines of evidence, as described in Sections 6.2-6.5.
In 6 (4 rat and 2 mouse) of the 14 animal carcinogenicity studies conducted with glyphosate, no
tumors were identified for evaluation. In the remaining 8 studies, although statistically
significant trends following adjustment for multiple comparisons were observed in 6 of these
studies for different individual tumor types, almost all of these lacked pairwise significance
Page 141 of 216
-------
following adjustment for multiple comparisons. Pairwise significance was only observed at the
highest dose tested for testicular tumors (Lankas, 1981) and hemangiomas (Sugimoto, 1997).
For testicular tumors, a closer examination of the incidence data across doses did not
demonstrate a monotonic dose response and the tumor findings were not reproduced in studies of
equal quality, including studies in the same animal species and strain at similar or higher doses.
For hemangiomas, the statistical significance was seen at a dose more than 4X the limit dose,
which would not be considered relevant for human health risk assessment. Furthermore, the
tumor findings were not reproduced in studies of equal quality, including studies in the same
animal species and strain at similar or higher doses. In all of the animal carcinogenicity studies,
there was no evidence of corroborating pre-neoplastic or related non-neoplastic lesions to
support a treatment-related effect, including the testicular tumors. In a limited number of cases,
the agency also considered historical control data to inform the relevance of tumor findings and
these data generally indicated that incidence rates in the concurrent controls were unusually low
and/or observed tumor incidences were within historical control ranges.
Although positive responses were observed in a limited number of genotoxicity assays
evaluating chromosomal and primary DNA damage, the overall weight-of-evidence indicates
that there is no convincing evidence that glyphosate induces mutations in vivo via the oral route.
When administered via i.p. injection the studies were predominantly negative. In the two cases
where an increase in micronuclei were reported via this route of administration, the results were
contradicted by numerous other studies at similar or higher doses using the same assays and
route of administration. Technical glyphosate was negative in all gene mutation studies. There
was limited evidence of positive findings in studies evaluating primary DNA damage; however,
the endpoints measured in these assays are less specific in regards to detecting permanent DNA
changes (mutations) and can be attributed to other factors, such as cytotoxicity or cell culture
conditions. Although some positive findings were reported for chromosomal alterations in vitro,
these findings were limited to a few studies and are not supported by the in vivo studies that are
the most relevant for human risk assessment.
In summary, considering the entire range of information for the weight-of-evidence, the evidence
outlined above to potentially support the "suggestive evidence of carcinogenic potential"
descriptor are contradicted by other studies of equal or higher quality and, therefore, the data do
not support this cancer classification descriptor.
For the "not likely to be carcinogenic to humans" descriptor, one of the considerations is
whether there is "convincing evidence that carcinogenic effects are not likely below a
defined dose range". In the case of glyphosate, the agency did not consider any of the
tumors observed in the animal carcinogenicity studies to be treatment-related; however,
some believe that the increased tumor incidences in various studies at the highest doses
tested are treatment-related. In all of these studies, the highest dose tested was
approximately equal to or greater than the limit dose (1000 mg/kg/day), except for the
testicular tumors observed in a single study that were not considered treatment-related and
were not reproduced in studies of equal quality, including studies in the same animal
species and strain at similar or higher doses. In genotoxicity studies, assays with oral
administration were negative except for one instance where an extremely high dose (5,000
mg/kg/day) was administered.
Page 142 of 216
-------
The 2005 EPA Guidelines for Carcinogen Risk Assessment also state that "weighing of the
evidence includes addressing not only the likelihood of human carcinogenic effects of the agent
but also the conditions under which such effects may be expressed". Increased tumor incidence
was typically observed at doses of 1,000 mg/kg/day or greater; however, none of these were
considered treatment-related by the agency based on the weight-of-evidence evaluations.
Additionally, the only in vivo positive assays seen in the genotoxicity studies were administered
via i.p. injection at doses of 200 mg/kg/day and 300 mg/kg/day or orally at 5,000 mg/kg/day.
These high doses are not considered relevant to human health risk assessment based on the
currently registered use pattern for glyphosate. Maximum potential glyphosate exposure in
residential and occupational settings have been estimated at 0.47 mg/kg/day and 7 mg/kg/day,
respectively, which are well-below the doses necessary to elicit the effects seen in these animal
carcinogenicity and genotoxicity studies. Additionally, non-linear kinetics may also be
occurring at the high doses. The carcinogenicity test guidelines (OCSPP 870.4200 and OCSPP
870.4300) and the 2005 Guidelines for Carcinogen Risk Assessment state that inappropriate
toxicokinetics (e.g., overwhelming absorption or detoxification mechanisms) should be avoided.
A well-conducted pharmacokinetic study evaluating the toxicokinetic profile of glyphosate is
needed to further investigate the toxicokinetic properties between high and low dose levels to
ensure that inappropriate toxicokinetics is avoided.
Overall, there is not strong support for the "suggestive evidence of carcinogenic potential"
cancer classification descriptor based on the weight-of-evidence, which includes the fact that
even small, non-statistically significant changes observed in animal carcinogenicity and
epidemiological studies were contradicted by studies of equal or higher quality. The strongest
support is for "not likely to be carcinogenic to humans".
6.7 Proposed Conclusions Regarding the Carcinogenic Potential of Glyphosate
Glyphosate is a non-selective, phosphonomethyl amino acid herbicide registered to control
weeds in various agricultural and non-agricultural settings. Labeled uses of glyphosate include
over 100 terrestrial food crops as well as other non-agricultural sites, such as greenhouses,
aquatic areas, and residential areas. Following the introduction of glyphosate-resistant crops in
1996, glyphosate use increased dramatically; however, glyphosate use has stabilized in recent
years due to the increasing number of glyphosate-resistant weed species.
Since its registration in 1974, numerous human and environmental health analyses have been
completed for glyphosate, which consider all anticipated exposure pathways. Glyphosate is
currently undergoing Registration Review. As part of this process, the hazard and exposure of
glyphosate are reevaluated to determine its potential risk to human and environmental health
using current practices and policies. The human carcinogenic potential of glyphosate has been
evaluated by the agency several times. As part of the current evaluation for Registration Review,
the agency has performed a comprehensive analysis of available data from submitted guideline
studies and the open literature. This includes epidemiological, animal carcinogenicity, and
genotoxicity studies.
Page 143 of 216
-------
An extensive database exists for evaluating the carcinogenic potential of glyphosate, including
63 epidemiological studies, 14 animal carcinogenicity studies, and nearly 90 genotoxicity studies
for the active ingredient glyphosate. These studies were evaluated for quality and results were
analyzed across studies within each line of evidence. The modified Bradford Hill criteria were
then used to evaluate multiple lines of evidence using such concepts as strength, consistency,
dose response, temporal concordance and biological plausibility. The available data at this time
do no support a carcinogenic process for glyphosate. Overall, animal carcinogenicity and
genotoxicity studies were remarkably consistent and did not demonstrate a clear association
between glyphosate exposure and outcomes of interest related to carcinogenic potential. In
epidemiological studies, there was no evidence of an association between glyphosate exposure
and numerous cancer outcomes; however, due to conflicting results and various limitations
identified in studies investigating NHL, a conclusion regarding the association between
glyphosate exposure and risk of NHL cannot be determined based on the available data.
Increases in tumor incidence were not considered treatment-related in any of the animal
carcinogenicity studies. In 6 of these studies, no tumors were identified for evaluation. In the
remaining studies, the tumors were not considered treatment-related due to lack of pairwise
statistical significance, lack of a monotonic dose response, absence of preneoplastic or related
non-neoplastic lesions, no evidence of tumor progression, and/or historical control information
(when available). Additionally, tumor findings seen in individual rat and mouse studies were
also not reproduced in other studies, including those conducted in the same animal species and
strain at similar or higher doses. Furthermore, data from epidemiological and animal
carcinogenicity studies do not reliably demonstrate expected dose-response relationships. In
genotoxicity studies, there was no convincing evidence that glyphosate is genotoxic in vivo via
the oral route.
For cancer descriptors, the available data and weight-of-evidence clearly do not support the
descriptors "carcinogenic to humans", "likely to be carcinogenic to humans", or "inadequate
information to assess carcinogenic potential". For the "suggestive evidence of carcinogenic
potential" descriptor, considerations could be looked at in isolation; however, following a
thorough integrative weight-of-evidence evaluation of the available data, the database would not
support this cancer descriptor. The strongest support is for "not likely to be carcinogenic to
humans".
This analysis integrating multiple lines of evidence highlights the need for mechanistic studies to
elucidate the MOA/AOP of glyphosate, as well as additional epidemiology studies and updates
from the AHS cohort study to further investigate the carcinogenic potential of glyphosate in
humans. This evaluation focused on studies on the active ingredient glyphosate; however,
additional research could also be performed to determine whether formulation components, such
as surfactants, influence the toxicity of glyphosate formulations. The agency has been working
on plans to initiate research given these identified data gaps and these plans are described in
Section 7.0.
Page 144 of 216
-------
7.0 Collaborative Research Plan for Glyphosate and Glyphosate Formulations
As previously mentioned, some have believed that glyphosate formulations may be more toxic
than glyphosate alone. Glyphosate has been studied in a multitude of studies and there are
studies that have been conducted on numerous formulations that contain glyphosate; however,
there are relatively few research projects that have attempted to directly compare glyphosate and
the formulations in the same experimental design. Furthermore, there are even less instances of
studies comparing toxicity across formulations.
The agency has been collaborating with the NTP Division of the National Institute of
Environmental FTealth Sciences to develop a research plan intended to evaluate the role of
glyphosate in product formulations and the differences in formulation toxicity. Four objectives
were identified that laid out how research by NTP might contribute to these research questions:
1) compare the toxici ty of glyphosate vs. formulations, as well as compare formulations vs.
formulations, 2) provide publicly available toxicology data on cancer-related endpoints, 3)
provide publicly available toxicology data on non-cancer endpoints, and 4) investigate the
mechanisms of how glyphosate and formulations cause toxic effects.
As part of the first objective, NTP will investigate the differential biological activity of
glyphosate, glyphosate formulations, and the individual components of formulations. The NTP
Laboratory Branch generated preliminary data by exposing human hepatoma cells (HepG2) to
five different glyphosate products bought off the shelf. The endpoint in the assay was cell
viability, measured by ATP levels. The data, presented in Figure 7.1, demonstrate at-a-glance
that formulations are not equally toxic and that the toxicity is not being driven by the amount of
glyphosate in the formulations, at least for the endpoint of cell viability. This observation
highlights how informative the data generated from this research can be to the overall research
questions.
a>
o
c
o
o
W
(1)
c
¦¦¦
E
3
107-.
106n
10^
104
103i
102'
iL :
~ ~ 5
ri i J
k* 1
s ~
~" ¦
T
TT
uiiq
mi'l
0.001 0.01 0.1 1 10 100 1000
Glyphosate (jjM)
Amount of glyphosate in the formulation
1.9% (0.12 M)
1.0% (0.06 M)
41% (2.40 M) A
41% (2.40 M)B
-A- 18% (1.10 M)
Figure 7.1. Results of HepG2 exposures following 24 hour incubation using different glyphosate
formulations. Note: some of the formulations included other active ingredients besides glyphosate.
Page 145 of 216
-------
For the second objective, NTP will contribute to the publicly available knowledge-base
describing the biological effects of glyphosate and formulations by conducting guideline studies
addressing genotoxicity and studies that evaluate the oxidative stress potential. In order to
organize publicly available data on glyphosate and formulations, IARC used 10 key
characteristics of carcinogens as a way to help inform their conclusion (Smith et al., 2016).
Their review concluded that data were only available for two of these characteristics
(genotoxicity and oxidative stress) and little to no information on the remaining characteristics
was available. However, when members of a NTP workgroup looked at the available data
included in the IARC review, the group did not agree with IARC that the data provided strong or
clear evidence for either genotoxicity or induction of oxidative stress given protocol deficiencies
that could produce questionable results.
Currently, the publicly available information regarding non-cancer endpoints for glyphosate and
glyphosate formulations is limited. To begin to address the third objective, NTP will conduct a
screening level analysis of the literature using text mining software, for studies regarding non-
cancer endpoints resulting from glyphosate exposure. The resulting scoping report will describe
the evidence base for health outcomes investigated in connection to glyphosate, as well as help
identify data gaps.
As discussed in Section 6.0, there is a need for mechanistic studies to elucidate the MOA/AOP of
glyphosate. Although there are data suggesting glyphosate may be genotoxic or cause oxidative
stress, there is little mechanistic information to support these observations. For the last
objective, NTP will use in vitro screening assays to gain mechanistic information on the effects
of glyphosate and different formulations for a variety of endpoints and allow for direct
comparisons among them. The screening approach will also allow for the identification of test
substances that would be good candidates for further in vivo testing. Since in vivo findings in
genetic toxicology testing are generally considered as having a greater relevance to humans than
in vitro findings, it is valuable to confirm the results seen at the cellular level at the whole animal
level.
Page 146 of 216
-------
8.0 References
Akanuma, M. (1995). HR-001: Reverse Mutation Test. Kodaira Laboratories. The Institute of
Environmental Toxicology, Tokyo, Japan. Laboratory Project ID: IET 94-0142.
April 3, 1995. MRID 50017102. Unpublished.
Alvarez-Moya et al. (2014) "Comparison of the in vivo and in vitro genotoxicity of glyphosate
isopropylamine salt in three different organisms". Genetics and Molecular Biology, 37, 1, 105-
110
Andreotti, G., Freeman, L.E., Hou, L., Coble, J., Rusiecki, J., Hoppin, J.A., Silverman, D.T., and
Alavanja, M.C. (2009). Agricultural pesticide use and pancreatic cancer risk in the Agricultural
Health Study Cohort. International Journal of Cancer 124, 2495-2500.
Andreotti G., Koutros S., Hofmann J.N., Sandler D.P., Lubin J.H., Lynch C.F., Lerro C.C., De
Roos A.J., Parks C.G., Alavanja M.C., Silverman D.T., Beane Freeman L.E. Glyphosate Use
and Cancer Incidence in the Agricultural Health Study. J Natl Cancer Inst. 110 (5). Nov. 2017.
Atkinson, C., Strutt, A., Henderson, W., et al. (1993a). 104-Week Chronic Feeding/
Oncogenicity study in rats with 52-week interim kill. MRID No. 49631701. Unpublished
Atkinson, C., Martin, T., Hudson, P., and Robb, D. (1993b). Glyphosate: 104 week dietary
carcinogenicity study in mice. Inveresk Research International, Tranent, EH33 2NE, Scotland.
IRI Project No. 438618. April 7, 1993. MRID 49631702. Unpublished.
Band, P.R., Abanto, Z., Bert, J., Lang, B., Fang, R., Gallagher, R.P., and Le, N.D. (2011).
Prostate cancer risk and exposure to pesticides in British Columbia farmers. The Prostate 77,
168-183.
Baris, D, Garrity, TJ, Telles, JL, Heineman, EF, Olshan, A, HoarZahm, S. 2001. American
Journal of Industrial Medicine. 39: 463-476.
Benbrook. (2016). Trends in glyphosate herbicide use in the United States and globally.
Environmental Sciences Europe. 28(3).
Benjamini, Y and Hochberg, Y. 1995. Controlling the false discovery rate: a practical and
powerful approach to multiple testing. Journal of the Royal Statistical Society B. 57: 289-300.
Bolognesi, C., Bonatti, S., Degan, P., Gallerani, E., Peluso, M., Rabboni, R., Roggieri, P., and
Abbondandolo, A. (1997). Genotoxic activity of glyphosate and its technical formulation
roundup. Journal of Agricultural and Food Chemistry 45, 1957-1962.
Brammer. (2001). Glyphosate Acid: Two Year Dietary Toxicity and Oncogenicity Study in
Wistar Rats. Central Toxicology Laboratory, Alderley Park Macclesfield, Cheshire, UK:
Syngenta. MRID 49704601. Unpublished.
Page 147 of 216
-------
Brown, L.M., Blair, A., Gibson, R., Everett, G.D., Cantor, K.P., Schuman, L.M., Burmeister,
L.F., Vanlier, S.F., and Dick, F. (1990). Pesticide Exposures and Other Agricultural Risk Factors
for Leukemia among Men in Iowa and Minnesota. Cancer Research 50, 6585-6591.
Brown, L.M., Burmeister, L.F., Everett, G.D., and Blair, A. (1993). Pesticide Exposures and
Multiple Myeloma in Iowa Men. Cancer Causes Control 4, 153-156.
Brayton etal., 2012. Pathology of aging mice and GEM background strains and experimental
design. Vet Path. 49 (1): 85-105.
Burnett, P., Borders, J.; Kush, J. (1979) Report to Monsanto Company: Two Year Chronic Oral
Toxicity Study with CP- 76100 in Albino Rats: IBT No. 8560-08924. (Unpublished study
received Jun 24, 1982 under 524-308; prepared by Industrial Bio-Test Laboratories, Inc.,
submitted by Monsanto Co., Washington, DC; CDL:247746-A; 247745; 247747; 247748;
247749; 247750; 247751; 247752)
Collander R.D. (1996). Glyphosate Acid: An Evaluation of Mutagenic Potential Using S.
typhimurium and E. coli. Central Toxicology Laboratory, Cheshire, UK. Laboratory Project ID:
CTL/P/4874 Study No. YV3611. February 16, 1996. MRID 44320617. Unpublished.
Cantor, K.P., Blair, A., Brown, L.M., Burmeister, L.F., and Everett, G. (1993). Pesticides and
Other Agricultural Risk Factors for Non-Hodgkin's Lymphoma among Men in Iowa and
Minnesota. Cancer Research 53, 2421-2421.
Carreon, T., Butler, M.A., Ruder, A.M., Waters, M.A., Davis-King, K.E., Calvert, G.M., Schulte,
P.A., Connally, B., Ward, E.M., Sanderson, W.T., etal. (2005). Gliomas and Farm Pesticide
Exposure in Women: The Upper Midwest Health Study. Environmental Health Perspectives 113,
546-551.
Cimino, M.C. (2006). Comparative overview of current international strategies and guidelines
for genetic toxicology testing for regulatory purposes. Environmental and Molecular
Mutagenesis 47 (9): 362-390.
Chang, E.T., and Delzell, E. (2016). Systematic review and meta-analysis of glyphosate
exposure and risk of lymphohematopoietic cancers. Journal of environmental science and health
Part B, Pesticides, food contaminants, and agricultural wastes 51, 402-434.
Chhabra et al. (1990). An over view of prechronic and chronic toxicity/carcinogenicity
experimental study designs and criteria used by the National Toxicology Program. Environ
Health Perspect. 86: 313-321.
Chruscielska etal. (2000). Glyphosate: evaluation of chronic activity and possible far-reaching
effects. Part 1. Studies on chronic toxicity. Pestycydy (Warsaw). 3-4: 11-20.
Collins, A.R., Oscoz, A.A., Brunborg, G., Gaivao, I., Giovannelli, L., Kruszewski, M., C.C.,
Stetina, R. (2008). The Comet assay: topical issues. Mutagenesis 23 (3): 143-151.
Page 148 of 216
-------
Cocco, P., Satta, G., Dubois, S., Pili, C., Pilleri, M., Zucca, M., t Mannetje, A.M., Becker, N.,
Benavente, Y., de Sanjose, S., eial. (2013). Lymphoma risk and occupational exposure to
pesticides: results of the Epilymph study. Occupational and environmental medicine 70, 91-98.
Cooke et al., (2003). Oxidative DNA damage: mechanisms, mutation, and disease. FASEB J.
17 (10): 1195-214.
Chruscielska, K. et al. 2000. Glyphosate Evaluation of chronic activity and possible far-reaching
effects. Part 2. Studies on mutagenic activity. Pestycydy, 2000, (3-4), 21-25. Published.
Dennis, L.K., Lynch, C.F., Sandler, D.P., and Alavanja, M.C. (2010). Pesticide use and
cutaneous melanoma in pesticide applicators in the agricultural heath study. Environ Health
Perspect 775, 812-817.
De Roos, A.J., Zahm, S.H., Cantor, K.P., Weisenburger, D.D., Holmes, F.F., Burmeister, L.F.,
and Blair, A. (2003). Integrative assessment of multiple pesticides as risk factors for non-
Hodgkin's lymphoma among men. Occupational and environmental medicine 60. 1-9.
De Roos, A. J., et al. (2005). "Cancer incidence among glyphosate-exposed pesticide applicators
in the Agricultural Health Study." Environ Health Perspect 113(1): 49-54.
Durward, R. (2006). Technical Glyphosate: Micronucleus Test in the Mouse. Safepharm
Laboratories Limited, Shardlow Business Park, Shardlow, Derbyshire DE72 2GD, UK, Study
No. 2060/014. February 8, 2006. MRID 49957411. Unpublished.
Engel, L.S., Hill, D.A., Hoppin, J.A., Lubin, J.H., Lynch, C.F., Pierce, J., Samanic, C., Sandler,
D.P., Blair, A., and Alavanja, M.C. (2005). Pesticide use and breast cancer risk among farmers'
wives in the agricultural health study. Am J Epidemiol 7(57, 121-135.
Enemoto, K. (1997), HR-001: 24-Month Oral Chronic Toxicity and Oncogenicity Study in Rats,
Vol. 1. The Institute of Environmental Toxicology, Kodaira-shi, Tokyo, Japan, ArystaLife
Sciences, Study No.: IET 94-0150. MRID 50017104, 50017105, 5001703. Unpublished.
Eriksson, M., Hardell, L., Carlberg, M., and Akerman, M. (2008). Pesticide exposure as risk
factor for non-Hodgkin lymphoma including histopathological subgroup analysis. International
journal of cancer 123, 1657-1663.
Excel (1997). Combined chronic toxicity/carcinogenicity study of glyphosate technical in
Sprague Dawley rats. Pune, India: Indian Institute of Toxicology.
Fisher, RA. (1950) Statistical methods for research workers. Edinburgh, Scotland: Oliver and
Boyd.
Flower, K.B., Hoppin, J.A., Lynch, C.F., Blair, A., Knott, C., Shore, D.L., and Sandler, D.P.
(2004). Cancer Risk and Parental Pesticide Application in Children of Agricultural Health Study
Participants. Environmental Health Perspectives 772, 631-635.
Flowers, L.J., and Kier, L.D., Hannah, L.H. (1978) Final Report on Salmonella Mutagenicity
Assay of Glyphosate: Test No. LF-78-161. MRID 00078620. Unpublished.
Page 149 of 216
-------
Fontana et al (1998). Incidence rates of lymphomas and environmental measurements of
phenoxy herbicides: ecological analysis and case-control study. Archives of Environmental
Health: An International Journal. 53: 384-387.
Fox, V. (1998). Glyphosate acid: In vitro cytogenetic assay in human lymphocytes. Central
Toxicology Laboratory, Cheshire, UK. Report CTL/P/6050. October 29, 1998. MRID
49961803. Unpublished
Fox and Mackay (1996). Glyphosate acid: Mouse bone marrow micronucleus test. Central
Toxicology Laboratory, Cheshire, UK. March 21, 1996. MRID 44320619. Unpublished.
Green and Owen, (2011). Herbicide-resistant crops: utilities and limitations for herbicide-
resisitant week management. J. AgricFood Chem. 59 (11): 5819-29.
Greim, H., et al. (2015). "Evaluation of carcinogenic potential of the herbicide glyphosate,
drawing on tumor incidence data from fourteen chronic/carcinogenicity rodent studies." Crit Rev
Toxicol 45(3): 185-208.
George, J., et al. (2010). "Studies on glyphosate-induced carcinogenicity in mouse skin: a
proteomic approach." JProteomics 73(5): 951-964.
Ghisi, N.d.C., de Oliveira, E.C., and Prioli, A.J. (2016). Does exposure to glyphosate lead to an
increase in the micronuclei frequency? A systematic and meta-analytic review. Chemosphere
145, 42-54.
Giknis, M. L. A., and Clifford, C. B. (2005). Spontaneous Neoplastic Lesions in the Crl:CDl
(ICR) Mouse in Control Groups from 18 Month to 2 Year Studies. Charles River.
fattp://www.givereom/fttes/pdfs/rros/cd 1/rm mi j ions cried I icr mouse.aspx
Hardell, L., Eriksson, M., and Nordstrom, M. (2002). Exposure to pesticides as risk factor for
non-Hodgkin's lymphoma and hairy cell leukemia: Pooled analysis of two Swedish case-control
studies. Leukemia & Lymphoma 43, 1043-1049.
Hardell, L., and Eriksson, M. (1999). A case-control study of non-Hodgkin lymphoma and
exposure to pesticides. Cancer 85, 1353-1360.
Haseman, JK. (1995) Data analysis: Statistical analysis and use of historical control data. Regul
Toxicol Pharmacol 21:52-59.
Hill AB (1965). The Environment and Disease: Association or Causation? Proc R Soc Med. May
1965; 58(5): 295-300.
Hoar et al. (1986). Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma.
JAMA. 256:1141-1147.
Page 150 of 216
-------
Hohenadel, K., Harris, S.A., McLaughlin, J.R., Spinelli, J.J., Pahwa, P., Dosman, J.A., Demers,
P.A., and Blair, A. (2011). Exposure to multiple pesticides and risk of non-Hodgkin lymphoma
in men from six Canadian provinces. International journal of environmental research and public
health 8, 2320-2330.
Hornarvar, N. (2008). Technical Glyphosate: Micronucleus Test of the Mouse. RCC, Cytotest
Cell Research GmbH (RCC-CCR0, In den Leppsteinswiesen 19, 64380 Rossdorf, Germany,
Study No. 1158500. June 9, 2008. MRID 49961802. Unpublished.
Hsu and Stedeford (2010). Cancer Risk Assessment: Chemical Carcinogenesis, Hazard
Evaluation, and Risk Quantification. John Wiley & Sons.
Jensen, J.C., 1991a. Mutagenicity test: Ames Salmonella Assay with Glyphosate, Batch 206-
JaK-25-1. Scantox A/S, Lemvig, Denmark. Laboratory No. 12323. October 9, 1991. MRID
49961502. Unpublished.
Jensen, J.C. (1991b). Mutagenicity test: In vitro mammalian cell gene mutation test with
glyphosate, batch 206-JaK-25-l. Scantox A/S, Lemvig, Denmark. Laboratory No. 12325.
October 9, 1991. MRID 49961504. Unpublished.
Jensen, JC (1991c). Mutagenicity test: Micronucleus test with Glyphosate, batch 206-JAK-25-
1. Scantox A/S, Lemvig, Denmark. Report number 12324. December 9, 1991. MRID
49961503.
Kachuri, L., Demers, P.A., Blair, A., Spinelli, J.J., Pahwa, M., McLaughlin, J.R., Pahwa, P.,
Dosman, J.A., and Harris, S.A. (2013). Multiple pesticide exposures and the risk of multiple
myeloma in Canadian men. International journal of cancer 133, 1846-1858.
Karipidis etal. 2007. Occupational exposure to ionizing and non-ionizing radiation and risk of
non-Hodgkin lymphoma. Int Arch Occup Environ Health. 80: 663-670.
Karunanayake, CP, McDuffie, HH, Dosman, JA, Spinelli, JJ, Pahwa, P. 2008. Occupational
exposures and non-Hodgkin's lymphoma: Canadian case-control study. Environmental Health.
7:44.
Karunanayake, C.P., Spinelli, J. J., McLaughlin, J.R., Dosman, J.A., Pahwa, P., and McDuffie,
H.H. (2012). Hodgkin lymphoma and pesticides exposure in men: a Canadian case-control study.
Journal of agromedicine 17, 30-39.
Kaufman, D.W., Anderson, T.E., and Issaragrisil, S. (2009). Risk factors for leukemia in
Thailand. Annals of hematology 88, 1079-1088.
Kato etal. (2005). Personal and occupational exposure to organic solvents and risk of non-
Hodgkin's lymphoma (NHL) in women (United States). Cancer Causes & Control. 16:1215-
1224.
Kier, L. D. (2015). "Review of genotoxicity biomonitoring studies of glyphosate-based
formulations." CritRev Toxicol 45(3): 209-218.
Page 151 of 216
-------
Kier and Kirkland (2013). Review of genotoxicity studies of glyphosate and glyphosate-based
formulations. Critical Reviews in Toxicology. 43: 283-315.
Knezevich, A.L and Hogan, G. K. (1983). A chronic feeding study of glyphosate in mice.
Unpublished report prepared by Bio/Dynamic Inc., dated July 21, 1983. Report No. 77-2011.
EPA Accession No. 251007-251009, and 251014. EPA Accession no. 251007-09, 251014.
Unpublished.
Koller, V.J., Furhacker, M., Nersesyan, A., Misik, M., Eisenbauer, M., and Knasmueller, S.
(2012). Cytotoxic and DNA-damaging properties of glyphosate and Roundup in human-derived
buccal epithelial cells. Archives of toxicology 86, 805-813.
Koureas, M., Tsezou, A., Tsakalof, A., Orfanidou, T., and Hadjichristodoulou, C. (2014).
Increased levels of oxidative DNA damage in pesticide sprayers in Thessaly Region (Greece).
Implications of pesticide exposure. The Science of the total environment 496, 358-364.
Koutros, S., Beane Freeman, L.E., Lubin, J.H., Heltshe, S.L., Andreotti, G., Barry, K.H.,
DellaValle, C.T., Hoppin, J. A., Sandler, D.P., Lynch, C.F., etal. (2013). Risk of total and
aggressive prostate cancer and pesticide use in the Agricultural Health Study. Am J Epidemiol
777, 59-74.
Kumar, D.P.S. (2001), Carcinogenicity Study with Glyphosate Technical in Swiss Albino Mice,
Toxicology Department Rallis Research Centre, Rallis India Limited. Study No. TOXI:
1559.CARCI-M. MRID 49987403. Unpublished.
Landgren, O., Kyle, R.A., Hoppin, J.A., Beane Freeman, L.E., Cerhan, J.R., Katzmann, J.A.,
Rajkumar, S.V., and Alavanja, M.C. (2009). Pesticide exposure and risk of monoclonal
gammopathy of undetermined significance in the Agricultural Health Study. Blood 773, 6386-
6391.
J. Langsdale etal. (2009). Glyphosate. Human Health Assessment Scoping Document in Support
of Registration Review. June 3, 2009. D362745.
Lankas, G, P. (1981) A Lifetime Study of Glyphosate in Rats. Report No. 77-2062 prepared by
Bio Dynamics, Inc. EPA Accession. No. 247617 - 247621. December 23, 1981. MRID
00093879. Unpublished.
Lee, W.J., Cantor, K.P., Berzofsky, J.A., Zahm, S.H., and Blair, A. (2004a). Non-Hodgkin's
lymphoma among asthmatics exposed to pesticides. International journal of cancer 777, 298-302.
Lee, W.J., Lijinsky, W., Heineman, E.F., Markin, R.S., Weisenburger, D.D., and Ward, M.H.
(2004b). Agricultural pesticide use and adenocarcinomas of the stomach and oesophagus.
Occupational and environmental medicine 61, 743-749.
Lee, W.J., Colt, J.S., Heineman, E.F., McComb, R., Weisenburger, D.D., Lijinsky, W., and
Ward, M.H. (2005). Agricultural pesticide use and risk of glioma in Nebraska, United States.
Occupational and environmental medicine 62, 786-792.
Page 152 of 216
-------
Lee, W.J., Sandler, D.P., Blair, A., Samanic, C., Cross, A.J., and Alavanja, M.C. (2007).
Pesticide use and colorectal cancer risk in the Agricultural Health Study. International journal of
cancer 121, 339-346.
Li, A.P. (1983a). CHO/HGPRT gene mutation assay with glyphosate. Environmental Heath
Lab, St. Louis, MO. Study Number T830044. October 20, 1983. MRID 00132681.
Unpublished.
Li, A.P. (1983b). In vivo bone marrow cytogenetic study of glyphosate in Sprague Dawley rats.
Environmental Health Laboratory, St. Louis, MO. October 20, 1983. MRID 00132683.
Unpublished.
Li, A. P. and T. J. Long (1988). An evaluation of the genotoxic potential of glyphosate. Fundam
Appl Toxicol 10(3): 537-546.
Lioi M.B., etal. (1998a). Genotoxicity and oxidative stress induced by pesticide exposure in
bovine lymphocyte cultures in vitro. Mutation Research- Fundamental and Molecular
Mechanisms of Mutagenesis. 403 (1-2), 13-20.
Lioi M.B., etal. (1998b). Cytogenetic Damage and Induction of Pro-Oxidant State in Human
Lymphocytes Exposed In Vitro to Gliphosate, Vinclozolin, Atrazine, and DPX-E9636.
Environmental and Molecular Mutagenesis 32: 39-46.
Lioi, M.B., Scarfi, M.R., Santoro, A., Barbieri, R., Zeni, O., Salvemini, F., Di Berardino, D., and
Ursini, M.V. (1998a). Cytogenetic damage and induction of pro-oxidant state in human
lymphocytes exposed in vitro to gliphosate, vinclozolin, atrazine, and DPX-EP636.
Environmental and Molecular Mutagenesis 32, 39-46.
Lioi, M.B., Scarfi, M.R., Santoro, A., Barbieri, R., Zeni, O., Di Berardino, D., and Ursini, M.V.
(1998b). Genotoxicity and oxidative stress induced by pesticide exposure in bovine lymphocyte
cultures in vitro. Mutation Research-Fundamental and Molecular Mechanisms of Mutagenesis
403,13-20.
Lowit et al. (2015). Literature Review on Neurodevelopmental Effects & FQPA Safety Factor
Determination for the Organophosphate Pesticides. September 15, 2015. D331251.
Maizlish, N, Beaumont, J, Singleton, J. 1998. Mortality among California highway workers.
American Journal of Industrial Medicine. 13: 363-379.
Majeska, J.; Zameck, R.; Matheson, D. (1982a) SC-0224 (Lot No. 7269- 10): Mutagenicity
Evaluation in Salmonella typhimurium: Report No. T-10847. MRID 00126612. Unpublished.
Majeska, J. (1982b) Morphological transformation of Balb/3T3 cells. Report No. T-10884.
MRID 00126616. Unpublished.
Majeska, J. (1982c) Mutagenicity evaluation of bone marrow cytogenetic analysis in rats. Report
No. T10884. MRID 00132176. Unpublished.
Page 153 of 216
-------
Majeska, J. (1985a) Mutagenicity Evaluation in Salmonella typhimurium: SC-0224: Report No.
T-12660. MRID 00155527. Unpublished.
Majeska, J. (1985b) Mutagenicity Evaluation in Mouse Lymphoma Multiple Endpoint Test
Forward Mutation Assay: SC-0224: Report No. T-12661. MRID 00155528. Unpublished.
Majeska, J. (1985c) Mutagenicity Evaluation in Chinese Hamster Ovary Cytogenetic Assay: SC-
0224: Report No. T-12663. MRID 00155530. Unpublished.
Majeska, J (1987). Mutagenic evaluation in bone marrow micronucleus. Environmental Health
Center, Farmington, CT. Study Report No. T12689/SC-0024. April 23, 1987. MRID
40214004. Unpublished.
Manas, F., Peralta, L., Raviolo, J., Ovando, H.G., Weyers, A., Ugnia, L., Cid, M.G., Larripa, I.,
and Gorla, N. (2009). Genotoxicity of glyphosate assessed by the comet assay and cytogenetic
tests. Environmental toxicology and pharmacology 28, 37-41.
Manas, F., Peralta, L., Ugnia, L., Weyers, A., Garcia Ovando, H., and Gorla, N. (2013).
Oxidative stress and comet assay in tissues of mice administered glyphosate and ampa in
drinking water for 14 days. BAG Journal of basic and applied genetics 24, 67-75.
Marques, M.F.C. (1999). A Micronucleus Study in Mice for Glifosate Tecnico Nufarm. Bioagri
Laboratarios Ltda. Study No: RF-G12.79/99. December 27, 1999. MRID 49957412.
Unpublished.
Matsumoto (1995). HR-001: In vitro cytogenetics test. The Institute of Environmental
Toxicology, Tokyo, Japan. Laboratory Project ID: IET 94-0143. May 29, 1995. MRID
50017106. Unpublished.
McConnell, EE; Solleveld, HA; Swenberg, JA; et al. (1986) Guidelines for combining
neoplasms for evaluation of rodent carcinogenesis studies. J Natl Cancer Inst 76:283-289.
McDuffie, H.H., Pahwa, P., McLaughlin, J.R., Spinelli, J.J., Fincham, S., Dosman, J.A., Robson,
D., Skinnider, L.F., and Choi, N.W. (2001). Non-Hodgkin's lymphoma and specific pesticide
exposures in men: Cross-Canada study of pesticides and health. Cancer Epidemiol Biomarkers
Prev 10, 1155-1163.
McDuffie, HH, Pahwa, P, McLaughlin, JR, Fincham, S, Robson, D, Dosman, JA, Hu, J. 2002.
Canadian male farm residents, pesticide safety handling practices, exposure to animals and non-
Hodgkin's lymphoma (NHL). American Journal of Industrial Medicine. 42: 54-61.
McDuffie, HH, P Pahwa, JJ Spinelli, JR McLaughlin, S. Fincham, D. Robson, JA Dosman, J.
Hu. 2002. Canadian Male Farm Residents, Pesticide Safety Handling Practices, Exposure to
Animals, and Non-Hodgkin's Lymphoma. American Journal of Industrial Medicine Supplement
2: 54-61
Page 154 of 216
-------
Meek ME, Boobis A, Cote I, Dellarco V, Fotakis G, Munn S, Seed J, Vickers C. 2014. New
developments in the evolution and application of the WHO/IPCS framework on mode of
action/species concordance analysis. JAppl Toxicol. 2014 Jan;34(l):l-18.
Mink, P. J., et al. (2012). "Epidemiologic studies of glyphosate and cancer: a review." Regul
Toxicol Pharmacol 63(3): 440-452.
Mladinic, M., Berend, S., Vrdoljak, A.L., Kopjar, N., Radic, B., and Zeljezic, D. (2009a).
Evaluation of genome damage and its relation to oxidative stress induced by glyphosate in
human lymphocytes in vitro. Environmental and molecular mutagenesis 50, 800-807.
Mladinic, M., Perkovic, P., and Zeljezic, D. (2009b). Characterization of chromatin instabilities
induced by glyphosate, terbuthylazine and carbofuran using cytome FISH assay. Toxicology
letters 189, 130-137.
Moriya, M. et al. 1983. Further mutagenicity studies on pesticides in bacterial reverse assay
systems. Mutation Research, 116(1983), 185-216.
Morton et al. (2014). Etiologic heterogeneity among non-Hodgkin lymphoma subtypes: the
interlymp non-Hodgkin lymphoma subtypes project. J Natl Cancer Inst Monogr 48: 130-144.
Nordstrom, M., Hardell, L., Magnuson, A., Hagberg, H., and Rask-Andersen, A. (1998).
Occupational exposures, animal exposure and smoking as risk factors for hairy cell leukaemia
evaluated in a case-control study. Br J Cancer 77, 2048-2052.
NTP (1992). NTP technical report on the toxicity studies of Glyphosate (CAS No. 1071-83-6)
Administered In Dosed Feed To F344/N Rats And B6C3F1 Mice. Toxic Rep Ser 16: l-d3.
OECD (2015). Guidance Document on Revisions to OECD Genetic Toxicology Test
Guidelines. August 31, 2015.
Olsson and Brandt (1988). Risk of non-Hodgkin's lymphoma among men occupationally
exposed to organic solvents. Scandinavian Journal of Work, Environment, & Health. 14:246-
251.
Orsi, L., Delabre, L., Monnereau, A., Delval, P., Berthou, C., Fenaux, P., Marit, G., Soubeyran,
P., Huguet, F., Milpied, N., et al. (2009). Occupational exposure to pesticides and lymphoid
neoplasms among men: results of a French case-control study. Occupational and environmental
medicine 66, 291-298.
Pahwa, P., Karunanayake, C.P., Dosman, J.A., Spinelli, J.J., McLaughlin, J.R., and Cross-
Canada, G. (2011). Soft-tissue sarcoma and pesticides exposure in men: results of a Canadian
case-control study. Journal of occupational and environmental medicine / American College of
Occupational and Environmental Medicine 53, 1279-1286.
Pahwa, P., Karunanayake, C.P., Dosman, J. A., Spinelli, J.J., McDuffie, H.H., and McLaughlin,
J.R. (2012). Multiple myeloma and exposure to pesticides: a Canadian case-control study.
Journal of agromedicine 17, 40-50.
Page 155 of 216
-------
Parajuli, K. R., etal. (2015). "Aminomethylphosphonic acid and methoxyacetic acid induce
apoptosis in prostate cancer cells." Int J Mol Sci 16(5): 11750-11765.
Parajuli, K. R., etal. (2016). "Aminomethylphosphonic acid inhibits growth and metastasis of
human prostate cancer in an orthotopic xenograft mouse model." Oncotarget 7(9): 10616-10626.
Pavkov, K.LI, Turnier, J.C. (1987) Two-Year Chronic Toxicity and Oncogenecity Dietary Study
with SC-0224 in Mice. Stauffer Chemical Company. MRID 40214006. Unpublished.
Pavkov, K.LI, Wyand, S. (1987) Two-Year Chronic Toxicity and Oncogenecity Dietary Study
with SC-0224 in Rats. Stauffer Chemical Company. MRID 40214006. Unpublished.
Peluso M. et al. (1998). 32P-Postlabeling detection of DNA adducts in mice treated with
herbicide Roundup. Environ and Mol Mutagen 31:55-59.
Piesova, E. (2004). The influence of different treatment length on the induction of micronuclei in
bovine lymphocytes after exposure to glyphosate. Folia Veterinaria 48, 130-134.
Piesova, E. (2005). The effect of glyphosate on the frequency of micronuclei in bovine
lymphocytes in vitro. Acta Veterinaria-Beograd 55, 101-109.
Portier, C.J., Armstrong, B.K., Baguley, B.C., Baur, X., Belyaev, I., Belle, R., Belpoggi, F.,
Biggeri, A., Bosland, M.C., Bruzzi, P., etal. (2016). Differences in the carcinogenic evaluation
of glyphosate between the International Agency for Research on Cancer (IARC) and the
European Food Safety Authority (EFSA). Journal of epidemiology and community health.
Rank, J., etal. (1993). "Genotoxicity testing of the herbicide Roundup and its active ingredient
glyphosate isopropylamine using the mouse bone marrow micronucleus test, Salmonella
mutagenicity test, and Allium anaphase-telophase test." MutatRes 300(1): 29-36.
Recore et al. (2014). Glyphosate: Tier II incident report. February 6, 2014. D417808.
Reyna, M.S. Gordon, D.E. (1973) 18-Month Carcinogenic Study with CP67573 in Swiss White
Mice: IBT No. B569. (Unpublished study, including sponsor's validation report dated Feb 1,
1978, received Jun 21, 1978 under 524-308; prepared by Industrial Bio-Test Laboratories, Inc.,
submitted by Monsanto Co., Washington, D.C.; CDL:234136-G). MRID 00061113.
Unpublished.
Reyna, M.S. Richter, W.R. Gordon, D.E. (1974) Two- Year Chronic Oral Toxicity Study with
CP67573 in Albino Rats: IBT No. B564. MRID 00062507. Unpublished.
Ribeiro do Val, R. (2007). Bacterial reverse mutation test (Ames test) for Glifosato Technico
Helm. TECAM Tecnologia Ambiental Ltda., Sao Paulo, Brasil. Study Number 3393/2007-
2.0AM, Report Number RL3393/2007-2.0 AM-B, December 13, 2007. MRID 50000903.
Unpublished.
Page 156 of 216
-------
Rodney, DE (1980). Dominant lethal study in mice. International Research and Development
Corp. May 23, 1980. MRID 0004634
Roustan, A., et al. (2014). "Genotoxicity of mixtures of glyphosate and atrazine and their
environmental transformation products before and after photoactivation." Chemosphere 108: 93-
100.
Ruder, A.M., Waters, M.A., Butler, M.A., Carreon, T., Calvert, G.M., Davis-King, K.E., Schulte,
P.A., Sanderson, W.T., Ward, E.M., Connally, L.B., et al. (2004). Gliomas and farm pesticide
exposure in men: the upper midwest health study. Archives of environmental health 59, 650-657.
Schinasi, L., and Leon, M.E. (2014). Non-Hodgkin lymphoma and occupational exposure to
agricultural pesticide chemical groups and active ingredients: a systematic review and meta-
analysis. International journal of environmental research and public health 77, 4449-4527.
Shirasu, Y.; Moriya, M.; Ohta, T. (1978) Microbial Mutagenicity Testing on CP67573
(Glyphosate). Apr 25, 1979 under 524-308; prepared by Institute of Environ- mental
Toxicology, Japan, submitted by Monsanto Co., Washington, D.C.; CDL:238233-A. MRID
00078619. Unpublished.
Sivikova, K. and J. Dianovsky (2006). "Cytogenetic effect of technical glyphosate on cultivated
bovine peripheral lymphocytes." Int J Hyg Environ Health 209(1): 15-20.
Smith et al., (2016). Key characteristics of carcinogens as a basis for organizing data on
mechanisms of carcinogenesis. Environmental Health Perspectives. 124: 713.
Snedecor, GW; Cochran, WG. (1967) Statistical methods, 6th ed. Ames, Iowa: Iowa State
University Press.
Sokolowski, A. (2007a). Salmonella typhimurium and Escherichia coli Reverse Mutation Assay
with Glyphosate Technical (NUP-05068). RCC Cytotest Cell Research GmbH, Rossdorf,
Germany. Study Number 1061401. March, 16, 2007. MRID 49957406. Unpublished.
Sokolowski, A. (2007b). Salmonella typhimurium and Escherichia coli Reverse Mutation Assay
with Glyphosate Technical (NUP-05070). RCC Cytotest Cell Research GmbH, Rossdorf,
Germany. Study Number 1061402. March, 16, 2007. MRID 49957407. Unpublished.
Sokolowski, A. (2007c). Salmonella typhimurium and Escherichia coli Reverse Mutation Assay
with Glyphosate Technical (NUP-05067). RCC Cytotest Cell Research GmbH, Rossdorf,
Germany. Study Number 1061403. March, 16, 2007. MRID 49957408. Unpublished.
Sokolowski, A. (2009b). Glyphosate technical- Salmonella Typhimurium and Escherichia Coli
reverse mutation assay. Cytotest Cell Research GmbH (Harlan CCR), Rossdorf, German. Study
and Report Number 1264500. December 18, 2009. MRID 49961801. Unpublished.
Page 157 of 216
-------
Sokolowski, A. (2010). Salmonella typhimurium and Escherichia coli Reverse Mutation Assay
with Glyphosate TC spiked with glyphosine. RCC Cytotest Cell Research GmbH, Rossdorf,
Germany. Study Number 1332300. April 7, 2010. MRID 500000902. Unpublished.
Sorahan, T. (2015). "Multiple Myeloma and Glyphosate Use: A Re-Analysis of US Agricultural
Health Study (AHS) Data." Int J Environ Res Public Health 12(2): 1548-1559.
Stout, L. D. and Ruecker, P. A. (1990). Chronic Study of Glyphosate Administered in Feed to
Albino Rats. MRID No. 41643801; Historical Controls. MRID 41728700. Unpublished.
Sugimoto, K. (1997), HR-001: 18-Month Oral Oncogenicity Study in Mice, Vol. 1 and 2. The
Institute of Environmental Toxicology, 2-772, Suzuki-cho, Kodaira-shi, Tokyo, 187, Japan,
Study No.: IET 94-0151. MRID 50017108, 50017109. Unpublished.
Suresh, T.P. (1993b). Mutagenicity-Micronucleus Test in Swiss Albino Mice. Rallis India
Limited. Study No: TOXI: 889-MUT.MN. May 6, 1993. MRID 49987407. Unpublished.
Suresh, T.P. (1994). Genetic Toxicology- In vivo mammalian bone marrow cytogenetic test-
Chromosomal analysis. Rallis Agrochemical Research Station, Bangalore, India. January 1,
1994. MRID 49987408. Unpublished.
Suresh, T.P. (1996) Combined Chronic Toxicity and Carcinogenicity Study with Glyphosate
Technical in Wistar Rats. Toxicology Department Rallis Research Centre, Rallis India Limited,
TOXI-1559, 002/1-GPT-CARCI-M. MRID 49987401. Unpublished.
Taddesse-Heath, L.; Chattopadhyay, S.K.; Dillehay, D.; L.; Lander, M.R.; Nagashfar, Z.; Morse
III, H.C.; Hartley, J.W. (2000): Lymphomas and high-level expression of murine leukemia
viruses in CFW mice Journal of Virology 74:6832-6837
Tarone, RE. (1982). The use of historical control information in testing for a trend in
proportions. Biometrics 38:215-220.
Thompson, P.W. (1996). Technical Glyphosate: Reverse Mutation Assay "Ames Test" using
Salmonella typhimurium and Escherichia coli. Safepharm Laboratories Limited, Derby, UK.
Study Number 434/014. February 20, 1996. MRID 49957409. Unpublished.
Wang etal. (2009). Occupational exposure to solvents and risk of non-Hodgkin lymphoma in
Connecticut women. American Journal of Epidemiology. 169:176-185.
Ward, J. M. (2006). Lymphomas and leukemias in mice. Experimental and Toxicologic
Pathology, 57 (5-6): 377-381.
Weisenburger, D.D. (1992). Pathological Classsification of Non-Hodgkin's Lymphoma for
Epidemiological Studies. Cancer Research 52, 5456S-5462S.
Page 158 of 216
-------
Wilderman, A.G. and Nazar, R.N. (1982). Significance of plant metabolism in the mutagenicity
and toxicity of pesticides. Canadian Journal of Genetics and Cytology 24(4): 437-449.
Williams, G. M., et al. (2000). "Safety Evaluation and Risk Assessment of the Herbicide
Roundup and Its Active Ingredient, Glyphosate, for Humans." Regulatory Toxicology and
Pharmacology 31(2): 117-165.
Wood, E., Dunster, J., Watson, P., and Brooks, P. (2009a) Glyphosate Technical: Dietary
Combined Chronic Toxicity/Carcinogenicity Study in the Rat. Harlan Laboratories Limited,
Shardlow Business Park, Shardlow, Derbyshire DE72 2GD, UK. Study No. 2060-012. April,
23, 2009. MRID 49957404. Unpublished.
Wood, E., Dunster, J., Watson, P., and Brooks, P. (2009b) Glyphosate Technical: Dietary
Carcinogenicity Study in the Mouse. Harlan Laboratories Limited, Shardlow Business Park,
Shardlow, Derbyshire DE72 2GD, UK. Study No. 2060-011. April, 22, 2009. MRID
49957402. Unpublished.
Wright, N.P. (1996). Technical Glyphosate: Chromosomal aberration test in CHL cells in vitro.
Safepharm Laboratories Limited, Derby, UK. Study Number 434/015. March 13, 1996, MRID
49957410. Unpublished.
Yauk et al., (2015). Approaches to identifying germ cell mutagens: Report of the 2013 IWGT
workshop on germ cell assays. Mutat Res Genet Toxicol Environ Mutagen, 783: 36-54.
Yiin, J.H., Ruder, A.M., Stewart, P.A., Waters, M.A., Carreon, T., Butler, M.A., Calvert, G.M.,
Davis-King, K.E., Schulte, P.A., Mandel, J.S., et al. (2012). The upper midwest health study: a
case-control study of pesticide applicators and risk of glioma. Environ Health 77, 13.
Zaccaria, C.B. (1996). A micronucleus study in mice for the product Glyphosate. BioAgri.
Study No: G.1.2 - 06/96. November 18, 1996. MRID 49961501. Unpublished.
Zahm et al (1990). A case-control study of non-Hodgkin's lymphoma and the herbicide 2,4-
dichlorophenoxyacetic acid (2,4-D) in Eastern Nebraska. Epidemiology. 1:349-356.
Zhang et al. (2007). Ultraviolet radiation exposure and risk of non-Hodgkin's lymphoma.
American Journal of Epidemiology. 165: 1255-1264.
Zoriki-Hosmi, (2007). Hosomi, R. (2007) Mammalian Erythrocyte Micronucleus Test for
Glifosato Tecnico Helm. Project Number: RL/3393/2007/3/0MN/B. MRID 50000901.
Unpublished.
Page 159 of 216
-------
Appendix A. Journal articles obtained from open literature search
Abstract Only
Cebollero, L. R., et al. (2011). "Glyphosate based herbicides toxicity, a new approach." Toxicology
Letters 205, Supplement: S233.
Abstract Only
Monroy, C. M., et al. (2004). "In vitro evaluation of glyphosate-induced DNA damage in fibrosarcoma
cells HT1080 and Chinese hamster ovary (CHO) cells." Environ Mol Mutagen 44(3): 216-216.
Abstract Only
Ramos-Morales, P., et al. (2008). "Combined use of multiple biomarkers to evaluate the genotoxic
activity of the herbicide Glyphosate." Environ Mol Mutagen 49(7): 577-577.
Abstract Only /Full article
already identified
Sorahan, T. (2015). "Multiple Myeloma and Glyphosate Use: A Re-Analysis of US Agricultural Health
Study (AHS) Data." Int J Environ Res Public Health 12(2): 1548-1559.
Article not in English
Kwiatkowska, M., et al. (2013). "GLYPHOSATE AND ITS FORMULATIONS - TOXICITY,
OCCUPATIONAL AND ENVIRONMENTAL EXPOSURE." Med Pr 64(5): 717-729.
Article not in English
Lawson, R. and E. Estrade-Chapellaz (1999). "Intoxication volontaire par le glufosinate (Basta®)."
Annales Francaiscs d'Anesthesie et de Reanimation 18(9): 1025-1026.
Article not in English
Manas, F., et al. (2009)."Aberraciones cromosomicas en trabajadores rurales de la Provincia de Cordoba
expuestos a plaguicidas." BAG. Journal of basic and applied genetics 20(1): 0-0.
Article not in English
Martinez, A., et al. (2007). "[Cytotoxicity of the herbicide glyphosate in human peripheral blood
mononuclear cells]." Biomedica 27(4): 594-604.
Article not in English
Monroy, C. M., et al. (2005). "[Cytotoxicity and genotoxicity of human cells exposed in vitro to
glyphosate]." Biomedica 25(3): 335-345.
Article not in English
Pieniazek, D., et al. (2003). "[Glyphosate~a non-toxic pesticide?]." Med Pr 54(6): 579-583.
Article not in English
Saratovskikh, E. A., et al. (2007). "Genotoxicity of the pestiside in Ames test and the possibility to
formate the complexeses with DNA." Ekologicheskaya genetika V(3): 46-54.
Article not in English
B CTaTbe npcjCTaB.iCHbi pe3yni>TaTi>i reHOTOKCHKO .lornHCCKHx. a xic p ro .to rn lic c k h x h
MMMV HO-lOrMMCCKMX HCC.1C JOB3HHH. npOBCJCHHblX B paMKaX MC JHKO-OHO.lOrHHCCKOH OUCHKH
6e30naCH0CTH rCHHO-HH>KCHCpHO-\IO;iH-(|)HHHpOBaHHOH KyKypy3bI JIHHHH MON 88017, yCTOHMMBOH K
rjiH(})ocaTy h >kykv Diabrotica spp. AHajiro jaHHbix. no.iyHCHHbix npn hsyhchhh ypoBHH noBpoiQCHHH
AHK h ypoBHJi xpoMOCOMHbix aoeppauHH. t;i>kccth aKTHBHoro a Ha (| w.ia kt hlicc ko ro niOKa h
HHTeHCHBHOCTH ryMopajibHoro HMMyHHoro OTBeTa, cocto;ihh;i ryMopanbHoro h K.iCTOHHoro 3BeHbeB
HMMyHHTeTa, He BbiHBHji KaKoro-jiH6o re hoto kc hlicc ko ro. ajinepreHHoro, h\i \iv ho mo jy .iupv K)luc ro h
c e h c h 6 h h's h py k) me ro jchctbhsi rM-KyKypy3bi jihhhh MON 88017 no cpaBHeHino c ee
TpanHijHOHHbiM aHanoroM.
Cancer treatment
Parajuli, K. R., et al. (2015). "Aminomethylphosphonic acid and methoxyacetic acid induce apoptosis in
prostate cancer cells." Int J Mol Sci 16(5): 11750-11765.
Cancer treatment
Li, Q., et al. (2013). "Glyphosate and AMPA inhibit cancer cell growth through inhibiting intracellular
glycine synthesis." Drug Des Devel Ther 7: 635-643.
Cancer treatment
Parajuli, K. R., et al. (2016). "Aminomethylphosphonic acid inhibits growth and metastasis of human
prostate cancer in an orthotopic xenograft mouse model." Oncotarget 7(9): 10616-10626.
Correspondence article
Belle, R., et al. (2012). "LETTER TO THE EDITOR: TOXICITY OF ROUNDUP AND
GLYPHOSATE." Journal of Toxicology and Environmental Health-Part B-Critical Reviews 15(4): 233-
235.
Correspondence article
Carrasco, A. E. (2011). "Reply to the Letter to the Editor Regarding Our Article (Paganelli et al., 2010)."
ChemRes Toxicol 24(5): 610-613.
Correspondence article
de Souza, L. andL. Macedo Oda (2013). "Letter to the editor." Food and Chemical Toxicology 53: 440.
Correspondence article
de Vendomois, J. S., et al. (2010). "Debate on GMOs Health Risks after Statistical Findings in Regulatory
Tests." Int J Biol Sci 6(6): 590-598.
Correspondence article
Farmer, D. R., etal. (2005). "Glyphosate results revisited." Environ Health Perspect 113(6): A365-A366.
Correspondence article
Grunewald, W. and J. Bury (2013). "Comment on "Long term toxicity of a Roundup herbicide and a
Roundup-tolerant genetically modified maize" by Seralini et al." Food and Chemical Toxicology 53: 447-
448.
Correspondence article
Huang, K. and W. Xu (2013). "Reply to letter to the editor." Food and Chemical Toxicology 59: 811-
812.
Correspondence article
Korsaeth, A., et al. (2015). "Comments on the recently published study: "Compositional differences in
soybeans on the market: Glyphosate accumulates in Roundup Ready GM soybeans", by T. Bohn, M.
Page 160 of 216
-------
Cuhra, T. Traavik, M. Sanden, J. FaganandR. Primicerio (Food Chemistry 2014, 153: 207-215)." Food
Chemistry 172: 921-923.
Correspondence article
Ollivier, L. (2013). "A Comment on "Seralini, G.-E., et al., Long term toxicity of a Roundup herbicide
and a Roundup-tolerant genetically modified maize. Food Chem. Toxicol. (2012),"
http://dx.doi.Org/10.1016/j.fct.2012.08.005." Food and Chemical Toxicology 53: 458.
Correspondence article
Portier, C. J., et al. (2016). "Differences in the carcinogenic evaluation of glyphosate between the
International Agency for Research on Cancer (IARC) and the European Food Safety Authority (EFSA)."
J Epidemiol Community Health.
Correspondence article
Roberfroid, M. (2014). "Letter to the editor." Food and Chemical Toxicology 66: 385.
Correspondence article
Sanders, D., et al. (2013). "Comment on "Long term toxicity of a Roundup herbicide and a Roundup-
tolerant genetically modified maize" by Seralini et al." Food and Chemical Toxicology 53: 450-453.
Correspondence article
Schorsch, F. (2013). "Serious inadequacies regarding the pathology data presented in the paper by
Seralini et al. (2012)." Food and Chemical Toxicology 53: 465-466.
Correspondence article
Wallace Hayes, A. (2014). "Editor in Chief of Food and Chemical Toxicology answers questions on
retraction." Food and Chemical Toxicology 65: 394-395.
Effects on cellular
processes
Benachour, N. and G.-E. Seralini (2009). "Glyphosate Formulations Induce Apoptosis and Necrosis in
Human Umbilical, Embryonic, and Placental Cells." Chem Res Toxicol 22(1): 97-105.
Effects on cellular
processes
Chaufan, G., et al. (2014). "Glyphosate commercial formulation causes cytotoxicity, oxidative effects,
and apoptosis on human cells: differences with its active ingredient." Int J Toxicol 33(1): 29-38.
Effects on cellular
processes
Coalova, I., et al. (2014). "Influence of the spray adjuvant on the toxicity effects of a glyphosate
formulation." Toxicology in Vitro 28(7): 1306-1311.
Effects on cellular
processes
George, J., et al. (2010). "Studies on glyphosate-induced carcinogenicity in mouse skin: a proteomic
approach." JProteomics 73(5): 951-964.
Effects on cellular
processes
George, J. and Y. Shukla (2013). "Emptying of Intracellular Calcium Pool and Oxidative Stress
Imbalance Are Associated with the Glyphosate-induced Proliferation in Human Skin Keratinocytes
HaCaT Cells." ISRNDermatol 2013: 825180.
Effects on cellular
processes
Heu, C., et al. (2012). "A step further toward glyphosate-induced epidermal cell death: Involvement of
mitochondrial and oxidative mechanisms." Environmental Toxicology and Pharmacology 34(2): 144-
153.
Effects on cellular
processes
Thongprakaisang, S., et al. (2013). "Glyphosate induces human breast cancer cells growth via estrogen
receptors." Food Chem Toxicol 59: 129-136.
Not Relevant to current
fit for purpose review
(1934). "The 1933 meeting of the American society of orthodontists at Oklahoma City." International
Journal of Orthodontia and Dentistry for Children 20(1): 102-105.
Not Relevant to current
fit for purpose review
(1938). "Supplementary report of region III: Reports of state chairmen." The Journal of Pediatrics
12(6): 846-850.
Not Relevant to current
fit for purpose review
(1939). "Meeting of the Executive Board of the American Academy of Pediatrics." The Journal of
Pediatrics 15(2): 294-315.
Not Relevant to current
fit for purpose review
(1939). "Proceedings Meeting of the Executive Board of the American Academy of Pediatrics." The
Journal of Pediatrics 14(2): 251-254.
Not Relevant to current
fit for purpose review
(1967). "AORN Proceedings." AORN Journal 5(1): 82-91.
Not Relevant to current
fit for purpose review
(1969). "AORN Proceedings." AORN Journal 9(5): 89-96.
Not Relevant to current
fit for purpose review
(1969). "Nurse Recruitment Program—Stage I Winners." AORN Journal 10(6): 71-76.
Not Relevant to current
fit for purpose review
(1970). "2036. Excretion of heliotrine in urine and bile: Jago, Maijorie, V., Lanigan, G. W., Bingley, J.
B., Piercy, D. W. T., Whittem, J. H.& Titchen, D. A. (1969). Excretion of the pyrrolizidine alkaloid
heliotrine in the urine and bile of sheep. J. Path. 98, 115." Food and Cosmetics Toxicology 8(5): 607-
608.
Not Relevant to current
fit for purpose review
(1970). "2037. A round-up of fungal toxins: Krogh, P. (1969). The pathology of mycotoxicoses. J. stored
Prod. Res. 5, 259." Food and Cosmetics Toxicology 8(5): 608.
Not Relevant to current
fit for purpose review
(1972). "News and comment." American Journal of Orthodontics 62(3): 319-334.
Not Relevant to current
fit for purpose review
(1976). "NHI proposals: a wide spectrum." AORN Journal 23(5): 814-818.
Page 161 of 216
-------
Not Relevant to current
fit for purpose review
(1977). "The SR periodicals miscellany." Serials Review 3(2): 11-19.
Not Relevant to current
fit for purpose review
(1980). "Appendix F — administrative procedures, facilities and standardization of testing." Survey of
Ophthalmology 24, Supplement: 594-597.
Not Relevant to current
fit for purpose review
(1981). "Calendar." Annals of Emergency Medicine 10(3): 5-6.
Not Relevant to current
fit for purpose review
(1981). "Reviews of Books." The Lancet 317(8232): 1239-1240.
Not Relevant to current
fit for purpose review
(1982). "Reviews of Books." The Lancet 320(8307): 1076-1078.
Not Relevant to current
fit for purpose review
(1983). "Product news." Journal of Molecular Graphics 1(3): 89-91.
Not Relevant to current
fit for purpose review
(1984). "24-27 October 1984 International meeting on artificial intelligence: Marseilles, France."
Computer Compacts 2(3-4): 121-122.
Not Relevant to current
fit for purpose review
(1985). "Newsview." Journal of Allergy and Clinical Immunology 76(2, Part 1): A36-A41.
Not Relevant to current
fit for purpose review
(1990). "Regional direct marketing clubs/associations support DM education at local
colleges/universities." Journal of Direct Marketing 4(1): 46-52.
Not Relevant to current
fit for purpose review
(1990). "Subject index." Toxicology 61(3): 313-316.
Not Relevant to current
fit for purpose review
(1990). "Volume contents." Toxicology 61(3): 319-320.
Not Relevant to current
fit for purpose review
(1991). "4971677 Fluorescence detection type electrophoresis apparatus: Hideki Kambara, Yoshiko
Katayama, Tetsuo Nishikawa, Hachiouji, Japan assigned to Hitachi Ltd." Biotechnology Advances 9(1):
89.
Not Relevant to current
fit for purpose review
(1991). "4971903 Pyrophosphate-based method and apparatus for sequencing nucleic acids: Edward
Hyman." Biotechnology Advances 9(1): 89.
Not Relevant to current
fit for purpose review
(1991). "4971908 Glyphosate-tolerant 5-enolpyruvyl 3-phosphoshikimate synthase: Ganesh M Kishore,
Dilip Shah assigned to Monsanto Company." Biotechnology Advances 9(1): 89.
Not Relevant to current
fit for purpose review
(1991). "Author Index." Journal of Chromatography A 540: 479-482.
Not Relevant to current
fit for purpose review
(1991). "Challenge to US computer fraud & abuse act." Computer Fraud & Security Bulletin
1991(3): 5-6.
Not Relevant to current
fit for purpose review
(1991). "Chemical index for volumes 16-17." Fundamental and Applied Toxicology 17(4): 848-850.
Not Relevant to current
fit for purpose review
(1991). "Court case round-ups." Computer Fraud & Security Bulletin 1991(3): 5.
Not Relevant to current
fit for purpose review
(1991). "FAT attacker!" Computer Fraud & Security Bulletin 1991(3): 5.
Not Relevant to current
fit for purpose review
(1991). "FIGO news." International Journal of Gynecology & Obstetrics 35(3): 265-267.
Not Relevant to current
fit for purpose review
(1991). "Local direct marketing clubs/associations support undergraduate/graduate level DM education."
Journal of Direct Marketing 5(3): 61-65.
Not Relevant to current
fit for purpose review
(1991). "Manipulation of molluscan haemocytes in vitro: +S.E. Fryer. Department of Zoology, Oregon
State University, Corvallis OR 97331, USA." Developmental & Comparative Immunology 15,
Supplement 1: S73.
Not Relevant to current
fit for purpose review
(1991). "Preparation of monoclonal antibodies against hemolymph of the kuruma shrimp Penaeus
japonicus (Crustacea: Decapoda): J. Rodriguez, V. Boulo, E. Bachere and E. Mialhe. IFREMER, Unite
Pathol. Immunol. Genet. Mol., BP 133, 17390 La Tremblade, France." Developmental & Comparative
Immunology 15, Supplement 1: S73.
Not Relevant to current
fit for purpose review
(1991). "Round-up report from college/university direct marketing centers." Journal of Direct Marketing
5(2): 59-64.
Not Relevant to current
fit for purpose review
(1992). "Forthcoming papers." Soil Biology and Biochemistry 24(1): 79.
Page 162 of 216
-------
Not Relevant to current
fit for purpose review
(1992). "Subject index volume 85 (1992)." Plant Science 85(2): 235-237.
Not Relevant to current
fit for purpose review
(1993). "Media watch." Physiotherapy 79(7): 496-498.
Not Relevant to current
fit for purpose review
(1993). "Meetings and Notices." Journal of Equine Veterinary Science 13(9): 481-536.
Not Relevant to current
fit for purpose review
(1993). "Meetings and notices." Journal of Equine Veterinary Science 13(8): 432-476.
Not Relevant to current
fit for purpose review
(1993). "ROUND-UP Publications." Reproductive Health Matters 1(1): 109-110.
Not Relevant to current
fit for purpose review
(1994). "Media Watch." Physiotherapy 80(11): 793-795.
Not Relevant to current
fit for purpose review
(1994). "ROUND-UP Research." Reproductive Health Matters 2(4): 121-122.
Not Relevant to current
fit for purpose review
(1994). "ROUND-UP Service Delivery." Reproductive Health Matters 2(3): 125-128.
Not Relevant to current
fit for purpose review
(1995). "ROUND-UP Law and Policy." Reproductive Health Matters 3(6): 164-166.
Not Relevant to current
fit for purpose review
(1995). "ROUND-UP Publications." Reproductive Health Matters 3(5): 146-151.
Not Relevant to current
fit for purpose review
(1995). "ROUND-UP Service Delivery." Reproductive Health Matters 3(6): 167-169.
Not Relevant to current
fit for purpose review
(1996). "5410871 Emission control device and method: Masters Ben F; Self James M Gastonia, NC,
United States Assigned to Unlimited Technologies Inc." Environment International 22(2): XVI-XVII.
Not Relevant to current
fit for purpose review
(1996). "5411697 Method for processing contaminated plastic waste: McGraw Peter S; Drake John; Hane
Thomas H Severna Park, MD, United States Assigned to The United States of America as represented by
the Secretary of the Navy." Environment International 22(2): XVII.
Not Relevant to current
fit for purpose review
(1996). "5411944 Glyphosate-sulfuric acid adduct herbicides and use: Young Donald C Fullerton, CA,
United States Assigned to Union Oil Company of California." Environment International 22(2): XVII.
Not Relevant to current
fit for purpose review
(1996). "5412544 Method of illuminating and providing emergency egress guidance for hazardous areas:
Derrick Donald E; Harris Hollis A; Marion Robert H; Tower William A; Towle L Christophe Hanover,
NH, United States Assigned to Loctite Luminescent Systems Inc; The MTL Instruments Group."
Environment International 22(2): XVII.
Not Relevant to current
fit for purpose review
(1996). "Animal breeding and infertility: M.J. Meredith (Editor), Blackwell Science, Oxford, 1995, 508
pp., £60.00, ISBN 0-632-04038-6." Animal Reproduction Science 44(2): 135-136.
Not Relevant to current
fit for purpose review
(1996). "Preliminary program of the ninety-sixth annual session, May 11-15,1996." American Journal of
Orthodontics and Dentofacial Orthopedics 109(2): 196-214.
Not Relevant to current
fit for purpose review
(1996). "Roundup of Federal Regulations on Food and Nutrition Issues." Journal of the American Dietetic
Association 96(7): 654.
Not Relevant to current
fit for purpose review
(1996). "ROUND-UP Research." Reproductive Health Matters 4(8): 149-153.
Not Relevant to current
fit for purpose review
(1997). "ROUND-UP Conferences." Reproductive Health Matters 5(10): 180.
Not Relevant to current
fit for purpose review
(1997). "ROUND-UP Research." Reproductive Health Matters 5(10): 162-167.
Not Relevant to current
fit for purpose review
(1998). "E. coli antiserum." Journal of Equine Veterinary Science 18(8): 507.
Not Relevant to current
fit for purpose review
(1998). "EIA in wild horses." Journal of Equine Veterinary Science 18(8): 507.
Not Relevant to current
fit for purpose review
(1998). "Exercise-induced endotoxemia." Journal of Equine Veterinary Science 18(8): 506-507.
Not Relevant to current
fit for purpose review
(1998). "Flamel Technologies surles starting-blocks." Biofutur 1998(176): 41.
Not Relevant to current
fit for purpose review
(1998). "Legislative roundup: protection of human subjects key issue for Congress." J Natl Cancer Inst
90(2): 97-98.
Page 163 of 216
-------
Not Relevant to current
fit for purpose review
(1998). "ROUND-UP Conferences." Reproductive Health Matters 6(12): 186.
Not Relevant to current
fit for purpose review
(1999). "Research poster summaries from the ENA 1999 Annual Meeting." Journal of Emergency
Nursing 25(6): 446-456.
Not Relevant to current
fit for purpose review
(1999). "ROUND-UP Publications." Reproductive Health Matters 7(14): 196-201.
Not Relevant to current
fit for purpose review
(2000). "Genome roundup." Nat Biotechnol 18(1): 9.
Not Relevant to current
fit for purpose review
(2000). "ROUND-UP Law and Policy." Reproductive Health Matters 8(15): 171-174.
Not Relevant to current
fit for purpose review
(2000). "ROUND-UP Research." Reproductive Health Matters 8(16): 190-192.
Not Relevant to current
fit for purpose review
(2000). "Trauma news today." International Journal of Trauma Nursing 6(3): 105-108.
Not Relevant to current
fit for purpose review
(2001). "ROUND-UP Research." Reproductive Health Matters 9(18): 196-198.
Not Relevant to current
fit for purpose review
(2002). "ROUND-UP Research." Reproductive Health Matters 10(19): 209-210.
Not Relevant to current
fit for purpose review
(2003). "New Web site lights childhood obesity with fun." Journal of the American Dietetic Association
103(6): 671-672.
Not Relevant to current
fit for purpose review
(2003). "ROUND-UP Law and Policy." Reproductive Health Matters 11(22): 199-203.
Not Relevant to current
fit for purpose review
(2003). "ROUND-UP Research." Reproductive Health Matters 11(21): 201-204.
Not Relevant to current
fit for purpose review
(2004). "Contemporary issues in women's health." International Journal of Gynecology & Obstetrics
87(2): 111-113.
Not Relevant to current
fit for purpose review
(2004). "Contents page, CD logo." The Journal of Men's Health & Gender 1(4): 283-284.
Not Relevant to current
fit for purpose review
(2004). "Editorial Board." European Journal of Oncology Nursing 8(4): i.
Not Relevant to current
fit for purpose review
(2004). "Editorial Board." European Journal of Oncology Nursing 8(3): i.
Not Relevant to current
fit for purpose review
(2004). "Editorial Board." European Journal of Oncology Nursing 8, Supplement 1: i.
Not Relevant to current
fit for purpose review
(2004). "Editorial Board." European Journal of Oncology Nursing 8, Supplement 2: i.
Not Relevant to current
fit for purpose review
(2004). "Research round-up." Complementary Therapies in Nursing and Midwifery 10(4): 262-263.
Not Relevant to current
fit for purpose review
(2004). "ROUND-UP Condoms." Reproductive Health Matters 12(23): 176-177.
Not Relevant to current
fit for purpose review
(2004). "ROUND-UP Publications." Reproductive Health Matters 12(24): 231-236.
Not Relevant to current
fit for purpose review
(2004). "ROUND-UP Service Delivery." Reproductive Health Matters 12(23): 191-200.
Not Relevant to current
fit for purpose review
(2004). "Volume Contents Index." The Journal of Men's Health & Gender 1(4): 413-415.
Not Relevant to current
fit for purpose review
(2004). "WebWatch." The Journal of Men's Health & Gender 1(2-3): 240.
Not Relevant to current
fit for purpose review
(2004). "You've Been Had! How the Media and Environmentalists Turned America into a Nation of
Hypochondriacs: Melvin A. Benarde. Rutgers University Press, 2002. xiv + 308 pp. $28.00. ISBN 0-
8135-3050-4." Chemical Health and Safety 11(2): 39-40.
Not Relevant to current
fit for purpose review
(2005). "Contents page, CD logo." The Journal of Men's Health & Gender 2(4): 387.
Not Relevant to current
fit for purpose review
(2005). "Contents page, CD logo." The Journal of Men's Health & Gender 2(2): 159-160.
Page 164 of 216
-------
Not Relevant to current
fit for purpose review
(2005). "Editorial Board." European Journal of Oncology Nursing 9(4): i.
Not Relevant to current
fit for purpose review
(2005). "Editorial Board." European Journal of Oncology Nursing 9(3): i.
Not Relevant to current
fit for purpose review
(2005). "Editorial Board." European Journal of Oncology Nursing 9(2): i.
Not Relevant to current
fit for purpose review
(2005). "Editorial Board." European Journal of Oncology Nursing 9(1): i.
Not Relevant to current
fit for purpose review
(2005). "Journal Watch." The Journal of Men's Health & Gender 2(3): 353-356.
Not Relevant to current
fit for purpose review
(2005). "News Round-up." The Journal of Men's Health & Gender 2(3): 360-363.
Not Relevant to current
fit for purpose review
(2005). "News Round-up." The Journal of Men's Health & Gender 2(1): 116-118.
Not Relevant to current
fit for purpose review
(2005). "News Round-up for jmhg vol 2 no 4 (Dec 2005)." The Journal of Men's Health & Gender 2(4):
444.
Not Relevant to current
fit for purpose review
(2005). "Research round-up: a brief summary of research publications in CAM." Complementary
Therapies in Clinical Practice 11(2): 139-141.
Not Relevant to current
fit for purpose review
(2006). "ACFAOM newsletter." The Foot 16(4): 226-227.
Not Relevant to current
fit for purpose review
(2006). "Aims & Scope/Editorial board." The Journal of Men's Health & Gender 3(1): iii.
Not Relevant to current
fit for purpose review
(2006). "Contents page, CD logo." The Journal of Men's Health & Gender 3(2): 119-120.
Not Relevant to current
fit for purpose review
(2006). "Contents page, CD logo." The Journal of Men's Health & Gender 3(4): 315-316.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board." European Journal of Oncology Nursing 10(5): i.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board." European Journal of Oncology Nursing 10(4): i.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board." European Journal of Oncology Nursing 10(3): i.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board." European Journal of Oncology Nursing 10(2): i.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board." European Journal of Oncology Nursing 10(1): i.
Not Relevant to current
fit for purpose review
(2006). "Editorial Board/Publication/Advertising info." Progress in Biophysics and Molecular Biology
92(1): C02.
Not Relevant to current
fit for purpose review
(2006). "News Round-up." The Journal of Men's Health & Gender 3(1): 93-96.
Not Relevant to current
fit for purpose review
(2006). "News Round-up for jmhg vol 3 no 3 (September 2006)." The Journal of Men's Health & Gender
3(3): 304-306.
Not Relevant to current
fit for purpose review
(2006). "Research." Reproductive Health Matters 14(28): 210-218.
Not Relevant to current
fit for purpose review
(2006). "Research problems." Discrete Applied Mathematics 154(3): 604-607.
Not Relevant to current
fit for purpose review
(2006). "Table of Contents." The American Journal of Emergency Medicine 24(4): A3-A4.
Not Relevant to current
fit for purpose review
(2006). "Volume Contents Index: Volume 3." The Journal of Men's Health & Gender 3(4): 427-431.
Not Relevant to current
fit for purpose review
(2006). "WebWatch for jmhg vol 3 no 1 (issue 9)." The Journal of Men's Health & Gender 3(1): 88.
Not Relevant to current
fit for purpose review
(2007). "Art exhibit powered by geothermal energy." Renewable Energy Focus 8(6): 14.
Page 165 of 216
-------
Not Relevant to current
fit for purpose review
(2007). "Contents." The Journal of Men's Health & Gender 4(2): 113-114.
Not Relevant to current
fit for purpose review
(2007). "Contents page." The Journal of Men's Health & Gender 4(1): 1-2.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11, Supplement 2: ii.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11(5): i.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11(4): i.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11(3): i.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11(2): i.
Not Relevant to current
fit for purpose review
(2007). "Editorial Board." European Journal of Oncology Nursing 11(1): i.
Not Relevant to current
fit for purpose review
(2007). "Inbrief." Renewable Energy Focus 8(6): 18-19.
Not Relevant to current
fit for purpose review
(2007). "North America." Renewable Energy Focus 8(5): 6.
Not Relevant to current
fit for purpose review
(2007). "US could "eliminate C02 emissions without nuclear power"." Renewable Energy Focus 8(5):
18.
Not Relevant to current
fit for purpose review
(2008). "Editorial Board." European Journal of Oncology Nursing 12(5): i.
Not Relevant to current
fit for purpose review
(2008). "Editorial Board." European Journal of Oncology Nursing 12(1): i.
Not Relevant to current
fit for purpose review
(2008). "Editorial Board." European Journal of Oncology Nursing 12(4): i.
Not Relevant to current
fit for purpose review
(2008). "Editorial Board." European Journal of Oncology Nursing 12(3): i.
Not Relevant to current
fit for purpose review
(2008). "Editorial Board." European Journal of Oncology Nursing 12(2): i.
Not Relevant to current
fit for purpose review
(2008). "In Brief." Renewable Energy Focus 9(3): 16.
Not Relevant to current
fit for purpose review
(2008). "In brief - projects." Renewable Energy Focus 9(6): 15.
Not Relevant to current
fit for purpose review
(2008). "Swiss FiT on slippery slope?" Renewable Energy Focus 9(5): 12.
Not Relevant to current
fit for purpose review
(2008). "US wind heavyweights unite over PTC." Renewable Energy Focus 9(3): 17.
Not Relevant to current
fit for purpose review
(2008). "Wartsila BioPower plant to use brewery spent grain." Renewable Energy Focus 9(2): 12.
Not Relevant to current
fit for purpose review
(2009). "Acknowledgements." European Journal of Agronomy 31(3): iv.
Not Relevant to current
fit for purpose review
(2009). "Carbon Trust calls for switch to biomass heating." Renewable Energy Focus 10(2): 15.
Not Relevant to current
fit for purpose review
(2009). "Editorial Board." European Journal of Oncology Nursing 13(4): i.
Not Relevant to current
fit for purpose review
(2009). "Editorial Board." European Journal of Oncology Nursing 13(3): i.
Not Relevant to current
fit for purpose review
(2009). "Editorial Board." European Journal of Oncology Nursing 13(2): i.
Not Relevant to current
fit for purpose review
(2009). "Editorial Board." European Journal of Oncology Nursing 13(1): i.
Page 166 of 216
-------
Not Relevant to current
fit for purpose review
(2009). "EU renewables in short." Renewable Energy Focus 9(7): 14.
Not Relevant to current
fit for purpose review
(2009). "Pickens scraps "mega" wind plans." Renewable Energy Focus 10(5): 18.
Not Relevant to current
fit for purpose review
(2009). "Policy still needed for solar growth." Renewable Energy Focus 10(6): 16.
Not Relevant to current
fit for purpose review
(2009). "Shell withdraws from wind and solar — focuses on biofuels." Renewable Energy Focus 10(3):
19.
Not Relevant to current
fit for purpose review
(2010). "Communications orales du 27e congres de la SFE." Annales d'Endocrinologie 71(5): 340-353.
Not Relevant to current
fit for purpose review
(2010). "TenGWUS wind in 2009." Renewable Energy Focus 11(1): 13.
Not Relevant to current
fit for purpose review
(2010). "UK introduces feed-in tariffs." Renewable Energy Focus 11(1): 8.
Not Relevant to current
fit for purpose review
(2011). "€51 billion offshore wind market." Renewable Energy Focus 12(6): 12.
Not Relevant to current
fit for purpose review
(2011). "December, 2010." The Lancet Neurology 10(3): 209.
Not Relevant to current
fit for purpose review
(2011). "Directions & Connections." Journal of the American Medical Directors Association 12(8):
A14-A15.
Not Relevant to current
fit for purpose review
(2011). "Directions & Connections." Journal of the American Medical Directors Association 12(7):
A14-A15.
Not Relevant to current
fit for purpose review
(2011). "Directions & Connections." Journal of the American Medical Directors Association 12(9):
A10-A11.
Not Relevant to current
fit for purpose review
(2011). "Editorial Board." Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 78(5):
C02.
Not Relevant to current
fit for purpose review
(2011). "EU: huge renewables growth if projections met." Renewable Energy Focus 12(6):
10.
Not Relevant to current
fit for purpose review
(2011). "Table of Contents." Regulatory Toxicology and Pharmacology 61(2): iii.
Not Relevant to current
fit for purpose review
(2011). "Table of Contents." The American Journal of Emergency Medicine 29(4): A5-A7.
Not Relevant to current
fit for purpose review
(2012). "Contents." The American Journal of Medicine 125(8): A5-A7.
Not Relevant to current
fit for purpose review
(2012). "Directions & Connections." Journal of the American Medical Directors Association 13(1):
A10-A11.
Not Relevant to current
fit for purpose review
(2012). "Directions & Connections." Journal of the American Medical Directors Association 13(2):
A14-A15.
Not Relevant to current
fit for purpose review
(2012). "Impact sanitaire de la pollution environnementale et solutions pour la depollution cellulaire." La
Revue d'Homeopathie 3(3): 115-117.
Not Relevant to current
fit for purpose review
(2012). "Revue de presse." La Revue d'Homeopathie 3(4): 159-160.
Not Relevant to current
fit for purpose review
(2013). "December, 2012." The Lancet Neurology 12(3): 230.
Not Relevant to current
fit for purpose review
(2015). "60 Seconds." New Scientist 225(3014): 7.
Not Relevant to current
fit for purpose review
(2015). "60 Seconds." New Scientist 226(3022): 7.
Not Relevant to current
fit for purpose review
(2015). "GLYPHOSATE UNLIKELY TO CAUSE CANCER, EU AGENCY FINDS." Chemical &
Engineering News 93(45): 26-26.
Not Relevant to current
fit for purpose review
(2016). "EDITORIAL BOARD." NeuroToxicology 52: IFC.
Not Relevant to current
fit for purpose review
Abdullahi, A. E. (2002). "Cynodondactylon control with tillage and glypho sate." Crop Protection 21(10):
1093-1100.
Page 167 of 216
-------
Not Relevant to current
fit for purpose review
Acar, A., et al. (2015). "The Investigation of Genotoxic, Physiological and Anatomical Effects of
Paraquat Herbicide on Allium cepa L." Cytologia 80(3): 343-351.
Not Relevant to current
fit for purpose review
Acquavella, J., et al. (1999). "A case-control study of non-Hodgkin lymphoma and exposure to
pesticides." Cancer 86(4): 729-731.
Not Relevant to current
fit for purpose review
Aicher, R. H. (1999). "Vicarious Liability." Aesthetic Surgery Journal 19(4): 341.
Not Relevant to current
fit for purpose review
Akaza, H. (2007). "Report from the 1st Japanese Urological Association-Japanese Society of Medical
Oncology joint conference, 2006: 'A step towards better collaboration between urologists and medical
oncologists'." Int J Urol 14(5): 375-383.
Not Relevant to current
fit for purpose review
Akert, K. (1959). "Morphologue und physiologie des nervensystems: Paul Glees Georg Thieme, Verlag,
Stuttgart, 1957, 445 pp. $13.50." Electroencephalography and Clinical Neurophysiology 11(1): 191-192.
Not Relevant to current
fit for purpose review
Albrecht, B., et al. (1991). "In vivo utilization of N-(phosphonomethyl)-anilines and related substances
by Pseudomonas spec. GS." J Basic Microbiol 31(6): 403-411.
Not Relevant to current
fit for purpose review
Altenhoff, A. M. and C. Dessimoz (2009). "Phylogenetic and functional assessment of orthologs
inference projects and methods." PLoS Comput Biol 5(1): el000262.
Not Relevant to current
fit for purpose review
Amaro, P. (2012). "O mancozebe e um optimo exemplo da diversidade de classificacao toxicologica dos
pesticidas em Portugal." Revista de Ciencias Agrarias 35(2): 118-125.
Not Relevant to current
fit for purpose review
Andersson, K.-E. (2011). "This Month in Investigative Urology." The Journal of Urology 185(4): 1176.
Not Relevant to current
fit for purpose review
Angelini, C. (2012). "Neuromuscular diseases: advances in therapy and diagnosis." The Lancet
Neurology 11(1): 15-17.
Not Relevant to current
fit for purpose review
Anklam, E. (1999). "The validation of methods based on polymerase chain reaction for the detection of
genetically modified organisms in food." Analytica Chimica Acta 393(1-3): 177-179.
Not Relevant to current
fit for purpose review
Antonini, A. (2012). "Movement disorders: towards new therapies in Parkinson's disease." The Lancet
Neurology 11(1): 7-8.
Not Relevant to current
fit for purpose review
Aijo, G., et al. (2013). "Plurality of opinion, scientific discourse and pseudoscience: an in depth analysis
of the Seralini et al. study claiming that Roundup (TM) Ready corn or the herbicide Roundup (TM) cause
cancer in rats." Transgenic Res 22(2): 255-267.
Not Relevant to current
fit for purpose review
Astudillo, O. (2012). "Clinical trials: round-up." The Lancet Oncology 13(10): 976.
Not Relevant to current
fit for purpose review
Atkinson, S. (2003). "Product developments in biotechnology and medical applications." Membrane
Technology 2003(2): 10-11.
Not Relevant to current
fit for purpose review
Avella, P., et al. (2007). "Metric inequalities and the Network Loading Problem." Discrete Optimization
4(1): 103-114.
Not Relevant to current
fit for purpose review
Axelrad, J. C., et al. (2002). "Enhanced in vitro toxicity of the herbicide glyphosate to neuroblastoma
cells chronically pre-treated with the organophosphate pesticide diazinon." Toxicology 178(1): 62-63.
Not Relevant to current
fit for purpose review
Axelrad, J. C., et al. (2003). "The effects of acute pesticide exposure on neuroblastoma cells chronically
exposed to diazinon." Toxicology 185(1-2): 67-78.
Not Relevant to current
fit for purpose review
Azevedo, L., et al. (2010). "In Vivo Antimutagenic Properties of Transgenic and Conventional
Soybeans." J Med Food 13(6): 1402-1408.
Not Relevant to current
fit for purpose review
Bababunmi, E. A., et al. (1979). "The uncoupling effect of N-(phosphonomethyl)glycine on isolated rat
liver mitochondria." Biochemical Pharmacology 28(6): 925-927.
Not Relevant to current
fit for purpose review
Badr, A., et al. (2013). "Cytophysiological impacts of Metosulam herbicide on Vicia faba plants." Acta
Physiologiae Plantarum 35(6): 1933-1941.
Not Relevant to current
fit for purpose review
Barker, W. F. (1995). "The Western Vascular Society: Its first ten years." Journal of Vascular Surgery
22(4): 505-506.
Not Relevant to current
fit for purpose review
Baron, R., et al. (2012). "Chronic pain: genes, plasticity, and phenotypes." The Lancet Neurology 11(1):
19-21.
Not Relevant to current
fit for purpose review
Bar-Or, A. (2016). "Multiple sclerosis and related disorders: evolving pathophysiologic insights." The
Lancet Neurology 15(1): 9-11.
Not Relevant to current
fit for purpose review
Bartelink, H., et al. (1993). "Foreword." Radiotherapy and Oncology 29(2): xi-xii.
Not Relevant to current
fit for purpose review
Bates, J. A. (2004). "Use of narrative interviewing in everyday information behavior research." Library
& Information Science Research 26(1): 15-28.
Page 168 of 216
-------
Not Relevant to current
fit for purpose review
Begna, S. H., et al. (2001). "Morphology and yield response to weed pressure by corn hybrids differing
in canopy architecture." European Journal of Agronomy 14(4): 293-302.
Not Relevant to current
fit for purpose review
Benachour, N., et al. (2007). "Time- and dose-dependent effects of roundup on human embryonic and
placental cells." Arch Environ Contam Toxicol 53(1): 126-133.
Not Relevant to current
fit for purpose review
Ben-Menachem, E. (2016). "Epilepsy in 2015: the year of collaborations for big data." The Lancet
Neurology 15(1): 6-7.
Not Relevant to current
fit for purpose review
Berdal, K. G. and A. Hoist-Jensen (2001). "Roundup Ready (R) soybean event-specific real-time
quantitative PCR assay and estimation of the practical detection and quantification limits in GMO
analyses." European Food Research and Technology 213(6): 432-438.
Not Relevant to current
fit for purpose review
Berer, M. (1996). "Men." Reproductive Health Matters 4(7): 7-10.
Not Relevant to current
fit for purpose review
Berkovic, S. F. (2010). "Epilepsy: insights into causes and treatment dilemmas." The Lancet Neurology
9(1): 9-11.
Not Relevant to current
fit for purpose review
Bermel, R. A. and J. A. Cohen (2011). "Multiple sclerosis: advances in understanding pathogenesis and
emergence of oral treatment options." The Lancet Neurology 10(1): 4-5.
Not Relevant to current
fit for purpose review
Bertheussen, K., et al. (1997). "A new sensitive cell culture test for tee assessment of pesticide toxicity."
Journal of Environmental Science and Health Part B-Pesticides Food Contaminants and Agricultural
Wastes 32(2): 195-211.
Not Relevant to current
fit for purpose review
Bertucci, A., et al. (2015). "Detection of unamplified genomic DNA by a PNA-based micro structured
optical fiber (MOF) Bragg-grating optofluidic system." Biosens Bioelectron 63: 248-254.
Not Relevant to current
fit for purpose review
Bf (1995). "Let's Get Things Moving: Overcoming constipation." Physiotherapy 81(5): 301.
Not Relevant to current
fit for purpose review
Bjerrum-Bohr, J. J., et al. (2012). "Hydrodynamics of a quark droplet." Nuclear Physics A 882: 90-106.
Not Relevant to current
fit for purpose review
Blumann, W. and D. Fauth (1993). "Expensive NM standards round-up: Network Management
Standards: OSI, SNMP and CMOL Protocols by Ulyses Black. Published by McGraw Hill, 1992, £38.95,
336 pp." Computer Communications 16(7): 449.
Not Relevant to current
fit for purpose review
Bogler, O. and R. Sawaya (2008). "Foreword." Current Problems in Cancer 32(3): 95-96.
Not Relevant to current
fit for purpose review
Bonn, D. (2005). "Web round-up: rare-tumour information." The Lancet Oncology 6(3):
144.
Not Relevant to current
fit for purpose review
Botha, G. M. and C. D. Viljoen (2009). "South Africa: A case study for voluntary GM labelling." Food
Chemistry 112(4): 1060-1064.
Not Relevant to current
fit for purpose review
Bournat, P. (1997). "Les techniques de transgenese vegetale." Biofutur 1997(172): 3-7.
Not Relevant to current
fit for purpose review
Boutet, I., et al. (2004). "Molecular identification and expression of two non-P450 enzymes, monoamine
oxidase A and flavin-containing monooxygenase 2, involved in phase I of xenobiotic biotransformation
in the Pacific oyster, Crassostrea gigas." Biochim Biophys Acta 1679(1): 29-36.
Not Relevant to current
fit for purpose review
Bowersox, T. W., et al. (1990). "Coppicing success of young Eucalyptus saligna in Hawaii." Biomass
23(2): 137-148.
Not Relevant to current
fit for purpose review
Bowman, W. C. (1983). "Drugs and anesthesia—Pharmacology for anesthesiologists." Trends in
Pharmacological Sciences 4: 358-359.
Not Relevant to current
fit for purpose review
Brandenberger, L. P., et al. (2005). "Preemergence weed control in direct-seeded watermelon." Weed
Technology 19(3): 706-712.
Not Relevant to current
fit for purpose review
Breaden, A. (1991). "Victoria ASCCN." Confederation of Australian Critical Care Nurses Journal 4(3):
6.
Not Relevant to current
fit for purpose review
Briscoe, M. and C. P. Wynne (1988). "State Round-Up Tasmanian Branch of A.S.C.C.N." CNSA Journal
1(1): 3-7.
Not Relevant to current
fit for purpose review
Burkhart, C. G. (1981). "Looking at eyelid lesions~a clinical roundup." Geriatrics 36(8): 91-95.
Not Relevant to current
fit for purpose review
Burr, T. J., et al. (1995). "SURVIVAL AND TUMORIGENICITY OF AGROBACTERIUM-VITIS IN
LIVING AND DECAYING GRAPE ROOTS AND CANES IN SOIL." Plant Disease 79(7): 677-682.
Not Relevant to current
fit for purpose review
Bushby, K. (2011). "Neuromuscular diseases: milestones in development of treatments." The Lancet
Neurology 10(1): 11-13.
Page 169 of 216
-------
Not Relevant to current
fit for purpose review
Butler, T. J., et al. (2002). "Broomsedge (Andropogon virginicus) response to herbicides and burning."
Weed Technology 16(1): 18-22.
Not Relevant to current
fit for purpose review
Cai, Z., et al. (2012). "3.102 PERINATAL LPS EXPOSURE INCREASES THE RISK FOR
DOPAMINERGIC DISORDERS IN ADULT LIFE." Parkinsonism & Related Disorders 18, Supplement
2: S190.
Not Relevant to current
fit for purpose review
Caloni, F., et al. (2015). "In vitro effects of glyphosate on cell proliferation and steroid production by
bovine granulosa cells." Toxicology Letters 238(2, Supplement): S290.
Not Relevant to current
fit for purpose review
Campbell, M. H. and A. R. Gilmour (1979). "EFFECT OF TIME AND RATE OF APPLICATION OF
HERBICIDES ON SERRATED TUSSOCK (NASSELLA-TRICHOTOMA) AND IMPROVED
PASTURE SPECIES .1. GLYPHOSATE AND 2,2-DPA." Australian Journal of Experimental
Agriculture 19(99): 472-475.
Not Relevant to current
fit for purpose review
Cao, G., et al. (2013). "Draft genome sequence of pseudomonas strain p818, isolated from glyphosate-
polluted soil." Genome Announc 1(6).
Not Relevant to current
fit for purpose review
Casorri, L., et al. (2010). "Testing Round-up Ready soybean along processing chain by traditional and
innovative methods." Journal of Biotechnology 150, Supplement: 536.
Not Relevant to current
fit for purpose review
Castillo, M., et al. (2014). "Seedbed Preparation Techniques and Weed Control Strategies for Strip-
Planting Rhizoma Peanut into Warm-Season Grass Pastures." Crop Science 54(4): 1868-1875.
Not Relevant to current
fit for purpose review
Cavusoglu, K., et al. (2011). "Protective effect of Ginkgo bilobaL. leaf extract against glyphosate toxicity
in Swiss albino mice." J Med Food 14(10): 1263-1272.
Not Relevant to current
fit for purpose review
Cebeci, O. and H. Budak (2009). "Global Expression Patterns of Three Festuca Species Exposed to
Different Doses of Glyphosate Using the Affymetrix GeneChip Wheat Genome Array." Comp Funct
Genomics: 505701.
Not Relevant to current
fit for purpose review
Chang, H. S., et al. (2001). "Allergenicity test of genetically modified soybean in Sprague-Dawley rats."
Archives of Pharmacal Research 24(3): 256-261.
Not Relevant to current
fit for purpose review
Chang, J.-S., et al. (1999). "Transformation of rat fibroblasts by phospholipase C-yl overexpression is
accompanied by tyrosine dephosphorylation of paxillin." FEBS Letters 460(1): 161-165.
Not Relevant to current
fit for purpose review
Charles, A. (2011). "Headache: new genes, new mechanisms, and new therapies." The Lancet Neurology
10(1): 13-14.
Not Relevant to current
fit for purpose review
Charlton, C. G. and G. V. Muthian (2012). "3.101 PRENATAL SENSITIZATION AND POSTNATAL
PRECIPITATION IN A MODEL OF PARKINSONISM." Parkinsonism & Related Disorders 18,
Supplement 2: S189-S190.
Not Relevant to current
fit for purpose review
Chee, P. P. and J. L. Slightom (1991). "TRANSFER AND EXPRESSION OF CUCUMBER MOSAIC-
VIRUS COAT PROTEIN GENE IN THE GENOME OF CUCUMIS-SATIVUS." Journal of the
American Society for Horticultural Science 116(6): 1098-1102.
Not Relevant to current
fit for purpose review
Chen, L., et al. (2009). "Modeling Multi-typed Structurally Viewed Chemicals with the UMLS Refined
Semantic Network." Journal of the American Medical Informatics Association 16(1): 116-131.
Not Relevant to current
fit for purpose review
Chid, A. and G. Lauria (2016). "Degenerative neuromuscular diseases: crucial gene and cell machinery
discoveries." The Lancet Neurology 15(1): 12-13.
Not Relevant to current
fit for purpose review
Chorfa, A., et al. (2013). "Specific Pesticide-Dependent Increases in alpha-Synuclein Levels in Human
Neuroblastoma (SH-SY5Y) and Melanoma (SK-MEL-2) Cell Lines." Toxicological Sciences 133(2):
289-297.
Not Relevant to current
fit for purpose review
Cichosz, G. and S. K. Wiackowski (2012). "[Genetically modified food-great unknown]." Pol Merkur
Lekarski 33(194): 59-63.
Not Relevant to current
fit for purpose review
Cooke, R. M. (1997). "Protein NMR extends into new fields of structural biology." Current Opinion in
Chemical Biology 1(3): 359-364.
Not Relevant to current
fit for purpose review
Cornish-Bowden, A. and M. L. Cardenas (2003). "Metabolic analysis in drug design." Comptes Rendus
Biologies 326(5): 509-515.
Not Relevant to current
fit for purpose review
Costa, M. d. D. L. d., et al. (2003). "Altcracocs de neuroimagem no parkinsonismo: estudo de cinco
casos." Arquivos De Neuro-Psiquiatria 61(2B): 381-386.
Not Relevant to current
fit for purpose review
Costello, J. (2015). "Research roundup." Int J Palliat Nurs 21(8): 410-411.
Not Relevant to current
fit for purpose review
Coyle, C. (2003). "A Practical Guide to Intensity - Modulated Radiation Therapy." Clinical Oncology
15(8): 507.
Page 170 of 216
-------
Not Relevant to current
fit for purpose review
Croft, S. M., et al. (1974). "Inhibition of chloroplast ribosome formation by N,N-
bis(phosphonomethyl)glycine." Biochimica et Biophysica Acta (BBA) - Nucleic Acids and Protein
Synthesis 335(2): 211-217.
Not Relevant to current
fit for purpose review
Cross, J. R. (1995). "Modern Chinese Acupuncture - A review of acupuncture techniques as practised in
China today." Physiotherapy 81(5): 302.
Not Relevant to current
fit for purpose review
Cui, J., et al. (2011). "Detecting biological network organization and functional gene orthologs."
Bioinformatics 27(20): 2919-2920.
Not Relevant to current
fit for purpose review
Culbreth, M. E., et al. (2012). "Comparison of chemical-induced changes in proliferation and apoptosis
in human and mouse neuroprogenitor cells." NeuroToxicology 33(6): 1499-1510.
Not Relevant to current
fit for purpose review
Cundiff, J. S., etal. (1997). "A linear programming approach for designing a herbaceous biomass delivery
system." Bioresource Technology 59(1): 47-55.
Not Relevant to current
fit for purpose review
da Costa, M. D. L., et al. (2003). "Neuroimaging abnormalities in parkinsonism: study of five cases."
Arquivos De Neuro-Psiquiatria 61(2B): 381-386.
Not Relevant to current
fit for purpose review
D'Amico, D. B. (1978). "Critics prompt plans to revise health planning guidelines." AORN Journal 27(4):
702-706.
Not Relevant to current
fit for purpose review
Dan, B. and P. Baxter (2014). "Paediatric neurology: a year of DNA technology." The Lancet Neurology
13(1): 16-18.
Not Relevant to current
fit for purpose review
Dan, Y. H., et al. (2006). "MicroTom-a high-throughput model transformation system for functional
genomics." Plant Cell Rep 25(5): 432-441.
Not Relevant to current
fit for purpose review
Daniel, R. K., et al. (1986). "Tissue transplants in primates for upper extremity reconstruction: A
preliminary report." The Journal of Hand Surgery 11(1): 1-8.
Not Relevant to current
fit for purpose review
Das, M., et al. (2010). "A composite transcriptional signature differentiates responses towards closely
related herbicides in Arabidopsis thaliana and Brassica napus." Plant Mol Biol 72(4-5): 545-556.
Not Relevant to current
fit for purpose review
Das, M., et al. (2010). "A composite transcriptional signature differentiates responses towards closely
related herbicides in Arabidopsis thaliana and Brassica napus." Plant Mol Biol 72(4-5): 545-556.
Not Relevant to current
fit for purpose review
Datta, R. and H. Daniell (1996). "Transformation of the tobacco chloroplast genome with the arcA gene
to confer glyphosate tolerance." Plant Physiol 111(2): 790-790.
Not Relevant to current
fit for purpose review
Davis, W. M. and I. W. Waters (1999). "A new-drug roundup." Rn 62(4): 54-60; quiz 62.
Not Relevant to current
fit for purpose review
De Marco, A., et al. (1992). "Importance of the type of soil for the induction of micronuclei and the
growth of primary roots of Vicia faba treated with the herbicides atrazine, glyphosate and maleic
hydrazide." MutatRes 279(1): 9-13.
Not Relevant to current
fit for purpose review
DeAngelis, L. M. (2010). "Neuro-oncology: new hope for patients with gliomas." The Lancet Neurology
9(1): 17-18.
Not Relevant to current
fit for purpose review
Defarge, N., et al. (2016). "Co-Formulants in Glyphosate-Based Herbicides Disrupt Aromatase Activity
in Human Cells below Toxic Levels." Int J Environ Res Public Health 13(3).
Not Relevant to current
fit for purpose review
Dehghan, R., et al. (2015). "Optimization of overlapping activities in the design phase of construction
projects." Automation in Construction 59: 81-95.
Not Relevant to current
fit for purpose review
DeLuca, T. F., et al. (2006). "Roundup: a multi-genome repository of orthologs and evolutionary
distances." Bioinformatics 22(16): 2044-2046.
Not Relevant to current
fit for purpose review
Demarco, A., et al. (1995). "INFLUENCE OF SOIL CHARACTERISTICS ON THE CLASTOGENIC
ACTIVITY OF MALEIC HYDRAZIDE IN ROOT-TIPS OF VICIA-FABA." Mutation Research-
Genetic Toxicology 344(1-2): 5-12.
Not Relevant to current
fit for purpose review
Demarco, A., et al. (1992). "IMPORTANCE OF THE TYPE OF SOIL FOR THE INDUCTION OF
MICRONUCLEI AND THE GROWTH OF PRIMARY ROOTS OF VICIA-FABA TREATED WITH
THE HERBICIDES ATRAZINE, GLYPHOSATE AND MALEIC HYDRAZIDE." Mutat Res 279(1):
9-13.
Not Relevant to current
fit for purpose review
Demeke, T. and I. Ratnayaka (2008). "Multiplex qualitative PCR assay for identification of genetically
modified canola events and real-time event-specific PCR assay for quantification of the GT73 canola
event." Food Control 19(9): 893-897.
Not Relevant to current
fit for purpose review
Deming, D. (1989). "Application of bottom-hole temperature corrections in geothermal studies."
Geothermics 18(5-6): 775-786.
Not Relevant to current
fit for purpose review
Desimone, C., et al. (1992). "GENOTOXIC EFFECT INDUCED BY HERBICIDES ATRAZINE
GLYPHOSATE IN PLANTS OF VICIA-FABA GROWN IN DIFFERENT SOILS." Science of the Total
Environment 123: 233-240.
Page 171 of 216
-------
Not Relevant to current
fit for purpose review
Diener, H.-C. (2004). "Important advances in headache." The Lancet Neurology 3(1): 12.
Not Relevant to current
fit for purpose review
Dong Kyu Kim, M. D., et al. (2015). "The Factors Associated with the Hypotension Development in
Acute Glyphosate-surfactant Herbicide Poisoning." Journal of The Korean Society of Emergency
Medicine 26(3): 248-255.
Not Relevant to current
fit for purpose review
Donnan, G. A. (2004). "Stroke: prediction, prevention, and outcome." The Lancet Neurology 3(1): 9.
Not Relevant to current
fit for purpose review
Dragutinovich, S. (1987). "Stimulus intensity reducers: Are they sensation seekers, extraverts, and strong
nervous types?" Personality and Individual Differences 8(5): 693-704.
Not Relevant to current
fit for purpose review
Druwe, I., et al. (2015). "Sensitivity of neuroprogenitor cells to chemical-induced apoptosis using a
multiplexed assay suitable for high-throughput screening." Toxicology 333: 14-24.
Not Relevant to current
fit for purpose review
Duncan, K., etal. (1984). "The complete amino acid sequence of Escherichia coli 5-enolpyruvylshikimate
3-phosphate synthase." FEBS Letters 170(1): 59-63.
Not Relevant to current
fit for purpose review
Dunn, S. V. (1999). "Recognising our limits: the murky shores where science meets values and beliefs."
Australian Critical Care 12(4): 130.
Not Relevant to current
fit for purpose review
Ehlert, A., et al. (2008). "Development of a modular system for detection of genetically modified
organisms in food based on ligation-dependent probe amplification." European Food Research and
Technology 227(3): 805-812.
Not Relevant to current
fit for purpose review
Elgorashi, E. E., et al. (2004). "Isolation of captan from Cyrtanthus suaveolens: the effect of pesticides
on the quality and safety of traditional medicine." South African Journal of Botany 70(4): 512-514.
Not Relevant to current
fit for purpose review
Elie-Caille, C., et al. (2010). "Morphological damages of a glyphosate-treated human keratinocyte cell
line revealed by a micro- to nanoscale microscopic investigation." Cell Biol Toxicol 26(4): 331-339.
Not Relevant to current
fit for purpose review
Endo, M. and S. Toki (2013). "Creation of herbicide-tolerant crops by gene targeting." Journal of
Pesticide Science 38(1-2): 49-59.
Not Relevant to current
fit for purpose review
Engel, K.-H., et al. (2006). "Quantification of DNA from genetically modified organisms in composite
and processed foods." Trends in Food Science & Technology 17(9): 490-497.
Not Relevant to current
fit for purpose review
Erickson, B. (2015). "GLYPHOSATE LINKED TO CANCER." Chemical & Engineering News 93(13):
5-5.
Not Relevant to current
fit for purpose review
Ewald, M., et al. (2014). "From surface to intracellular non-invasive nanoscale study of living cells
impairments." Nanotechnology 25(29).
Not Relevant to current
fit for purpose review
Farrington, J. W. (1979). "Newsletters." Serials Review 5(2): 55-57.
Not Relevant to current
fit for purpose review
Favata, A., et al. (2014). "A nonlinear theory of prestressed elastic stick-and-spring structures."
International Journal of Engineering Science 80: 4-20.
Not Relevant to current
fit for purpose review
Feng, Z., et al. (2005). "The influence of GFP-actin expression on the adhesion dynamics of HepG2 cells
on a model extracellular matrix." Biomaterials 26(26): 5348-5358.
Not Relevant to current
fit for purpose review
Feriotto, G., et al. (2002). "Biosensor technology and surface plasmon resonance for real-time detection
of genetically modified Roundup Ready soybean gene sequences." J Agric Food Chem 50(5): 955-962.
Not Relevant to current
fit for purpose review
Ferrara, G., et al. (2000). "Evaluation of antimutagenic and desmutagenic effects of humic and fulvic
acids on root tips of Viciafaba." Environmental Toxicology 15(5): 513-517.
Not Relevant to current
fit for purpose review
Ferrari, M. D. (2013). "Headache: the changing migraine brain." The Lancet Neurology 12(1): 6-8.
Not Relevant to current
fit for purpose review
Fiers, T. (2004). "Lab-on-a-Chip. Miniaturised systems for (bio)chemical analysis and synthesis: (E.
Oosterbroek, A. van den Berg) Elsevier. ISBN 0-444 51100-8." Clinica Chimica Acta 343(1-2): 245.
Not Relevant to current
fit for purpose review
Figenschau, Y., et al. (1997). "A sensitive serum-free colorimetric assay for the detection of cytotoxic
effects of pesticides." Journal of Environmental Science and Health Part B-Pesticides Food Contaminants
and Agricultural Wastes 32(2): 177-194.
Not Relevant to current
fit for purpose review
Fischer, G., et al. (2016). "Superficial Temporal Artery to Middle Cerebral Artery Bypass via a
Minimized Approach: Operative Nuances and Problem-Solving Aspects." World Neurosurgery 88: 97-
103.
Not Relevant to current
fit for purpose review
Flora, T. and Z. Simon (1982). "Thermische zersetzung des wirkstoffes von herbizid glyphosat."
Thermochimica Acta 59(2): 125-132.
Not Relevant to current
fit for purpose review
Forgacs, A. L., et al. (2012). "BLTK1 murine Leydig cells: a novel steroidogenic model for evaluating
the effects of reproductive and developmental toxicants." Toxicol Sci 127(2): 391-402.
Page 172 of 216
-------
Not Relevant to current
fit for purpose review
Forsyth, R. (2005). "Paediatrics: genetic insights and long-term follow-up." The Lancet Neurology 4(1):
8.
Not Relevant to current
fit for purpose review
Fox, R. J. and R. A. Rudick (2004). "Multiple sclerosis: disease markers accelerate progress." The Lancet
Neurology 3(1): 10.
Not Relevant to current
fit for purpose review
Fox, S. (1999). "European Roundup - Positive news for Marimastat - British Biotech's drug shows
promise against nonoperable gastric cancer." Genetic Engineering News 19(16): 20-+.
Not Relevant to current
fit for purpose review
Fox, S. (2000). "European roundup - Genome research initiative launches in European union -
Networking, training and mobility to be focus of integrated projects." Genetic Engineering News 20(21):
41-+.
Not Relevant to current
fit for purpose review
Francisco Rossi, L., et al. (2016). "Chaetophractus villosus as a sentinel organism: Baseline values of
mitotic index, chromosome aberrations and sister chromatid exchanges." Mutation Research-Genetic
Toxicology and Environmental Mutagenesis 796: 40-45.
Not Relevant to current
fit for purpose review
Fu, G. M., et al. (2016). "Optimization of liquid-state fermentation conditions for the glyphosate-
degradation enzyme production of strain Aspergillus oryzae by ultraviolet mutagenesis." Prep Biochem
Biotechnol.
Not Relevant to current
fit for purpose review
Garry, V. F., et al. (2002). "Birth defects, season of conception, and sex of children born to pesticide
applicators living in the Red River Valley of Minnesota, USA." Environ Health Perspect 110: 441-449.
Not Relevant to current
fit for purpose review
Gasnier, C., et al. (2010). "Digl protects against cell death provoked by glyphosate-based herbicides in
human liver cell lines." J Occup Med Toxicol 5: 29.
Not Relevant to current
fit for purpose review
Gasnier, C., et al. (2009). "Glyphosate-based herbicides are toxic and endocrine disrupters in human cell
lines." Toxicology 262(3): 184-191.
Not Relevant to current
fit for purpose review
Gatty, A. (1990). "South Australian branch ASCCN." Confederation of Australian Critical Care Nurses
Journal 3(3): 10.
Not Relevant to current
fit for purpose review
Gayen, D., et al. (2013). "Comparative analysis of nutritional compositions of transgenic high iron rice
with its non-transgenic counterpart." Food Chemistry 138(2-3): 835-840.
Not Relevant to current
fit for purpose review
Giakoumaki, E., et al. (2003). "Combination of amplification and post-amplification strategies to improve
optical DNA sensing." Biosensors and Bioelectronics 19(4): 337-344.
Not Relevant to current
fit for purpose review
Giuffrida, M. C., et al. (2015). "Isothermal circular-strand-displacement polymerization of DNA and
microRNA in digital microfluidic devices." Anal Bioanal Chem 407(6): 1533-1543.
Not Relevant to current
fit for purpose review
Goldstein, C. and R. Chervin (2016). "Waking up to sleep research in 2015." The Lancet Neurology
15(1): 15-17.
Not Relevant to current
fit for purpose review
Gonzalez, F. R. and A. M. Nardillo (1997). "Retention in multistep programmed-temperature gas
chromatography and flow control Linear head pressure programs." Journal of Chromatography A 757(1-
2): 109-118.
Not Relevant to current
fit for purpose review
Gonzalez, N. V., et al. (2011). "A combination of the cytokinesis-block micronucleus cytome assay and
centromeric identification for evaluation of the genotoxicity of dicamba." Toxicology Letters 207(3):
204-212.
Not Relevant to current
fit for purpose review
Graus, F. and J. Dalmau (2012). "CNS autoimmunity: new findings and pending issues." The Lancet
Neurology 11(1): 17-19.
Not Relevant to current
fit for purpose review
Graus, F. and A. Tortosa (2006). "Neuro-oncology: setting new standards of management." The Lancet
Neurology 5(1): 8-9.
Not Relevant to current
fit for purpose review
Gregg, J. (1995). "Discipline, control, and the school mathematics tradition." Teaching and Teacher
Education 11(6): 579-593.
Not Relevant to current
fit for purpose review
Gress, S., et al. (2015). "Cardiotoxic Electrophysiological Effects of the Herbicide Roundup((R)) in Rat
and Rabbit Ventricular Myocardium In Vitro." Cardiovasc Toxicol 15(4): 324-335.
Not Relevant to current
fit for purpose review
Grosicka-Maciag. E., et al. (2012). "Dithiocarbamate fungicide zineb induces oxidative stress and
apoptosis in Chinese hamster lung fibroblasts." Pesticide Biochemistry and Physiology 102(1): 95-101.
Not Relevant to current
fit for purpose review
Gubbiga, N. G., et al. (1996). "Root/rhizome exudation of nicosulfuron from treated johnsongrass
(Sorghum halepense) and possible implications for corn (Zea mays)." Weed Science 44(3): 455-460.
Not Relevant to current
fit for purpose review
Guchi, T., et al. (2008). "Development of event-specific quantitation method for GA21 maize, which is
a GM event without CaMV35S promoter." Journal of the Food Hygienic Society of Japan 49(1): 16-22.
Not Relevant to current
fit for purpose review
Guo, B., et al. (2015). "Co-expression of G2-EPSPS and glyphosate acetyltransferase GAT genes
conferring high tolerance to glyphosate in soybean." Front Plant Sci 6: 847.
Not Relevant to current
fit for purpose review
Guyton, K. Z., et al. (2015). "Carcinogenicity of tetrachlorvinphos, parathion, malathion, diazinon, and
glyphosate." Lancet Oncology 16(5): 490-491.
Page 173 of 216
-------
Not Relevant to current
fit for purpose review
Hammond, B., et al. (2004). "Results of a 13 week safety assurance study with rats fed grain from
glyphosate tolerant corn." Food Chem Toxicol 42(6): 1003-1014.
Not Relevant to current
fit for purpose review
Hanberry, B. B., et al. (2013). "Small Mammal Responses to Intensively Established Pine Plantations in
Coastal Plain Mississippi." Southern Journal of Applied Forestry 37(1): 53-58.
Not Relevant to current
fit for purpose review
Haritou, M., et al. (2000). "Agents facilitating the electric field-induced fusion of intact rabbit
erythrocytes." Bioelectrochemistry 52(2): 229-238.
Not Relevant to current
fit for purpose review
Harkin, D. G., et al. (2000). "A novel approach to studying the migratory morphology of embryonic
mesenchymal cells." Biology of the Cell 92(7): 537-543.
Not Relevant to current
fit for purpose review
Harlow, S. D. and O. M. R. Campbell (2000). "Menstrual dysfunction: A missed opportunity for
improving reproductive health in developing countries." Reproductive Health Matters 8(15): 142-147.
Not Relevant to current
fit for purpose review
Hassannayebi, E. and S. H. Zegordi "Variable and adaptive neighbourhood search algorithms for rail
rapid transit timetabling problem." Computers & Operations Research.
Not Relevant to current
fit for purpose review
Hatamie, S., et al. (2015). "Curcumin-reduced graphene oxide sheets and their effects on human breast
cancer cells." Mater Sci Eng C Mater Biol Appl 55: 482-489.
Not Relevant to current
fit for purpose review
Haupt, S. E. (2006). "A quadratic empirical model formulation for dynamical systems using a genetic
algorithm." Computers & Mathematics with Applications 51(3-4): 431-440.
Not Relevant to current
fit for purpose review
Healy, C., et al. (2008). "Results of a 13-week safety assurance study with rats fed grain from corn
rootworm-protected, glyphosate-tolerant MON 88017 corn." Food and Chemical Toxicology 46(7):
2517-2524.
Not Relevant to current
fit for purpose review
Heck, G. R., et al. (2005). "Development and characterization of a CP4 EPSPS-Based, glyphosate-
tolerant corn event." Crop Science 45(1): 329-339.
Not Relevant to current
fit for purpose review
Henderson, N., et al. (2016). "PCR-Based Detection and Quantification of a Transgenic Glyphosate-
Tolerant Canola Using a Novel Reference Gene System." Food Analytical Methods 9(2): 353-361.
Not Relevant to current
fit for purpose review
Heraux, F. M. G., et al. (2005). "Combining Trichoderma virens-inoculated compost and a rye cover crop
for weed control in transplanted vegetables." Biological Control 34(1): 21-26.
Not Relevant to current
fit for purpose review
Hernandez, A. F., et al. (2006). "Influence of exposure to pesticides on serum components and enzyme
activities of cytotoxicity among intensive agriculture farmers." Environmental Research 102(1): 70-76.
Not Relevant to current
fit for purpose review
Herrmann, K. M. and L. M. Weaver (1999). "The shikimate pathway." Annual Review of Plant
Physiology and Plant Molecular Biology 50: 473-503.
Not Relevant to current
fit for purpose review
Heu, C., et al. (2012). "Glyphosate-induced stiffening of HaCaT keratinocytes, a Peak Force Tapping
study on living cells." J Struct Biol 178(1): 1-7.
Not Relevant to current
fit for purpose review
Hill, H. F. (1962). "Nakajima's Glaucoma Operation." American Journal of Ophthalmology 53(2): 391-
392.
Not Relevant to current
fit for purpose review
Hohzaki, R. (2006). "Search allocation game." European Journal of Operational Research 172(1): 101-
119.
Not Relevant to current
fit for purpose review
Hokanson, R., et al. (2007). "Alteration of estrogen-regulated gene expression in human cells induced by
the agricultural and horticultural herbicide glyphosate." Hum Exp Toxicol 26(9): 747-752.
Not Relevant to current
fit for purpose review
Hothorn, L. A. andR. Oberdoerfer (2006). "Statistical analysis used in the nutritional assessment of novel
food using the proof of safety." Regulatory Toxicology and Pharmacology 44(2): 125-135.
Not Relevant to current
fit for purpose review
Hour,B. T., etal. (2012). "Herbicide Roundup Intoxication: Successful Treatment with Continuous Renal
Replacement Therapy." The American Journal of Medicine 125(8): el-e2.
Not Relevant to current
fit for purpose review
Hsu, M.-H., et al. (2007). "2-(3-Fluorophenyl)-6-methoxyl-4-oxo-l,4-dihydroquinoline-3-carboxylic
acid (YJC-1) induces mitotic phase arrest in A549 cells." European Journal of Pharmacology 559(1): 14-
20.
Not Relevant to current
fit for purpose review
Hu, B.-b., et al. (2014). "Energy efficiency of massive MIMO wireless communication systems with
antenna selection." The Journal of China Universities of Posts and Telecommunications
21 (6): 1-8.
Not Relevant to current
fit for purpose review
Hu, T., et al. (2011). "Isolation and characterization of a rice glutathione S-transferase gene promoter
regulated by herbicides and hormones." Plant Cell Rep 30(4): 539-549.
Not Relevant to current
fit for purpose review
Huang, C. C., et al. (2004). "Development and application of a nested polymerase chain reaction method
for the detection of genetically modified soybean in Chinese traditional fermented soy food-sufu." Journal
of Food and Drug Analysis 12(3): 266-272.
Not Relevant to current
fit for purpose review
Huang, J. L., et al. (2016). "Enhancement of the genotoxicity of benzo[a]pyrene by arecoline through
suppression of DNA repair in HEp-2 cells." Toxicology in Vitro 33: 80-87.
Page 174 of 216
-------
Not Relevant to current
fit for purpose review
Huang, Y.-T., et al. (2014). "The depletion of securin enhances butein-induced apoptosis and tumor
inhibition in human colorectal cancer." Chemico-Biological Interactions 220: 41-50.
Not Relevant to current
fit for purpose review
Hiibner, P., et al. (1999). "Quantitative competitive PCR for the detection of genetically modified
organisms in food." Food Control 10(6): 353-358.
Not Relevant to current
fit for purpose review
Hubner, P., et al. (2001). "Validation of PCR methods for quantitation of genetically modified plants in
food." J AOAC Int 84(6): 1855-1864.
Not Relevant to current
fit for purpose review
Huisman, W., et al. (1997). "Costs of supply chains of Miscanthus giganteus." Industrial Crops and
Products 6(3-4): 353-366.
Not Relevant to current
fit for purpose review
Hultberg, M. (2007). "Cysteine turnover in human cell lines is influenced by glyphosate." Environmental
Toxicology and Pharmacology 24(1): 19-22.
Not Relevant to current
fit for purpose review
Huynh, Q. K., et al. (1988). "Site-directed mutagenesis of Petunia hybrida 5-enolpyruvylshikimate-3-
phosphate synthase: Lys-23 is essential for substrate binding." J Biol Chem 263(24): 11636-11639.
Not Relevant to current
fit for purpose review
Jackson, E. (1933). "Progressive Ophthalmology." American Journal of Ophthalmology 16(2): 154-155.
Not Relevant to current
fit for purpose review
Jacobelli, J., et al. (2004). "New views of the immunological synapse: variations in assembly and
function." Current Opinion in Immunology 16(3): 345-352.
Not Relevant to current
fit for purpose review
Jank, B. and A. Haslberger (2000). "Recombinant DNA insertion into plant retrotransposons." Trends in
Biotechnology 18(8): 326-327.
Not Relevant to current
fit for purpose review
Jankovic, J. (2008). "Parkinson's disease and movement disorders: moving forward." The Lancet
Neurology 7(1): 9-11.
Not Relevant to current
fit for purpose review
Jennings, J. C., et al. (2003). "Determining whether transgenic and endogenous plant DNA and transgenic
protein are detectable in muscle from swine fed Roundup Ready soybean meal." J Anim Sci 81(6): 1447-
1455.
Not Relevant to current
fit for purpose review
Jiang, L. X., et al. (2013). "Glyphosate effects on the gene expression of the apical bud in soybean
(Glycine max)." Biochem Biophys Res Commun 437(4): 544-549.
Not Relevant to current
fit for purpose review
Johnson, K. A. (2013). "A century of enzyme kinetic analysis, 1913 to 2013." FEBS Letters 587(17):
2753-2766.
Not Relevant to current
fit for purpose review
Jones, W. E. (1986). "Genetic recreation of wild horses." Journal of Equine Veterinary Science
6(5): 246-249.
Not Relevant to current
fit for purpose review
Kang, M., et al. (1993). "Interactive partitioning criteria set method for multiple objective linear
programming." Computers & Operations Research 20(4): 435-446.
Not Relevant to current
fit for purpose review
Kang, S. (2014). "Research round-up." The Lancet Psychiatry 1(7): 502.
Not Relevant to current
fit for purpose review
Kang, S. (2014). "Research round-up." The Lancet Psychiatry 1(4): 261.
Not Relevant to current
fit for purpose review
Kang, S. (2015). "Research round-up." The Lancet Psychiatry 2(1): 17.
Not Relevant to current
fit for purpose review
Kang, S. (2016). "Research round-up." The Lancet Psychiatry 3(4): 323.
Not Relevant to current
fit for purpose review
Kang, Y., et al. (2011). "Knockout and pullout recombineering for naturally transformable Burkholderia
thailandensis and Burkholderia pseudomallei." Nature Protocols 6(8): 1085-1104.
Not Relevant to current
fit for purpose review
Keenan, R. J., et al. (2005). "DNA shuffling as a tool for protein crystallization." Proc Natl Acad Sci U
S A 102(25): 8887-8892.
Not Relevant to current
fit for purpose review
Kelly, D. (2003). "Research Round-up." European Journal of Oncology Nursing 7(3): 210-212.
Not Relevant to current
fit for purpose review
Kiernan, M. C. (2014). "ALS and neuromuscular disease: in search of the Holy Grail." The Lancet
Neurology 13(1): 13-14.
Not Relevant to current
fit for purpose review
Kim, H. J., et al. (2007). "Engineering and characterization of the isolated C-terminal domain of 5-
enolpyruvylshikimate-3-phosphate (EPSP) synthase." Journal of Microbiology and Biotechnology 17(8):
1385-1389.
Not Relevant to current
fit for purpose review
Kim, I., et al. (2012). "Cloud computing for comparative genomics with windows azure platform." Evol
Bioinform Online 8: 527-534.
Not Relevant to current
fit for purpose review
Kim, J.-H., et al. (2015). "Detection of eight genetically modified canola events using two event-specific
pentaplexPCR systems." Food Control 51: 183-189.
Page 175 of 216
-------
Not Relevant to current
fit for purpose review
Kim, J.-H., et al. (2015). "A simplified and accurate detection of the genetically modified wheat
MON71800 with one calibrator plasmid." Food Chemistry 176: 1-6.
Not Relevant to current
fit for purpose review
Kim, Y. H., et al. (2013). "Mixtures of glyphosate and surfactant TN20 accelerate cell death via
mitochondrial damage-induced apoptosis and necrosis." Toxicol In Vitro 27(1): 191-197.
Not Relevant to current
fit for purpose review
Kim, Y. H., et al. (2014). "Heart rate-corrected QT interval predicts mortality in glyphosate-surfactant
herbicide-poisoned patients." The American Journal of Emergency Medicine 32(3): 203-207.
Not Relevant to current
fit for purpose review
Kira, J.-i. (2008). "MS: prevention of neural damage by early intervention." The Lancet Neurology 7(1):
3-5.
Not Relevant to current
fit for purpose review
Kirkham, F. (2004). "Paediatric neurology: genes and the environment." The Lancet Neurology 3(1): 18.
Not Relevant to current
fit for purpose review
Klein, E. T. (1977). "J. Lyndon Carman (1905-1977)." American Journal of Orthodontics 72(2): 213-
214.
Not Relevant to current
fit for purpose review
Knopper, L. D. and D. R. S. Lean (2004). "Carcinogenic and genotoxic potential of turf pesticides
commonly used on golf courses." Journal of Toxicology and Environmental Health-Part B-Critical
Reviews 7(4): 267-279.
Not Relevant to current
fit for purpose review
Kohn, P. M., et al. (1987). "Relationships between psychometric and experimental measures of
arousability." Personality and Individual Differences 8(2): 225-231.
Not Relevant to current
fit for purpose review
Kordowska, J., et al. (2006). "Phosphorylated 1-caldesmon is involved in disassembly of actin stress fibers
and postmitotic spreading." Experimental Cell Research 312(2): 95-110.
Not Relevant to current
fit for purpose review
Kovalic, D., et al. (2012). "The Use of Next Generation Sequencing and Junction Sequence Analysis
Bioinformatics to Achieve Molecular Characterization of Crops Improved Through Modern
Biotechnology." Plant Genome 5(3): 149-163.
Not Relevant to current
fit for purpose review
Kresge, K. J. (2005). "Women and HIV. Women's research roundup." Beta 17(4): 30-33.
Not Relevant to current
fit for purpose review
Kristensen, T. (2004). "Lethal pictures." Network Security 2004(10): 19-20.
Not Relevant to current
fit for purpose review
Kristensen, T. (2004). "Microsoft leaves Win2000, XPSP1 users in lurch." Network Security 2004(11):
16-17.
Not Relevant to current
fit for purpose review
Kristensen, T. (2005). "More holes than a phishing net." Network Security 2005(1): 18-19.
Not Relevant to current
fit for purpose review
Kudtarkar, P., et al. (2010). "Cost-effective cloud computing: a case study using the comparative
genomics tool, roundup." Evol Bioinform Online 6: 197-203.
Not Relevant to current
fit for purpose review
Kulesa, P. M., et al. (2008). "Neural crest invasion is a spatially-ordered progression into the head with
higher cell proliferation at the migratory front as revealed by the photoactivatable protein, KikGR."
Developmental Biology 316(2): 275-287.
Not Relevant to current
fit for purpose review
Kuramochi, Y., et al. (2006). "Neuregulin activates erbB2-dependent src/FAK signaling and cytoskeletal
remodeling in isolated adult rat cardiac myocytes." Journal of Molecular and Cellular Cardiology 41(2):
228-235.
Not Relevant to current
fit for purpose review
Kuribara, H., et al. (2002). "Novel reference molecules for quantitation of genetically modified maize
and soybean." J AOAC Int 85(5): 1077-1089.
Not Relevant to current
fit for purpose review
Kutateladze, T., et al. (2007). "Detection of biotechnological crops in Georgia." Journal of Biotechnology
131(2, Supplement): S37-S38.
Not Relevant to current
fit for purpose review
Kwon, K., et al. (2012). "3.103 PREDICTION AND CHARACTERIZATION OF ARABIDOPSIS
THALIANA HOMOLOGS OF PARKINSON DISEASE-ASSOCIATED GENES BASED ON
SEQUENCE SIMILARITIES AND MOLECULAR MODELING." Parkinsonism & Related Disorders
18, Supplement 2: S190.
Not Relevant to current
fit for purpose review
LaMura, M.,etal. (2011). "Application of QUIZ for GM quantification in food." Food Chemistry 125(4):
1340-1344.
Not Relevant to current
fit for purpose review
Lam, S. H. F., et al. "Use of Intravenous Fat Emulsion in the Emergency Department for the Critically
111 Poisoned Patient." The Journal of Emergency Medicine.
Not Relevant to current
fit for purpose review
Lamb, G. C., et al. (2000). "Prolonging the MGA-prostaglandin F2a interval from 17 to 19 days in an
estrus synchronization system for heifers." Theriogenology 53(3): 691-698.
Not Relevant to current
fit for purpose review
Larkin, M. (2001). "Websites in brief." The Lancet 358(9284): 850.
Page 176 of 216
-------
Not Relevant to current
fit for purpose review
Larkin, M. (2003). "Technology and Public Health." The Lancet Infectious Diseases 3(5): 314-315.
Not Relevant to current
fit for purpose review
Larkin, M. (2004). "Sites accelerate global response to HIV." The Lancet Infectious Diseases 4(7): 472-
473.
Not Relevant to current
fit for purpose review
Larsen, R. J. and D. W. Baggs (1986). "Some psychophysical and personality correlates of the Strelau
temperament inventory." Personality and Individual Differences 7(4): 561-565.
Not Relevant to current
fit for purpose review
Larsen, R. J. and M. A. Zarate (1991). "Extending reducer/augmenter theory into the emotion domain:
The role of affect in regulating stimulation level." Personality and Individual Differences 12(7): 713-723.
Not Relevant to current
fit for purpose review
Lee, A. H. (2008). "A pilot intervention for pregnant women in Sichuan, China on passive smoking."
Patient Education and Counseling 71(3): 396-401.
Not Relevant to current
fit for purpose review
Lee, C.-H., et al. (2008). "The early prognostic factors of glyphosate-surfactant intoxication." The
American Journal of Emergency Medicine 26(3): 275-281.
Not Relevant to current
fit for purpose review
Lee, D., et al. (2010). "Quantitation using informative zeros (QUIZ): Application for GMO detection and
quantification without recourse to certified reference material." Food Chemistry 118(4): 974-978.
Not Relevant to current
fit for purpose review
Lee, E.-J., et al. (2004). "G2/M cell cycle arrest and induction of apoptosis by a stilbenoid, 3,4,5-
trimethoxy-4'-bromo-cis-stilbene, in human lung cancer cells." Life Sciences 75(23): 2829-2839.
Not Relevant to current
fit for purpose review
Lee, H. J., et al. (2010). "Cloning and Characterization of a 5-Enolpyruvyl Shikimate 3-Phosphate
Synthase (EPSPS) Gene from Korean Lawn Grass (Zoysia japonica)." Korean Journal of Horticultural
Science & Technology 28(4): 648-655.
Not Relevant to current
fit for purpose review
Lee, S. R., et al. (2007). "Neuroprotective effects of kobophenol A against the withdrawal of tropic
support, nitrosative stress, and mitochondrial damage in SH-SY5Y neuroblastoma cells." Bioorganic &
Medicinal Chemistry Letters 17(7): 1879-1882.
Not Relevant to current
fit for purpose review
Lees, K. R. (2009). "Stroke: success for extending acute treatment." The Lancet Neurology 8(1): 2-4.
Not Relevant to current
fit for purpose review
Lerat, S., et al. (2005). "Real-time polymerase chain reaction quantification of the transgenes for roundup
ready corn and roundup ready soybean in soil samples." J Agric Food Chem 53(5): 1337-1342.
Not Relevant to current
fit for purpose review
Leva, A. (2015). "Computing systems and the network as a control education arena." IFAC-
PapersOnLine 48(29): 265-276.
Not Relevant to current
fit for purpose review
Levy, M. L. (2006). "What's in this issue." Primary Care Respiratory Journal 15(6): 321-
322.
Not Relevant to current
fit for purpose review
Li, G., et al. (2015). "A genetic algorithm-based decomposition approach to solve an integrated
equipment-workforce-service planning problem." Omega 50: 1-17.
Not Relevant to current
fit for purpose review
Li, H., et al. (2014). "Chaetoglobosins from Chaetomium globosum, an endophytic fungus in Ginkgo
biloba, and their phytotoxic and cytotoxic activities." J Agric Food Chem 62(17): 3734-3741.
Not Relevant to current
fit for purpose review
Li, L., et al. (2015). "A transcriptomic analysis for identifying the unintended effects of introducing a
heterologous glyphosate-tolerant EPSP synthase into Escherichia coli." Molecular Biosystems 11(3):
852-858.
Not Relevant to current
fit for purpose review
Li, R. and R. S. El-Mallahk (2000). "A novel evidence of different mechanisms of lithium and valproate
neuroprotective action on human SY5Y neuroblastoma cells: caspase-3 dependency." Neuroscience
Letters 294(3): 147-150.
Not Relevant to current
fit for purpose review
Lian, Y., et al. (2016). "Direct and simultaneous quantification of ATP, ADP and AMP by 1H and 3 IP
Nuclear Magnetic Resonance spectroscopy." Talanta 150: 485-492.
Not Relevant to current
fit for purpose review
Liang, J., et al. (2009). "The Role of Seminal Vesicle Motion in Target Margin Assessment for Online
Image-Guided Radiotherapy for Prostate Cancer." International Journal of Radiation
Oncology*Biology*Physics 73(3): 935-943.
Not Relevant to current
fit for purpose review
Liberatore, M., et al. (2010). "Forensic investigation of peer-to-peer file sharing networks." Digital
Investigation 7, Supplement: S95-S103.
Not Relevant to current
fit for purpose review
Lin, N. and V. F. Garry (2000). "In vitro studies of cellular and molecular developmental toxicity of
adjuvants, herbicides, and fungicides commonly used in Red River Valley, Minnesota." J Toxicol
Environ Health A 60(6): 423-439.
Not Relevant to current
fit for purpose review
Lin, Y.-C., et al. (2011). "744 ZOLEDRONIC ACID INDUCES AUTOPHAGIC CELL DEATH IN
HUMAN PROSTATE CANCER CELLS." The Journal of Urology 185(4, Supplement): e298-e299.
Not Relevant to current
fit for purpose review
Lin, Y.-K, et al. (2013). "Multiple-objective heuristics for scheduling unrelated parallel machines."
European Journal of Operational Research 227(2): 239-253.
Page 177 of 216
-------
Not Relevant to current
fit for purpose review
Lipton, R. B. and M. E. Bigal (2005). "Headache: triumphs in translational research." The Lancet
Neurology 4(1): 11-12.
Not Relevant to current
fit for purpose review
Liu, M. (2008). "Stroke: encouragement and disappointment in clinical trials." The Lancet Neurology
7(1): 5-7.
Not Relevant to current
fit for purpose review
Liu, S., et al. (2003). "A prospective study of the association between APOE genotype and the risk of
myocardial infarction among apparently healthy men." Atherosclerosis 166(2): 323-329.
Not Relevant to current
fit for purpose review
Lluis, M., et al. (2008). "Severe acute poisoning due to a glufosinate containing preparation without
mitochondrial involvement." Hum Exp Toxicol 27(6): 519-524.
Not Relevant to current
fit for purpose review
Lodge, G. M., et al. (1994). "EFFECTS OF GLYPHOSATE, FLUPROPANATE AND 2,2-DPA ON
HYPARRHENIA-HIRTA (L) STAPF (COOLATAI GRASS)." Australian Journal of Experimental
Agriculture 34(4): 479-485.
Not Relevant to current
fit for purpose review
Lombardo, M. and G. Lombardo (2010). "Wave aberration of human eyes and new descriptors of image
optical quality and visual performance." Journal of Cataract & Refractive Surgery 36(2): 313-331.
Not Relevant to current
fit for purpose review
Loredano, P., et al. (2010). "CASCAT: The Power of the Combined Protein Engineering Approach:
Evolution of A Glycine Oxidase for A Novel Mechanism of Glyphosate Tolerance." Journal of
Biotechnology 150: S122-S123.
Not Relevant to current
fit for purpose review
Lucentini, J. (2008). "Technology roundup." Scientist: 76-82.
Not Relevant to current
fit for purpose review
Machado, A. F. L., et al. (2008). "Anatomical characterization of the leaf, stem and rhizome of Digitaria
insularis." PlantaDaninha26(l): 1-8.
Not Relevant to current
fit for purpose review
Mahmood, A. (2009). "Renewables in Asia - a roundup." Renewable Energy Focus 9(7): 19.
Not Relevant to current
fit for purpose review
Main, C. L., et al. (2004). "Sulfentrazone persistence in southern soils: Bioavailable concentration and
effect on a rotational cotton crop." Weed Technology 18(2): 346-352.
Not Relevant to current
fit for purpose review
Makowski, M., et al. (2009). "Oestrogen receptor alpha polymorphisms in women with breast cancer and
their clinical significance." Przeglad Menopauzalny 8(1): 40-44.
Not Relevant to current
fit for purpose review
Malatesta, M., et al. (2008). "Hepatoma tissue culture (HTC) cells as a model for investigating the effects
of low concentrations of herbicide on cell structure and function." Toxicol In Vitro 22(8): 1853-1860.
Not Relevant to current
fit for purpose review
Malecot, M., et al. (2013). "Specific proteomic response of Unio pictorum mussel to a mixture of
glyphosate and microcystin-LR." JProteomeRes 12(11): 5281-5292.
Not Relevant to current
fit for purpose review
Malhotra, R. C., et al. (2010). "Glyphosate-surfactant herbicide-induced reversible encephalopathy."
Journal of Clinical Neuroscience 17(11): 1472-1473.
Not Relevant to current
fit for purpose review
Mariana, A., et al. (2009). "The impact of simultaneous intoxication with agrochemicals on the
antioxidant defense system in rat." Pesticide Biochemistry and Physiology 94(2-3): 93-99.
Not Relevant to current
fit for purpose review
Marques, M. R., et al. (2007). "The inhibition of 5-enolpyruvylshikimate-3-phosphate synthase as a
model for development of novel antimicrobials." Current Drug Targets 8(3): 445-457.
Not Relevant to current
fit for purpose review
Marrelli, M., et al. (2013). "A comparative study of phytochemical composition of genetically and non-
genetically modified soybean (Glycine max L.) and evaluation of antitumor activity." Nat Prod Res 27(6):
574-578.
Not Relevant to current
fit for purpose review
Marriott, G. and J. W. Walker (1999). "Caged peptides and proteins: new probes to study polypeptide
function in complex biological systems." Trends in Plant Science 4(8): 330-334.
Not Relevant to current
fit for purpose review
Mastaglia, F. L. (2005). "Neuromuscular disorders: molecular and therapeutic insights." The Lancet
Neurology 4(1): 6-7.
Not Relevant to current
fit for purpose review
Mavunganidze, Z., et al. (2014). "The impact of tillage system and herbicides on weed density, diversity
and yield of cotton (Gossipium hirsutum L.) and maize (Zea mays L.) under the smallholder sector." Crop
Protection 58: 25-32.
Not Relevant to current
fit for purpose review
Mayeux, R. and P. S. G. Hyslop (2008). "Alzheimer's disease: advances in trafficking." The Lancet
Neurology 7(1): 2-3.
Not Relevant to current
fit for purpose review
McAlister, B. G. and J. van Staden (1995). "Effect of artificially induced stress conditions on the growth
of the medicinal plant Hypoxis hemerocallidea." South African Journal of Botany 61(2): 85-89.
Not Relevant to current
fit for purpose review
McCleneghan, J. S. (1994). "The 1993 newspaper science reporter: Contributing, creative, and
responsible." The Social Science Journal 31(4): 467-477.
Not Relevant to current
fit for purpose review
McDowell, L. M., et al. (2004). "Rotational-echo double-resonance NMR-restrained model of the ternary
complex of 5-enolpyruvylshikimate-3-phosphate synthase." J Biomol NMR 28(1): 11-29.
Page 178 of 216
-------
Not Relevant to current
fit for purpose review
McKeone, D. H. (1973). "Liquid crystals for display applications." Optics & Laser Technology 5(1): 39-
40.
Not Relevant to current
fit for purpose review
Meacham, M. (1976). "Development of school libraries around the world." International Library Review
8(4): 453-459.
Not Relevant to current
fit for purpose review
Merlis, J. K. (1959). "Fundamentals of clinical neurophysiology: Paul 0. Chatfield Charles C. Thomas,
Springfield, Illinois, 1957, 392 pp. $8.50." Electroencephalography and Clinical Neurophysiology 11(1):
192.
Not Relevant to current
fit for purpose review
Mesnage, R., et al. (2013). "Ethoxylated adjuvants of glyphosate-based herbicides are active principles
of human cell toxicity." Toxicology 313(2-3): 122-128.
Not Relevant to current
fit for purpose review
Milan, M. and S. M. Cohen (1999). "Notch Signaling Is Not Sufficient to Define the Affinity Boundary
between Dorsal and Ventral Compartments." Molecular Cell 4(6): 1073-1078.
Not Relevant to current
fit for purpose review
Miller, J. W., et al. (1991). "Phthalocyanine Photodynamic Therapy of Experimental Iris
Neovascularization." Ophthalmology 98(11): 1711-1719.
Not Relevant to current
fit for purpose review
Miller, S. P. and D. M. Ferriero (2013). "Paediatric neurology: improved care of the developing brain."
The Lancet Neurology 12(1): 16-18.
Not Relevant to current
fit for purpose review
Milligan, A. L., et al. (2003). "A field assessment of the role of selective herbicides in the restoration of
British moorland dominated by Molinia." Biological Conservation 109(3): 369-379.
Not Relevant to current
fit for purpose review
Mink, P. J., et al. (2011). "Epidemiologic studies of glyphosate and non-cancer health outcomes: a
review." Regul Toxicol Pharmacol 61(2): 172-184.
Not Relevant to current
fit for purpose review
Minshull, J., et al. (2005). "Predicting enzyme function from protein sequence." Current Opinion in
Chemical Biology 9(2): 202-209.
Not Relevant to current
fit for purpose review
Mo, Q. and X. Zhuang (2012). "Matrix splitting with symmetry and dyadic framelet filter banks over
algebraic number fields." Linear Algebra and its Applications 437(10): 2650-2679.
Not Relevant to current
fit for purpose review
Monsen, E. R. (1998). "ADA'S Nutrition Campaigns Link Dietitians and the Public's Health." Journal of
the American Dietetic Association 98(7): 746.
Not Relevant to current
fit for purpose review
Monsen, E. R. (1998). "Dietetics Professionals Fight Obesity." Journal of the American Dietetic
Association 98(10): 1098.
Not Relevant to current
fit for purpose review
Monsen, E. R. (1999). "The Past Dietetics Century and the Future Dietetics Century." Journal of the
American Dietetic Association 99(12): 1498.
Not Relevant to current
fit for purpose review
Monsen, E. R. (1999). "What's Fresh in this Month's Journal?" Journal of the American Dietetic
Association 99(3): 274.
Not Relevant to current
fit for purpose review
Moreano, F., et al. (2006). "Ligation-dependent probe amplification for the simultaneous event-specific
detection and relative quantification of DNA from two genetically modified organisms." European Food
Research and Technology 222(5-6): 479-485.
Not Relevant to current
fit for purpose review
Morgensztern, D. and R. Govindan (2010). "Best of the month: a roundup of articles published in recent
months." J Thorac Oncol 5(8): 1305-1307.
Not Relevant to current
fit for purpose review
Morgensztern, D. and R. Govindan (2011). "A Roundup of Recently Published Articles Relevant to
Thoracic Oncology." Journal of Thoracic Oncology 6(7): 1295-1297.
Not Relevant to current
fit for purpose review
Morgensztern, D. and R. Govindan (2011). "A roundup of recently published articles relevant to thoracic
oncology." J Thorac Oncol 6(7): 1295-1297.
Not Relevant to current
fit for purpose review
Morris, R. T., et al. (2011). "RiceRBP: A database of experimentally identified RNA-binding proteins in
Oryza sativaL." Plant Science 180(2): 204-211.
Not Relevant to current
fit for purpose review
Muntoni, F. and J. H. Cross (2015). "Paediatric neurology: from molecular mechanisms to targeted
treatments." The Lancet Neurology 14(1): 16-18.
Not Relevant to current
fit for purpose review
Myllynen, P. and K. Vahakangas (2013). "Placental transfer and metabolism: An overview of the
experimental models utilizing human placental tissue." Toxicology in Vitro 27(1): 507-512.
Not Relevant to current
fit for purpose review
Nam, H.-J., et al. (2008). "The ERK-RSK1 activation by growth factors at G2 phase delays cell cycle
progression and reduces mitotic aberrations." Cellular Signalling 20(7): 1349-1358.
Not Relevant to current
fit for purpose review
Naydenova, E., et al. (2007). "Synthesis, cytotoxicity and clastogenicity of novel alpha-aminophosphonic
acids." Amino Acids 33(4): 695-702.
Not Relevant to current
fit for purpose review
Naydenova, E. D., et al. (2008). "Novel N-(phosphonomethyl) glycine derivatives: Design,
characterization and biological activity." Eur J Med Chem 43(6): 1199-1205.
Not Relevant to current
fit for purpose review
Neumann, G., et al. (2011). "Assessment of the genetic stability of GMOs with a detailed examination of
MON810 using Scorpion probes." European Food Research and Technology 233(1): 19-30.
Page 179 of 216
-------
Not Relevant to current
fit for purpose review
Newton, F. H. (1962). "Appraising New Techniques." American Journal of Ophthalmology 53(2): 392.
Not Relevant to current
fit for purpose review
Norrving, B. (2006). "Stroke: the catch-up of underinvestigated topics." The Lancet Neurology 5(1): 10-
11.
Not Relevant to current
fit for purpose review
Novis, P. M., et al. (2009). "Identification and Characterization of Freshwater Algae from a Pollution
Gradient Using rbcL Sequencing and Toxicity Testing." Arch Environ Contam Toxicol 57(3): 504-514.
Not Relevant to current
fit for purpose review
Ocana, K. A. C. S., et al. (2013). "Designing a parallel cloud based comparative genomics workflow to
improve phylogenetic analyses." Future Generation Computer Systems 29(8): 2205-2219.
Not Relevant to current
fit for purpose review
Oeding, L. and G. Ottaviani (2013). "Eigenvectors of tensors and algorithms for Waring decomposition."
Journal of Symbolic Computation 54: 9-35.
Not Relevant to current
fit for purpose review
Oguchi, T., et al. (2008). "Development of event-specific quantitation method for GA21 maize, which is
a gm event without CaMV35S promoter." Shokuhin Eiseigaku Zasshi 49(1): 16-22.
Not Relevant to current
fit for purpose review
Okamoto, T. (2015). "SecondDEP: Resilient Computing that Prevents Shellcode Execution in Cyber-
Attacks." Procedia Computer Science 60: 691-699.
Not Relevant to current
fit for purpose review
Olanow, C. W. (2006). "Movement disorders: a step in the right direction." The Lancet Neurology 5(1):
3-5.
Not Relevant to current
fit for purpose review
Omrani, D. (1993). "Doesn't quite live up to expectations: Enterprise-Wide Computing: How to
Implement and Manage LANs by Thomas W. Madron. Published by John Wiley, 1991, £25.95, 372 pp."
Computer Communications 16(7): 449-450.
Not Relevant to current
fit for purpose review
Onori, R., et al. (2013). "Traceability of genetically modified Roundup Ready soybean: A case study on
sampling and analytical uncertainty along processing chain." Food Control 34(2): 494-501.
Not Relevant to current
fit for purpose review
O'Rourke, K. (2004). "Genome roundup." J Am Vet Med Assoc 224(4): 491-492.
Not Relevant to current
fit for purpose review
Ortiz, S., et al. (2006). "Acute cytotoxicity in mammal cells exposed in vitro to a glyphosate-based
formulation." EnvironMol Mutagen 47(6): 473-473.
Not Relevant to current
fit for purpose review
Park, S., et al. (2015). "The effects of nonyl phenoxypolyethoxyl ethanol on cell damage pathway gene
expression in SK-N-SH cells." Korean Journal of Internal Medicine 30(6): 873-883.
Not Relevant to current
fit for purpose review
Pasqualone, A., et al. (2007). "Detection of soft wheat in semolina and durum wheat bread by analysis of
DNA micro satellites." J Agric Food Chem 55(9): 3312-3318.
Not Relevant to current
fit for purpose review
Patterson, P. (1980). "Nurses hold own in federal battle for education funds." AORN Journal 32(1): 116-
118.
Not Relevant to current
fit for purpose review
Patton, M. L. and R. F. Grove (1992). "Statolith hair movements and the regulation of tonic gravity
reflexes in the lobster, Homarus americanus." Comparative Biochemistry and Physiology Part A:
Physiology 101(2): 259-268.
Not Relevant to current
fit for purpose review
Perisse, P., et al. (2011). "Seed and seedling morphology of Dicliptera squarrosa Nees (Acanthaceae) as
a character identification source, and its relationship with survival structures." Phyton-International
Journal of Experimental Botany 80: 73-78.
Not Relevant to current
fit for purpose review
Perret, R. (2012). "Wanted dead or alive? Western genre items in the 21st century United States library."
Library Collections, Acquisitions, and Technical Services 36(1-2): 39-52.
Not Relevant to current
fit for purpose review
Perumal, L., et al. (2016). "Analysis of thin plates with holes by using exact geometrical representation
within XFEM." Journal of Advanced Research 7(3): 445-452.
Not Relevant to current
fit for purpose review
Pervaiz, S. (2013). "Conference roundup MAC 2011." Mitochondrion 13(3): 153-154.
Not Relevant to current
fit for purpose review
Petersen, R. C. (2010). "Alzheimer's disease: progress in prediction." The Lancet Neurology 9(1): 4-5.
Not Relevant to current
fit for purpose review
Poltl, D., et al. (2012). "3.104 CELL-DEATH PATHWAYS DOWNSTREAM OF MITOCHONDRIA
IN HUMAN DOPAMINERGIC NEURON DEGENERATION." Parkinsonism & Related Disorders 18,
Supplement 2: SI90.
Not Relevant to current
fit for purpose review
Posch, K. C. and R. Posch (1990). "Approaching encryption at ISDN speed using partial parallel modulus
multiplication." Microprocessing and Microprogramming 29(3): 177-184.
Not Relevant to current
fit for purpose review
Potrebic, O., et al. (2009). "Acute glyphosate-surfactant poisoning with neurological sequels and fatal
outcome." Vojnosanitetski Pregled 66(9): 758-762.
Not Relevant to current
fit for purpose review
Powell, S. L. (2000). "Herbal Supplements: What Works, What Doesn't. 1999. From National Health
Video, Inc. 12021 Wilshire Blvd., Suite 550, Los Angeles, CA 90025, 1-800-543-6803, videotape, 15
minutes, $89.00." Journal of Nutrition Education 32(5): 292-293.
Page 180 of 216
-------
Not Relevant to current
fit for purpose review
Printz, C. (2013). "A roundup of news and information from our community." Clin Transl Sci 6(2): 83-
84.
Not Relevant to current
fit for purpose review
Printz, C. (2014). "In the news: a roundup of news and information from our community." Clin Transl
Sci 7(4): 287-288.
Not Relevant to current
fit for purpose review
Printz, C. (2015). "In the news: a roundup of news and information from our community." Clin Transl
Sci 8(1): 1-2.
Not Relevant to current
fit for purpose review
Qin, F., et al. (2012). "Composition of Transgenic Soybean Seeds with Higher gamma-Linolenic Acid
Content Is Equivalent to That of Conventional Control." J Agric Food Chem 60(9): 2200-2204.
Not Relevant to current
fit for purpose review
Radenski, A. and L. Ehwerhemuepha (2014). "Speeding-up codon analysis on the cloud with local
MapReduce aggregation." Information Sciences 263: 175-185.
Not Relevant to current
fit for purpose review
Radtke, P. J., et al. (2010). "An evaluation of overhead laser scanning to estimate herbage removals in
pasture quadrats." Agricultural and Forest Meteorology 150(12): 1523-1528.
Not Relevant to current
fit for purpose review
Rao, C. D., et al. (2015). "An enzyme-linked immuno focus assay for rapid detection and enumeration,
and a newborn mouse model for human non-polio enteroviruses associated with acute diarrhea." Journal
of Virological Methods 224: 47-52.
Not Relevant to current
fit for purpose review
Rappold, P., et al. (2012). "3.106 PARAQUAT NEUROTOXICITY IS MEDIATED BY THE
DOPAMINE TRANSPORTER AND ORGANIC CATION TRANSPORTER-3." Parkinsonism &
Related Disorders 18, Supplement 2: S190-S191.
Not Relevant to current
fit for purpose review
Rau, F. J. (1995). "Video review roundup." Adolescent and Pediatric Gynecology 8(3): 165.
Not Relevant to current
fit for purpose review
Rau, F. J. (1997). "Comprehensive Review of Colposcopy 1996." Journal of Pediatric and Adolescent
Gynecology 10(1): 49.
Not Relevant to current
fit for purpose review
Rau, F. J. and K. M. Mulchahey (1997). "Milner-Fenwick Video Counseling Library: Colposcopy and
Treatment of Dysplasia." Journal of Pediatric and Adolescent Gynecology 10(2): 103-104.
Not Relevant to current
fit for purpose review
Rau, F. J. and E. E. Yordan (1996). "Sexual Assault: The Collection of Evidence for Forensic Analysis."
Journal of Pediatric and Adolescent Gynecology 9(3): 169.
Not Relevant to current
fit for purpose review
Reeve, J. C. (1995). "Core Textbook of Respiratory Care Practice." Physiotherapy 81(5): 301.
Not Relevant to current
fit for purpose review
Reis, L. F., et al. (2006). "GMOs: building the future on the basis of past experience." An Acad Bras
Cienc 78(4): 667-686.
Not Relevant to current
fit for purpose review
Richard, S., et al. (2005). "Differential effects of glyphosate and roundup on human placental cells and
aromatase." Environ Health Perspect 113(6): 716-720.
Not Relevant to current
fit for purpose review
Richter, L. (1979). "Summary and guideline to the status reports." Euromicro Newsletter 5(3): 111-112.
Not Relevant to current
fit for purpose review
Ring, M. (2007). "Driven Out by Hitler, a Dental Historian Enriches America: The Story of Curt
Proskauer." Alpha Omegan 100(1): 25-29.
Not Relevant to current
fit for purpose review
Robinson, D. J. (1995). "Study Package." Physiotherapy 81(5): 302.
Not Relevant to current
fit for purpose review
Rogoli, R. P., et al. (2008). "Resposta de plantas de beterraba (Beta vulgaris) e de cenoura (Daucus carota)
a deriva simulada de glyphosate e clomazone." Planta Daninha 26(2): 451-456.
Not Relevant to current
fit for purpose review
Rosculete, E., et al. (2015). "Detection and quantification of genetically modified soybean in food and
feed traded in the Romanian market." Journal of Biotechnology 208, Supplement: S73.
Not Relevant to current
fit for purpose review
Roskelley, C. D., et al. (1995). "A hierarchy of ECM-mediated signalling regulates tissue-specific gene
expression." Current Opinion in Cell Biology 7(5): 736-747.
Not Relevant to current
fit for purpose review
Rouquie, D., et al. (2010). "Investigation of endogenous soybean food allergens by using a 2-dimensional
gel electrophoresis approach." Regulatory Toxicology and Pharmacology 58(3, Supplement): S47-S53.
Not Relevant to current
fit for purpose review
Ryd, W. and B. Hagmar (1977). "In vitro Effects of Cytochalasin B on TA3 Tumor Cells." Beitrage zur
Pathologie 161(2): 131-141.
Not Relevant to current
fit for purpose review
Sakamoto, Y., et al. (2008). "A 104-week feeding study of genetically modified soybeans in F344 rats."
Journal of the Food Hygienic Society of Japan 49(4): 272-282.
Not Relevant to current
fit for purpose review
Sakr, J., et al. (2014). "First comprehensive GMOs testing in Lebanon: Screening, identification and
quantification of GM soybean imports." Food Control 36(1): 146-152.
Not Relevant to current
fit for purpose review
Sanfilippo, J. S. (1998). ""Get A Life"-My Old Kentucky Home." Journal of Pediatric and Adolescent
Gynecology 11(1): 1.
Page 181 of 216
-------
Not Relevant to current
fit for purpose review
Sarraj, A. and J. C. Grotta (2014). "Stroke: new horizons in treatment." The Lancet Neurology 13(1): 2-
3.
Not Relevant to current
fit for purpose review
Sauer, N. J., et al. (2016). "Oligonucleotide-mediated genome editing provides precision and function to
engineered nucleases and antibiotics in plants." Plant Physiol.
Not Relevant to current
fit for purpose review
Sbruzzi, F. A., et al. (2013). "Transgenic and conventional Brazilian soybeans don't cause or prevent
preneoplastic colon lesions or oxidative stress in a 90-day in vivo study." Revista De Nutricao-Brazilian
Journal of Nutrition 26(4): 443-453.
Not Relevant to current
fit for purpose review
Schapira, A. H. V. (2010). "Movement disorders: advances in cause and treatment." The Lancet
Neurology 9(1): 6-7.
Not Relevant to current
fit for purpose review
Scherlen, A. (2008). "Local to Global: The Importance of State-Level Journals to Library Literature."
Serials Review 34(2): 129-136.
Not Relevant to current
fit for purpose review
Schoenen, J. and G. Coppola (2010). "Headache: spreading from molecules to patients." The Lancet
Neurology 9(1): 11-12.
Not Relevant to current
fit for purpose review
Scotter, E. L. and C. E. Shaw (2013). "Neuromuscular disease: new insights and avenues for therapy."
The Lancet Neurology 12(1): 13-15.
Not Relevant to current
fit for purpose review
Selvapandiyan, A., et al. (1996). "Evidence for the shikimate-3-phosphate interacting site in the N-
terminal domain of 5-enolpyruvyl shikimate-3-phosphate synthase of Bacillus subtilis." Biochemistry
and Molecular Biology International 40(3): 603-610.
Not Relevant to current
fit for purpose review
Selvapandiyan, A. and R. K. Bhatnagar (1994). "CLONING OF GENES ENCODING FOR C-P LYASE
FROM PSEUDOMONAS ISOLATES PG2982 AND GLC11 - IDENTIFICATION OF A CRYPTIC
ALLELE ON THE CHROMOSOME OF PSEUDOMONAS-AERUGINOSA." Current Microbiology
29(5): 255-261.
Not Relevant to current
fit for purpose review
Selvapandiyan, A., et al. (1995). "Point mutation of a conserved arginine (104) to lysine introduces
hypersensitivity to inhibition by glyphosate in the 5-enolpyruvylshikimate-3-phosphate synthase of
Bacillus subtilis." FEBS Lett 374(2): 253-256.
Not Relevant to current
fit for purpose review
Shi, G., et al. (2011). "MultiMSOAR 2.0: an accurate tool to identify ortholog groups among multiple
genomes." PLoS One 6(6): e20892.
Not Relevant to current
fit for purpose review
Shields, R. and K. Pollock (1974). "The adhesion of BHK and PyBHK cells to the substratum." Cell 3(1):
31-38.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A. and J. N. Evans (1994). "Site-directed mutagenesis and NMR studies of histidine-
385 mutants of 5-enolpyruvylshikimate-3-phosphate synthase." Biochemistry 33(23): 7062-7068.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A. and J. N. Evans (1996). "The H385N mutant of 5-enolpyruvylshikimate-3-phosphate
synthase: kinetics, fluorescence, and nuclear magnetic resonance studies." Arch Biochem Biophys
334(1): 37-42.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A. and J. N. S. Evans (1994). "SITE-DIRECTED MUTAGENESIS AND NMR-
STUDIES OF HISTIDINE-385 MUTANTS OF 5-ENOLPYRUVYLSHIKIMATE-3-PHOSPHATE
SYNTHASE." Biochemistry 33(23): 7062-7068.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A. and J. N. S. Evans (1996). "The H385N Mutant of 5-Enolpyruvylshikimate-3-
phosphate Synthase: Kinetics, Fluorescence, and Nuclear Magnetic Resonance Studies." Archives of
Biochemistry and Biophysics 334(1): 37-42.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A., et al. (1992). "Over-production of 5-enolpyruvylshikimate-3-phosphate synthase in
Escherichia coli: use of the T7 promoter." Protein Eng 5(5): 461-466.
Not Relevant to current
fit for purpose review
Shuttleworth, W. A., et al. (1999). "Site-directed mutagenesis of putative active site residues of 5-
enolpyruvylshikimate-3-phosphate synthase." Biochemistry 38(1): 296-302.
Not Relevant to current
fit for purpose review
Sienkiewicz, Z. (2007). "Rapporteur report: Roundup, discussion and recommendations." Progress in
Biophysics and Molecular Biology 93(1-3): 414-420.
Not Relevant to current
fit for purpose review
Sikorski, J. A. andK. J. Grays (1997). "Understanding Glyphosate's molecular mode of action with EPSP
synthase: Evidence favoring an allosteric inhibitor model." Accounts of Chemical Research 30(1): 2-8.
Not Relevant to current
fit for purpose review
Simonetti, E., et al. (2015). "An Interlaboratory Comparative Study on the Quantitative Determination of
Glyphosate at Low Levels in Wheat Flour." J AO AC Int 98(6): 1760-1768.
Not Relevant to current
fit for purpose review
Singh, N. and A. Srivastava (2014). "Biomonitoring of Genotoxic Effect of Glyphosate and
Pendimethalin in Vigna mungo Populations." Cytologia 79(2): 173-180.
Not Relevant to current
fit for purpose review
Singh, S. K. and S. Kumar (2011). "Novel adaptive color space transform and application to image
compression." Signal Processing: Image Communication 26(10): 662-672.
Page 182 of 216
-------
Not Relevant to current
fit for purpose review
Sinha, R. P., et al. (2003). "Wavelength-dependent induction of a mycosporine-like amino acid in a rice-
field cyanobacterium, Nostoc commune: role of inhibitors and salt stress." Photochemical &
Photobiological Sciences 2(2): 171-176.
Not Relevant to current
fit for purpose review
Sliseris, J., et al. (2013). "Optimal Design of GFRP-Plywood Variable Stiffness Plate." Procedia
Engineering 57: 1060-1069.
Not Relevant to current
fit for purpose review
Smith, S. O. (1993). "Magic angle spinning NMR methods for internuclear distance measurements."
Current Opinion in Structural Biology 3(5): 755-759.
Not Relevant to current
fit for purpose review
Smitt, P. S. (2004). "Neuro-oncology: diagnosis in the spotlight." The Lancet Neurology 3(1): 14.
Not Relevant to current
fit for purpose review
Somasiri, A., et al. (2000). "Phosphatidylinositol 3-Kinase is required for adherens junction-dependent
mammary epithelial cell spheroid formation." Differentiation 66(2-3): 116-125.
Not Relevant to current
fit for purpose review
Sorensen, P. S. (2005). "Multiple sclerosis: pathophysiology revisited." The Lancet Neurology 4(1): 9-
10.
Not Relevant to current
fit for purpose review
Spaeth, S. C. and T. R. Sinclair (1983). "Variation in nitrogen accumulation and distribution among
soybean cultivars." Field Crops Research 7: 1-12.
Not Relevant to current
fit for purpose review
Spector, L. S. (1990). "Not-so-open skies." Space Policy 6(1): 9-18.
Not Relevant to current
fit for purpose review
Sperling, R. A. and K. A. Johnson (2012). "Dementia: new criteria but no new treatments." The Lancet
Neurology 11(1): 4-5.
Not Relevant to current
fit for purpose review
Sroka, J., et al. (2002). "Phenotype modulation in non-adherent and adherent sublines of Walker
carcinosarcoma cells: the role of cell-substratum contacts and microtubules in controlling cell shape,
locomotion and cytoskeletal structure." The International Journal of Biochemistry & Cell Biology 34(7):
882-899.
Not Relevant to current
fit for purpose review
Stanton, T. L., et al. (1990). "The Effects of Wet on Growing Heifer Brewers Grain Performance." The
Professional Animal Scientist 6(2): 31-34.
Not Relevant to current
fit for purpose review
Stauffer, M. E., et al. (2001). "Shikimate-3-phosphate binds to the isolated N-terminal domain of 5-
enolpyruvylshikimate-3-phosphate synthase." Biochemistry 40(13): 3951-3957.
Not Relevant to current
fit for purpose review
Stobiecka, M., et al. (2007). "Piezoelectric sensor for determination of genetically modified soybean
roundup ready((R)) in samples not amplified by PCR." Sensors 7(8): 1462-1479.
Not Relevant to current
fit for purpose review
Stoessl, A. J. (2011). "Movement disorders: new insights into Parkinson's disease." The Lancet
Neurology 10(1): 5-7.
Not Relevant to current
fit for purpose review
Stovner, L. J. (2008). "Headache: new concepts, models, and treatments." The Lancet Neurology 7(1):
11-12.
Not Relevant to current
fit for purpose review
Struyker-Boudier, H. A. J. (1983). "Clinical hypertension and hypotension." Trends in Pharmacological
Sciences 4: 359.
Not Relevant to current
fit for purpose review
Subramanian, J., et al. (2012). "A roundup of articles published in recent months." J Thorac Oncol 7(3):
626-628.
Not Relevant to current
fit for purpose review
Sutou, S. and H. Shindo (1975). "Concanavalin a induces endoreduplication in cultured mammalian
cells." Biochemical and Biophysical Research Communications 67(1): 79-84.
Not Relevant to current
fit for purpose review
Suzuki, N., et al. (1996). "Purification and characterization of a cytosolic isozyme of 3-Deoxy-D-arabino-
heptulosonate 7-phosphate synthase from cultured carrot cells." Journal of Plant Physiology 149(1-2):
19-22.
Not Relevant to current
fit for purpose review
Swaney, S. (1995). "Spasmodic Torticollis." Physiotherapy 81(5): 301.
Not Relevant to current
fit for purpose review
Tai, A., et al. (2008). "Estimate of Radiobiologic Parameters from Clinical Data for Biologically Based
Treatment Planning for Liver Irradiation." International Journal of Radiation Oncology*Biology *Physics
70(3): 900-907.
Not Relevant to current
fit for purpose review
Tai, H.-K., et al. (2011). "A one-time inducible transposon for terminating selectable markers in
transgenic plants." Botanical Studies 52(4): 375-381.
Not Relevant to current
fit for purpose review
Takabatake, R., et al. (2010). "Development and evaluation of event-specific quantitative per method for
genetically modified soybean MON89788." Journal of Biotechnology 150, Supplement: 482-483.
Not Relevant to current
fit for purpose review
Talbot, K. (2007). "Neuromuscular disorders: therapeutic advances." The Lancet Neurology
6( 1): 18-19.
Not Relevant to current
fit for purpose review
Tan, A. and G. D. Badhwar (1997). "Detecting the dynamical state of the atmosphere from the orbital
decay of the ODERACS spheres." Journal of Atmospheric and Solar-Terrestrial Physics 59(4): 431-437.
Page 183 of 216
-------
Not Relevant to current
fit for purpose review
Tan, J. H., et al. (2013). "An interactive lung field segmentation scheme with automated capability."
Digital Signal Processing 23(3): 1022-1031.
Not Relevant to current
fit for purpose review
Tan, S., et al. (2006). "Herbicidal inhibitors of amino acid biosynthesis and herbicide-tolerant crops."
Amino Acids 30(2): 195-204.
Not Relevant to current
fit for purpose review
Tardieu, M. (2010). "Paediatric neurology: brain development at an interface between genetics, the
environment, and the immune system." The Lancet Neurology 9(1): 13-14.
Not Relevant to current
fit for purpose review
Taverniers, I., et al. (2004). "Cloned plasmid DNA fragments as calibrators for controlling GMOs:
different real-time duplex quantitative PCR methods." Anal Bioanal Chem 378(5): 1198-1207.
Not Relevant to current
fit for purpose review
Thakur, K. T., et al. (2015). "CNS infections in 2014: guns, germs, and will." The Lancet Neurology
14(1): 20-22.
Not Relevant to current
fit for purpose review
Tong, X. H., et al. (2009). "Physiological and molecular mechanisms of glyphosate tolerance in an in
vitro selected cotton mutant." Pesticide Biochemistry and Physiology 94(2-3): 100-106.
Not Relevant to current
fit for purpose review
Tonkin, J. (1991). "South Australia CACCN." Confederation of Australian Critical Care Nurses Journal
4(3): 6.
Not Relevant to current
fit for purpose review
Toyota, A., et al. (2006). "Quantification of genetically modified soybeans using a combination of a
capillary-type real-time PCR system and a plasmid reference standard." Biosci Biotechnol Biochem
70(4): 821-827.
Not Relevant to current
fit for purpose review
Trojano, M. and C. Tortorella (2015). "MS and related disorders: looking for markers of phenotypes."
The Lancet Neurology 14(1): 11-13.
Not Relevant to current
fit for purpose review
Trusler, C. S., et al. (2007). "Italian ryegrass (Lolium multiflorum) management options in winter wheat
in Oklahoma." Weed Technology 21(1): 151-158.
Not Relevant to current
fit for purpose review
Truta, E., et al. (2011). "Evaluation of Roundup-induced toxicity on genetic material and on length
growth of barley seedlings." Acta Biol Hung 62(3): 290-301.
Not Relevant to current
fit for purpose review
Turk, J. V. andD. L. Snedeker (1986). "Intro public relations texts: A round-up review." Public Relations
Review 12(4): 48-55.
Not Relevant to current
fit for purpose review
Tyshko, N. V., et al. (2008). "Medical and biological safety assessment of genetically modified maize
event MON 88017. Report 2. Genotoxicologic, immunologic and allergologic examinations." Voprosy
pitaniya 77(5): 13-17.
Not Relevant to current
fit for purpose review
Ugarte, R. (2014). "Interaction between glyphosate and mitochondrial succinate dehydrogenase."
Computational and Theoretical Chemistry 1043: 54-63.
Not Relevant to current
fit for purpose review
Underwood, M. (1991). "Western Australia ASCCN." Confederation of Australian Critical Care Nurses
Journal 4(3): 6.
Not Relevant to current
fit for purpose review
Uppal, R. K. and M. J. Gooding (2013). "Gibberellin-responsive and -insensitive dwarfing alleles on
wheat performance in contrasting tillage systems." Field Crops Research 141: 55-62.
Not Relevant to current
fit for purpose review
Urda, A., et al. (2012). "Final voluntary assessment for Traumatology and Orthopaedic Surgery medical
residents: A report on the results and a look at the future." Revista Espanola de Cirugia Ortopedica y
Traumatologia (English Edition) 56(3): 188-196.
Not Relevant to current
fit for purpose review
Valencia, E., et al. (1999). "Yield and botanical composition of rhizoma peanut-grass swards treated with
herbicides." Agronomy Journal 91(6): 956-961.
Not Relevant to current
fit for purpose review
Van Caneghem, T. (2002). "EARNINGS MANAGEMENT INDUCED BY COGNITIVE REFERENCE
POINTS." The British Accounting Review 34(2): 167-178.
Not Relevant to current
fit for purpose review
van Duijn, G., et al. (1999). "Detection methods for genetically modified crops." Food Control 10(6):
375-378.
Not Relevant to current
fit for purpose review
van Haselen, R. and R. Jiitte (2013). "The placebo effect and its ramifications for clinical practice and
research. Villa La Collina at Lake Como, Italy, 4-6 May 2012." Complementary Therapies in Medicine
21(2): 85-93.
Not Relevant to current
fit for purpose review
Van Kampen, J. and J. D. Hundleby (1994). "An investigation of the associations between augmenting-
reducing, reactivity, sensation seeking and body image." Personality and Individual Differences 16(3):
373-377.
Not Relevant to current
fit for purpose review
Van Loon, A. J. (1991). "Glaciotectonic landforms and structures: J.S. Aber, D.G. Croot and M.M.
Fenton. Kluwer Academic Publishers Group, Dordrecht, 1989, ix + 200 pp., Dfl.155.-, US$79.-, £49.-,
ISBN 0-7923-0100-5 (hardcover)." Sedimentary Geology 75(1-2): 164-165.
Page 184 of 216
-------
Not Relevant to current
fit for purpose review
Venancio, V. P., et al. (2012). "Conventional (MG-BR46 Conquista) and Transgenic (BRS Valiosa RR)
Soybeans Have No Mutagenic Effects and May Protect Against Induced-DNA Damage In Vivo."
Nutrition and Cancer-an International Journal 64(5): 725-731.
Not Relevant to current
fit for purpose review
Vyn, T. J., et al. (2002). "Corn response to potassium placement in conservation tillage." Soil and Tillage
Research 67(2): 159-169.
Not Relevant to current
fit for purpose review
Walker, D. C., et al. (2004). "The epitheliome: agent-based modelling of the social behaviour of cells."
Biosystems 76(1-3): 89-100.
Not Relevant to current
fit for purpose review
Wallace, K., et al. (2015). "A multiplexed assay for determination of neurotoxicant effects on
spontaneous network activity and viability from microelectrode arrays." NeuroToxicology 49: 79-85.
Not Relevant to current
fit for purpose review
Wander, J. G. N. and H. J. Bouwmeester (1998). "Effects of nitrogen fertilization on dill (Anethum
graveolens L.) seed and carvone production." Industrial Crops and Products 7(2-3): 211-216.
Not Relevant to current
fit for purpose review
Wang, F., et al. (2008). "The microtubule plus end-binding protein EB1 is involved in Sertoli cell
plasticity in testicular seminiferous tubules." Experimental Cell Research 314(1): 213-226.
Not Relevant to current
fit for purpose review
Wang, G., et al. (2006). "Methyl protodioscin induces G2/M cell cycle arrest and apoptosis in HepG2
liver cancer cells." Cancer Letters 241(1): 102-109.
Not Relevant to current
fit for purpose review
Wang, G., et al. (2012). "Damage to DNA caused by UV-B radiation in the desert cyanobacterium
Scytonemajavanicumandthe effects of exogenous chemicals on the process." Chemosphere 88(4): 413-
417.
Not Relevant to current
fit for purpose review
Wang, G., et al. (2012). "3.105 GLYPHOSATE INDUCED CELL DEATH THROUGH APOPTOTIC
AND AUTOPHAGIC MECHANISMS." Parkinsonism & Related Disorders 18, Supplement 2: S190.
Not Relevant to current
fit for purpose review
Wang, M., et al. (2015). "Gene editing by co-transformation of TALEN and chimeric RNA/DNA
oligonucleotides on the rice OsEPSPS gene and the inheritance of mutations." PLoS One 10(4):
e0122755.
Not Relevant to current
fit for purpose review
Wang, S.-A., et al. (2009). "Heat Shock Protein 90 Is Important for Spl Stability during Mitosis." Journal
of Molecular Biology 387(5): 1106-1119.
Not Relevant to current
fit for purpose review
Wang, X., et al. (2005). "Effects of Phe-to-Trp mutation and fluorotryptophan incorporation on the
solution structure of cardiac troponin C, and analysis of its suitability as a potential probe for in situ NMR
studies." Protein Science 14(9): 2447-2460.
Not Relevant to current
fit for purpose review
Wang, X., et al. (2011). "Construction of a reference plasmid molecule containing eight targets for the
detection of genetically modified crops." Appl Microbiol Biotechnol 90(2): 721-731.
Not Relevant to current
fit for purpose review
Weng, H., et al. (2005). "Novel reference gene, High-mobility-group protein I/Y. used in qualitative and
real-time quantitative polymerase chain reaction detection of transgenic rapeseed cultivars." J AO AC Int
88(2): 577-584.
Not Relevant to current
fit for purpose review
White, A. K. and W. W. Metcalf (2004). "Two C-P lyase Operons inPseudomonas stutzeri and their roles
in the oxidation of phosphonates, phosphite, and hypophosphite." Journal of Bacteriology 186(14): 4730-
4739.
Not Relevant to current
fit for purpose review
Willenborg, C. J. and E. N. Johnson (2013). "Influence of seeding date and seeding rate on cow cockle,
a new medicinal and industrial crop." Industrial Crops and Products 49: 554-560.
Not Relevant to current
fit for purpose review
Williams, M. J., et al. (2002). "No-till establishment of rhizoma peanut." Agronomy Journal 94(6): 1350-
1354.
Not Relevant to current
fit for purpose review
Williamson, P. R., et al. (2016). "CNS infections in 2015: emerging catastrophic infections and new
insights into neuroimmunological host damage." The Lancet Neurology 15(1): 17-19.
Not Relevant to current
fit for purpose review
Wilson, R. (1970). "AORN RECRUITMENT PROGRAM ROUND-UP." AORN Journal 11(1): 66-68.
Not Relevant to current
fit for purpose review
Wiznitzer, M. (2006). "Paediatric neurology: clinical advances with low technology." The Lancet
Neurology 5(1): 14-15.
Not Relevant to current
fit for purpose review
Wong, P. K. Y., et al. (1973). "Rapid, selective procedure for isolation of spontaneous temperature-
sensitive mutants of Moloney leukemia virus." Virology 51(2): 424-431.
Not Relevant to current
fit for purpose review
Wood, E. J. (1985). "From cells to atoms: an illustrated introduction to molecular biology: by A. R. Rees
and M. J. E. Sternberg, Blackwell, 1984. £6.80 (viii + 94 pages) ISBN 0 632 00888 1." Trends in
Biochemical Sciences 10(5): 214-215.
Not Relevant to current
fit for purpose review
Wood, E. J. (1985). "Lecture notes on biochemistry: by J. K. Candish, Blackwell, 1984. £9.80 (vi + 290
pages) ISBN 0 632 01253 6." Trends in Biochemical Sciences 10(5): 214-215.
Not Relevant to current
fit for purpose review
Wurz, A., et al. (1999). "Quantitative analysis of genetically modified organisms (GMO) in processed
food by PCR-based methods." Food Control 10(6): 385-389.
Page 185 of 216
-------
Not Relevant to current
fit for purpose review
Xia, J., et al. (2015). "Retraction of a study on genetically modified corn: Expert investigations should
speak louder during controversies over safety." Bioscience Trends 9(2): 134-137.
Not Relevant to current
fit for purpose review
Xu, J., et al. (2002). "Cloning of genomic DNA of rice 5-enolpyruvylshikimate 3-phosphate synthase
gene and chromosomal localization of the gene." Sci China C Life Sci 45(3): 251-259.
Not Relevant to current
fit for purpose review
Xu, J. W., et al. (2002). "Isolation of rice EPSP synthase cDNA and its sequence - Analysis and copy
number determination." ActaBotanica Sinica44(2): 188-192.
Not Relevant to current
fit for purpose review
Xu, W. T., et al. (2011). "Establishment and evaluation of event-specific qualitative and quantitative PCR
method for genetically modified soybean DP-356043-5." European Food Research and Technology
233(4): 685-695.
Not Relevant to current
fit for purpose review
Yan, H.-Q., et al. (2011). "Novel AroA from Pseudomonas putida Confers Tobacco Plant with High
Tolerance to Glyphosate." PLoS One 6(5).
Not Relevant to current
fit for purpose review
Yang, L. T., et al. (2005). "Screening and construct-specific detection methods of transgenic Huafan No
1 tomato by conventional and real-time PCR." Journal of the Science of Food and Agriculture 85(13):
2159-2166.
Not Relevant to current
fit for purpose review
Yang, R., et al. (2007). "Event-specific qualitative and quantitative PCR detection of roundup ready event
GT73 based on the 3'-integration junction." Plant Cell Rep 26(10): 1821-1831.
Not Relevant to current
fit for purpose review
Yang, Z. and X. Dong (2015). "Earnings roundup in private and public bank holding companies."
Advances in Accounting 31(1): 96-99.
Not Relevant to current
fit for purpose review
Yao, P., et al. (2015). "Improvement of glycine oxidase by DNA shuffling, and site-saturation
mutagenesis of F247 residue." Int J Biol Macromol 79: 965-970.
Not Relevant to current
fit for purpose review
Yilmaz, R., et al. "Verification of GMO analytical methods: DNA isolation and quantitative detection
methods for Roundup Ready soy flour." New Biotechnology 29, Supplement: S196-S197.
Not Relevant to current
fit for purpose review
You, Y., et al. (2012). "Effect of intravenous fat emulsion therapy on glyphosate-surfactant-induced
cardiovascular collapse." The American Journal of Emergency Medicine 30(9): 2097.e2091-2097.e2092.
Not Relevant to current
fit for purpose review
Yousef, M. I., et al. (1996). "A sensitive sperm-motility test for the assessment of cytotoxic effect of
pesticides." Journal of Environmental Science and Health Part B-Pesticides Food Contaminants and
Agricultural Wastes 31(1): 99-115.
Not Relevant to current
fit for purpose review
Zanis, D. A., et al. (1997). "A comparison of three methods of measuring the type and quantity of services
provided during substance abuse treatment." Drug and Alcohol Dependence 49(1): 25-32.
Not Relevant to current
fit for purpose review
Zanoli, L. M., et al. (2013). "Peptide nucleic acid molecular beacons for the detection of PCR amplicons
in droplet-based microfluidic devices." Anal Bioanal Chem 405(2-3): 615-624.
Not Relevant to current
fit for purpose review
Zawahir, S., et al. (2009). "Acute intentional self-poisoning with a herbicide product containing
fenoxaprop-P-ethyl, ethoxysulfuron, and isoxadifen ethyl: a prospective observational study." Clinical
Toxicology 47(8): 792-797.
Not Relevant to current
fit for purpose review
Zhan, T., et al. (2013). "Improving Glyphosate Oxidation Activity of Glycine Oxidase from Bacillus
cereus by Directed Evolution." PLoS One 8(11).
Not Relevant to current
fit for purpose review
Zhang, H., et al. (2008). "Development of one novel multiple-target plasmid for duplex quantitative PCR
analysis of roundup ready soybean." Journal of Biotechnology 136, Supplement: S250.
Not Relevant to current
fit for purpose review
Zhang, J., et al. (1998). "A transformation technique from RGB signals to the Munsell system for color
analysis of tobacco leaves." Computers and Electronics in Agriculture 19(2): 155-166.
Not Relevant to current
fit for purpose review
Zhang, M., et al. (2007). "Detection of Roundup Ready soy in highly processed products by triplex nested
PCR." Food Control 18(10): 1277-1281.
Not Relevant to current
fit for purpose review
Zhang, Q., et al. (2014). "Characterization of cytochalasins from the endophytic Xylaria sp. and their
biological functions." J Agric Food Chem 62(45): 10962-10969.
Not Relevant to current
fit for purpose review
Zheng, Q. and D. C. Chang (1991). "High-efficiency gene transfection by in situ electroporation of
cultured cells." Biochimica et Biophysica Acta (BBA) - Gene Structure and Expression 1088(1): 104-
110.
Not Relevant to current
fit for purpose review
Zhou, H., et al. (1995). "Glyphosate-tolerant CP4 and GOX genes as a selectable marker in wheat
transformation." Plant Cell Rep 15(3-4): 159-163.
Not Relevant to current
fit for purpose review
Zhu, X., et al. (2012). "A high-throughput fluorescence resonance energy transfer (FRET)-based
endothelial cell apoptosis assay and its application for screening vascular disrupting agents." Biochemical
and Biophysical Research Communications 418(4): 641-646.
Not Relevant to current
fit for purpose review
Ziegelberger, G., etal. (2006). "International commission on non-ionizing radiation protection." Progress
in Biophysics and Molecular Biology 92(1): 1-3.
Page 186 of 216
-------
Not Relevant to current
fit for purpose review
Zulet, A., et al. (2015). "Fermentation and alternative oxidase contribute to the action of amino acid
biosynthesis-inhibiting herbicides." J Plant Physiol 175: 102-112.
Not Relevant to current
fit for purpose review
o et al. (2006). "Prognostic Predictors of Outcome for Poisoning by Glyphosate-containing
Herbicides, Based on Initial Findings." Journal of The Korean Society of Emergency Medicine 17(6):
630-636.
Not Relevant to current
fit for purpose review
Boyle, W. S. and J. O. Evans (1974). "EFFECTS OF GLYPHOSATE AND ETHEPHON ONMEIOTIC
CHROMOSOMES OF SECALE-CEREALE L." Journal of Heredity 65(4): 250-250.
Not Relevant to current
fit for purpose review
Sherwood, M. M. and W. C. Davison (1957). "Correspondence." The Journal of Pediatrics 51(4): 486-
487.
Not Relevant to current
fit for purpose review
Belle, R., et al. (2007). "[Sea urchin embryo, DNA-damaged cell cycle checkpoint and the mechanisms
initiating cancer development]." J Soc Biol 201(3): 317-327.
Not Relevant to current
fit for purpose review
Gehin, A., et al. (2005). "Vitamins C and E reverse effect of herbicide-induced toxicity on human
epidermal cells HaCaT: abiochemometric approach." Int J Pharm 288(2): 219-226.
Not Relevant to current
fit for purpose review
Gehin, A., et al. (2006). "Glyphosate-induced antioxidant imbalance in HaCaT: The protective effect of
vitamins C and E." Environmental Toxicology and Pharmacology 22(1): 27-34.
Not Relevant to current
fit for purpose review
Lueken, A., et al. (2004). "Synergistic DNA damage by oxidative stress (induced by H202) and
nongenotoxic environmental chemicals in human fibroblasts." Toxicol Lett 147(1): 35-43.
Not Relevant to current
fit for purpose review
Baurand, P. E., et al. (2015). "Genotoxicity assessment of pesticides on terrestrial snail embryos by
analysis of random amplified polymorphic DNA profiles." J Hazard Mater 298: 320-327.
Not Relevant to current
fit for purpose review
Guilherme, S., et al. (2009). "Tissue specific DNA damage in the European eel (Anguilla anguilla)
following a short-term exposure to a glyphosate-based herbicide." Toxicology Letters 189: S212-S212.
Not Relevant to current
fit for purpose review
Marc, J., et al. (2004). "Glyphosate-based pesticides affect cell cycle regulation." Biol Cell 96(3): 245-
249.
Not Relevant to current
fit for purpose review
Marc, J., et al. (2003). "Embryonic cell cycle for risk assessment of pesticides at the molecular level."
Environmental Chemistry Letters 1(1): 8-12.
Not Relevant to current
fit for purpose review
Nwani, C. D., et al. (2014). "Induction of micronuclei and nuclear lesions in Channa punctatus following
exposure to carbosulfan, glyphosate and atrazine." Drug Chem Toxicol 37(4): 370-377.
Not Relevant to current
fit for purpose review
Owczarek, M., et al. (1999). "Evaluation of toxic and genotoxic activity of some pesticides in a soil-plant
system." Human and Environmental Exposure to Xenobiotics: 755-762.
Not Relevant to current
fit for purpose review
Perez-Iglesias, J. M., et al. (2016). "Effects of glyphosate on hepatic tissue evaluating
melanomacrophages and erythrocytes responses in neotropical anuran Leptodactylus latinasus." Environ
Sci Pollut Res Int.
Not Relevant to current
fit for purpose review
Poletta, G. L., et al. (2009). "Genotoxicity of the herbicide formulation Roundup (glyphosate) in broad-
snouted caiman (Caiman latirostris) evidenced by the Comet assay and the Micronucleus test." Mutat Res
672(2): 95-102.
Not Relevant to current
fit for purpose review
Siddiqui, S., etal. (2012). "Glyphosate, alachor and maleic hydrazide have genotoxic effect on Trigonella
foenum-graecum L." Bull Environ Contam Toxicol 88(5): 659-665.
Not Relevant to current
fit for purpose review
Song, H.-Y., et al. (2012). "Cellular Toxicity of Surfactants Used as Herbicide Additives." J Korean Med
Sci 27(1): 3-9.
Not Relevant to current
fit for purpose review
Unver, T., et al. (2010). "Genome-wide profiling and analysis of Festuca arundinacea miRNAs and
transcriptomes in response to foliar glyphosate application." Mol Genet Genomics 283(4): 397-413.
Relevant- Cancer Epi
Acquavella, J. F., et al. (2006). "Exposure misclassification in studies of agricultural pesticides - Insights
frombiomonitoring." Epidemiology 17(1): 69-74.
Relevant- Cancer Epi
Acquavella, J. F., et al. (2005). "Implications for epidemiologic research on variation by pesticide in
studies of farmers and their families." Scandinavian Journal of Work Environment & Health 31: 105-
109.
Relevant- Cancer Epi
Baker, B. A., et al. (2005). "Farm Family Exposure Study: methods and recruitment practices for a
biomonitoring study of pesticide exposure." Journal of Exposure Analysis and Environmental
Epidemiology 15(6): 491-499.
Relevant- Cancer Epi
Chang, E. T. and E. Delzell (2016). "Systematic review and meta-analysis of glyphosate exposure and
risk of lymphohematopoietic cancers." J Environ Sci Health B 51(6): 402-434.
Relevant- Cancer Epi
De Roos, A. J., et al. (2005). "Cancer incidence among glyphosate-exposed pesticide applicators in the
Agricultural Health Study." Environ Health Perspect 113(1): 49-54.
Page 187 of 216
-------
Relevant- Cancer Epi
Firth, H. M., et al. (2007). "Chemical exposure among NZ farmers." International Journal of
Environmental Health Research 17(1): 33-43.
Relevant- Cancer Epi
Lash, T. L. (2007). "Bias analysis applied to Agricultural Health Study publications to estimate non-
random sources of uncertainty." J Occup Med Toxicol 2: 15.
Relevant- Cancer Epi
Mink, P. J., et al. (2012). "Epidemiologic studies of glyphosate and cancer: a review." Regul Toxicol
Pharmacol 63(3): 440-452.
Relevant- Cancer Epi
Sorahan, T. (2012). "Multiple myeloma and glyphosate use: A re-analysis of US Agricultural Health
Study data." Toxicology Letters 211, Supplement: S169.
Relevant-
Carcinogenicity
Greim, H., et al. (2015). "Evaluation of carcinogenic potential of the herbicide glyphosate, drawing on
tumor incidence data from fourteen chronic/carcinogenicity rodent studies." CritRev Toxicol 45(3): 185-
208.
Relevant-
Carcinogenicity
Williams, G. M., et al. (2000). "Safety Evaluation and Risk Assessment of the Herbicide Roundup and
Its Active Ingredient, Glyphosate, for Humans." Regulatory Toxicology and Pharmacology 31(2): 117-
165.
Relevant- Genotoxicity
Alvarez-Moya et al. (2014) "Comparison of the in vivo and in vitro genotoxicity of glyphosate
isopropylamine salt in three different organisms". Genetics and Molecular Biology, 37, 1, 105-110
Relevant- Genotoxicity
Chan, P. and J. Mahler (1992). "NTP technical report on the toxicity studies of Glyphosate (CAS No.
1071-83-6) Administered In Dosed Feed To F344/N Rats And B6C3F1 Mice." Toxic Rep Ser 16: l-d3.
Relevant- Genotoxicity
Bakry, F. A., et al. (2015). "Glyphosate herbicide induces genotoxic effect and physiological disturbances
inBulinus truncatus snails." Pestic BiochemPhysiol 123: 24-30.
Relevant- Genotoxicity
Bolognesi, C., et al. (1997). "Genotoxic activity of glyphosate and its technical formulation roundup." J
Agric Food Chem 45(5): 1957-1962.
Relevant- Genotoxicity
Bolognesi, C., et al. (2009). "Biomonitoring of genotoxic risk in agricultural workers from five Colombian
regions: association to occupational exposure to glyphosate." J Toxicol Environ Health A 72(15-16): 986-
997.
Relevant- Genotoxicity
Da Silva, F. R., et al. (2014). "Genotoxic assessment in tobacco farmers at different crop times." Science
of the Total Environment 490: 334-341.
Relevant- Genotoxicity
Dimitrov, B. D., et al. (2006). "Comparative genotoxicity of the herbicides Roundup, Stomp andReglone
in plant and mammalian test systems." Mutagenesis 21(6): 375-382.
Relevant- Genotoxicity
El-Shenawy, N. S. (2009). "Oxidative stress responses of rats exposed to Roundup and its active
ingredient glyphosate." Environ Toxicol Pharmacol 28(3): 379-385.
Relevant- Genotoxicity
Fortes, C., et al. (2016). "Occupational Exposure to Pesticides With Occupational Sun Exposure Increases
the Risk for Cutaneous Melanoma." J Occup Environ Med 58(4): 370-375.
Relevant- Genotoxicity
Ghisi Nde, C., et al. (2016). "Does exposure to glyphosate lead to an increase in the micronuclei
frequency? A systematic and meta-analytic review." Chemosphere 145: 42-54.
Relevant- Genotoxicity
Heydens, W. F., et al. (2008). "Genotoxic potential of glyphosate formulations: mode-of-action
investigations." J Agric Food Chem 56(4): 1517-1523.
Relevant- Genotoxicity
Kale, P. G., et al. (1995). "MUTAGENICITY TESTING OF 9 HERBICIDES AND PESTICIDES
CURRENTLY USED IN AGRICULTURE." Environ Mol Mutagen 25(2): 148-153.
Relevant- Genotoxicity
Kier, L. D. (2015). "Review of genotoxicity biomonitoring studies of glyphosate-based formulations."
Crit Rev Toxicol 45(3): 209-218.
Relevant- Genotoxicity
Koller, V. J., et al. (2012). "Cytotoxic and DNA-damaging properties of glyphosate and Roundup in
human-derived buccal epithelial cells." Arch Toxicol 86(5): 805-813.
Relevant- Genotoxicity
Li, A. P. and T. J. Long (1988). "An evaluation of the genotoxic potential of glyphosate." Fundam Appl
Toxicol 10(3): 537-546.
Relevant- Genotoxicity
Lioi, M. B., et al. (1998). "Genotoxicity and oxidative stress induced by pesticide exposure in bovine
lymphocyte cultures in vitro." MutatRes 403(1-2): 13-20.
Relevant- Genotoxicity
Manas, F., et al. (2009). "Genotoxicity of AMP A, the environmental metabolite of glyphosate, assessed
by the Comet assay and cytogenetic tests." Ecotoxicol Environ Saf 72(3): 834-837.
Relevant- Genotoxicity
Manas, F., et al. (2009). "Genotoxicity of glyphosate assessed by the comet assay and cytogenetic tests."
Environmental Toxicology and Pharmacology 28(1): 37-41.
Relevant- Genotoxicity
Mandel, J. S., et al. (2005). "Biomonitoring for farm families in the farm family exposure study."
Scandinavian Journal of Work Environment & Health 31: 98-104.
Relevant- Genotoxicity
Mladinic, M., et al. (2009). "Evaluation of genome damage and its relation to oxidative stress induced by
glyphosate in human lymphocytes in vitro." Environ Mol Mutagen 50(9): 800-807.
Page 188 of 216
-------
Relevant- Genotoxicity
Mladinic, M. and D. Zeljezic (2008). "Assessment of oxidative DNA damage by glyphosate applying
hOGGl modified comet and micronucleus assay." Toxicology Letters 180: S170-S171.
Relevant- Genotoxicity
Paz-y-Mino, C., et al. (2011). "Baseline determination in social, health, and genetic areas in communities
affected by glyphosate aerial spraying on the northeastern Ecuadorian border." Rev Environ Health 26(1):
45-51.
Relevant- Genotoxicity
Paz-y-Mino, C., et al. (2007). "Evaluation of DNA damage in an Ecuadorian population exposed to
glyphosate." Genetics and Molecular Biology 30(2): 456-460.
Relevant- Genotoxicity
Peluso, M., et al. (1998). "32P-postlabeling detection of DNA adducts in mice treated with the herbicide
Roundup." EnvironMol Mutagen 31(1): 55-59.
Relevant- Genotoxicity
Piesova, E. (2005). "The effect of glyphosate on the frequency of micronuclei in bovine lymphocytes in
vitro." Acta Veterinaria-Beograd 55(2-3): 101-109.
Relevant- Genotoxicity
Prasad, S., et al. (2009). "Clastogenic effects of glyphosate in bone marrow cells of swiss albino mice."
J Toxicol 2009: 308985.
Relevant- Genotoxicity
Rank, J., et al. (1993). "Genotoxicity testing of the herbicide Roundup and its active ingredient glyphosate
isopropylamine using the mouse bone marrow micronucleus test, Salmonella mutagenicity test, and
Allium anaphase-telophase test." MutatRes 300(1): 29-36.
Relevant- Genotoxicity
Roustan, A., et al. (2014). "Genotoxicity of mixtures of glyphosate and atrazine and their environmental
transformation products before and after photoactivation." Chemosphere 108: 93-100.
Relevant- Genotoxicity
Silva Kahl, V. F., et al. (2016). "Telomere measurement in individuals occupationally exposed to
pesticide mixtures in tobacco fields." Environ Mol Mutagen 57(1): 74-84.
Relevant- Genotoxicity
Sivikova, K. and J. Dianovsky (2006). "Cytogenetic effect of technical glyphosate on cultivated bovine
peripheral lymphocytes." Int J Hyg Environ Health 209(1): 15-20.
Relevant- Genotoxicity
Vigfusson, N. V. and E. R. Vyse (1980). "The effect of the pesticides, Dexon, Captan and Roundup, on
sister-chromatid exchanges in human lymphocytes in vitro." MutatRes 79(1): 53-57.
Retracted Article
Seralini, G.-E., et al. (2014). "Retraction notice to "Long term toxicity of a Roundup herbicide and a
Roundup-tolerant genetically modified maize" [Food Chem. Toxicol. 50 (2012) 4221-4231]." Food and
Chemical Toxicology 63: 244.
Page 189 of 216
-------
Appendix B
De Roos et al. (2003)
Hoar etal. (1986)
Lee et al. (2004a)
Cantor etal. (1992)
Zahmetal. (1990)
Figure B.l. Visual representation of studies included in De Roos et al. (2003).
Hardell etal. (2002)
Nordstrom et al. (1998)
Hardell arid Eriksson (1999)
Figure B.2. Visual representation of studies included in Hardell et al. (2002).
McDuffie et al. (2001)
(51 exposed cases/
133 exposed controls)
Hohenadal et al. (2011)
(19 exposed cases/
78 exposed controls)
Figure B.3. Visual representation of the association between McDuffie et al. (2001) and the follow-up analysis
by Hohenadal et al. (2011).
Page 190 of 216
-------
Yiin etal, (2012)
(51 exposed cases/
76 exposed controls)
Carreon et al. (2005)
(18 exposed cases/
41 exposed controls)
Figure B.4. Visual representation of the association between Carreon et al. (2005), which investigated gliomas
in women only, and Yiin et al. (2012), which investigated both sexes.
Page 191 of 216
-------
Appendix C
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
Alavanja et al.
(2003)
USA: Iowa and
North Carolina
Enrollment (1993-
1997) through 2001
Males enrolled in AHS;
licensed private and
commercial applicators
Males enrolled in AHS;
licensed private and
commercial applicators
566 cases
54,766 controls
not reported
No
Andreotti et al.
(2009)
USA: Iowa and
North Carolina
Enrollment (1993-
1997) through 2004
Participants enrolled in
AHS; licensed private and
commercial applicators
and spouses
Participants enrolled in
AHS; licensed private and
commercial applicators and
spouses
93 cases(64
applicators, 29
spouses)
82,503 controls
(52,721 applicators,
29,782 spouses)
55 cases
48,461 controls
No
Bandef al. (2011)
Canada: British
Columbia
1983-1990
Male residents in British
Columbia identified as
cancer patients in British
Columbia Cancer Registry
(excluding farmers that
worked all outside British
Columbia)
Male residents in British
Columbia identified as
cancer patients in British
Columbia Cancer Registry
(excluding farmers that
worked all outside British
Columbia) with other
cancer sites excluding lung
cancer and cancers of
unknown primary site
1,153 cases
3,999 controls
25 cases
60 controls
Yes (included
in adjustment)
Brown et al. (1990)
USA: Iowa and
Minnesota
Iowa: 1981-1983;
Minnesota: 1980-
1982
Initial interview
1981-1984 and
supplemental
interviews (Iowa
only) in 1987
White males (30 years or
older) residing in Iowa or
Minnesota diagnosed with
leukemia
White males without
lymphatic or hematopoietic
cancer selected by random
digit dialing (< age 65),
Medicare records (age >
65) and state death
certificate files (deceased
controls) - frequency
matched for 5-year age
group, vital status, and state
of residence
Initial: 578 cases;
1245 controls
Supplemental: 92
cases; 211 controls
15 cases
49 controls
Yes (not
evaluated)
Brown et al. (1993)
USA: Iowa
Iowa: 1981-1983;
Interview 1981-
1984
White males (30 years or
older) residing in Iowa
diagnosed with multiple
myeloma
White males without
lymphatic or hematopoietic
cancer selected by random
digit dialing (< age 65),
Medicare records (age >
173 cases
650 controls
11 cases
40 controls
Yes (not
evaluated)
Page 192 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
65) and state death
certificate files (deceased
controls) - frequency
matched for 5-year age
group, vital status, and state
of residence
Cocco etal. (2013)
Czech Republic,
France, Germany,
Italy, Ireland, and
Spain
1998-2004
Adult patients first
diagnosed with lymphoma
residing in the referral
area of the participating
centers
Controls from Germany
and Italy were randomly
selected by sampling from
the general population,
matched to cases on sex, 5-
year age-group, and
residence area. The rest of
the centers used matched
hospital controls, with
eligibility criteria limited to
diagnoses other than
cancer, infectious diseases,
and immunodeficient
diseases
2,348 cases
2,462 controls
4 cases
2 controls
No
De Roos et al.
(2003)
USA: Nebraska,
Iowa, Minnesota,
and Kansas
Nebraska: 1983-
1986
Iowa: 1981-1983
Minnesota: 1980-
1982
Kansas: 1979-1981
White males diagnosed
with NHL in one of the 4
states (21 years or older in
Nebraska and Kansas; 30
years or older in Iowa and
Minnesota)
Males living in same
geographic area obtained
by random digit dialing,
Medicare records and state
mortality files - frequency
matched for race, sex, age,
and vital status
870 cases
2,569 controls
36 cases
61 controls
Yes (not
significant in
covariate
analysis)
De Roos et al.
(2005)
USA: Iowa and
North Carolina
Enrollment (1993-
1997) through 2001
Participants enrolled in
AHS; licensed private and
commercial applicators
and spouses
Participants enrolled in
AHS; licensed private and
commercial applicators and
spouses
54,315 subjects
included in this
analysis
All cancers - 358 cases
Lung - 26 cases
Oral cavity - 10 cases
Colon - 15 cases
Rectum - 14 cases
Pancreas - 7 cases
Kidney - 9 cases
Bladder - 17 cases
Prostate - 145 cases
Melanoma - 14 cases
All lymphohematopoietic
cancers - 36 cases
NHL - 17 cases
Leukemia - 9 cases
No
Page 193 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
Multiple myeloma - 6
cases
(13,280 subjects not
exposed to glyphosate
used for comparison
population)
Engel etal. (2005)
USA: Iowa and
North Carolina
Enrollment (1993-
1997) through 2000
Wives of applicators
enrolled in AHS study
with no history of breast
cancer
Wives of applicators
enrolled in AHS study with
no history of breast cancer
309 cases
30,145 controls
82 cases; 10,016 controls
No
Eriksson et al.
(2008)
Sweden
1999-2002
Patients (18-74 years of
age) residing in Sweden
and diagnosed with NHL
Swedish residents randomly
selected living in same
health service regions as
cases - frequency matched
for age (in 10 years) and
sex
910 cases
1,016 controls
29 cases
18 controls
No
Flower et al. (2004)
USA: Iowa
1993-1997
Children (born after 1975)
of participants enrolled in
AHS study who were
diagnosed with childhood
cancer up to 19 years of
age
Children (born after 1975)
of participants enrolled in
AHS study not diagnosed
with childhood cancer up to
19 years of age
50 cases out of 17,357
total study population
Maternal use: 13 cases of
6075 total exposed
Paternal use: 6 cases of
3231 total exposed
No
Hardell et al. (2002)
Sweden
NHL: 1987-1990
HCL: 1987-1992
NHL: Male residents of
one of four northern or
three middle counties in
Sweden age 25 years and
older diagnosed with
NHL; identified from
regional cancer registries
HCL: Living male
residents of
Sweden age 25 years and
older
diagnosed with HC1;
identified from
the Swedish Cancer
Registry
NHL: Two male controls
for each case matched by
age, year of death if
deceased, and county HCL:
Four male controls for each
case matched by age and
county
515 cases
1,141 controls
8 cases
8 controls
Yes (not
evaluated)
Kachuri et al.
(2013)
Canada: Alberta,
British Columbia,
Manitoba, Ontario,
1991-1994
Men aged >19 years (>30
years in analysis) - pulled
from hospital records in
Quebec,
Men aged >19 years (30
years in analysis) - pulled
from provincial health
insurance records in
342 cases
1,357 controls
32 cases
121 controls
Yes (included
in adjustment)
Page 194 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
Quebec, and
Saskatchewan
cancer registries in all
other
provinces
Alberta, Saskatchewan,
Manitoba, and Quebec;
computerized telephone
listings in Ontario; voter
lists in British Columbia
Karunanayake et al.
(2012)
Canada: Alberta,
British Columbia,
Manitoba, Ontario,
Quebec, and
Saskatchewan
1991-1994
Men aged >19 years -
pulled from hospital
records in Quebec,
cancer registries in all
other
provinces
Men aged >19 years -
pulled from provincial
health insurance records in
Alberta, Saskatchewan,
Manitoba, and Quebec;
computerized telephone
listings in Ontario; voter
lists in British Columbia
316 cases
1,506 controls
38 cases
133 controls
No
Koureas et al.
(2014)
Greece
2010
Inhabitants
of the city of Larissa;
Eligibility criteria for
pesticide sprayers were
1) to personally apply
pesticides systematically,
and 2) to have recently
applied pesticides (no
longer than 7 days
between last application
and
sampling).
The rural residents group
were occupied in
administrative services,
public order services, health
services, education or trade.
Inclusion criteria for this
group: absence of any
involvement in agricultural
activities either as a
primary or secondary
occupation by participant or
any member of household.
Also recruited urban
residents (mainly blood
donors) from the city of
Larissa.
80 pesticide sprayers,
85 rural residents, and
121 individuals
Not reported
No
Koutros et al.
(2013)
USA: Iowa and
North Carolina
Enrollment (1993-
1997) through 2007
Males enrolled in AHS;
licensed private and
commercial applicators
Males enrolled in AHS;
licensed private and
commercial applicators
1,962 incident cases
(including 919
aggressive prostate
cancers) among
54,412 applicators
1464 cases
42,420 controls
No
Landgren et al.
(2009)
USA: Iowa and
North Carolina
Exposure
information:
enrollment (1993-
1997) and 5-year
follow-up interview
Males enrolled in AHS;
licensed private and
commercial applicators
Males enrolled in AHS;
licensed private and
commercial applicators
678 participants
27 cases out of 570 total
exposed
No
Page 195 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
Blood samples:
2006-2007 (Iowa)
and 2008 (North
Carolina)
White residents of 1 of 66
Nebraska counties age 21
years or older with a
newly confirmed case of
adenocarcinoma of the
Frequency matched by age
and sex to the combined
distribution of glioma,
stomach, and esophageal
cancer cases from a control
group from a previous
study (1986-1987) that
selected controls from the
general population by
random digit dialing for
those under 65 years,
Health Care Financing
Administration Medicare
files for those over 65
years, mortality records for
deceased and matched by
race, sex, vital status (or
year of death if deceased)
Stomach: 170 cases
Lee et al. (2004b)
USA: Nebraska
1988-1993
stomach or Cases
identified from the
Nebraska Cancer Registry
(1988-1990) or from
discharge diagnosis and
pathology records from 14
Nebraska hospitals (1991—
1993)
Esophagus: 137 cases
502 Controls
12 cases
46 controls
Yes (analysis
showed
differences)
Lee etal. (2005)
USA: Nebraska
1988-1993
White residents of 1 of 66
Nebraska counties age 21
years or older with
confirmed adult glioma.
Cases identified from
Nebraska Cancer Registry
or from participating
hospitals in Lincoln and
Omaha, Nebraska
Frequency matched by age,
sex, and vital status to the
combined distribution of
glioma, stomach, and
esophageal cancer cases
from a control group from a
previous study (1986-1987)
that selected controls from
the general population by
random digit dialing for
those under 65 years,
Medicare files for those
over 65 years, mortality
records for deceased and
matched by race, sex, vital
status (or year of death if
deceased), and 5-year age
groups to the overall case
distribution. Additional
251 glioma cases
498 controls
17 cases
32 controls
Yes (analysis
showed
differences,
included in
adjustment)
Page 196 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
younger controls were
brought into the study
through random digit
dialing and from death
certificates
Lee et al. (2007)
USA: Iowa and
North Carolina
1993-97; follow-up
to 2002
Agricultural Health Study
participants: private and
commercial applicators
licensed in Iowa or North
Carolina with no history
of colorectal cancer at
enrollment. Followed
through 2002 for incident
colorectal cancer
Agricultural Health Study
participants: private and
commercial applicators
licensed in Iowa or North
Carolina with no history of
colorectal cancer at
enrollment. Followed
through 2002 for incident
colorectal cancer
56,813 licensed
pesticide applicators
305 incident colorectal
cancer cases(212
colon, 93 rectum)
56,508 controls
Colon - 151 cases;
49 controls
Rectum - 74 cases;
18 controls
Colorectal - 225 cases;
67 controls
No
McDuffie et al.
(2001)
Canada: Alberta,
British Columbia,
Manitoba, Ontario,
Quebec, and
Saskatchewan
1991-1994
Male residents of six
Canadian provinces age
19 years and older
diagnosed with STS, HD,
NHL, or MM; this study
only evaluated those with
NHL. Cases were
identified from Canadian
Cancer Registries; in
Quebec, hospital
ascertainment was used
Random control subject
selection using Health
Insurance records,
computerized telephone
listings, and voters' lists;
males 19 years and older
from same six Canadian
provinces as cases matched
by age (within 2 years)
517 cases
1506 controls
Univariate analysis:
51 cases; 133 controls
(multivariate analysis
also conducted -
glyphosate exposed
numbers not reported)
No
Orsi et al. (2009)
France
2000-2004
Men aged 20-75 years
living in the catchment
areas of the main hospitals
in Brest, Caen, Nantes,
Lille, Toulouse, and
Bordeaux, with no history
of immunosuppression or
taking immunosuppressant
drugs. Cases ascertained
from hospital records.
Patients from the same
hospital catchment area as
the cases. Patients were
hospitalized for orthopedic
or rheumatological
conditions (89.3%),
gastrointestinal or
genitourinary tract diseases
(4.8%), cardiovascular
diseases (1.1%), skin and
subcutaneous tissue disease
(1.8%), and infections
(3.0%), excluding patients
admitted for cancer or a
disease
directly related to
491 cases
456 controls
NHL: 12 cases
24 controls
HL: 6 cases
15 controls
Lymphoproliferative
syndromes: 4 cases
18 controls
Multiple myeloma:
5 cases; 18 controls
Lymphoid neoplasms: 27
cases; 24 controls
No
Page 197 of 216
-------
Table C.l. Design Characteristics of Epidemiological Studies Evaluated for Study Quality.
Study
Location
Study Years
Case Population
Control Population
Total Number of
Subjects
Number of Glyphosate
Exposed Cases
Proxy Use
occupation,
smoking, or alcohol abuse
Pahwaefa/. (2011)
Canada (Alberta,
British Columbia,
Manitoba, Ontario,
Quebec, and
Saskatchewan)
1991-1994
Men aged >19 years -
pulled from hospital
records in Quebec,
cancer registries in all
other
provinces
Men aged >19 years -
pulled from provincial
health insurance records in
Alberta, Saskatchewan,
Manitoba, and Quebec;
computerized telephone
listings in Ontario; voter
lists in British Columbia
342 cases
1,506 age/resident
matched controls
32 cases
133 controls
No
Pahwa etal. (2012)
Canada (Alberta,
British Columbia,
Manitoba, Ontario,
Quebec, and
Saskatchewan)
1991-1994
Men aged >19 years -
pulled from hospital
records in Quebec,
cancer registries in all
other
provinces
Men aged >19 years -
pulled from provincial
health insurance records in
Alberta, Saskatchewan,
Manitoba, and Quebec;
computerized telephone
listings in Ontario; voter
lists in British Columbia
342 cases
1506 age/resident
matched controls
32 cases
133 controls
No
Controls age 18-64
randomly selected from
state driver's
license/nondriver ID
Yiin et al. (2012)
USA: Upper
Midwest Health
Study (Iowa,
Michigan,
Minnesota and
Wisconsin)
1995-1997
Age 18-80 (at
ascertainment or diagnosis
in 1995 through January
1997) residing in counties
where the largest
population center had
fewer than 250,000
residents. Referral by
physicians or through state
cancer registries with
cases verified by
histological evaluation.
records, and those age 65-
80 were selected from
Health Care Financing
Administration's (HCFA)
Medicare data within 10-
year age group strata, with
the proportion/stratum
determined by the age
distribution of glioma cases
in that state from 1992 to
1994. Controls were
frequency-matched within a
state but not by county of
residence. Selected even if
they had a self-reported
history of cancer other than
glioma.
798 glioma cases;
1,175 controls
12 cases
19 controls
Yes (analysis
showed no
differences)
Page 198 of 216
-------
Appendix D. List of studies assigned a low quality ranking and not evaluated in detail
As described in Section 3.2, if studies did not collect exposure information on glyphosate from
all subjects, did not assess an outcome (e.g., biomonitoring studies), and/or did not provide a
quantitative measure of an association between glyphosate and a cancer outcome, then these
studies were assigned a low quality ranking and were not further evaluated in detail. These
studies included the following 32 studies:
Acquavella et al. 2006; Andre et al., 2007; Baker et al. 2005; Benedetti et al., 2013; Bolognesi et
al., 2002; Bolognesi et al., 2004; Bolognesi et al. 2009; Bortoli et al., 2009; Costa et al., 2006;
Da Silva et al. 2014; Dennis et al. 2010; Firth et al. 2007; Gomez-Arroyo et al., 2013; Gregio
D'Arce et al., 2000; El-Zaemey et al., 2013; Fortes et al., 2016; Fritschi et al., 2005; Hernandez
et al., 2006; Kaufman et al. 2009; Khayat et al., 2013; Lebailly et al., 2003; Mandel et al. 2005;
Martinez-Valenzuela et al., 2009; Monge et al., 2007; Pastor et al., 2003; Paz-y Mino et al.,
2007; Paz-y Mino et al. 2011; Ruder et al. 2004; Shaham et al., 2001; Silva Kahl et al. 2016;
Simoniello et al., 2008; Vlastos et al., 2006.
Page 199 of 216
-------
Appendix E
Chronic Dietary Exposure
The agency uses Dietary Exposure Evaluation Model- Food Consumption Intake Database
(DEEM-FCID; version 3.16), which incorporates consumption data from United States
Department of Agriculture (USDA) National Health and Nutrition Examination Survey, What
We Eat in America (NHANES/WWEIA; 2003-2008) to calculate potential chronic dietary
exposures. In an unrefined chronic dietary analysis, several conservative assumptions are used
to generate high end estimates of potential exposure. These assumptions include tolerance-level
residues for all registered commodities, 100% crop treated, and drinking water values from a
direct application to water scenario, as well as DEEM default processing factors. For
glyphosate, the highest exposure value for any population subgroup in an unrefined chronic
dietary analysis would be 0.23 mg/kg/day for children 1-2 years old (Table E.l; see T. Bloem,
30-NOV-2017, D429229 for DEEM inputs and results).
Table E.l. Chronic dietary exposure estimates
Population Subgroup
Exposure (mg/kg/day)
General U.S. Population
0.089771
All Infants (< 1 year old)
0.138338
Children 1-2 years old
0.228379
Children 3-5 years old
0.212036
Children 6-12 years old
0.147749
Youth 13-19 years old
0.088362
Adults 20-49 years old
0.074650
Adults 50-99 years old
0.061258
Females 13-49 years old
0.069318
Post-application Incidental Oral and Dermal Exposure
Glyphosate has residential uses, including application to turf, which would result in the highest
potential post-application exposures; therefore, there is potential for children to be exposed via
incidental oral and dermal routes from playing on treated lawns. For this assessment, the agency
evaluates exposures from hand-to-mouth behavior, object-to-mouth behavior, incidental soil
ingestion, and dermal contact using the 2012 Standard Operating Procedures for Residential
Pesticide Exposure Assessment29. Incidental oral exposures from hand-to-mouth, object-to-
mouth, and incidental soil ingestion are considered inter-related and, therefore, not combined.
To calculate high end estimates of exposures, the following is assumed according to the 2012
SOP to be health-protective: 1) maximum label rates are applied to the turf, 2) exposures are
assumed to occur every day to the residue values on the day of application (i.e., no dissipation),
and 3) individuals engage in post-application activities for the maximum amount of time
represented by data for children spending time outdoors and not specifically engaged in activities
29 Available: http://www2.epa.gov/pesticide-science-and-assessing-pesticide-risks/standard-operating-procedures-
residential-pesticide
Page 200 of 216
-------
on turf, when in actuality children do not spend all of their outdoor time on turf. The highest
exposure from incidental oral scenarios using the maximum application rate for turf applications
would be 0.16 mg/kg/day from hand-to-mouth behaviors by children (1 to <2 years old). Dermal
post-application to children 1 to <2 years old would be 0.08 mg/kg/day.
Table E.2. Post-application Exposure Estimates for Application of Glyphosate to Turf1.
Lifestage
Post-application Exposure Scenario
Exposure (mg/kg/day)
Children 1 to <2 year old
Turf - sprays
Hand-to-Mouth
0.16
Object-to-Mouth
0.005
Incidental Soil Ingestion
0.0003
Dermal (high contact activities)
0.08
1 Based on Roundup® Weed & Grass Super Concentrate, EPA Reg. No. 71995-25.
Page 201 of 216
-------
Appendix F
Genotoxicity Studies with Glyphosate Based Formulations
While the focus of this analysis to determine the genotoxic potential of glyphosate, the agency
has identified numerous studies conducted with glyphosate-based formulations that contain
various concentrations of the glyphosate as well as other components of the end use products and
are presented in Tables F.1-F.5.
Page 202 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint Test System
Concentrations Test Material/
Results
Reference
Comments
Concentration
Bacterial
Reverse
Mutation
S. typhimurium
TA1535,
TA1537,
TA1538,
TA98 and
TA100; E. coli
WP2 uvrA
pKMlOl ± S9
1.6-5000
Hg/plate ± S9
(plate
incorporation)
ICIA 0224 57.6%
in water
Negative ±
S9
Callander
(1988)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537; E.
coli WP2P and
uvrA ± S9
100-5000
Hg/plate ±S9
plate
incorporation &
pre-incubation
protocols
TMSC (tri-
methyl-sulfonium
cliloride) 95%
purity
Negative ±
S9
Callander
(1993)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
26, 43, 72, 120,
200 (ig/plate
Glyphosate liquid
formulation (480
g/L
isopropylainine
salt)
Negative ±
S9
Camolesi
(2009)1
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
26, 43, 72, 120,
200 (ig/plate
MON 77280
equivalent of
glyphosate acid:
495 g/L
Negative ±
S9
Camolesi
(2010)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
0.2-2000
(ig/plate
MON 76190
53.2% glyphosate
Negative ±
S9
Catoyra (2009)1
Bacterial
Reverse
Mutation
S. typhimurium
TA97a, TA98,
TA100 and
TA102± S9
2 (ig/plate (toxic)
Perzocyd 10 SL
formulation
Negative ±
S9
Cliruscielska et
al. (2000)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
0.03-3.0 (iL/plate
MON 8080
(87.6%)
Negative ±
S9
Flowers (1981)
Page 203 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint Test System
Concentrations Test Material/
Results
Reference
Comments
Concentration
TA1535, and
TA1537 ± S9
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
3.16-1000
(ig/plate
TROPM
(Glyphosate 480);
35.84% purity
based on acid,
48.46% pure
based on IPA salt
Negative ±
S9
Fliigge (2010a)1
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
0.316-100
Glyphosate 757
g/kg granular
formulation
(76.1%
monoammonium
glyphosate salt)
Negative ±
S9
Fliigge (2010d)1
Bacterial
Reverse
Mutation
S. tvphimurium
TA97a, TA98,
TA100, and
TA1535 ± S9
1-5000 (ig/plate
Roundup WG
784 g/kg
ammonium salt
equivalent
Negative ±
S9
Gava (1998)
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA1535,
TA1537± S9
50-5000 (ig/plate
Rodeo®
(containing IPA
salt and water
only); 40%
glyphosate (acid
equivalent)
Negative ±
S9
Kier et a/.,
(1992)
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA1535,
TA1537 ± S9
5-500 (ig/plate
(-S9)/15-1500
(ig/plate (+S9)
MON 2139
(Roundup®) 31%
Glyphosate (acid
equivalent)
Negative ±
S9
Kier etal.,
(1992)
Cytotoxic at top
concentrations
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA1535,
TA1537 ± S9
5-500 (ig/plate
(-S9)/15-1500
(ig/plate (+S9)
MON 14445
(Direct®); 75%
Glyphosate (acid
equivalent)
Negative ±
S9
Kier et al.,
(1992)
Cytotoxic at the top
concentrations,
occasionally at lower
concentrations
Bacterial
Reverse
Mutation
S. tvphimurium
TA98, TA100,
TA1535,
TA1537 ± S9
0.2-2000
(ig/plate
MON 79672
(Roundup
Ultramax); 74.7%
monoammonium
Negative ±
S9
Lope (2008)1
Page 204 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint Test System
Concentrations Test Material/
Results
Reference
Comments
Concentration
glyphosate salt;
68.2% glyphosate
Bacterial
Reverse
Mutation
S. typhimurium
TA1535,
TA1537,
TA1538,
TA98 and
TA100 ± S9
0.617-50
(iL/plate ± S9
SC-0224, 19.2%
purity
Negative ±
S9
Majeska (1982)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
TA strains: 10 -
5000 (ig/plate
(+S9); 3.33-3330
(ig/plate (-S9); E.
coli: 33.3-5000
(ig/plate (+/- S9)
MON 78239
36.6% glyphosate
(44.9% potassium
salt of glyphosate)
Negative
Mecclii (2003a)
Increase in revertants
seen in TA98 and
TA1535 -S9 on first
trial, not conc-dep;
however no increase
in revertants seen in
repeat in those strains;
overall negative.
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
TA strains: 3.33-
3330 (ig/plate
(+S9); 1.0-1000
(ig/plate (-S9); E.
coli: 33.3-5000
(ig/plate (+/- S9)
MON 78634
65.2% w/w
glyphosate
(71.8% w/w as
monoaimnonium
salt of glyphosate)
Negative
Mecchi (2003b)
Bacterial
Reverse
Mutation
S. tvphimurium
TA 98,
TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
10 - 5000
(ig/plate (+/-S9)
MON 79864
38.7% glyphosate
acid (wt %)
Negative
Mecclii (2008a)
Inhibited growth seen
at >2000 -S9
Bacterial
Reverse
Mutation
S. tvphimurium
TA 98,
TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
33.3-5000
(ig/plate
MON 76313
30.9% glyphosate
acid
Negative
Mecclii (2008b)
Page 205 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint
Test System
Concentrations
Test Material/
Concentration
Results
Reference
Comments
Bacterial
Reverse
Mutation
S. tvphimurium
TA 98,
TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
10-5000 (ig/plate
(+/-S9)
MON 76171
31.1% glyphosate
Negative
Mecclii
(2008c)1
Bacterial
Reverse
Mutation
S. tvphimurium
TA 98,
TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
10-5000 (ig/plate
(+/-S9)
MON 79991
71.6% glyphosate
acid
Negative
Mecclii (2009a)
Bacterial
Reverse
Mutation
S. tvphimurium
TA 98,
TA100,
TA1535,
TA1537 and
E. coli WP2
uvrA ± S9
10-5000 (ig/plate
(+/-S9)
MON 76138
38.5% glyphosate
Negative
Mecclii
(2009b)1
Bacterial
Reverse
Mutation
S. tvphimurium
TA97a, TA98,
TA100, and
TA1535 ± S9
1-5000 (ig/plate
MON 77280
646.4 g/L salt
equivalent
Negative
Perina (1998)
Bacterial
Reverse
Mutation
S. tvphimurium
TA98 and
TA100 ± S9
0-1440 (ig/plate
(calculated as
glyphosate IPA
salt)
Roundup, 480 g/L
glyphosate
isopropylainine
salt
Negative -
S9,
Equivocal
+S9
Rank el al.
(1993)
Stat significant
increase at 360
(ig/plate for TA98 (-
S9) and 720 (ig/plate
for TA100 (+S9). Not
significant at higher
concentrations and
were not replicated.
Effects occurred at
close to toxic levels.
Page 206 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint Test System
Concentrations Test Material/
Results
Reference
Comments
Concentration
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
500-5000
(ig/plate;
495 g/L
glyphosate
isopropylainine
salt; 371.0 g/L
(equivalent of
glyphosate acid)
Negative ±
S9
Silvino (2011)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
1.5-5000
(ig/plate
MON 8709
495 g/L
glyphosate
isopropylainine
salt; 371.0 g/L
(equivalent of
glyphosate acid)
Negative ±
S9
Silvino (2011)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535, and
TA1537 ± S9
15-5000 (ig/plate
MON 76313
468 g/L
glyphosate
isopropylainine
salt (351 g/L
glyphosate acid
equivalent)
Negative ±
S9
Silvino (2012)
Cytotoxic at 5000
Hg/plate for some
strains
Bacterial
Reverse
Mutation
S. typhimurium
TA97a, TA98,
TA100 and
TA1535 ± S9
1-5000 (ig/plate
Glifos
formulation
(glyphosate
isopropylaimnoni
um salt, Berol 907
and water)
Negative ±
S9
Vargas (1996)
Cytotoxic at the two
upper concentrations
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA102,
TA1535,
TA1537± S9
3.16-316
Hg/plate
FSG 3090-H1
360 g/L
Negative ±
S9
Uhde (2004)1
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100
±S9
0.01-100
Hg/plate
64% (glyphosate
Isopropylaimnoni
um salt)
Negative ±
S9
Wang el al.
(1993)
Bacterial
Reverse
Mutation
S. typhimurium
TA98, TA100,
TA1535,
All strains: 33.3-
5000 (ig/plate
MON 78910
30.3% glyphosate
acid
Negative ±
S9
Xu (2006)
Cytotoxic >1000
(ig/plate (-S9)
Page 207 of 216
-------
Table F.l. In vitro Test for Gene Mutations in Bacteria: Glyphosate Formulations.
Test/Endpoint Test System
Concentrations
Test Material/
Concentration
Results
Reference
Comments
TA1537 and
E. coli WP2
uvrA ± S9
(+S9); 10-3330
(ig/plate (-S9)
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline, test material purity,
control chemicals and summary data tables.
Table F.2. In Vitro Tests for Chromosome Damage in Mammalian Cells- Glyphosate Formulations
Test/Endpoint
Test System
Concentrations
Test Material/
Concentration
Results
Reference
Comments
In vitro
Chromosomal
Aberration using
fluorescent in
situ
hybridization
(FISH)
Bovine lymphocytes
(from two 6-8 month old
calves)
-whole chromosome (1)
painting probe
28-1120 nM
24 h exposure
62%
Isopropylamine
salt of glyphosate
(38% inert
ingredients)
Negative.
Holeckova
(2006)
Small but significant
increase in polyploidy
seen at 56
No positive control
reported.
In vitro
Cytokinesis
Block
Micronucleus
Assay
(with FISH
analysis)
TR146 cells (human-
derived buccal
epithelial
cell line)
0, 10, 15 and 20
mg/L;
20 minute
exposure.
Roundup Ultra
Max (450 g/1
glyphosate acid)
Positive
Increase in
MN at all
test
concentratio
ns
Koller el al.
(2012)
No apoptosis observed at
any conc.
Necrosis reported at 20
mg/L.
Increase in NB and NPB
seen at all concentrations
MI= mitotic index. FISH= fluorescent in situ hybridization, MN= micronuclei; NB= nuclear buds; NPB= nucleoplasms bridges.
Page 208 of 216
-------
Table F.3. In Vivo Tests for Chromosomal Aberrations in Mammals- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Administration
Doses
Test Material
Purity
Results
Reference
Comments
Bone Marrow
Chromosomal
Aberration
Swiss albino mice
(males only)
Vehicle: DMSO
Intraperitoneal
injection;
sampling 24, 48
and 72 h
0, 25 and 50
mg/kg
(5/dose)
Roundup
(>41%
isopropylamine
glyphosate)
Positive
Increase in MN
at all time points
at both doses
Prasad et al.
(2009)
Significant decrease
in mitotic index seen
at all doses and time
points
Bone Marrow
Chromosomal
Aberration
C57BL mice
(males only)
Vehicle: water
Oral
administration;
sampling 6, 24,
48, 72, 96 and
120 h
0.05,0.01,
0.5 and
1.0%
(8/dose)
Roundup
Negative
Dimitrov et al.
(2006)
Bone Marrow
Chromosomal
Aberration
New Zealand white
rabbits
(males only)
Vehicle:
Drinking water
for 60 days
0, 750 ppm
(5/dose)
Roundup
Positive
Helal and
Moussa (2005)
BM= bone marrow, SC= spermatocyte.
Page 209 of 216
-------
Table F.4. In Vivo Tests for Micronuclei Induction in Mammals- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
Swiss CD1 mice
(males only)
Intraperitoneal
injection; 2
injections of half
the dosage of
135 mg/kg 24 h
apart; sampling
at 6 and 24 h
0, 450 mg/kg
roundup, equiv.
to 135 kg
glyphosate
(3/dose)
Roundup,
30.4%
glyphosate
Positive
Bolognesi et
al. (1997)
Stat significant
increase in MN
at 6 and 24 h
Bone Marrow
Micronucleus
Test
C3H mice
(males only)
Vehicle: water
Intraperitoneal
Injection
(single
treatment);
sampling after
24, 48 and 72 h
0, 90 mg/kg
Not
reported
Negative
Chruscielska et
al. (2000)
Bone Marrow
Micronucleus
Test
Swiss mice
(males and females)
Vehicle: water
Intraperitoneal
Injection
(2 treatments, 24
h apart);
sampling after 24
h (last treatment)
0, 50, 100 and
200 mg/kg
480g/L
Isopropyla
mine salt of
glyphosate
Negative
Grisolia (2002)
Bone Marrow
Micronucleus
Test
CD-I mice
(males and females)
Intraperitoneal
injection;
sampling 24, 48
and 72 h
0, 140, 280, and
555 mg/kg
Roundup
(31%
glyphosate
salt)
Negative
Kier (1992)
Some deaths observed
at high dose (HD),
jPCE/NCE ratio at
HD at 48 h in males.
Bone Marrow
Micronucleus
Test
Swiss albino mice
(males and females)
Intraperitoneal
Injection
(2 treatments, 24
h apart);
sampling after 24
h (last treatment)
0, 212.5, 425 and
637.5 mg/kg
MON
77280
646.4 g/L
glyphosate
salt
equivalent
Negative
Momna (1998)
Doses tested
corresponded to 25%,
50% and 75% LD50
Page 210 of 216
-------
Table F.4. In Vivo Tests for Micronuclei Induction in Mammals- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
NMRI-Bom mice
Intraperitoneal
Injection (single
treatment);
sampling after 24
h
0, 133 and 200
mg/kg
(4/sex/dose)
Roundup,
480 g
glyphosate
isopropyla
mine salt
per liter
Negative
Rank et al.
(1993)
BM toxicity indicated
by %PCE decreased
at 200 mg/kg
Bone Marrow
Micronucleus
Test
Swiss albino mice
(males only2)
Vehicle: water
Oral gavage (two
treatments, 24 h
apart); sampled
at 18 and 24 h
after last dose
0, 2000 mg/kg
MON
8709494.7
g/L salt of
isopropyla
mine (371.0
glyphosate
acid)
Negative
Claro (2011)
OECD 474 Guideline
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
C57BL mice
(males only)
Vehicle: water
Oral
administration;
sampling 6, 24,
48, 72, 96 and
120 h
0.05,0.01,0.5
and 1.0%
(1%=1080
mg/kg)
(8/dose)
Roundup
Negative
Dimitrov et al.
(2006)
Toxicity seen in 1.0%
dose group
Bone Marrow
Micronucleus
Test
Crl:CD-l(ICR) BR
mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
78239
(36.6%
glyphosate)
Negative
Erexson
(2003a)
EPA Guideline (84-2)
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
Crl:CD-l(ICR) BR
mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
78634
(65.2%
glyphosate)
Negative
Erexson
(2003b)
EPA Guideline (84-2)
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
Crl:CD-l(ICR) BR
mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
78910
(30.3%
glyphosate)
Negative
Erexson (2006)
EPA Guideline (84-2)
No significant signs
of toxicity observed
in main study.
Page 211 of 216
-------
Table F.4. In Vivo Tests for Micronuclei Induction in Mammals- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Administration
Doses
Test
Material
Purity
Results
Reference
Comments
Bone Marrow
Micronucleus
Test
NMRI mice
(males and females)
Vehicle: 0.8%
hydro xypropylmethyl
cellulose
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 mg/kg
(5/sex/dose)
TROPM
(Glyphosate
480); 358.4
g/L
glyphosate
acid; 483.6
g/L IPA salt
Negative
Fliigge
(2010c)1
OECD Guideline 474
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
Swiss mice
(males only2)
Vehicle: water
Oral gavage
(2 treatments, 24
h apart);
sampling after 24
h (last treatment)
0, 2000 mg/kg
(6/dose)
A17035A
289.7 g/L
glyphosate
Negative
Negro Silva
(2009)1
OECD Guideline 474
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
Swiss mice
(males only2)
Vehicle: water
Oral gavage
(2 treatments, 24
h apart);
sampling after 24
h (last treatment)
0, 2000 mg/kg
(6/dose)
Glyphosate
SL (499.35
g/L
glyphosate)
Negative
Negro Silva
(2011)1
OECD Guideline 474
No significant signs
of toxicity observed
in main study
Bone Marrow
Micronucleus
Test
Hsd: CD-I (ICR) mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
79864
(38.7%
glyphosate)
Negative #
Xu (2008a)
EPA Guideline (84-2)
/OECD 474
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
CD-1(ICR)BR mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
76171
(31.1%
glyphosate)
Negative
Xu (2008b)
EPA Guideline (84-2)
/OECD 474
No significant signs
of toxicity observed
in main study.
Bone Marrow
Micronucleus
Test
CD-1(ICR)BR mice
(males only2)
Vehicle: water
Oral gavage
(single
treatment);
sampling after 24
and 48 h (high
dose only)
0, 500, 1000, and
2000 (mg/kg)
(5/dose)
MON
79991
(71.6%
glyphosate)
Negative
Xu (2009a)
EPA Guideline (84-2)
/OECD 474
No significant signs
of toxicity observed
in main study.
Page 212 of 216
-------
Table F.4. In Vivo Tests for Micronuclei Induction in Mammals- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Doses
Test
Results
Reference
Comments
Administration
Material
Purity
Bone Marrow
CD-1(ICR)BR mice
Oral gavage
0, 500, 1000, and
MON
Negative
Xu (2009b)1
EPA Guideline (84-2)
Micronucleus
(males only2)
(single
2000 (mg/kg)
76138
/OECD 474
Test
Vehicle: water
treatment);
(5/dose)
(38.5%
No significant signs
sampling after 24
glyphosate)
of toxicity observed
and 48 h (high
in main study.
dose only)
Bone Marrow
Hsd:CD-l(ICR)BR
Oral gavage
0, 500, 1000, and
MON
Negative
Xu (2009c)1
EPA Guideline (84-2)
Micronucleus
mice
(single
2000 (mg/kg)
76313
/OECD 474
Test
(males only2)
treatment);
(5/dose)
(30.9%
No significant signs
Vehicle: water
sampling after 24
glyphosate)
of toxicity observed
and 48 h (high
in main study.
dose only)
Bone Marrow
CD rats
Oral gavage
0, 500, 1000, and
757 g/kg
Negative
Fliigge
OECD Guideline 474
Micronucleus
(males and females)
(single
2000 mg/kg
granular
(2010c)1
No significant signs
Test
Vehicle: 0.8%
treatment);
(5/sex/dose)
formulation
of toxicity observed
hydro xypropylmethyl
sampling after 24
(69.1%
in main study
cellulose
and 48 h (high
glyphosate
dose only)
acid)
1 Study was cited in Kier and Kirkland (2013). Supplementary information about the study was provided online including test guideline, test material purity,
control chemicals and summary data tables.
2 Only males tested; report indicated that there was no difference between sexes seen in range finding study.
BM= bone marrow, CA= chromosomal aberrations, MN= micronucleated erythrocytes, NCE= normochromatic erythrocytes, PCE=polychromatic erythrocytes.
Page 213 of 216
-------
Table F.5. Other Assays for Detecting DNA Damage- Glyphosate Formulations.
Test/Endpoint Test System
Route of
Doses/
Test Material/ Results
Reference
Comments
Administration
Concentrations
Concentration
Bacterial SOS
Chromotest
Escherichia coli PQ37
strain
NA (in vitro)
0.25|ig/samplc
Roundup BIO
formulation;
Positive
Raipulis et al.
(2009)
DNA Adducts
32p_
postlabeling
Swiss CD1 mice
(males and females)
Liver and kidney
evaluated
Intraperitoneal
injection
0, 400, 500 and
600 mg/kg,
corresponding
to 122, 152 and
182 mg/kg
glyphosate salt
Roundup
(30.4%
isopropylammo
nium salt of
glyphosate)
Positive
(liver and
kidney)
Peluso et al.
(1998)
DNA oxidative
damage:
8-OHdG
formation
Swiss CD-I mice
(males)
liver and kidney
evaluated
Intraperitoneal
injection (single
dose); sampling
4 and 24 h after
injection
900 mg/kg
corresponding
to 270 mg/kg
glyphosate
(3/dose)
900 mg/kg
corresponding
to 270 mg/kg
glyphosate
Kidney:
positive at
8 and 24 h
Liver:
negative
Bolognesi et
al. (1997)
Single-cell gel
electrophoresis
(SCGE) assays-
COMET assay
TR146 cells (human-
derived buccal
epithelial
cell line). Alkaline
conditions
NA (in vitro)
Roundup Ultra
Max (450 g/1
glyphosate acid)
Induced
DNA
migration
at >20
mg/L
Koller et al.
(2012)
Also measured
multiple cellular
integrity parameters to
assess cytotoxicity.
Formulation was more
toxic than technical.
Significant increase in
LDHe at all
concentrations tested.
Cytotoxic > 60 mg/L
Sister
Chromatid
Exchange
(SCE)
Bovine lymphocytes
NA (in vitro)
28- 1112 nM;
±S9; sampling
at 24 and 48 h
62%
Isopropylamine
salt of
glyphosate
Positive
Sivikova &
Dianovsky
(2006)
Page 214 of 216
-------
Table F.5. Other Assays for Detecting DNA Damage- Glyphosate Formulations.
Test/Endpoint
Test System
Route of
Administration
Doses/
Concentrations
Test Material/
Concentration
Results
Reference
Comments
Sister
Chromatid
Exchange
(SCE)
Human lymphocytes
(2 donors)
NA (in vitro)
250, 2500 and
25000 |ig/mL
Roundup;
Isopropylamine
salt of
glyphosate
(purity not
stated)
Stat.
significant
increase
(p<0.001)
at 250
Hg/mL in
both
donors,
and in one
donor at
2500
Hg/mL
Vigfusson and
Vyse (1980)
No growth seen at
highest concentration
(25 mg/mL)
Sister
Chromatid
Exchange
(SCE)
Human lymphocytes
NA (in vitro)
-S9: 0,0.1 and
0.33 mg/mL; 72
h exposure
Roundup,
30.4%
glyphosate
Positive
Bolognesi et
al. (1997)
Stat significant
increase in SCE/cell
at >0.1 mg/mL
Alkaline
elution assay-
DNA single
strand breaks
Swiss CD-I mice
(males)
liver and kidney
evaluated
Intraperitoneal
injection (single
dose); sampling
4 and 24 h after
injection
900 mg/kg
corresponding
to 270 mg/kg
glyphosate
(3/dose)
900 mg/kg
corresponding
to 270 mg/kg
glyphosate
Positive
(Increased
elution
rate) at 4
hours in
liver and
kidney
At 24 h,
returned to
control
levels
Bolognesi et
al. (1997)
Return to control
values at 24 h may
indicate DNA repair
or reflect rapid
elimination of
compound
h= hour, NA= not applicable, SCE= sister chromatid exchange, LDHe= extracellular lactate dehydrogenase
Page 215 of 216
-------
Appendix G
The following studies were considered during the systematic review, but were excluded from the
analysis.
Amer S.M. et al (2006). In vitro and in vivo evaluation of the genotoxicity of the herbicide
glyphosate in mice. Bulletin of the National Research Centre (Cairo) 31 (5): 427-446.
Aboukila, R.S. et al. (2014). Cytogenetic Study on the Effect of Bentazon and Glyphosate
Herbicide on Mice. Alexandria Journal of Veterinary Sciences, 41: 95-101.
Majeska (1982d) MRID 00126616
Majeska (1982e) MRID 00126614
Majeska (1982f) MRID 00126615
Page 216 of 216
------- |