EPA/635/R-08/016F
www.epa.gov/iris
vvEPA
TOXICOLOGICAL REVIEW
OF
HYDROGEN CYANIDE
AND
CYANIDE SALTS
(CAS No. various)
In Support of Summary Information on the
Integrated Risk Information System (IRIS)
September 2010
U.S. Environmental Protection Agency
Washington, DC
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection
Agency policy and approved for publication. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
11
-------
CONTENTS —TOXICOLOGICAL REVIEW OF HYDROGEN CYANIDE AND
CYANIDE SALTS
CONTENTS —TOXICOLOGICAL REVIEW OF HYDROGEN CYANIDE AND
CYANIDE SALTS iii
LIST OF TABLES v
LIST OF FIGURES vi
LIST OF FIGURES vi
LIST OF ACRONYMS AND ABBREVIATIONS vii
FOREWORD ix
AUTHORS, CONTRIBUTORS, AND REVIEWERS x
1. INTRODUCTION 1
2. CHEMICAL AND PHYSICAL INFORMATION 3
3. TOXICOKINETICS 7
3.1. ABSORPTION 7
3.2. DISTRIBUTION 8
3.3. METABOLISM 11
3.4. ELIMINATION 16
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS 17
4. HAZARD IDENTIFICATION 19
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS 19
4.1.1. Acute Oral, Inhalation, and Dermal Studies 19
4.1.2. Subchronic and Chronic Oral Studies 21
4.1.3. Subchronic and Chronic Inhalation Studies 21
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS—ORAL AND INHALATION 29
4.2.1. Oral Studies 29
4.2.2. Inhalation Studies 41
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES 42
4.3.1. Oral Studies 42
4.3.2. Inhalation Studies 45
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES 46
4.4.1. Acute Oral Studies 46
4.4.2. Acute Inhalation Studies 48
4.4.3. Neurotoxicity Studies 49
4.4.4. Immune Endpoints 50
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE
MODE OF ACTION 51
4.5.1. Genotoxicity 51
4.5.2. Acute Neurotoxicity 52
4.5.3. Thyroid Disruption 52
4.5.4. Reproductive Effects 53
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS AND MODE OF
ACTION 54
4.7. EVALUATION OF CARCINOGENICITY 62
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES 62
4.8.1. Possible Childhood Susceptibility 62
4.8.2. Possible Gender Differences 64
in
-------
4.8.3. Other Susceptible Populations 64
5. DOSE RESPONSE ASSESSMENTS 66
5.1. ORAL REFERENCE DOSE (RfD) 66
5.1.1. Choice of Principal Study and Critical Effect 66
5.1.2. Method of Analysis 70
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs) 72
5.1.4. RfD Comparison Information 75
5.1.5. Previous RfD Assessment 77
5.2. INHALATION REFERENCE CONCENTRATION (RfC) 78
5.2.1. Choice of Principal Study and Critical Effect 78
5.2.2. Method of Analysis 81
5.2.3. RfC Derivation—Including Application of Uncertainty Factors (UFs) 82
5.2.4. Previous RfC Assessment 84
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND
INHALATION REFERENCE CONCENTRATION 84
5.4. CANCER ASSESSMENT 89
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF HAZARD AND DOSE
RESPONSE 91
6.1. HUMAN HAZARD POTENTIAL 91
6.2. DOSE RESPONSE 93
6.2.1. Noncancer—Oral 93
6.2.2. Noncancer—Inhalation 95
6.2.3. Cancer 96
7. REFERENCES 97
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION A-l
APPENDIX B. BENCHMARK DOSE MODELING RESULTS B-l
IV
-------
LIST OF TABLES
2-1. Physical and chemical properties of cyanide compounds 4
4-1. Thyroid uptake of 131Iin electroplating workers 24
4-2. Thyroid parameters in former silver-reclaiming workers 26
4-3. Thyroid parameters in HCN-exposed and unexposed electroplating workers 28
4-4. Reproductive effects in male rats administered NaCN in drinking water for 13 weeks 31
4-5. Reproductive effects in mice administered NaCN in drinking water for 13 weeks 34
4-6. Summary of subchronic and chronic oral toxicity studies for cyanide in animals 56
4-7. Summary of subchronic and chronic inhalation toxicity studies for cyanide in humans 59
5-1. Reproductive endpoints in male rats and mice observed following administration of
NaCN in drinking water for 13 weeks 71
5-2. BMD modeling results for observed reproductive endpoints 72
5-3. Potential PODs with applied UFs and resulting potential reference values 75
B-l. Decreased cauda epididymis weight in F344 rats following administration of NaCN
in drinking water for 13 weeks B-l
B-2. BMD modeling results for decreased cauda epididymis weight in rats B-l
B-3. Decreased epididymis weight in F344 rats following administration of NaCN in
drinking water for 13 weeks B-6
B-4. BMD modeling results for decreased epididymis weight in rats B-6
B-5. Decreased testis weight in F344 rats following administration of NaCN in drinking
water for 13 weeks B-ll
B-6. BMD modeling results for decreased testis weight in rats B-ll
B-7. Decreased testicular spermatid concentration in F344 rats following administration
of NaCN in drinking water for 13 weeks B-l 5
B-8. BMD modeling results for decreased testicular spermatid concentration in rats B-l5
-------
LIST OF FIGURES
3-1. Cyanide primary metabolic pathways 12
4-1. Urinary SGST of exposed workers plotted against individual breathing concentrations of
HCN 23
5-1. Potential reference value comparison array 76
B-l. Observed and predicted decrease in cauda epididymis weight in F344 rats following
administration of NaCNin drinking water for 13 weeks B-2
B-2. Observed and predicted decrease in epididymis weight in F344 rats following
administration of NaCNin drinking water for 13 weeks B-7
B-3. Observed and predicted decrease in testis weight in F344 rats following administration of
NaCN in drinking water for 13 weeks B-ll
B-4. Observed and predicted decrease in testicular spermatid concentration in F344 rats
following administration of NaCNin drinking water for 13 weeks B-15
VI
-------
LIST OF ACRONYMS AND ABBREVIATIONS
ACH acetone cyanohydrin
ADP adenosine diphosphate
AIC Akaike's Information Criterion
ALP alkaline phosphatase
ALT alanine aminotransferase
ANOVA analysis of variance
AST aspartate aminotransferase
ATP adenosine triphosphate
ATSDR Agency for Toxic Substances and Disease Registry
AUC area under the curve
BMD benchmark dose
BMDL 95% lower confidence limit on the BMD
BMDS BMD software
BMR benchmark response
CAP compound action potential
CASRN Chemical Abstracts Service Registry Number
CI confidence interval
CN~ cyanide ion
(CN)i cyanogen
CNS central nervous system
DPO diphenyl oxide
ECETOC European Centre for Ecotoxicology and Toxicology of Chemicals
EGL external granular layer
FEV forced expiratory volume
FVC forced vital capacity
GD gestation day
GFAP glial fibrillary acid protein
GGT y-glutamyl transferase
HCN hydrogen cyanide
i.p. intraperitoneal or intraperitoneally
IPCS International Programme on Chemical Safety
IRIS Integrated Risk Information System
i.v. intravenous or intravenously
KCN potassium cyanide
KSCN potassium thiocyanate
LDso median lethal dose
LDH lactate dehydrogenase
LOAEL lowest-observed-adverse-effect level
MCH mean corpuscular hemoglobin
MCHC mean corpuscular hemoglobin concentration
ML molecular layer
MPST mercaptopyruvate sulfurtransferase
NaCN sodium cyanide
NHANES National Health and Nutrition Examination Survey
NIS sodium-iodide symporter
NOAEL no-observed-adverse-effect level
vn
-------
NRC National Research Council
NTP National Toxicology Program
OCN~ cyanate
OSOT hypothiocyanate
PBPK physiologically based pharmacokinetic
PND postnatal day
POD point of departure
RBC red blood cell
RfC reference concentration
RfD reference dose
RR relative risk
SOT thiocyanate
SD standard deviation
SE standard error
SEM standard error of the mean
TS triiodothyronine
T4 thyroxine
TSH thyroid-stimulating hormone
TWA time-weighted average
UF uncertainty factor
U.S. EPA U.S. Environmental Protection Agency
Vmax maximum velocity
WBC white blood cell
Vlll
-------
FOREWORD
The purpose of this Toxicological Review is to provide scientific support and rationale
for the hazard and dose-response assessment in IRIS pertaining to chronic exposure to hydrogen
cyanide and cyanide salts. It is not intended to be a comprehensive treatise on the chemical or
toxicological nature of hydrogen cyanide and cyanide salts.
The intent of Section 6, Major Conclusions in the Characterization of Hazard and Dose
Response, is to present the major conclusions reached in the derivation of the reference dose,
reference concentration and cancer assessment, where applicable, and to characterize the overall
confidence in the quantitative and qualitative aspects of hazard and dose response by addressing
the quality of data and related uncertainties. The discussion is intended to convey the limitations
of the assessment and to aid and guide the risk assessor in the ensuing steps of the risk
assessment process.
For other general information about this assessment or other questions relating to IRIS,
the reader is referred to EPA's IRIS Hotline at (202) 566-1676 (phone), (202) 566-1749 (fax), or
hotline.iris@epa.gov (email address).
IX
-------
AUTHORS, CONTRIBUTORS, AND REVIEWERS
CHEMICAL MANAGER/AUTHOR
Kathleen Newhouse, M.S.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
CO-AUTHORS
Nancy Chiu, Ph.D.
Office of Water/HECD
U.S. Environmental Protection Agency
Washington, DC
CONTRACTOR SUPPORT
Lynne Haber, Ph.D.
Toxicology Excellence for Risk Assessment
Joan Strawson, M.S.
Toxicology Excellence for Risk Assessment
Bonnie Ransom Stern, Ph.D., M.P.H.
Consulting in Health Sciences and Risk Assessment
BR Stern and Associates, ICF Consulting, Inc.
REVIEWERS
This document has been provided for review to EPA scientists, interagency reviewers
from other federal agencies and White House offices, and the public, and peer reviewed by
independent scientists external to EPA. A summary and EPA's disposition of the comments
received from the independent external peer reviewers and from the public is included in
Appendix A.
INTERNAL EPA REVIEWERS
TedBerner, M.S.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Glinda Cooper, Ph.D.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
-------
Ralph Cooper, Ph.D.
National Health and Environmental Effects Research Laboratory
Office of Research and Development
Lynn Flowers, Ph.D.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Martin Gehlhaus, MHS
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
Susan Griffin, Ph.D.
Region VIII
U.S Environmental Protection Agency
Denver, CO
Sally Perreault Darney, Ph.D.
National Health and Environmental Effects Research Laboratory
Office of Research and Development
Jamie Strong, Ph.D.
National Center for Environmental Assessment
U.S. Environmental Protection Agency
Washington, DC
EXTERNAL PEER REVIEWERS
Cheryl B. Bast, Ph.D.
Oak Ridge National Laboratory
George P. Daston, Ph.D.
The Procter & Gamble Company
Michael J. DiBartolomeis, Ph.D.
Toxicology Research International
Jeffrey W. Fisher, Ph.D. (chair)
University of Georgia
John D. Meeker, Sc.D.
University of Michigan
XI
-------
1. INTRODUCTION
This document presents background information and justification for the Integrated Risk
Information System (IRIS) Summary of the hazard and dose-response assessment of hydrogen
cyanide (HCN) and cyanide salts. IRIS Summaries may include oral reference dose (RfD) and
inhalation reference concentration (RfC) values for chronic and other exposure durations, and a
carcinogenicity assessment.
The RfD and RfC, if derived, provide quantitative information for use in risk assessments
for health effects known or assumed to be produced through a nonlinear (presumed threshold)
mode of action. The RfD (expressed in units of mg/kg-day) is defined as an estimate (with
uncertainty spanning perhaps an order of magnitude) of a daily exposure to the human
population (including sensitive subgroups) that is likely to be without an appreciable risk of
deleterious effects during a lifetime. The inhalation RfC (expressed in units of mg/m3) is
analogous to the oral RfD, but provides a continuous inhalation exposure estimate. The
inhalation RfC considers toxic effects for both the respiratory system (portal-of-entry) and for
effects peripheral to the respiratory system (extrarespiratory or systemic effects). Reference
values are generally derived for chronic exposures (up to a lifetime), but may also be derived for
acute (< 24 hours), short-term (>24 hours up to 30 days), and subchronic (>30 days up to 10% of
lifetime) exposure durations, all of which are derived based on an assumption of continuous
exposure throughout the duration specified. Unless specified otherwise, the RfD and RfC are
derived for chronic exposure duration.
The carcinogenicity assessment provides information on the carcinogenic hazard
potential of the substance in question and quantitative estimates of risk from oral and inhalation
exposure may be derived. The information includes a weight-of-evidence judgment of the
likelihood that the agent is a human carcinogen and the conditions under which the carcinogenic
effects may be expressed. Quantitative risk estimates may be derived from the application of a
low-dose extrapolation procedure. If derived, the oral slope factor is a plausible upper bound on
the estimate of risk per mg/kg-day of oral exposure. Similarly, an inhalation unit risk is a
plausible upper bound on the estimate of risk per |ig/m3 air breathed.
Development of these hazard identification and dose-response assessments for HCN and
cyanide salts has followed the general guidelines for risk assessment as set forth by the National
Research Council (NRC, 1983). U.S. Environmental Protection Agency (U.S. EPA) Guidelines
and Risk Assessment Forum Technical Panel Reports that may have been used in the
development of this assessment include the following: Guidelines for the Health Risk
Assessment of Chemical Mixtures (U.S. EPA, 1986a), Guidelines for Mutagenicity Risk
Assessment (U.S. EPA, 1986b), Recommendations for and Documentation of Biological Values
for Use in Risk Assessment (U.S. EPA, 1988), Guidelines for Developmental Toxicity Risk
-------
Assessment (U.S. EPA, 1991), Interim Policy for Particle Size and Limit Concentration Issues in
Inhalation Toxicity (U.S. EPA, 1994a), Methods for Derivation of Inhalation Reference
Concentrations and Application of Inhalation Dosimetry (U.S. EPA, 1994b), Use of the
Benchmark Dose Approach in Health Risk Assessment (U.S. EPA, 1995), Guidelines for
Reproductive Toxicity Risk Assessment (U.S. EPA, 1996), Guidelines for Neurotoxicity Risk
Assessment (U.S. EPA, 1998a), Science Policy Council Handbook: Risk Characterization (U.S.
EPA, 2000a), Benchmark Dose Technical Guidance Document (U.S. EPA, 2000b),
Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures (U.S.
EPA, 2000c), A Review of the Reference Dose and Reference Concentration Processes (U.S.
EPA, 2002), Guidelines for Carcinogen Risk Assessment (U.S. EPA, 2005a), Supplemental
Guidance for Assessing Susceptibility from Early-Life Exposure to Carcinogens (U.S. EPA,
2005b), Science Policy Council Handbook: Peer Review (U.S. EPA, 2006a), and A Framework
for Assessing Health Risks of Environmental Exposures to Children (U.S. EPA, 2006b).
The literature search strategy employed for this compound was based on the Chemical
Abstracts Service Registry Numbers (CASRNs) and at least one common name. Any pertinent
scientific information submitted by the public to the IRIS Submission Desk was also considered
in the development of this document. The relevant literature was reviewed through
January 2010.
-------
2. CHEMICAL AND PHYSICAL INFORMATION
The term cyanide refers to any compound that contains the cyanide ion (CIST), consisting
of a carbon atom triple bonded to a nitrogen atom. Hydrogen cyanide (HCN) is a colorless or
pale blue liquid or gas with a faint bitter almond-like odor, while sodium cyanide (NaCN) and
potassium cyanide (KCN) are white crystalline powders. HCN is a weak acid with a pKa of 9.2;
therefore, HCN and GST can interconvert based on pH and temperature. In solution under
physiological conditions, the majority of HCN is present in the undissociated form. The simple
cyanide salts, KCN and NaCN, are very soluble in water and mildly soluble in ethanol. These
compounds readily dissociate in water, and therefore, exposure to any of these compounds in
aqueous media results in exposure to CIST. For the sake of comparability, doses in this review
are given as CIST unless stated otherwise. Physical properties for HCN and other simple cyanide
salts are summarized in Table 2-1.
The dissociation constants of metallocyanides vary significantly depending on oxidation
states, pH, temperature, and photodegradation (Beck, 1987). As noted above, some of these
compounds, such as NaCN and KCN, dissociate completely when dissolved in water, whereas
others do not.
Other inorganic and organic compounds containing the CN~ group include the nitriles, in
which the CN~ group is covalently bound to the rest of the molecule (e.g., acetone cyanohydrin
[ACH]), the cyanogens (i.e., compounds of the form NC-CN or X-CN, where X is a halogen),
such as cyanogen chloride, and "cyanogenic" substances in some plant-based foods. These
cyanogenic compounds contain cyanogen glycosides that can undergo hydrolysis following
ingestion to produce HCN and other cyanide-containing compounds.
Cyanide originates primarily from anthropogenic sources in the environment, but cyanide
is also released from biomass burning, volcanoes, and natural biogenic processes from higher
plants, bacteria, and fungi (Agency for Toxic Substances and Disease Registry [ATSDR], 2006).
Cyanide is also a component of tobacco smoke and can be present at high concentrations in
structural fires (Steinmaus et al., 2007; Brauer et al., 2006; Tsuge et al., 2000; Brandt-Rauf et al.,
1988). Cyanide compounds are used in a number of industrial processes, including mining,
metallurgy, manufacturing, and photography, due to their ability to form stable complexes with a
range of metals. Cyanide has been employed extensively in electroplating, in which a solid
metal object is immersed in a plating bath containing a solution of another metal with which it is
to be coated, in order to improve the durability, electrical resistance, and/or conductivity of the
solid. HCN has also been used in gas chamber executions and in chemical warfare. NaCN and
KCN are also used as rodenticides. Conversion factors for HCN air concentrations are 1 mg/m3
= 0.90 ppm and 1 ppm =1.11 mg/m3.
-------
Table 2-1. Physical and chemical properties of cyanide compounds
CASRN #
Synonyms
Molecular weight
Form
Chemical formula
Boiling point (°C)
Melting point (°C)
Density (g/mL)/specific gravity
(unitless)
Solubility
Hydrogen
cyanide
74-90-8
Prussic acid,
hydrocyanic acid,
Cyclone B
27
Colorless gas or
liquid
HCN
25.7
-13.4
0.6884
Ethanol, ether
Sodium cyanide
143-33-9
Cyanogran, Cymag,
Cyanobrik, white
cyanide
49
White crystalline
powder
NaCN
1,496
563.7
1.6
Water, ethanol
Potassium
cyanide
151-50-8
Hydrocyanic acid,
potassium salt
65
White lumps or
crystals
KCN
1,625
634.5
1.52
Water, ethanol
Calcium cyanide
592-01-8
Calcyanide,
calcyan, cyanogas,
black cyanide
92
White powder
Ca(CN)2
Not applicable
640
1.85
Water, ethanol,
weak acid
Potassium silver
cyanide
506-61-6
Potassium
dicyanoargentate
199
White crystals
AgK(CN)2
Not found
Not found
2.36
Water, ethanol
Cyanogen
460-19-5
Dicyanogen,
ethanedinitrile,
oxalonitrile
52
Colorless gas
(CN)2
-21.17
-27.9
0.9537
Water, ethanol
-------
Cyanide or cyanogenic compounds are found in many foods. Cyanide compounds occur
naturally as part of sugars or other naturally occurring compounds in certain plant-derived foods,
including almonds, millet sprouts, lima beans, soy, spinach, bamboo shoots, sorghum, and
cassava roots. The parts of these plants that are eaten in the United States, however, contain
relatively low amounts of cyanide (ATSDR, 2006).
Around the world, cassava is a vital staple for about 500 million people. Cassava is a
major source of carbohydrate in parts of Africa, South America, and Southeast Asia. Its starchy
roots produce more food energy per unit of land than any other staple crop, and it can be dried
and ground into flour. Its leaves, commonly eaten as a vegetable in tropical regions, provide
vitamins and protein. However, the storage root of the cassava plant contains linamarin, a
cyanogenic glycoside that is easily hydrolyzed by the enzyme linamarase (a p-glucosidase) to
release HCN. Although HCN can be readily removed during processing of cassava, cyanide
liberated from residual linamarin is associated with goiter in iodine-deficient populations with
chronic intake of cassava-based food products (Taga et al., 2008; Teles, 2002; Abuye et al.,
1998).
Information on the concentration of cyanide in drinking water is available from the
National Drinking Water Contaminant Occurrence Database (U.S. EPA, 2003). In this database,
a cross-sectional study of 16 states was used to develop a statistical estimation indicative of the
national occurrence of contaminants in drinking water. Based on these data, the overall mean
cyanide concentrations in treated surface water and groundwater systems were 2 and 8 |ig/L,
respectively. Cyanide was detected infrequently; the average among the public water systems
that detected cyanide was 60 jig/L (parts per billion), although some systems had levels in the
parts per million range.
Although concentrations of HCN in foods are expected to be low, one study (Fiksel et al.,
1981) estimated that HCN intake from inhalation of air and ingestion of drinking water would be
less than the intake from food. Estimates of the HCN concentration in the total diet of U.S.
populations are not available in the peer-reviewed literature. An average daily SCN~ intake of
16.3 jimol/day was estimated for Koreans based on vegetable intake surveys obtained from the
Korean National Health and Nutrition Examination Survey (Han and Kwon, 2009). ATSDR
(2006) estimated an atmospheric concentration of 170 ppt (188 ng/m3), corresponding to an
inhalation exposure to the general U.S. nonurban, nonsmoking population of 3.8 jig HCN/day,
corresponding to 54 ng/kg-day HCN.
Smokers and those exposed to second-hand tobacco smoke make up a subset of the
general population that may be exposed to elevated levels of HCN. Smokers could be exposed
to 10-400 jig HCN per cigarette, whereas nonsmokers exposed to sidestream smoke could be
exposed to 0.06 to 108 jig HCN per cigarette (ATSDR, 2006). Serum and urinary levels of
thiocyanate (SCIST), the primary metabolite of HCN, are generally about two- to fivefold higher
-------
in smokers versus nonsmokers, indicating significantly elevated cyanide exposure through
tobacco smoke (Steinmaus et al., 2007; Brauer et al., 2006; Tsuge et al., 2000).
-------
3. TOXICOKINETICS
3.1. ABSORPTION
The available data show that cyanide is rapidly and extensively absorbed via the oral,
inhalation, and dermal routes, although quantitative data on the percent or extent of absorption
are limited. Oral absorption has been reported as being lower at lethal doses. Some cyanide
salts, including potassium cyanide (KCN) and sodium cyanide (NaCN), rapidly dissociate in
water. Because hydrogen cyanide (HCN) is a weak acid (pKa of 9.2), the acidic environment in
the stomach favors the nonionized form (HCN) (U.S. EPA, 1992). The nonionized form is also
favored under neutral conditions. Thus, HCN and the dissociated sodium and potassium salts are
predominantly present as HCN at the acidic pH levels of the stomach and lower gastrointestinal
tract. Accordingly, these compounds are presumably absorbed by passive diffusion across the
lipid matrix of the intestinal microvilli. The moderate lipid solubility and small size of the HCN
molecule also indicate that HCN crosses mucous membranes rapidly. HCN is absorbed rapidly
after inhalation, and it penetrates the epidermis. KCN and NaCN are corrosive to the skin, which
can increase dermal absorption. In the absence of such corrosion, however, these ionic forms of
cyanide are absorbed less completely than HCN via the dermal route.
Limited data are available on oral absorption of cyanide in humans. In a case report
(Liebowitz and Schwartz, 1948) of a suicide attempt by an 80-kg male who ingested an
estimated 15-25 mg/kg cyanide ion (CIST) as KCN, the authors estimated that, 2 hours after
ingestion, the patient had 2.5 g CN~ in the body, of which 1.2 g was in the blood, based on a
concentration of 200 mg HCN/L in the blood at that time. This study does not provide any
information on the disposition of the remaining cyanide.
Gettler and Baine (1938) reported data indicating that, at doses above historical lethal
doses, absorption decreases with increasing dose levels. Absorption was estimated at 19.5, 18.1,
and 15.7% in people estimated to have ingested 297, 557, and 1,450 mg HCN in suicide
attempts. The low absorption at the highest dose may have been due to death occurring before
absorption was complete. The assumption that absorption is lower at higher lethal doses is
supported by a case report where absorption was approximately 82% in an individual, estimated
as having ingested 30 mg HCN, who died more than 3 hours after exposure.
Only limited data are available on absorption of inhaled cyanide by humans. Landahl
and Herrmann (1950) measured the pulmonary retention of HCN in 10 volunteers exposed to
concentrations of 0.0005-0.02 mg/L (0.5-20 mg/m3) for up to 3 minutes. All subjects breathed
through their mouths. The percent retained in the lung (and, presumably, the percent absorbed)
was approximately 60% and ranged from 58 to 77% among people who were breathing
normally. Rapid, shallow breathing appeared to decrease absorption.
-------
Dermal absorption of HCN gas has also been observed in humans. Drinker (1932)
reported that three workers who entered an atmosphere containing 2% HCN (20,000 ppm
[22,100 mg/m3]) became dizzy and weak and were on the verge of unconsciousness, despite
wearing gas masks providing respiratory protection. The observed effects were attributed to
dermal absorption of the gas. Potter (1950) reported on a worker, wearing respiratory protection
and protective clothing, who was accidentally exposed to liquid HCN. Within 5 minutes, the
worker became dizzy, had difficulty breathing, and fell unconscious.
Animal data also indicate extensive absorption. In male Sprague-Dawley rats treated by
gavage with 1 mg/kg KCN, a peak blood level of 6.2 nmol/mL CN~ (160 |ig/L) was observed
2 minutes after treatment, indicating rapid absorption (Leuschner et al., 1991). As described
above for oral poisonings in humans, absorption by dogs decreased as the dose increased well
above lethal levels (Gettler and Baine, 1938). Oral absorption was essentially comparable
(16.6 and 15.7%) in dogs that ingested 100 and 50 mg HCN, respectively. However, absorption
increased to 72% in a dog that ingested 20 mg HCN (1.5 mg/kg).
No animal studies were located that quantitatively evaluated the rate or extent of
absorption via the inhalation or dermal routes. However, Walton and Witherspoon (1926)
reported toxic effects and death in guinea pigs exposed by holding the open end of a test tube
containing liquid HCN against their shaved stomachs and concentration-related signs of toxicity
(including death) in dogs given whole-body exposure to HCN (excluding the head). These
results support the human data, demonstrating that absorption via the dermal route occurs in
animal species and can produce toxic effects, including death.
3.2. DISTRIBUTION
Cyanide distributes rapidly and uniformly throughout the body following absorption.
HCN enters the systemic circulation when inhaled or dermally absorbed (Yamamoto et al., 1982;
Potter, 1950; Drinker, 1932). Limited qualitative and quantitative data are available regarding
the tissue distribution of cyanide in humans from inhalation exposure studies to high doses of
cyanide. For example, cyanide was found in the lung, heart, blood, kidneys, and brain of humans
who died following cyanide inhalation (Gettler and Baine, 1938). In addition, Knowles and Bain
(1968) evaluated the relationship between blood concentrations of cyanide and short-term
accidental exposures to lethal levels of HCN in human case reports. Air concentrations of
>300 ppm (333 mg/m3 HCN), >200 ppm (222 mg/m3 HCN), >100 ppm (111 mg/m3 HCN), and
>50 ppm (55 mg/m3 HCN) corresponded to blood concentrations of >10, >8-10, >3-8, and >2-
4 mg/L, respectively. The authors noted that there is considerable variability in this relationship,
presumably reflecting both interindividual variability and uncertainty of exposure duration and
concentrations estimated or measured retrospectively.
Limited data on distribution of cyanide in humans, following oral exposure, are available.
Immediately following oral cyanide exposure, the stomach contents appear to contain the highest
-------
concentration of cyanide. Other tissues containing cyanide included the liver, brain, spleen,
blood, kidneys, and lungs (Ansell and Lewis, 1970; Gettler and Baine, 1938).
Several animal studies are available that demonstrate the tissue distribution of cyanide
following both inhalation and oral exposures. In dogs exposed to lethal concentrations of
cyanide by inhalation, the highest concentrations of cyanide were found in the lungs, blood, and
heart (Gettler and Baine, 1938), while in rabbits exposed to 2,714 ppm HCN (3,000 mg/m3) for
5 minutes by inhalation, the highest tissue concentrations were detected in the heart, lung, and
brain, with lower levels in the spleen and kidneys (Ballantyne, 1983, as cited in ATSDR, 2006).
In rats and rabbits exposed by the oral route, the highest tissue concentrations of cyanide
were in the liver, lung, blood, spleen, and brain (Ballantyne, 1983, as cited in ATSDR, 2006;
Ahmed and Farooqui, 1982; Yamamoto et al., 1982). Yamamoto et al. (1982) compared cyanide
distribution in rats following gavage exposure to NaCN (7 or 21 mg/kg CN) or inhalation
exposure to HCN (average of 356 or 1,180 ppm, equivalent to 393 or 1,303 mg/m3 HCN,
respectively). These exposure levels resulted in death in <10 minutes. Elevated concentrations
were found in all tissues evaluated following exposure via either route, but the relative
concentrations were route dependent. There was also some dose dependence, which may have
been related to the faster time to death at higher exposures for each route (approximately 10
minutes at the lower exposure levels vs. 3-5 minutes at the higher levels). Focusing on the lower
oral dose, the highest tissue concentration of cyanide following exposure was in the liver,
followed by the blood and lungs, and then the spleen and brain. After inhalation exposure, the
highest concentration was in the lungs, followed by the blood and liver, and then spleen and
brain. The kidney was not evaluated for either exposure route. The route-specific difference
may be related to first-pass metabolism in the liver, following oral dosing, and initial deposition
at the portal of entry, following inhalation exposure.
Okoh and Pitt (1982) investigated the tissue distribution of cyanide in rats fed KCN in the
diet at 77 |imol/day (approximately 5.5 mg/kg-day CN) for 3 weeks and then injected
intraperitoneally (i.p.) with radiolabeled NaCN. Radioactivity was widely distributed, with the
highest concentrations in the gastrointestinal tract, blood, kidneys, lungs, spleen, and liver.
Radioactivity appeared in the stomach as early as 10 minutes after injection, with 18% of the
injected dose found in the stomach contents within 60 minutes of dosing. More than 80% of the
radioactivity in the stomach was present as thiocyanate (SCIST), with small portions present as
cyanide and radiolabeled carbon dioxide.
In a subchronic study, male Sprague-Dawley rats (26-40/group) received KCN in their
drinking water at doses of 0, 40, 80, or 140/160 mg/kg-day for 13 weeks (Leuschner et al.,
1991). Blood was collected every 2 weeks for analysis of CN" and SCIST levels; both were found
to be dose related. Within each dose group, however, the levels of both cyanide and thiocyanate
remained fairly constant over the 13-week exposure period. Cyanide levels in the blood were
16-25 nmol/mL CN" (420-650 |ig/L); thiocyanate levels were 341-877 nmol/mL SCN~ (20-
-------
51 mg/L). Small amounts of thiocyanate were also detected in the control animals at
concentrations of 11-53 nmol/mL SCIST (0.64-3.1 mg/L) in plasma.
Howard and Hanzal (1955) exposed rats to HCN in the diet at an average daily dose of up
to 10.8 mg/kg-day CN~ for 2 years and found virtually no cyanide in the plasma or kidneys.
Cyanide was detectable in the red blood cells (RBCs) of less than half of the rats at an average
concentration of 2 jig/100 g tissue. However, as the food was treated with HCN gas, some
volatilization likely occurred.
McMillan and Svoboda (1982) incubated cyanide with washed erythrocytes, resuspended in
phosphate-buffered saline containing glucose and bovine serum albumin, and found that cyanide
concentrated in the RBCs. The major portion of cyanide in blood is sequestered in the
erythrocytes, and a relatively small proportion is transported via the plasma to target organs
(International Programme on Chemical Safety [IPCS], 2004). The RBC:plasma ratio for cyanide
is approximately 200:1 (IPCS, 1992). Binding to RBCs is primarily due to cyanide reacting with
ferric iron (Fe3+) in methemoglobin to form the nontoxic complex cyanomethemoglobin (Chen
and Rose, 1952).
In rats treated orally with [14C]KCN (5 mg/kg), radioactivity levels in plasma and whole
blood were initially (at 3 hours) much higher than levels in RBCs (Farooqui and Ahmed, 1982).
Levels in plasma and whole blood decreased rapidly, and red cell levels increased slightly, so
that at 24 hours, plasma levels were only slightly higher than whole blood levels. Most of the
radioactivity in the red cells was in the heme fraction of hemoglobin rather than the membranes.
The reason for the finding of higher levels in the plasma than blood in this study is not clear, but
it may have been due to differences in sampling times. Approximately 60% of the cyanide in
plasma is bound to protein (IPCS, 1992).
Limited information indicates that cyanide can cross the placenta. Petti grew et al. (1977)
compared cyanide and urinary thiocyanate levels in a small group of smoking and nonsmoking
pregnant women matched for age, height, parity, and social class. Maternal plasma and urinary
thiocyanate levels were statistically significantly increased in smokers during gestation at
weeks 28, 32, and 36; at delivery, only plasma thiocyanate was measured and was also
statistically higher in mothers who smoked. Mean urinary thiocyanate levels of neonates of
smoking mothers were elevated compared to those of nonsmokers (40.6 compared to
23.1 |imol/L, respectively), although the difference was not statistically significant, probably due
to the small sample size (n = 10).
Lactational transfer of cyanide and thiocyanate has been shown to occur in goats. Soto-
Blanco and Gorniak (2003) dosed lactating goats with 0, 1.0, 2.0, or 3.0 mg/kg-day KCN
(equivalent to 0, 0.4, 0.8, or 1.2 mg/kg-day CIST) from lactation days 0 to 90 and measured whole
blood cyanide and thiocyanate concentrations on lactation days 30, 60, and 90. Both whole
blood cyanide and plasma thiocyanate concentrations were increased in a dose-dependent
manner in treated mothers, with mixed results regarding time dependence. In the offspring, both
10
-------
blood cyanide and plasma thiocyanate increased with increasing maternal cyanide dose at
lactation day 30 and decreased with lactation time. By lactation day 90, the concentration of
these compounds in the blood/plasma of the offspring was low or undetectable. The study
authors attributed these findings to a decrease in milk consumption, accompanied by a
concomitant increase in solid food (grass and feed) during the latter part of lactation.
Small levels of cyanide are normally present in blood plasma at 0-140 jig/L and in other
tissues at <0.5 mg/kg CN~ (ATSDR, 2006; Feldstein and Klendshoj, 1954). Chandra et al.
(1980) found that nonsmokers with no occupational exposure to cyanide had an average of
3.2 jig/100 mL CW (32 |ig/L) in blood; smokers had average blood cyanide levels of
4.8 jig/100 mL (48 |ig/L). The background level is attributed to exposure to cyanogenic food,
vitamin 812, and passive tobacco smoke. Cyanide preferentially binds to hemoglobin in RBCs
but does not appear to accumulate in tissues after chronic oral exposure to inorganic cyanides
(Leuschner et al., 1991; Chen and Rose, 1952).
3.3. METABOLISM
The major metabolic pathway for cyanide is conversion to the less acutely toxic
compound, thiocyanate, primarily by rhodanese, with some conversion occurring via
3-mercaptopyruvate sulfur transferase. Conversion to thiocyanate accounts for 60-80% of a
cyanide dose. Minor pathways include incorporation into a 1-carbon metabolic pool or
conversion to 2-aminothiazoline-4-carboxylic acid (ATSDR, 2006). Conversion to
2-aminothiazoline-4-carboxylic acid via reaction with cystine accounted for approximately 15%
of an injected dose of cyanide in rats (Wood and Cooley, 1956). Small amounts are also
converted to carbon dioxide in exhaled air or excreted unchanged as HCN in exhaled air. These
pathways are shown in Figure 3-1.
Rhodanese, a mitochondrial enzyme that converts cyanide to thiocyanate, facilitates
transfer of a sulfur atom to cyanide from a sulfane-sulfur donor. Because these donors must
contain an S-S bond, glutathione, thiosulfate, and cystine are sulfur donors for rhodanese,
whereas the thiols, cysteine, and reduced glutathione are not donors. Rhodanese is widely
distributed throughout the body. Using immunohistochemical staining techniques, rhodanese in
rabbits has been located in the liver, where it is most abundant in the hepatocytes near blood
vessels (Sylvester and Sander, 1990). It was also found in the lung, localized in epithelial cells
that formed the barrier between inhaled air and blood vessels. In the kidneys, rhodanese was
present in tubules closest to the glomeruli. The authors concluded that sites with the greatest
abundance of rhodanese are located to maximize conversion of cyanide to thiocyanate following
both oral and inhalation exposure. The presence of rhodanese in these tissues also indicates the
importance of first-pass metabolism in determining the toxicity of inhaled and ingested cyanide.
11
-------
CN-
Major Path (80%) _,.
Cyanide ^ —*- Thloc^nate
J - (SON')
Minor Path
2-Aminothiazoline-4-carboxylic acid &
2-Iminothiazolidine-4-carboxylic acid
HCN
Hydrogen Cyanide
(in expired air)
HCN
(Pool)
CNO-
Cyanate
CO,
z
Carbon dioxide
•Urinary Excretion
HCOOH
Formic Acid
Some
excreted
in urine
Metabolism of one-
carbon compounds
Formates
Source: Adapted from Ansell and Lewis (1970).
Figure 3-1. Cyanide primary metabolic pathways.
Metabolism of cyanide by rhodanese exhibits zero-order kinetics relative to cyanide; the
concentration of sulfur-containing donor molecules is the rate-limiting factor. The primary
endogenous sulfur donor is thiosulfate; others include glutathione and cystine. Schulz et al.
(1982) evaluated the metabolism of cyanide in humans continuously infused with the
hypotensive drug sodium nitroprusside, Na2[Fe(CN)sNO]-2H2O, which completely releases the
CIST in the blood. The authors estimated that the detoxification rate of cyanide in humans
(in the absence of antidotes) is about 1 |ig/kg-minute. McNamara (1976) estimated the
detoxification rate in humans as 17 |ig/kg-minute based on a study in men injected intravenously
(i.v.) with HCN. Lawrence (1947, reported in an extended abstract) found that continuous i.v.
infusion of NaCN into dogs at a rate of 0.013 mg/kg-minute (apparently as milligrams of CIST
but not explicitly stated) was "tolerated over 37 hours" and speculated that this rate of infusion
could be tolerated indefinitely. Infusion rates of 0.028 mg/kg-minute or higher resulted in
lethality. Based on these findings, the whole-body rate of cyanide detoxification in dogs can be
estimated to be approximately 13 jig/kg-minute. The actual rate may be lower because the
12
-------
portion of cyanide that remained unmetabolized at this dose may have been insufficient to cause
lethality. These findings suggest that the rate of cyanide detoxification in humans and dogs may
be similar.
Devlin et al. (1989a) evaluated rhodanese activity in rat liver and skeletal muscle. Using
histochemical staining techniques, the authors determined that only low levels of rhodanese
activity were present in the blood vessels. In contrast, high levels of rhodanese activity were
detected in the liver and skeletal muscle. Although the activity of rhodanese in muscle was
lower than in the liver, the authors concluded that the total skeletal muscle mass makes a
significant contribution to whole-body metabolism of cyanide. In a follow-up study in perfused
liver and hind-limb muscle, Devlin et al. (1989b) observed that the liver cleared 80% of the
available cyanide compared to 18% for the hind limbs. However, when the hind-limb data were
extrapolated to total muscle mass, muscle cleared cyanide 2.6-fold faster than did liver in the
absence of exogenous thiosulfate. When thiosulfate was included in the perfusion medium, liver
clearance was dependent on flow rate, but muscle tissue clearance was unaffected. Westley
(1981) found that purified bovine liver rhodanese has a high turnover rate of almost
20,000/minute in vitro (i.e., 1 mol of rhodanese could convert 20,000 mols of cyanide to
thiocyanate in 1 minute). This high turnover rate, coupled with the basal amount of rhodanese in
the liver and other tissues, means that the rate of cyanide metabolism should not depend critically
on the enzyme content of the tissue. Therefore, the limiting factor for cyanide metabolism is the
availability of the sulfur donor rather than the rhodanese metabolic capacity. Similarly,
differences between muscle and liver in ability to detoxify cyanide appear to be related to the
availability of sulfur donors.
Lewis et al. (1991) observed the presence of rhodanese in the epithelium of human nasal
tissue. Rhodanese activity in human nasal epithelium was higher in nonsmokers than smokers.
Individual enzyme kinetic data (Vmax and Km) suggested that decreased activity in smokers may
be due to decreased affinity. In kinetic studies with adequate sulfur present, rhodanese in human
nasal tissue exhibited a higher affinity (lower Km) for cyanide and a lower maximum velocity
(lower Vmax), compared to rhodanese in human liver. Human rhodanese exhibited a higher Km
and lower Vmax than did rat rhodanese. Dahl (1989) also found that rat nasal tissue exhibited
high levels of rhodanese activity, particularly in the olfactory region, which had almost sevenfold
more activity on a per milligram mitochondrial protein basis than did rhodanese in rat liver.
Rhodanese activity was also observed in the respiratory tracts of dogs, particularly in the nasal
cavity (Aminlari et al., 1994).
The tissue distribution and activity of rhodanese is highly variable among species. In
dogs, Himwich and Saunders (1948) observed that the highest activity of rhodanese was
observed in the adrenal glands, followed by the liver. The brain, spinal cord, kidneys, and testes
also had large amounts of rhodanese. In general, monkeys, rats, and rabbits had much higher
rhodanese activity (mg CN~ converted to SGST per g tissue) than dogs, with the liver and
13
-------
kidneys containing the highest activity. Drawbaugh and Marrs (1987) also studied the tissue
distribution of rhodanese in several species, including marmosets, rats, hamsters, rabbits, guinea
pigs, dogs, and pigeons. The highest rhodanese activities were found in rats, hamsters, and
guinea pigs; the lowest were found in pigeons, marmosets, and dogs. Except for rabbits,
rhodanese activity was higher in the liver than in the kidneys of the species studied. However,
the authors noted that the biological significance of species differences in rhodanese activity is
unclear.
The study by Drawbaugh and Marrs (1987) has been used to suggest that the dog is not
an appropriate animal model for cyanide toxicity in humans, due to significantly lower levels of
rhodanese in this species as compared with humans. However, as noted above, the amount of
sulfur donor, not the amount of rhodanese itself, is the rate-limiting factor for detoxification of
cyanide by rhodanese, even at bolus doses resulting in high acute toxicity. Schulz (1984)
reported that the rate of cyanide detoxification in humans is slower than the rate in rodents or
dogs despite the higher levels of rhodanese in humans. Other data (McNamara, 1976) suggest
that the cyanide detoxification rate in humans is slightly higher than in dogs. Furthermore,
urinary concentrations of thiocyanate have been shown to be higher in dogs than rats (NTP,
1993; Kamalu, 1993). Lower thiocyanate levels would be expected if metabolism via the
rhodanese pathway were limited in this species. Although this analysis did not normalize by
urine specific gravity or other factors, it suggests that cyanide was metabolized to a similar
degree in dogs and rats.
Chronic exposure to cyanide resulted in increased rhodanese levels in rabbits, suggesting
that rhodanese is inducible, at least in this species (Okolie and Osagie, 1999). Alternatively, the
increased levels could be due to other factors, such as increased protein stability. Data were not
located regarding whether chronic exposure increases rhodanese levels in other species.
Several polymorphisms in rhodanese have been identified in human populations,
although only a minimal effect on cyanide detoxification was detected (Billaut-Laden et al.,
2006). A second enzyme that converts cyanide to thiocyanate is mercaptopyruvate
sulfurtransferase (MPST). This enzyme differs from rhodanese in that it catalyzes the transfer of
sulfur from an organic thiol to cyanide (Wing and Baskin, 1992). Therefore, this enzyme breaks
a carbon-sulfur bond to facilitate transfer of sulfur to cyanide, whereas rhodanese breaks a
sulfur-sulfur bond. MPST is most active at pH 9.5, while rhodanese is most active at pH 8.6,
which is closer to physiological pH. MPST also appears to have a different tissue distribution
from that of rhodanese; this enzyme has been reported as being located in the RBCs and kidneys.
MPST is located in both the mitochondria and the cytosol, making it more accessible for
conversion of cyanide than rhodanese, which occurs only in the mitochondria (Wing and Baskin,
1992). Support for the role of MPST in cyanide detoxification was provided in in vitro studies
by Huang et al. (1998). These authors demonstrated that addition of L- or D-cysteine to
hepatocytes in cell culture prevented cyanide cytotoxicity and enhanced the formation of
14
-------
thiocyanate. Mercaptopyruvate and thiocystine, metabolites of L- and D-cysteine, are substrates
of MPST. Huang et al. (1998) observed that, when formation of these metabolites in isolated
hepatocytes was prevented, the formation of thiocyanate was also inhibited. However, it is not
clear whether MPST directly transfers sulfur to cyanide or whether it acts indirectly by
transferring sulfur to albumin in the liver. The modified albumin could then be excreted to form
a sulfane-sulfur pool that is available to react with cyanide via rhodanese (Wing and Baskin,
1992).
Although the reaction of rhodanese with cyanide is irreversible, thiocyanate can be
converted back to cyanide and sulfate by the action of thiocyanate oxidase located in the RBCs,
lymphocytes, mammary gland, and thyroid (Wood, 1975). Thiocyanate oxidase has been found
in the erythrocytes of humans, dogs, rabbits, and rats (Goldstein and Rieders, 1953). This
enzyme catalyzes the reaction of hydrogen peroxide and thiocyanate to form cyanide and sulfate.
In addition, these enzymes produce an intermediate oxidation product of thiocyanate, the OSCIST
ion known as hypothiocyanate, which reacts with cyanide to form cyanate (CNCT), which is then
hydrolyzed to ammonia and carbon dioxide (Wood, 1975).
The minor pathway shown in Figure 3-1 involves the spontaneous reaction of cyanide
with cystine to yield 2-aminothiazoline-4-carboxylic acid, which tautomerizes to
2-iminothiazolidine-4-carboxylic acid. This pathway accounted for approximately 15% of the
cyanide dose in a female rat receiving daily i.p. injections of NaCN for 8 days (Wood and
Cooley, 1956). In another experiment in the same publication, the percentage metabolism via
this pathway was higher when rats were injected i.v. with labeled cystine and subsequently were
administered NaCN subcutaneously.
The mean blood thiocyanate level in smokers with untreated tobacco amblyopia (a
condition causing visual defects that has been attributed to cyanide exposure) was significantly
lower than the concentration in smokers overall, suggesting that people with this condition have
a decreased ability to convert cyanide to thiocyanate (Pettigrew and Fell, 1973). Blood cyanide
levels were low in both smokers and nonsmokers in this study, with no significant effect of
smoking or tobacco amblyopia, perhaps because approximately 1-2 hours had elapsed between
the time the last cigarette was smoked and when cyanide levels were measured. The authors
suggested that the excess cyanide was bound up as cyanocobalamin (one form of vitamin 812),
but they did not investigate this hypothesis. Cyanide also reacts with methemoglobin
(hemoglobin that has been oxidized either by normal metabolism or by xenobiotic oxidant
stressors) in RBCs to form cyanomethemoglobin. Schulz (1984) noted that, theoretically, 1 g of
methemoglobin can bind approximately 60 jimol of HCN and that 1 L of erythrocytes should be
able to bind approximately 50-200 jimol (1.4-5.4 mg) HCN at physiological levels of
methemoglobin (0.25-1%). This readily reversible reaction is considered to be a naturally
occurring detoxification pathway for low levels of cyanide in the blood (Lundquist et al., 1985)
and forms the basis of the first phase of treatment for acute cyanide poisoning, which consists of
15
-------
administering amyl nitrite or sodium nitrite to cyanide-poisoned individuals (Klaassen, 2001).
Amyl nitrite and sodium nitrite are oxidants that increase the conversion of hemoglobin to
methemoglobin, thus providing a sink for CIST to reduce interaction with tissue cytochrome c
oxidase. Under this treatment approach, cyanide is then detoxified by slow release from
cyanomethemoglobin and cytochrome c oxidase and subsequent conversion by the enzyme
rhodanese to SGST, which has much lower acute toxicity than cyanide. Sodium thiosulfate
administered as the second phase of treatment for acute cyanide poisoning accelerates
detoxification by supplying a sulfur substrate for the reaction. Other substances used to detoxify
cyanide include hydroxycobalamin (vitamin Bi2a), an antidote used outside the United States that
binds cyanide to form cyanocobalamin (vitamin 812), and cobalt edetate, which is used as an
antidote in some countries due to the high affinity of cobalt for cyanide (Klaassen, 2001).
3.4. ELIMINATION
Data in humans and animals indicate that cyanide is primarily excreted in the urine as
thiocyanate following both inhalation and oral exposure. Smaller amounts are excreted as
urinary cyanide or as HCN or carbon dioxide in exhaled air. Following occupational exposure to
0.19-0.75 ppm HCN, urinary thiocyanate levels in nonsmoking exposed workers were
approximately 7 times the levels in nonsmoking controls (Chandra et al., 1980). Urinary cyanide
levels were also elevated in the exposed workers, but they were approximately 2 orders of
magnitude lower than thiocyanate levels.
Following a single subcutaneous injection of rats with [14C] KCN, 89% of the excreted
radioactivity was detected in urine within 24 hours; about 4% of the excreted radioactivity was
expired in air, primarily as carbon dioxide (Okoh, 1983). The authors found that 71-79% of the
urinary activity was in the form of thiocyanate. The excretion pattern was not affected by prior
exposure to cyanide in diet for 6 weeks.
In a related study of rats injected i.p. with radiolabeled NaCN after being fed KCN in the
diet at 77 jimol/day (approximately 5.5 mg/kg-day CN) for 3 weeks (Okoh and Pitt, 1982), 86%
of the radioactivity in the expired air was present as carbon dioxide and 14% was present as
HCN. Boxer and Rickards (1952) also observed that exhaled air contained both radiolabeled
HCN and carbon dioxide after dogs were injected subcutaneously with radiolabeled NaCN.
However, the primary path of excretion was still in the form of urinary thiocyanate, although
small amounts of cyanide and cyanocobalamin were also found in the urine. Sylvester et al.
(1983) treated dogs with i.v. NaCN and found that <1% of the total cyanide dose was eliminated
through exhaled air.
Leuschner et al. (1991) evaluated the elimination of KCN following both acute and
subchronic exposure. For the acute study, three male Sprague-Dawley rats were treated by
gavage with 1 mg/kg KCN. Blood was collected at regular intervals for up to 1 hour following
administration. A peak blood level of 6.2 nmol/mL (160 |ig/L) CN" was observed 2 minutes
16
-------
after treatment; by 60 minutes, the blood levels had dropped to the analytic detection limit. The
authors calculated an elimination half-life of 14 minutes.
In the subchronic portion of Leuschner et al. (1991), male Sprague-Dawley rats (26-
40/group) received KCN in their drinking water at doses of 0, 40, 80, or 160 mg/kg-day for
13 weeks. Blood was collected every 2 weeks for analysis of cyanide and thiocyanate levels.
Urine was collected over a 16-hour period during weeks 6 and 13 of the study to determine
cyanide and thiocyanate levels. Similar patterns of excretion were observed for both urinary
levels of cyanide and thiocyanate. A dose-response relationship was observed for the
concentration of both cyanide and thiocyanate in urine, and a small amount of thiocyanate was
observed in the urine of the controls. The levels of cyanide in the urine were much lower than
the thiocyanate levels; the ratio of cyanide to thiocyanate was about 1 to 1,000. Approximately
11% of the administered cyanide was eliminated per day as urinary thiocyanate during the dosing
period, while only about 0.003% was excreted per day unchanged. The study authors did not
report how they estimated the percent of total dose eliminated; radiolabeled material was not
used. Elimination half-life was not calculated for the subchronic study. Blood levels of cyanide
and thiocyanate were fairly consistent with time. Some elimination may have occurred as
exhaled HCN or carbon dioxide, but data indicate that this route accounts for <10% of a dose of
cyanide following acute dosing (Okoh, 1983). The study authors (Leuschner et al., 1991) also
noted that the percent of administered cyanide excreted via the urine was unchanged between
weeks 6 and 13, indicating that detoxification pathways were not saturated and the mode of
cyanide excretion was not affected over this period.
In cynomolgus monkeys exposed via inhalation for up to 30 minutes to approximately
100-170 mg/m3 HCN, levels of cyanide in the blood remained nearly constant after
approximately the first 10-15 minutes of exposure and for 60 minutes following termination of
exposure (Purser et al., 1984). Thus, the half-life under these conditions was longer than in the
Leuschner et al. (1991) gavage study in rats. The difference may have been due to saturation of
metabolism, first-pass metabolism following oral exposure, or species-related differences.
3.5. PHYSIOLOGICALLY BASED TOXICOKINETIC MODELS
A pharmacokinetic analysis of the distribution and metabolism of cyanide was conducted
in dogs following a single i.v. dose (Sylvester et al., 1983). Dogs (six/group) were administered
i.v. saline, NaCN (20.4 |imol/kg), or sodium thiocyanate (12.3 |imol/kg); cyanide concentration
was determined in whole blood, and thiocyanate concentration was determined in plasma. Blood
levels of CIST and SCIST measured after administration were used to develop a pharmacokinetic
model in dogs. The conversion of cyanide to thiocyanate was found to follow first-order
kinetics. Three hours following i.v. dosing with cyanide, 90% of the total dose had been
converted to thiocyanate. The half-life of thiocyanate was determined to be 29 hours. The
17
-------
authors also found that <8% of the cyanide dose was eliminated through non-thiocyanate routes
and only 1% of the total cyanide dose was eliminated through exhalation.
Additionally, some data exist on the comparative toxicokinetics of cyanide and
thiocyanate in several species (Sousa et al., 2003). Rats (n = 42), pigs (n = 6), and goats (n = 6)
were studied up to 24 hours after a single gavage dose of 3.0 mg/kg KCN. Cyanide was quickly
absorbed in all species. The peak plasma concentration of cyanide was highest in goats,
followed by rats and pigs. Goats also had the highest volume of distribution, highest area under
the curve (AUC), and slowest elimination compared with the other two species. The similarities
in absorption data between species indicated that pH differences between the monogastric
stomachs of rats and pigs (pH 1-2) and ruminant stomachs (pH 6.8) did not noticeably impact
absorption of cyanide. Toxicokinetic parameters for thiocyanate indicated the peak plasma
concentrations and AUC to be greatest in rats, followed by goats and pigs. Blood levels of
cyanide in each species indicate rapid decreases in cyanide blood concentration by 3 hours
following dosing, with the half-life of elimination for thiocyanate for all species about 9-
11 times longer.
18
-------
4. HAZARD IDENTIFICATION
4.1. STUDIES IN HUMANS—EPIDEMIOLOGY, CASE REPORTS, CLINICAL
CONTROLS
4.1.1. Acute Oral, Inhalation, and Dermal Studies
The effects of acute, high-level exposure to cyanide (CIST) are well characterized
(reviewed in ATSDR, 2006; IPCS, 2004; U.S. EPA, 1992; and Hall and Rumack, 1990).
Although acute oral doses of cyanide cause cardiovascular, respiratory, and neurophysiological
changes, the brain appears to be the organ most sensitive to acute cyanide toxicity (IPCS, 2004).
Several studies of the acute effects of cyanide in humans, following suicide attempts or
accidental poisoning by the oral and inhalation routes, provide additional details, although most
such studies include only a limited characterization of exposure. Symptoms of severe cyanide
poisoning include vomiting, nausea, weakness, confusion, lethargy, cyanosis, weak and ataxic
movements, increased respiratory and heart rates progressing to coma with respiratory
depression, seizures, cardiovascular collapse, and death. The principal feature of the acute
toxicity profile for cyanide includes lethality by all routes of administration, with a steep rate-
dependent dose-response curve. Death from cyanide poisoning is believed to result from central
nervous system (CNS) depression, subsequent to inhibition of brain cytochrome oxidase activity
(Way, 1984). The toxic effects and lethality associated with acute exposure to CW in humans
and animals are generally similar and are believed to result from inactivation of cytochrome
oxidase and inhibition of cellular respiration during the terminal reaction of the electron transport
chain. This inhibition prevents the formation of adenosine triphosphate (ATP) via oxidative
phosphorylation. The IPCS (2004) noted that the lowest reported oral lethal dose in humans is
0.54 mg/kg body weight; the average absorbed dose at the time of death was estimated at
1.4 mg/kg body weight (calculated as hydrogen cyanide [HCN]). Rapid recovery from relatively
low, short-term inhalation exposures often occurs once the exposed individual is moved to fresh
air. Individuals suffering from higher oral and inhalation exposures may benefit from
supplemental oxygen and the use of antidotes. Oral exposure results in slower absorption,
passage to the liver, and faster detoxification. If the patient responds to treatment and survives,
recovery is usually prompt and complete; however, delayed neurological symptoms, including
neuropsychiatric manifestations and Parkinson-type disease, can occur (IPCS, 2004). Exposure
to lower levels can cause flushing, light-headedness, dizziness, headache, and other symptoms
indicative of hypoxia (Wolfsie and Shaffer, 1959).
Liebowitz and Schwartz (1948) reported a case of cyanide poisoning following ingestion
of an estimated 38-63 mg/kg of potassium cyanide (KCN) (15-25 mg/kg CIST). The patient was
comatose with muscular rigidity and a thready pulse on admission. By 8 hours after admission,
the patient was alert and the symptoms had begun to subside (with the exception of weakness,
19
-------
nausea, and an enlarged heart). The authors suggested that the reason that the patient recovered
from exposure to a dose that was about 30 times the estimated lethal dose may have been
because he was a chemist who frequently immersed his hands in thiosulfate. Therefore,
exposure to thiosulfate may have had an antidotal effect.
Several authors (Grandas et al., 1989; Rosenberg et al., 1989; Carella et al., 1988; Uitti et
al., 1985) reported the development of symptoms of Parkinsonism in patients who recovered
from ingestion of a single dose of cyanide. The four cases included an 18-year-old male who
ingested 5.6-7.6 mg/kg cyanide in a suicide attempt (Uitti et al., 1985), a 46-year-old woman
who ingested an unreported amount of cyanide by accidental poisoning (Carella et al., 1988), a
46-year-old man who ingested 8.6 mg/kg cyanide in a suicide attempt (Rosenberg et al., 1989),
and a 39-year-old man who ingested an unknown amount of cyanide in a suicide attempt
(Grandas et al., 1989). In all cases, the patients recovered from the acute symptoms of cyanide
poisoning with treatment, and neurologic examinations immediately following the poisoning
were normal. Follow-up neurologic examination at times of 3 weeks (Rosenberg et al., 1989),
4 months (Uitti et al., 1985), or 1 year (Grandas et al., 1989; Carella et al., 1988), however,
revealed that the patients had developed symptoms of parkinsonism, including generalized
rigidity, bradykinesia, tremors of tongue and eyelids, slow-shuffling gait, and a weak dysphonic
voice. A computerized tomography scan or magnetic resonance imaging showed lesions in the
putamen and globus pallidus regions of the brain (Grandas et al., 1989; Rosenberg et al., 1989;
Carella et al., 1988; Uitti et al., 1985). Similarly, Lam and Lau (2000) reported mild impairment
of recent memory and concentration, which was confirmed by neurological testing, a year after a
19-year-old woman experienced an episode of acute inhalation exposure to cyanide.
Potter (1950) reported a case history of a worker accidentally exposed via inhalation to an
undetermined concentration of HCN. Early symptoms were dizziness, dyspnea, and weakness of
the legs followed by a period of deep unconsciousness accompanied by absent reflexes,
stertorous respiration, rapid pulse, fixed and unreactive pupils, and convulsions. The subject
recovered with treatment, and no after-effects were reported. In another case report, a worker
was found unconscious, lying in tank sludge after working without protective gear in a plating
tank containing silver-cyanide sludge (Singh et al., 1989). The duration of exposure was
unknown, but the tank air was later measured to contain 200 ppm HCN (220 mg/m3 HCN). He
had dermal evidence of chemical burns and was not breathing, with a rapid pulse, fixed and
dilated pupils, and no recorded blood pressure or response to pain. Blood cyanide was
804 |imol/L within 0.5 hours of hospital arrival and decreased to 15 |imol/L at 18 hours after
arrival and following detoxification efforts. The patient died within 3 days despite extensive
treatment.
Cyanide is readily absorbed through the skin (Lam and Lau, 2000; Potter, 1950);
therefore, systemic toxicity can readily result from dermal exposure to cyanide fumes or direct
dermal contact with HCN. A case report of a worker accidentally exposed to a brief stream of
20
-------
liquid HCN on his hand (amount not specified) reported that the worker became deeply
unconscious within 5 minutes of exposure (Potter, 1950). Breathing was hoarse, his face was
flushed, and reflexes were absent. The subject recovered with sodium nitrite and sodium
thiosulphate treatment.
Dizziness, weakness, and a throbbing pulse were reported when three workers wearing
gas masks entered an atmosphere containing 2% HCN gas (Drinker, 1932). These effects were
attributed to the dermal absorption of the gas. The men developed symptoms of dizziness,
weakness, and throbbing pulse after about 10 minutes of exposure and eventually became
unconscious. The symptoms persisted for several hours following exposure.
Relatively mild symptoms of cyanide poisoning (flushing, dizziness, headache, throat
discomfort, chest tightness, skin itchiness, and eye irritation) were reported in firefighters who
were wearing self-contained breathing apparatus when they responded to a HCN gas release
(Lam and Lau, 2000). The effects were attributed to dermal cyanide absorption and direct
contact with skin and eyes.
4.1.2. Subchronic and Chronic Oral Studies
No subchronic or chronic dose-response studies of human exposure to cyanide by the oral
route were located. However, a number of studies have examined populations exposed to
cyanogenic compounds in foods, particularly cassava root, which can be dried and ground into
flour and is a primary source of carbohydrate in many parts of Africa and Southeast Asia
(Bonmarin et al., 2002; Okafor et al., 2002; Makene and Wilson, 1972). Due to concern about
effects of cyanogenic compounds, many of the recent studies of chronic cyanide toxicity have
been conducted in the developing world, where exposure to cyanogenic compounds in food is a
significant public health and agricultural (for livestock) concern. Symptoms reported in these
populations include ataxic tropic neuropathy, spastic paraparesis (paralysis, particularly of the
lower extremities), optic atrophy, and decreased nerve conduction velocity. Effects seen in these
studies are often confounded by dietary deficiencies (particularly low dietary intake of protein,
vitamin Bi2, and/or iodine) and overall malnutrition. In addition, animal studies comparing the
effects of cassava ingestion and ingestion of cyanide indicate that some of the observed effects
are due to compounds besides cyanide in cassava (Banea-Mayambu et al., 1997; Kamalu, 1993;
Olusi et al., 1979). Because of the confounding factors of dietary deficiencies in the studied
populations and the presence of other potentially toxic compounds in cassava, human and animal
studies of cassava are of limited use for the hazard assessment of cyanide.
4.1.3. Subchronic and Chronic Inhalation Studies
Several reports of occupationally exposed workers indicate that chronic exposure to low
concentrations of cyanide can cause alterations of thyroid function and neurological symptoms
(Banerjee et al., 1997; Blanc et al., 1985; El Ghawabi et al., 1975). Occupational exposure to
21
-------
cyanide occurs primarily via inhalation, although dermal and limited oral exposure also can
occur. HCN is noted for its systemic toxicity, which would be expected to occur at
concentrations below those at which any direct respiratory tract effects would be anticipated.
El Ghawabi et al. (1975) evaluated the effects of long-term occupational exposure to
cyanide in 36 male workers employed in the electroplating sections of three factories in Egypt.
Cyanide exposure was from a plating bath that contained 3% copper cyanide, 3% sodium
cyanide (NaCN), and 1% sodium carbonate. Individual breathing zone air samples were taken to
determine the levels of airborne cyanide to which the men were exposed. Fifteen-minute air
samples were collected by using a Midget impinger. Twenty male volunteers of the same age
group and socioeconomic status who had no occupational exposure to cyanide were chosen as
controls. Information on how or from where the controls were recruited was not provided. None
of the exposed or control workers were cigarette smokers at the time of the study. Participants
were prohibited from ingesting cyanide-containing foods during the course of the investigation.
Cyanide-exposed workers and controls were given medical examinations (with special focus on
thyroid abnormalities), interviewed regarding medical history, and questioned regarding
symptoms experienced. Thyroid function (as measured by uptake of radiolabeled iodide) was
assayed, and urinary levels of SOT were recorded over a 2-month period and reported as
average daily excretion (in mg). No investigation of thyroid hormone levels was reported. Of
the 36 workers, 14 had been exposed for 5 years, 14 for 5-10 years, 7 for 10-15 years, and 1 for
>15 years. The mean and median exposure times for the worker population were not reported.
The mean cyanide air concentrations in the breathing zones of workers at each of the three plants
were 10.4, 6.4, and 8.1 ppm (11.5, 7.1, and 8.9 mg/m3, respectively, with a range of 4.2-12.4
ppm (4.6-13.7 mg/m3) HCN. The authors reported that workers were exposed to other
chemicals during the electroplating process (e.g., gasoline, alkali, and acid), although
concentrations to these other chemicals were not quantified. Urinary SCN~ concentrations from
exposed workers were measured during 2 successive months. Graphically presented data of
mean individual urinary SCN~ levels plotted against the concentration of HCN in the air
indicated a strong positive linear relationship between urinary SCN~ and HCN concentration in
the air (Figure 4-1).
22
-------
10-
8-
O
CO 4-
>,
CO
I 2H
o-
T T
•
•
T
Factory A
Factory B
Factory C
0
5 10
HCN in air (ppm)
15
Source: El Ghawabi et al. (1975).l
Figure 4-1. Urinary SCN~ of exposed workers plotted against individual
breathing concentrations of HCN.
Twenty of the 36 exposed workers (56%) had thyroid enlargement rated as being mild to
moderate; however, there was no correlation between duration of exposure and either incidence
or magnitude of enlargement. The authors reported that none of the workers showed clinical
symptoms of either hypo- or hyperthyroidism, although the basis of this assessment was not
provided. Radioactive iodine uptake measured following a 2-day break in HCN exposure
indicated statistically significantly elevated iodide uptake after 4 hours (38.7 compared to 22.4%)
and 24 hours (49.3 compared to 39.9%) as compared with controls (Table 4-1). In a separate
assay, blood samples taken 72 hours after administration of the iodide tracer indicated that
protein-bound 131I in the blood was similar between controls and workers (0.11 ± 0.041
compared to 0.12 ± 0.039).
Reproduced by EPA from a graph published in El Ghawabi et al. (1975) by estimation of original data points and
regeneration of graph. EPA's reproduced linear regression line had a slope of 0.63 compared to the slope of
0.65 reported in the published study and a statistically significant Pearson correlation coefficient of 0.5 (p < 0.001).
23
-------
Table 4-1. Thyroid uptake of 131I in electroplating workers
Percent 131I thyroid uptake (mean ± standard deviation [SD])
Controls (n = 20)
Exposed (n = 36)
After 4 hrs
22.42 ±7.21
38.722 ± 6.63a
After 24 hrs
39.95 ±4.80
49.33 ± 10.61"
Significant difference (p < 0.001) by Student's t-test.
Source: El Ghawabi et al. (1975).
The authors noted that the radioactive iodide uptake test was conducted after the workers
had been away from work for 2 days, allowing time for thiocyanate (SGST), with a 3-day
elimination half-life (as established by Schulz et al. [1979]), to be partially cleared from their
systems. The authors suggested that the sudden cessation of cyanide exposure may have caused
the thyroid gland to rapidly accumulate iodine. Increased 24-hour uptake of radioactive iodide
has been reported to occur in hyperthyroidism, iodine deficiency, and goiter (NLM, 2008a;
Ravel, 1995).
Other findings noted in this study included significantly higher hemoglobin (14.8 vs.
13.4 g/dL) and lymphocyte counts (42 vs. 30%) and punctate basophilia of erythrocytes in 78%
of workers. The authors indicated that the observed punctuate basophilia was not characteristic
of HCN exposure and may be related to other concomitant chemical exposures. Symptoms
reported more frequently in the exposed workers than in the controls included (in decreasing
order of frequency) headache, weakness, and changes in the senses of taste and smell.
Incidences of symptoms at individual plants were not reported, and no evaluation of symptoms
by exposure concentration was presented. Based on the observed thyroid effects, the lowest
mean concentration recorded in the three factories of 6.4 ppm (7.07 mg/m3) HCN was designated
as a lowest-observed-adverse-effect level (LOAEL) for this review.
An unpublished study by Leeser et al. (1990) compared the health of 63 male cyanide salt
(NaCN, KCN, and Cu(CN)2) production workers with a control group of 100 British workers
from a diphenyl oxide (DPO) plant in a cross-sectional study. Cyanide workers were exposed
for periods ranging from 1 to 32 years with a mean exposure duration of 12.6 years. Air cyanide
was monitored with static floor monitors that would set off an alarm at >11 mg/m3 (the floor
monitor alarms were never triggered), with Draeger pump tests of area samples, and with
personal monitoring. Personal samples were collected on four to five occasions on different
people for each of the eight job categories in NaCN production (34 samples total). The
geometric means for the eight job categories ranged from 0.03 to 1.05 mg/m3 cyanide. Draeger
pump samples were reported to range from 1.1 to 3.3 mg/m3 cyanide. Blood samples were
collected from workers to measure hematological parameters and serum levels of cyanide,
carboxyhemoglobin, vitamin 812, and thyroxine (T/j). Each cyanide worker had a complete
24
-------
medical examination and was given a self-administered questionnaire that included questions
addressing 15 symptoms (feeling unwell, gaining weight, losing weight, pain in chest, short of
breath, headaches, smell problem, sleep problem, hand shake, lacking energy, dizzy spells,
nausea, indigestion, nose bleeds, and taste problem). Analysis of continuous variables (e.g., T4,
hematological parameters) was conducted using linear regression adjusting for age, alcohol use,
and smoking status; body mass index (kg/m2) was also included in the models for blood pressure.
The authors noted that log-transformation was used for some of the dependent variables (but did
not specify which variables). Results of two sets of questionnaires (given at different time
periods) showed that cyanide workers had more self-reported symptoms than the control group
(66.6 vs. 50.0% reporting one or more symptom). Symptoms with a greater frequency in the
cyanide group than in the control group included gaining weight (25.4 vs. 12.0%), shortness of
breath (14.3 vs. 7.0%), headaches (6.4 vs. 3.0%), smell problems (9.5 vs. 3.0%), sleep problems
(12.7 vs. 8.0%), shaky hands (6.4 vs. 1.0%), lack of energy (14.3 vs. 5.0%), dizzy spells (7.9 vs.
2.0%), nausea (3.2 vs. 0%), and taste problems (3.2 vs. 1.0%). The difference in sleep problems
was attributed to differences in the proportion of shift workers in the two groups (89 and 43%,
respectively, in cyanide and control worker groups).
Mean (± standard error [SE]) cyanide levels of blood collected prior to the block of shifts
were higher in nonsmoking exposed workers than in non-smoking controls (3.32 ± 1.25 vs.
1.14 ± 1.11 jimol/100 mL), and in ex-smokers (2.16 ± 1.13 vs. 1.46 ± 1.10 |imol/100 mL), but
not among current smokers (2.94 ± 1.11 vs. 3.14 ±1.11 |imol/100 mL). Thus, the blood cyanide
levels in nonsmoking exposed workers were similar to that of smokers. Blood SCIST was
measured, but the authors reported that it could not be analyzed due to unspecified technical
reasons. Mean hemoglobin levels in cyanide workers were statistically significantly increased
compared to controls (15.57 ± 0.14 vs. 15.08 ± 0.10 g/dL). Ratios associated with hemoglobin,
such as mean corpuscular hemoglobin (MCH) and mean corpuscular hemoglobin concentration
(MCHC) were also statistically significantly elevated, although the difference in these
parameters was low (about 3%). Lymphocytes were statistically significantly elevated in
cyanide workers compared to controls (mean 2.87 ± 0.11 vs. 2.55 ± 0.08 x 109/L). Serum T4
levels in cyanide exposed workers were decreased in controls, but the difference was not
statistically significant according to the study authors (mean 85.13 ± 2.51 vs. 89.04 ±1.81
nmol/L). The number of workers for which serum T4 was measured was not reported; therefore,
the lack of statistical significance could not be confirmed for this review. Serum T4 was below
the clinical reference range (60-160 nmol/L) in 3 cyanide-exposed workers.. The authors
claimed that these three workers were otherwise normal and other thyroid functional tests
showed that there was no functional problem, although the authors did not state which additional
tests were conducted to confirm normal thyroid function. A LOAEL of 1 mg/m3 cyanide for
increased lymphocyte count and increased hemoglobin concentration was established for this
25
-------
review. A no-observed-adverse-effect level (NOAEL) for thyroid effects was not identified for
this study based on the lack of measurement of sensitive thyroid parameters.
A group of 36 male workers who had been exposed to HCN fumes in a silver-reclaiming
facility in Illinois were retrospectively studied by Blanc et al. (1985), following the death of one
employee from cyanide overexposure and the closure of the factory due to health and safety
violations. The authors attempted to recruit all previous employees. In this study, data
collection for the former workers included physical examinations (including examination of
neurological effects), serum biochemistry, hematology, urinalysis, serum enzymes (aspartate
aminotransferase [AST] and alanine aminotransferase [ALT]), and thyroid hormone analysis
(thyroid-stimulating hormone [TSH] and T/i), and a questionnaire designed to determine
exposure, symptoms during employment, and current symptoms. Workers were qualitatively
categorized into low-, moderate-, or high-exposure groups based on their primary job activities.
The median time elapsed since last employment at the facility was 10.5 months; the median
duration of employment was only 8.5 months. Environmental monitoring conducted the day
after the plant was shut down found that the 24-hour time-weighted average (TWA) exposure
was 15 ppm (16.6 mg/m3) HCN. None of the former workers were found to have palpable
thyroid gland abnormalities, mucosal erosion, or focal neurological deficits. Clinical tests
revealed decreases in the absorption of vitamin 812 (a possible factor protecting against cyanide
toxicity) and decreased folate levels that may have been secondary to the decrease in vitamin
Bi2. Serum TSH levels were also reported as being significantly elevated in workers relative to
laboratory controls (concurrent controls were not used in this study design). Triiodothyronine
(Ts) uptake in the highest exposed workers (n = 9) was statistically significantly elevated
compared to that in laboratory controls (Table 4-2). The authors reported that this elevation may
reflect a postinhibitory response.
Table 4-2. Thyroid parameters in former silver-reclaiming workers
Population
Laboratory controls (n = 100)
All workers (n= 33)
Low-exposure workers (n = 24)
High-exposure workers (n = 9)
Percent T3 uptake"
30.0 ±2.8
30.9 ±2.6
30.4 ±2.3
32.4±2.4C
TSHa (uU/mL)
1.7 ±1.2
2.2±1.6b
2.2 ±1.7
2.4 ±1.3
"Values are mean ± SD.
bStatistically significant by Student's t-test atp < 0.05 compared to laboratory controls.
Statistically significant by Student's t-test atp < 0.01 compared to laboratory controls.
Source: Blanc et al. (1985).
A statistically significant positive trend for self-reported weight loss was demonstrated
against the exposure index, supporting an exposure-response relationship. A statistically
26
-------
significant trend was also found between the incidence of symptoms reported during active
employment (headache, dizziness, nausea, and bitter almond taste), as well as those reported at
the time of the survey (after adjustment for time elapsed since exposure) and the qualitative
index of exposure, providing evidence of another exposure-response relationship. Some of the
symptoms were reported as persisting for >7 months following exposure termination. The
reported CNS effects suggest the occurrence of neurotoxicity associated with exposure to
cyanide or its metabolite, thiocyanate. Dermal exposure to cyanide was reported by half of the
workers, and additional exposure by ingestion was likely due to poor general hygiene in the
factory in addition to inadequate personal protective equipment and worker training. Because
there were multiple possible routes of cyanide exposure, including dermal exposure and
contamination of food, data do not support for the selection of a LOAEL for inhalation. This
study does demonstrate, however, the occurrence of nontransient effects of thyroid function (as
measured by percentage of T3 uptake) from occupational exposure to HCN.
In a study of electroplating workers in a factory in India, Banerjee et al. (1997) compared
levels of the thyroid hormones T3, T4, and TSH in 35 male workers who had been exposed to
cyanide via inhalation for >5 consecutive years to a randomly selected control group of
35 unexposed male workers matched for age and dietary habits. None of the subjects used
tobacco products or had a prior history of thyroid disease. No environmental monitoring data
were provided on HCN levels in the factory. However, serum SCIST levels, a measure of internal
dose, were reported in both workers and controls. The average serum thiocyanate level in the
exposed workers was 316 |imol/L compared with 90.8 |imol/L in the controls, a difference that
was statistically significant (p < 0.01). The exposed workers had significantly lower levels of T3
and T4 (48 and 37% lower, respectively) and significantly higher levels of TSH (142%)
compared with controls (Table 4-3). In addition, there was a significant negative correlation
between serum T4 and thiocyanate concentrations (r = -0.363., p < 0.05) and a significant positive
correlation between TSH and thiocyanate concentrations (r = -0.354,p < 0.05). There was also
an apparent negative correlation between T3 and thiocyanate (r = -0.245), but this difference was
not statistically significant. Levels of T3 and T4 in exposed workers were outside the reported
normal range for these endpoints, indicating a potentially clinically relevant alteration of thyroid
hormone levels.
27
-------
Table 4-3. Thyroid parameters in HCN-exposed and unexposed
electroplating workers
Controls (n= 3 5)
Exposed (n= 35)
SCN (jimol/L)a
90.8 ±9.02
316±15.0b
T4 Oig/dL)3
6.09 ±0.601
3.81±0.3181C
T3 Oig/dL)a
111.0±9.3
87.2 ±8.1°
TSH (jiU/mL)a
1.2 ±0.301
2.91±0.201C
aValues are mean ± SD.
bStatistically significant by Student's t-test atp < 0.01 compared to unexposed workers.
Statistically significant by Student's t-test atp < 0.05 compared to unexposed workers.
Source: Banerjee et al. (1997).
As part of a report on excretion of cyanide and its metabolites, Chandra et al. (1980)
reported on a group of 23 electroplating workers chronically exposed to HCN fumes at 0.2-
0.8 mg/m3, with a mean value of 0.45 mg/m3. The title of this report indicated that workers were
exposed chronically, although exposure durations were not provided. The concentration in the
breathing zone was reported as 0.1-0.2 mg/m3, with a mean of 0.15 mg/m3. The authors noted
that the workers complained of symptoms typical of cyanide poisoning but provided no
additional information on specific symptoms or further analysis. In the absence of further
information, no independent assessment of this study is possible.
Chatgtopadhyay et al. (2000) investigated the effect of exposure to cyanide fumes on
pulmonary function in workers at a metal-tempering plant. The authors evaluated 24 workers in
an initial assessment and conducted a follow-up study on 17 of these workers 2 years later. The
control group for the initial study consisted of 14 unexposed workers matched for socioeconomic
status and race. The follow-up study did not include a concurrent control group; data from the
control group in the initial assessment were used for comparison. No information was provided
on cyanide concentrations in the air. The mean duration of exposure was 21.0 ± 5.03 years at the
time of the first assessment. In the initial study, there were statistically significant decreases in
pulmonary function as assessed by reduced peak expiratory flow rate, forced vital capacity
(FVC), and forced expiratory volume (FEV) in 1 second as a percentage of FVC (FEVi<>/0);
decreases in other pulmonary function parameters were not statistically significant. In the
follow-up study, statistically significant decreases were observed in all measured pulmonary
function parameters. However, concurrent controls were not utilized in the follow-up study.
Furthermore, adjustments for smoking and other co-occurring chemical exposures as sources of
potential confounding were not conducted.
Population-based studies examining inhalation of cyanide at ambient levels and potential
health outcomes are limited. However, studies examining smokers, a subgroup with higher
inhalation exposure to cyanide through tobacco smoke, have indicated an association between
smoking and thyroid disorders. Specifically, a meta-analysis of eight studies found a statistically
significant association between smoking and the development of goiter in women (odds ratio =
28
-------
1.29, 95% confidence interval [CI] 1.01-1.65) (Vestergaard, 2002). A more recent
epidemiological study of a population in an industrialized area of Germany with relatively low
intake of iodine has indicated that SOT urinary excretion is a cofactor or indicator for goiter in
nonsmokers as well as in smokers. These same authors found urinary ratios of iodide to SGST to
be predictive of increased risk for development of goiter as compared to iodide status alone
(Brauer et al., 2006). This study population was believed to have high exposure to HCN due to
industrial activities in the area, although exposure levels of HCN were not presented.
4.2. SUBCHRONIC AND CHRONIC STUDIES AND CANCER BIOASSAYS IN
ANIMALS—ORAL AND INHALATION
4.2.1. Oral Studies
The National Toxicology Program (NTP, 1993) reported the results of a subchronic
bioassay of NaCN administered in drinking water to rats and mice. F344 rats were administered
NaCN in drinking water at concentrations of 0, 3, 10, 30, 100, or 300 ppm for 13 weeks. These
concentrations are equivalent to the following doses, estimated by the study authors and based on
measured body weights and water consumption (converted to CN~ equivalents for this
assessment): 0, 0.16, 0.48, 1.4, 4.5, or 12.5 mg/kg-day CIST, respectively, in male rats and 0,
0.16, 0.53, 1.7, 4.9, or 12.5 mg/kg-day CIST, respectively, in female rats. The parameters
evaluated included body weight, clinical signs, water consumption, clinical chemistry,
hematology, urinalysis, extensive histopathology, selected organ weights (heart, kidneys, liver,
lungs, thymus gland, testes, epididymis, cauda epididymis), testicular sperm measures (spermatid
count and spermatid heads), epididymal sperm measures (spermatozoa count and motility), and
vaginal cytology. Thyroid weight or levels of thyroid hormones were not evaluated in this study.
In rats, no treatment-related effects on mortality or clinical signs of toxicity were seen in
either males or females. Body weight was statistically significantly decreased by 6% in high-
dose males, but this was not considered to be biologically significant by the study authors. No
body weight changes were observed in females. There was a dose-related decrease in water
consumption that was >10% in both sexes exposed to 100 or 300 ppm NaCN, compared with
controls. Decreased urine volume and increased urine specific gravity were observed in the
high-dose male rats and were attributed to decreased water consumption. Urinalysis data were
not reported for females. Urinary thiocyanate concentration was statistically significantly
increased at drinking water concentrations of >30 ppm NaCN at the end of the study. There
were no observed effects on nonreproductive organ weights in males, but there was a statistically
significant increase in absolute (16%) and relative (12%) liver weights in high-dose females
relative to controls. For all examined organs, there were no histopathologic changes that were
attributed to cyanide exposure. In particular, no histologic effects were observed in either the
thyroid or the brain. Female rats in the 100 and 300 ppm dose groups (4.9 and 12.5 mg/kg-day
29
-------
CIST, respectively) spent significantly more time in proestrus and diestrus compared with
controls, but there was no clear dose response and the authors did not consider these results to be
exposure related.
Male reproductive endpoints in the testis and epididymis were evaluated only in rats
exposed to >30 ppm NaCN (>1.4 mg/kg-day CIST). All reproductive parameter measurements
were conducted with the left reproductive organ (Table 4-4). In addition to evaluation of
epididymis weight, the weight of the cauda subsection of the epididymis was also measured.
This section of the epididymis functions as a site of sperm maturation and storage. Because the
cauda is part of the epididymis, these weights are not independent endpoints. Reproductive
organ weights were reported by NTP (1993) as absolute organ weights. For this review, relative
weights of reproductive organs were also calculated based on the individual animal data
downloaded from NTP's web site (http://ntp-server.niehs.nih.gov/) and were statistically tested
by using analysis of variance (ANOVA) followed by Dunnett's test.
30
-------
Table 4-4. Reproductive effects in male rats administered NaCN in drinking
water for 13 weeks
Study parameter
Number of animals3
Dose (mg/kg-d CN")
0 ppm
10
0
30 ppm
10
1.4
100 ppm
10
4.5
300 ppm
10
12.5
Weights (g)
Body weight3
Epididymis, absolute
(relative weight, 10~4)c
Cauda epididymis, absolute
(relative weight, 10~4)c
Testis, absolute
(relative weight, 10~4)c
338 ±5
0.448 ± 0.006
(13.7 ±0.14)
0.162 ±0.003
(4.92 ± 0.05)
1.58 ±0.03
(48.2 ± 0.64)
335 ±5
0.437 ±0.005
(13. 3 ±0.21)
0.150±0.004b
(4.54 ± 0.07)b
1.56 ±0.02
(47.3 ±0.58)
338 ±4
0.425 ± 0.007
(13.0 ±0.18)
0.148±0.004b
(4.51±0.12)d
1.52 ±0.02
(46.4 ± 0.40)
319 ±5b
0.417 ±0.005d
(13.3 ±0.41)
0.141 ± 0.003d
(4.49 ± 0.09)d
1.46±0.02d
(46.6 ± 0.80)
Testicular sperm at id measurements
Spermatid heads (107/g testis)
Spermatid heads (107/testis)
Spermatid count (mean/10"4 mL
suspension)
11.35±0.38
17.86 ±0.61
89.28 ±3.05
10.88 ±0.53
16.94 ±0.81
84.68 ±4.03
10.92 ±0.37
16.58 ±0.63
82.90 ±3. 16
10.57 ±0.33
15.42 ±0.44b
77.10±2.20b
Epididymal spermatozoal measurements
Motility (%)
Concentration (106/g cauda
epididymal tissue)
Spermatozoa count
(106/cauda epididymis)0
94.24 ±0.58
615 ±42
99.4 ±6.8
90.67 ±1.25b
684 ± 40
102.9 ±7.5
92.09 ±0.85b
699 ±33
102.8 ±4.9
90.66 ± 1.46b
709 ± 45
99.4 ±5.8
3Data reported as mean ± standard error of the mean [SEM]. Statistical significance determined by NTP, using
Dunnett's test (for body weight only) or Shirley's test.
bStatistically different from control atp < 0.05.
Calculated for this assessment based on individual animal data available at
http://ntp-apps.niehs.nih.gov/ntp_tox/index.cfm. Statistical significance tested by one-way ANOVA followed by
Dunnett's test.
dStatistically different from control atp < 0.01.
Source: NTP (1993).
NTP (1993) reported statistically significant decreases in several reproductive parameters
including epididymis weight, cauda epididymis weight, testis weight, number of spermatid
heads, testicular spermatid concentration, and epididymal spermatozoa motility. Absolute and
relative cauda epididymis weights were statistically significantly decreased at all doses examined
(>1.4 mg/kg-day CN"). In contrast, absolute epididymis weight was statistically significantly
depressed only at the highest dose (12.5 mg/kg-day). At the highest dose tested (12.5 mg/kg-
day), cauda epididymis weight (absolute) was decreased 13% below controls. The whole
epididymis weight (absolute) was significantly depressed 7% at this dose, and absolute testis
weight was significantly depressed 8%. Relative epididymis and testis weights were not
31
-------
significantly different at any dose level. Additionally, standard histopathology did not
demonstrate any morphologic effects in any reproductive organ.
Testicular spermatid parameters, including spermatid count and spermatid heads per
testis, were statistically significantly depressed at the highest dose tested (12.5 mg/kg-day). No
effect was seen on epididymal spermatozoa concentration; however, spermatozoa motility was
statistically significantly reduced at all tested concentrations (>1.4 mg/kg-day), although motility
did not exhibit a clear dose-related trend. At the lowest and highest dose, the percent of mobile
spermatozoa was reduced 4%, a magnitude of change within the range of historical controls and
not considered by the study authors to be biologically significant. Because fertility may be a
function of total spermatozoa count, rather than the concentration per gram cauda epididymal
tissue, and because decreased cauda epididymis weight can mask changes in spermatozoa
content, the number of total spermatozoa/cauda epididymis was also calculated for this review
(U.S. EPA, 1996). This number did not vary with dose. The unaltered spermatozoa count,
coupled with the decreased cauda epididymal weight, explained the slight dose-related (but not
statistically significant) increase in cauda spermatozoa concentration (see Table 4-4). For the
purpose of this review, a LOAEL of 1.4 mg/kg-day was identified, based on significantly
decreased relative and absolute cauda epididymis weights in male rats.
NTP (1993) also conducted a subchronic bioassay of NaCN administered in drinking
water in mice. B6C3Fi mice (10/sex/group) were administered NaCN in drinking water at
concentrations of 0, 3, 10, 30, 100, or 300 ppm for 13 weeks. These concentrations are
equivalent to the following doses, estimated by the study authors based on measured body
weights and water consumption (converted to CN~ equivalents for this assessment): 0, 0.26, 0.96,
2.7, 8.6, or 24.4 mg/kg-day, respectively, in male mice and 0, 0.32, 1.1, 3.3, 10.1, or 28.8 mg/kg-
day, respectively, in female mice. The parameters evaluated were identical to rats and included
body weight, clinical signs, water consumption, clinical chemistry, hematology, urinalysis,
extensive histopathology, selected organ weights (heart, kidneys, liver, lungs, thymus gland,
testes, epididymis, cauda epididymis), testicular sperm measures (spermatid count, spermatid
heads), epididymal sperm measures (spermatozoa count and motility), and vaginal cytology.
Thyroid weight and level of thyroid hormones were not evaluated.
In mice, no significant treatment-related effects on mortality, body weight, or clinical
endpoints were observed. Water consumption in both males and females was decreased in the
mid- and high-dose groups. Absolute and relative liver weights were significantly increased by
18 and 23%, respectively, in the high-dose females, and relative liver weight (but not absolute
liver weight) was significantly increased in high-dose males (12%). However, there was no clear
dose response. No treatment-related effects were observed in clinical chemistry, hematology,
urinalysis, nonreproductive organ weights, or histopathology in any of the assessed organs.
Reproductive effects were only evaluated in mice exposed to the highest three doses
(>2.7 mg/kg-day). As in the rat component of this study, relative organ weights were not
32
-------
originally reported by NTP, but were calculated for this review and statistically analyzed by
using ANOVA followed by Dunnett's test.
In male mice, NTP (1993) found that the absolute weights of the epididymis and cauda
epididymis were statistically significantly decreased in the high-dose group (24.3 mg/kg-day)
relative to controls (Table 4-5). At the high dose, absolute epididymis and cauda epididymis
weights were reduced 10 and 18%, respectively. Relative cauda epididymis weight was
significantly decreased (18%) at 8.6 mg/kg-day. Neither relative epididymal and testis weights
(relative and absolute) nor sperm parameters (spermatozoa per gram cauda epididymis, total
spermatozoa per cauda epididymis, and spermatozoa motility) were statistically significantly
decreased. No reproductive effects were reported at any of the dose levels tested for female
mice. For this review, a LOAEL of 8.6 mg/kg-day was determined based on a statistically
significant decrease in relative cauda epididymis weight.
33
-------
Table 4-5. Reproductive effects in mice administered NaCN in drinking
water for 13 weeks
Study parameter
Number of animals3
Dose (mg/kg-d CN")
0 ppm
9
0
30 ppm
10
2.7
100 ppm
10
8.6
300 ppm
9
24.3
Weights (g)
Body weight
Epididymis
(relative weight, 10"4)b
Cauda epididymis
(relative weight, 10~4)b
Testis
(relative weight, 10"4)b
37± 1.0
0.049 ±0.001
(13.5 ±0.54)
0.017 ±0.001
(4.74 ± 0.24)
0.121 ±0.002
(33.4 ±1.8)
39.2 ± 1.3
0.047 ± 0.002
(12.1 ±0.52)
0.016 ±0.000
(4.12 ±0.15)
0.113 ±0.008
(29.2 ± 2.2)
38.6 ±1.1
0.047 ±0.001
(12.1 ±0.42)
0.015 ± 0.001
(3.88±0.22)c
0.117 ±0.002
(30.3 ±0.90)
35.5 ±1.1
0.044 ±0.001C
(11.8 ±0.29)
0.014 ± o.oor
(3.68±0.17)d
0.118 ±0.003
(3 1.7 ±0.92)
Testicular sperm at id measurements
Spermatid heads (107/g testis)
Spermatid heads (107/testis)
Spermatid count (mean/10"4 mL
suspension)
18.47 ±1.13
2.24 ±0.14
69.94 ±4.34
21.48 ±2.34
2.26 ±0.14
70.80 ±4.25
17.42 ±1.34
2.03 ±0.15
63.28 ±4.53
18.17 ±1.62
2.11±0.16
66.06 ± 4.87
Epididymal spermatozoal measurements
Motility (%)
Concentration (106/g cauda
epididymal tissue)
Spermatozoa count
(106/cauda epididymis)b
92.38 ±0.81
1,235 ± 82
21.2 ±1.2
90.63 ± 1.34
1,393 ±70
22.3 ± 1.3
91.43 ±0.55
1,386 ±70
20.5 ±1.1
89.52 ±0.96
1,462 ±101
19.6 ±0.85
"Data reported as mean ± SE. Statistical significance determined by NTP using Dunnett's test (for body weight
only) or Shirley's test.
bCalculated for this assessment based on individual animal data available at
http://ntp-apps.niehs.nih.gov/ntp_tox/index.cfm. Statistical significance tested by one way ANOVA, followed by
Dunnett's test.
Statistically different from control atp < 0.05.
dStatistically different from control atp < 0.01.
Source: NTP (1993).
As part of a study evaluating the effects of cyanogenic compounds in cassava, Kamalu
(1993) evaluated the toxicity of inorganic cyanide administered in a rice diet to male dogs
(six/group) for 14 weeks. The diet was supplemented at feeding time with NaCN at a dose
calculated to release 10.8 mg HCN per kg cooked food. Based on a reported daily food
consumption of 0.1 kg/kg body weight, this corresponds to 1.08 mg/kg-day HCN or 1.04 mg/kg-
day CN". Further information about the study animals was not provided in this study, but animal
selection and study pretreatment were described in an earlier publication by the same author
(Kamalu, 1991), apparently describing the same study. That publication reported that dogs of
mixed breeds were purchased from local African markets at 6 weeks of age; treatment was
initiated when the dogs were approximately 22 weeks old. The authors noted that the dogs were
34
-------
repeatedly treated for ecto-and endoparasites. It is unclear what impact the compromised health
status and repeated treatment for parasites had on the observed effects in the dogs. The basal diet
used rice as the carbohydrate source, supplemented with pork, bone meal, and a vitamin and
mineral supplement that included iodine.
Blood was obtained from each dog at study weeks 1,3, and 14; urine was collected at
weeks 1, 3, 5, 7, and 14. Plasma and urinary thiocyanate concentrations were determined for
each collection period. Serum enzymes (including y-glutamyl transferase [GGT], ALT,
isocitrate dehydrogenase, total serum protein, serum albumin, serum globulin, and urinary
protein), as well as sodium, magnesium, and phosphorus, were measured. Histopathologic
evaluation was performed on the liver, kidneys, heart, testes, and adrenal glands of each dog.
The thyroid gland was not evaluated. At all time points evaluated, both plasma and urinary
thiocyanate concentrations were significantly increased in the treated dogs compared with
controls. Relative to controls, treated dogs had significantly increased urinary protein
concentration at weeks 5 and 14. No treatment-related effects were observed in serum enzymes,
total serum protein, albumin, or globulin or in sodium, magnesium, and phosphorus
concentrations. No histopathologic changes were observed in the liver or heart of treated dogs;
however, treatment-related effects were observed in the kidneys, testes, and adrenal glands.
Kidneys of the treated dogs had casts in the lumens of the renal tubules, accompanied by
occasional desquamation. In the testes, specialized reproductive morphologic analysis indicated
that the treated dogs had a significantly decreased percentage of tubules in stage VIII of the
spermatogenic cycle (characterized by elongated spermatids lining the lumen of the seminiferous
tubules) as compared with controls (p < 0.01). This percentage was 1.6 ± 1.07% (mean ±
standard error of the mean [SEM]) in the treated group compared with 14.4 ± 0.94% in the
controls. Treated dogs also had an increased incidence of abnormal cells and sloughing of germ
cells in the seminiferous tubules. Hyperplasia and hypertrophy were observed in the adrenal
gland. The adrenal medulla was unaffected by inorganic cyanide treatment. Although the width
of the adrenal cortex did not differ significantly between the cyanide-treated and control groups,
the zona glomerulosa (the most superficial layer of the adrenal cortex) was significantly wider in
treated dogs. The results of this study indicate that cyanide may be a reproductive toxicant in
male dogs. Based on histopathologic changes in the kidneys, testes, and adrenal glands, the only
dose tested (1.04 mg/kg-day) was considered to be a LOAEL.
An evaluation of the thyroid from this study was presented in Kamalu and Agharanya
(1991). At week 14, serum T3 was significantly decreased by 55%, and thyroid weight was
significantly increased by 23% in the cyanide-exposed group (compared to control animals at the
same time point). A histopathologic evaluation of the thyroid gland found decreased colloid
content compared to that of controls.
A 40-week study in New Zealand white rabbits that reported both liver and kidney
lesions supports the kidney as a possible target organ for toxicity following exposure to cyanide
35
-------
(Okolie and Osagie, 1999). In this study, groups of six male rabbits were fed a diet of growers'
mash only or mash containing 702 ppm CIST as KCN for 10 months. Based on daily food
consumption and weekly body weight measurements, the study authors estimated that the
average CN~ intake was 0.39 mg/day per rabbit in the control group and 36.5 mg/day per rabbit
in the exposure group. Initial and final body weights were averaged to estimated daily doses of
0.2 and 20 mg/kg-day CN~, respectively. Decreased body weight (33%) and decreased food
efficiency were observed in the high-dose group (33%). At the end of 10 months of treatment,
serum levels of ALT, alkaline phosphatase (ALP), lactate dehydrogenase (LDH), and sorbitol
dehydrogenase were increased. ALP levels were reduced in the lung but not in the heart. LDH
was increased in the liver and kidneys, a finding that the study authors interpreted as indicative
of a shift from aerobic to anaerobic metabolism, thereby increasing the production of lactic acid.
Biochemical evidence of tissue injury in the liver and kidney was supported by histopathologic
findings of focal areas of hepatic necrosis and congestion and renal tubular and glomerular
necrosis. No abnormal histopathology was reported for the pancreas or the heart. Neither a full
list of tissues examined nor additional information on histopathologic changes in other organs
was provided in the study. However, the occurrence of focal pulmonary edema and necrosis in
treated rabbits was reported in a second paper on the same study (Okolie and Osagie, 2000).
Based on necrosis in the liver and kidney, the only dose tested (20 mg/kg-day CIST ) is
considered for this review to be a LOAEL.
Manzano et al. (2007) examined the effects of subchronic (70-day) KCN ingestion in
45-day-old Lanrace-Large white pigs. The number and sex of animals used in this study are
unclear since study details in the published report are conflicting, with indications of 6 animals
per group in the materials and methods section, but 5-10 animals indicated in the tables.
Animals were administered KCN in the diet, twice per day, for total daily doses of 0, 2, 4, or
6 mg/kg-day (0, 0.8, 1.6, or 2.4 mg/kg-day GST). Blood samples were collected prior to the
experimental period and then every week thereafter and analyzed for ALT, glucose, cholesterol,
blood urea nitrogen, creatine, TS, 14, and thiocyanate. At the conclusion of the experiment,
thyroid glands were weighed, and tissues from the CNS, thyroid, pancreas, liver, and kidneys
were examined histologically. Significantly decreased serum ALT was observed at>0.8 mg/kg-
day. Additionally, significantly increased urea and creatinine were observed at doses
>1.6 mg/kg-day. Thyroid weight was significantly increased (24%) in animals in the highest
dose group, although significant alterations in thyroid hormones were not observed. Histological
alterations of the thyroid gland, characterized by numerous vacuoles in the colloid of the thyroid
follicles, were observed in all dosed animals. The authors also reported histologic alterations in
the liver, kidney, and CNS. Liver lesions were reported as karyolysis and pyknosis (nuclear
DNA changes denoting cell death) and distortion of the normal lobular architecture.
Degeneration of the renal tubular epithelial cells was reported in the kidney. In the brain,
minimal degeneration of Purkinje cells and loss of cerebellar white matter were reported. All
36
-------
histologic lesions were reported by the authors to occur in a dose-related manner, although
neither incidence nor statistical analysis of these findings was presented. A LOAEL of
2.4 mg/kg-day based on increased thyroid weight was determined for this review.
Jackson (1988) evaluated the effects of oral administration of KCN on behavior and
thyroid function in miniature pigs. Doses of KCN equivalent to 0, 0.4, 0.7, or 1.2 mg/kg-day
CIST were administered to three pigs per group via gavage. A total of five females and seven
males were used; each dose group contained both male and female animals (one dose group
contained two females and one male while the others contained two males and one female). The
solutions were administered once daily for 24 weeks, prior to feeding, in order to increase the
gastrointestinal absorption of cyanide. Regularly measured serum thiocyanate levels were
positively correlated with cyanide dose. Serum levels of T3, 14, and glucose were measured
every 6 weeks. Behavioral evaluations were conducted daily. Two categories of behavior were
evaluated: performance measures, including innate behavior, and learning measures, including
the acquisition and retention of new behaviors. No other endpoints were evaluated. Changes in
thyroid hormones were portrayed graphically as means without reporting variance (SEM or
standard deviation [SD]) or individual animal data. Individual animal data were requested from
the study authors by EPA, but were not provided. Both T3 and T4 demonstrated a dose-related
decrease (23 and 13%, respectively) that was statistically significant by week 18 of the study.
Thyroid histopathology was not evaluated in this study. A variety of behaviors were
significantly altered in treated animals, including a decrease in dominance behavior (high-dose
group), a decrease in fighting (mid- and high-dose group), an increase in flight response (all
treated groups), a decrease in exploratory behaviors (all groups), and less aggressive feeding
patterns (high-dose group). The authors concluded that the overall pattern of behavioral changes
in the group administered 1.2 mg/kg-day was different from that of the control animals, but that
the changes at lower doses were inconsistent. This study supports the large body of evidence
demonstrating that the thyroid is a target organ for cyanide toxicity. Based on reported
behavioral changes and decreased thyroid hormones, the LOAEL and NOAEL values identified
for this review are 1.2 and 0.7 mg/kg-day CIST, respectively.
Philbrick et al. (1979) evaluated the long-term health effects of oral exposure to cyanide
in rats. Male rats from Woodlyn Laboratories (10/group, strain not specified) received diets
(10% casein supplemented with 0.3% methionine, potassium iodide, and vitamin B12) containing
either 0 or 1,500 ppm KCN or an equal molar amount (2,240 ppm) of potassium thiocyanate
(KSCN) for 11.5 months. Parallel studies were conducted with rats provided a diet deficient in
r\
methionine, iodine, and vitamin B12, containing 0 or 1,500 ppm KCN (44 mg/kg-day CIST) . At
4 and 11 months, plasma T4 levels, T4 secretion rates, and urinary thiocyanate levels were
measured in five animals per group. After sacrifice, brain, heart, liver, and thyroid weights were
2Based on the average food intake across rat strains (U.S. EPA, 1988a) and adjusting for the molecular weight ratio
of cyanide to potassium cyanide.
37
-------
recorded. Histopathologic evaluation was conducted on the brain, optic and sciatic nerves, spinal
cord, and thyroid gland. This study design, although limited by the use of only one dose level,
allowed for the comparison of effects mediated directly through cyanide versus the primary
metabolite, thiocyanate. It also allowed for the comparison of cyanide and thiocyanate treatment
in control rats compared to rats fed nutritionally restricted diets.
Body weight gains of the KCN-treated animals were significantly lower than those of
controls, beginning at week 8; administration of KSCN did not affect body weight. Urinary
thiocyanate levels were reported as |ig/g food ingested (instead of |ig/mL urine) and were
significantly higher than in controls in all treated groups. Urinary thiocyanate levels were lower
at 11 months than at 4 months of treatment in KCN-treated groups but not in KSCN-treated
groups. This appears to indicate reduced metabolism of KCN with chronic exposure.
Additionally, although animals were fed an equal molar amount of KCN and KSCN, at
11 months, urinary levels of SCIST in the KCN-treated animals were only approximately one
quarter of the urinary excretion of SCN~ in KSCN-treated animals. This appears to indicate only
partial metabolism of cyanide into thiocyanate. Administration of cyanide altered serum T4
levels at 4 months but not at 11 months; thiocyanate altered serum T4 levels at both 4 and
11 months. After 4 months of treatment, rats in the cyanide-exposed groups had significantly
decreased plasma T4 levels (53%) and decreased T4 secretion rates (68%) compared to controls;
however, after 11 months of cyanide treatment, T4 levels no longer differed from those of
controls, although T4 secretion rates were depressed 27%. KSCN-treated animals also showed a
significant reduction (62%) in T4 secretion rates (at 4 months but not at 11 months) and
decreased T4 levels (55% at 4 months and 26% at 11 months). At the termination of the study,
relative thyroid weights were significantly increased in both KCN- and KSCN-treated animals by
43 and 33%, respectively. In the nutritionally restricted control animals, levels and T4 secretion
rates were lower compared to controls fed a standard diet. However, alterations of T4 levels, T4
secretion rates, and thyroid weights in animals on the restricted diet treated with KCN and KSCN
were of similar magnitude compared to treated animals on the standard diet. No histopathologic
lesions were observed by light microscopy in the optic or sciatic nerves or thyroid gland of any
group. Increased vacuolation was observed in the spinal cord white matter of treated animals
(with both KCN and KSCN) receiving sufficient or deficient methionine compared to controls,
and spinal cord demyelination induced by methionine deficiency was exacerbated by treatment.
No information on incidence or severity of the observed histologic lesions was reported by the
authors. No measurable differences were reported in spinal cord pathology between cyanide-
and thiocyanate-treated rats. Based on increased thyroid gland weight, decreased thyroid
hormone levels, and histopathologic changes in the spinal cord, the single dose tested (44 mg/kg-
day CIST) is considered to be a LOAEL.
Howard and Hanzal (1955) conducted a 2-year dietary study in which 10 Carworth Farms
rats per sex per group were administered food fumigated with HCN. The authors indicated that
38
-------
only rats surviving to the end of the study were analyzed histologically because the accuracy of
necropsies performed on animals that died early were compromised by autolysis. It appears that
seven, five, and nine males and six, seven, and six females were examined histologically in the 0,
4.3, and 10.8 mg/kg-day CW dose groups, respectively. Although special feeding jars were used
to minimize air circulation and evaporation, the study authors noted that it was necessary to
measure the loss of HCN due to evaporation from the chow and to prepare fresh rations every
other day to keep the HCN concentration near the target values of 100 and 300 ppm (mg HCN
per kg diet). The average daily concentrations were 73 and 183 mg CIST per kg diet. These
average concentrations of cyanide in the food were estimated based on Howard and Hanzal's
(1955) data for concentrations at the beginning and end of each food preparation period and
assuming a first-order rate of loss for the intervening period (U.S. EPA, 1992). From the data
reported on food consumption and body weight, estimated doses were 0, 4.3, and 10.8 mg/kg-
day. There were no treatment-related effects on growth rate, no gross signs of toxicity, no
hematologic effects, and no histopathologic lesions in the tissues evaluated from an undisclosed
subset of animals (heart, lungs, liver, kidneys, spleen, stomach, small and large intestines,
adrenals, thyroid, testes or uterus and ovaries, cerebrum, and cerebellum). Histopathology of the
spinal cord was not examined. Howard and Hanzal (1955) also reported that there appeared to
be no effect on relative organ weight of the liver, kidneys, spleen, brain, heart, adrenals, testes, or
ovaries. Thyroid weight was not investigated. The highest dose tested, 10.8 mg/kg-day CIST, is
considered to be the NOAEL.
An unpublished study by Leuschner and Neumann (1989) administered KCN to male
Sprague-Dawley rats (26-40/group) in drinking water for 13 weeks. Administered doses were 0,
40, 80, and 160 mg KCN/kg-day or 0, 16, 32, and 64 mg/kg-day CW. Doses were lowered in
the highest dose group at week 12 due to excessive toxicity/mortality. Water consumption was
decreased in all dose groups, prompting the authors to add a control group matching the water
consumption observed in the highest KCN treatment group. Examinations included hematology,
clinical chemistry, and urinalysis. After 3 months of exposure, organ histology was performed
on the kidney, heart, liver, testes, thyroid, and brain. Organ weights were measured for the
following organs: adrenals, heart, kidneys, lungs, thymus, brain, liver, pituitary, testis, and
thyroid (only the left lobe). Epididymis weight was not weighed independently, but was
included as part of testicular weight.
Early mortality was observed at the high dose, with 11 animals dying prematurely.
Statistically significant body weight decreases were observed in the highest two doses. Body
weight was statistically significantly decreased 42% at the high dose and 15% in the mid-level
dose. Average body weight decrease (4%) at the low dose was not statistically different than
controls. Water consumption was decreased in all dose groups (17, 21, and 32% in the low, mid,
and high dose, respectively) compared to controls. No changes in hematology or clinical
biochemistry were reported. Urinalysis demonstrated increased protein, which appeared to be
39
-------
dose-related. No dose-related histopathological changes were reported. Organ weight changes
were not observed at the lowest dose level tested (16 mg CN7day). At the mid-dose level
(32 mg ClST/day), absolute thymus weight was statistically significantly decreased (20%).
Statistically significant relative and absolute organ weight changes were observed at the highest
dose level (64 mg CN7kg-day), although these changes were inconsistent. At the highest dose,
absolute heart, liver, spleen, kidney, and brain weights were statistically significantly decreased
compared to the controls. However, when relative weights were calculated, all organs showed
increased weight compared to controls (except for the thymus, which was decreased). For this
review, a LOAEL of 32 mg/kg-day and a NOAEL of 16 mg/kg-day were identified based on
decreased body weight in male rats treated with KCN for 13 weeks.
Two studies by Soto-Blanco et al. (2002a, b) provide evidence of neurological changes
associated with cyanide ingestion in rats and goats. In the first study, Soto-Blanco et al. (2002a)
evaluated the toxicity of KCN administered daily by gavage (vehicle not stated) to male Wistar
rats for 12 weeks. Administered doses of KCN were 0, 0.15, 0.3, or 0.6 mg/kg-day, equivalent
to 0, 0.06, 0.12, or 0.24 mg/kg-day CIST, respectively. The number of animals included in each
group was seven, six, six, and seven for the control, low-, mid-, and high-dose groups,
respectively. Endpoints evaluated included clinical signs of toxicity, body weight, food
consumption, serum cholesterol, glucose, T3, and T4. Histopathologic examination was limited
to the CNS, thyroid gland, and pancreas. No treatment-related effects were reported for clinical
signs of toxicity, body weight gain, food consumption, serum T3 and T4, or serum glucose.
Plasma cholesterol was significantly decreased in the high-dose group. No histopathologic
changes were observed in the thyroid gland or the pancreas. Reported CNS effects in the high-
dose group included neuron loss in the hippocampus, damaged Purkinje cells (further details not
reported) and loss of white matter in the cerebellum, and the occurrence of a dose-related
increase in spheroid bodies on white matter in the spinal cord. EPA was unsuccessful in
obtaining incidence data from the study authors for the observed histologic lesions. Because
quantitative information was not reported on these histologic observations, neither a LOAEL nor
a NOAEL could be identified from this study.
Soto-Blanco et al. (2002b) also evaluated the neurotoxicity of cyanide, administered daily
as KCN, to male Alpine-Saanen goats in milk or water for 5 months at doses of 0, 0.3, 0.6, 1.2,
or 3.0 mg/kg-day (equivalent to 0, 0.12, 0.24, 0.48, and 1.2 mg/kg-day CIST). The test compound
was administered twice daily in milk (one-half of the daily dose per treatment) for the first
3 months and in water for the remainder of the treatment period. The number of animals per
dose group ranged from six to eight. The CNS was evaluated histologically for the presence of
glial fibrillary acid protein (GFAP), a marker for glial cells. The only clinical signs of toxicity
were transient muscular tremors and ataxia in one high-dose animal on days 121-123 of the
study. Neuropathology, including congestion, hemorrhage, and gliosis in the cerebellum, spinal
cord, and pons, as well as spheroids on the gray matter of the spinal cord, was observed at
40
-------
0.58 and 1.2 mg/kg-day CN~. Additional findings in the high-dose group included damage and
loss of Purkinje cells in the cerebellum, spongiosis in the pons, and spheroids, axonal swelling,
gliosis, spongiosis, and ghost cells in the medulla oblongata. GFAP immunostaining confirmed
the gliosis observed by histopathology. While this study confirmed that the CNS is a target
organ of subchronic cyanide administration, no information on the incidence or severity of
histologic findings was reported. Therefore, a NOAEL or LOAEL could not be identified from
this study.
4.2.2. Inhalation Studies
No chronic or subchronic animal studies of cyanide inhalation exposure were located.
However, several subchronic inhalation studies of related compounds, including cyanogen (CN)2
and ACH, are available. A 6-month inhalation study in monkeys (5 males/group) and rats
(30 males/group) exists for the gas (CN)2. (CN)2 is thought to break down in aqueous solution to
CIST and cyanate (CNCT) ions (Cotton and Wilkinson, 1980). Lewis et al. (1984) exposed three
groups of male Rhesus monkeys or male albino rats (Sprague-Dawley) to 0, 11, or 25 ppm (CN)2
6 hours/day, 5 days/week for 6 months. This would be equivalent to 0, 12, or 28 mg/m3 HCN
(based on the creation of 1 mol CN~ per mol (CNh in water). Pathology evaluated for both
monkeys and rats included gross and microscopic examination of heart, liver, kidney,
cerebellum, lungs, thyroid, spleen, and bone marrow. Hemoglobin, hematocrit, TS, and T4 were
also evaluated. Additionally, behavioral tests and electrocardiograms were administered to the
monkeys. No significant changes were seen in monkeys other than decreased lung moisture
content in both dose groups. The only effect noted in rats was significantly depressed body
weight (13%) in the high-dose group.
Inhalation studies of subchronic ACH exposure in male and female rats are available.
ACH is a liquid at room temperature, with a boiling point of 95°C and a vapor pressure of
0.75 mm Hg. In neutral to basic aqueous environments, ACH is reported to dissociate readily to
acetone and cyanide (U.S. EPA, 1985). Sprague-Dawley rats (15/sex/group) were exposed by
inhalation at average concentrations of 10.1, 28.6, or 57.7 mg/m3 ACH 6 hours/day, 5 days/week
for 14 weeks (Monsanto Co., 1985a, b). This dose corresponds to the molecular equivalent of
HCN concentrations of 3.2, 8.8, and 18.2 mg/m3. Endpoints analyzed included hematology,
clinical chemistry (including TS and T4 levels), and gross and microscopic histopathology on a
wide range of organs and tissues. No effects on mortality, body weight, or behavior were
observed in treated animals. Blood and urine levels of thiocyanate were elevated in a dose-
dependent manner, although no alterations in TS or T4 were observed. No significant gross or
microscopic histology was observed in treated animals compared with controls. In summary, the
study authors found no gross signs of toxicity attributable to subchronic inhalation exposure to
ACH in rats. Reproductive and developmental studies by the inhalation route have also been
conducted for ACH (Monsanto Co., 1985a, b; IRDC, 1984) and are described in Section 4.3.2.
41
-------
4.3. REPRODUCTIVE/DEVELOPMENTAL STUDIES
4.3.1. Oral Studies
Imosemi et al. (2005) fed 20 pregnant female Wistar rats 0 or 500 ppm KCN (equivalent
to 20 mg/kg-day CIST) in the diet during gestation and up to postnatal day (PND) 50. Offspring
(five/group) were sacrificed on PNDs 1, 9, 14, 21, 28, and 50, and the cerebellar tissues were
examined grossly. Parameters examined included body weight, brain weight, cerebellar weight,
maximum vermis length (length between cerebellar hemispheres), maximum side-to-side
dimensions of the cerebellum, and maximum thickness (anteroposterior dimension) of the
cerebellum. Aggressive and restless behavior was noted in the exposed dams but not in controls.
Additionally, significantly decreased body weight (6%) and brain weight (19%) were observed
in the treated pups on PNDs 14 and 9, respectively. No significant changes in body weight or
brain weight were found at the additional five time points examined. Cerebellar weight was
significantly reduced on PNDs 14, 21, and 28. The maximum vernal length was significantly
reduced on day 50 and the maximum side-to-side width of the cerebellum was reduced on day
29. In a separate publication, the authors also reported on microscopic parameters of the
cerebellum (Malomo et al., 2004). A significantly thicker external granular layer (EGL) was
seen in the experimental group on PNDs 14 and 21. Reduced thickness of the molecular layer
(ML) was also observed on PNDs 28 and 50. The density and size of the Purkinje cells were not
different between groups. Additionally, staining of the white matter was similar between groups,
suggesting normal myelination. The authors concluded that maternal consumption of 20 mg/kg-
day CN~ did not significantly affect microscopic indicators of cerebellar development, but
caused mild changes later in postnatal life. The authors noted that the presence of a thicker EGL
layer in the experimental group suggested delayed maturation and migration of cells in the
cerebellum. A LOAEL of 20 mg/kg-day CIST was identified for this review based on altered
maturation of the cerebellum.
Soto-Blanco and Gorniak (2004) evaluated effects of gestational exposure to cyanide in
pregnant mixed-breed goats (six per group). Starting on day 24 of pregnancy, goats were
administered, by gavage, 0, 1, 2, or 3 mg/kg-day KCN (equivalent to 0, 0.4, 0.8, or 1.2 mg/kg-
day CIST) until parturition (day 150). Blood samples were collected every other week and
analyzed for plasma glucose, cholesterol, and thiocyanate. T4 and T? concentrations in plasma
were measured from the offspring at birth and at 1 week old. One control dam and one dam
from the highest dose group were sacrificed at day 120. Three months after birth, the male
offspring and one dam from each group were sacrificed and the pancreas, thyroid, and entire
CNS (including spinal cord) were collected for histologic examination. Two dams from the
highest dose group experienced clinical signs of cyanide intoxication, specifically ataxia and
convulsions. Cyanide treatment did not significantly alter the number of live offspring or the
length of gestation, although average length of gestation in all treated groups was about 2.5 days
42
-------
shorter than controls. T3 levels in dams and offspring tested at birth were significantly elevated
over controls in the highest dose group, while 14 levels did not appear to be different. The dam
sacrificed at day 120 of pregnancy revealed increased reabsorption vacuoles in the thyroid
follicular colloid and severe spongiosis of the cerebral, internal capsule, and cerebellar peduncle
white tracts, suggestive of myelin edema of the white matter. The histopathologic study of dams
and offspring 3 months after birth revealed no lesions. Due to the lack of incidence and severity
data for the observed histologic effects, a LOAEL could not be identified from this study.
Soto-Blanco and Gorniak (2003) dosed mixed-breed, lactating goats (six per group) with
0, 1.0, 2.0, or 3.0 mg/kg-day KCN (equivalent to 0, 0.4, 0.8, or 1.2 mg/kg-day CIST) by gavage
(in water) from lactation days 0 to 90 and measured whole-blood cyanide and thiocyanate
concentrations in dams and offspring on lactation days 30, 60, and 90. On the 90th day, glucose,
cholesterol, plasma urea nitrogen, creatinine, T4, T3, AST, ALT, and GOT were also determined.
After 90 days, one dam from each dose group and every male goat from all litters was sacrificed.
The pancreas, thyroid glands, liver, kidneys, and the whole CNS were collected for histologic
examination. No clinical signs of toxicity were seen in any group, although one dam in the
highest dose group died on the 55* day of lactation. Both whole-blood cyanide and plasma
thiocyanate concentrations were increased in a dose-dependent manner in treated dams. In the
offspring, both blood cyanide and plasma thiocyanate increased with increasing maternal cyanide
dose, peaking at lactation day 30, and decreased with lactation time. Plasma parameters in all
groups of dams and offspring appeared to be unaffected by KCN treatment, except for the level
of T4 in dams, which was significantly elevated (20%) over controls (p < 0.01 by a t-test
conducted for this review). T3 and T4 levels also appeared elevated in the high-dose animals,
although these differences were not significant. In the thyroid, histopathologic changes,
characterized by an increased number of reabsorption vacuoles on the colloid of the thyroidal
follicles were observed in dams and offspring. Additionally, histologic changes in the liver and
kidney were noted, characterized by hepatocellular vacuolization and degeneration and mild
vacuolization of tubular epithelial cells, in dams and offspring. The authors noted that observed
histologic lesions were most intense in the highest KCN dose group. No histologic lesions were
noted in other examined tissues. In the absence of incidence data or statistical analysis on any
histologic changes, a LOAEL was not identified for this study.
Tewe and Maner (1981) fed female rats (20 per group, strain not specified) either a basal
diet prepared from low-HCN cassava meal or the basal diet supplemented with 500 ppm of KCN
throughout mating, gestation, and lactation. In addition, two female weanling rats per litter were
maintained on each diet for 28 days following weaning. Adult rats on the basal diet alone
received a dose of 1.2 mg/kg-day (based on a dietary HCN concentration of 12 mg/kg and
average food intake among female rats of 0.102 kg/kg body weight). Adult rats on the basal diet
plus 500 ppm KCN (high-cyanide diet) received a total CIST dose of 21.6 mg/kg-day, including
the 1.2 mg/kg-day from the basal diet and 20.4 mg/kg-day as KCN. For the weanling rats, the
43
-------
corresponding doses were approximately 1.9 mg/kg-day CW for the basal diet and 34.3 mg/kg-
day CIST for the basal + KCN diet, based on average food intake (0.162 kg/kg body weight) for
female weanling rats (U.S. EPA, 1988). As compared with controls, the high-cyanide diet had
no effect on body weight of pregnant rats, food consumption, maternal liver or kidney weights,
litter size, birth weight of pups, or pup mortality. In the weanling rats, the high-cyanide diet
resulted in significant decreases in food consumption and growth rate and an increase in the ratio
of food consumption to body weight gain, indicating that the decreased weight gain was not due
solely to poor palatability. The high-cyanide diet also resulted in a significant increase in serum
thiocyanate in both dams and weanlings compared with animals on the basal diet alone. The
activity of rhodanese, the enzyme that metabolizes cyanide to thiocyanate, in the liver and
kidneys was comparable in all groups. A LOAEL of 34.3 mg/kg-day was identified from this
study, based on decreased daily weight gain in weanlings; a NOAEL of 1.2 mg/kg-day (in
adults) and 1.9 mg/kg-day (in weanlings) was identified (for this review) based on cyanide
content of the basal cassava diet.
Teratogenicity has been reported in some nontraditional developmental studies that
administered cyanide or cyanogenic foods to animals. Doherty et al. (1982) administered NaCN
to hamsters subcutaneously via osmotic minipumps at doses around 0.13 mmol/kg-hour
(approximately 80 mg/kg-day) on gestation days (GDs) 6-9 and observed fetotoxic effects,
including significantly increased fetal resorptions and malformations and decreased crown-rump
length. Doses utilized in this study covered a narrow range from about 78 to 81 mg/kg-day. At
these doses, clinical signs of toxicity were evident in the dams, including weight loss, ataxia, and
dyspnea. At the lowest dose tested (78 mg/kg-day), there were 63% resorptions as compared to
10% in controls. Additionally, at this dose, 62% of fetuses were malformed vs. 5% of controls.
The majority of malformations in treated groups were characterized as neural tube defects.
Coadministration of cyanide with the cyanide poisoning antidote sodium thiosulfate, which
serves as a sulfur donor in the conversion of cyanide to thiocyanate by the enzyme rhodanese,
protected against maternal toxicity and teratogenic effects. Another developmental study by
Frakes et al. (1986) observed reduced ossification and decreased body weight in offspring of
hamsters administered a protein-deficient cassava diet containing low- or high-cyanide levels
during days 3-14 of gestation. The developmental effects of dietary cyanide were not evaluated
in animals fed a protein-sufficient diet. Low (mean: 0.65 mmol GST/kg food) and high (mean:
8 mmol CN7kg food) cassava-containing diets equaled daily doses of approximately 1.3 or
14 mg/kg-day cyanide, respectively, and averaged only 4% protein (the standard laboratory diet
contained 25% protein). Body weight in the cassava-fed dams was approximately 30% lower
than in control dams fed a standard, protein-sufficient diet, regardless of whether they were in
the high- or low-cyanide group. The numbers of implantations, resorptions, live fetuses, and
malformed fetuses in cyanide-treated groups were not statistically different from those in
controls. Fetal body weight was significantly decreased by 14 and 8% in low- and high-cyanide
44
-------
treatment groups, respectively (compared with the low-protein controls). Significantly decreased
ossification centers (28-37%) were observed in portions of the fetal skeletons, including the
sacrocaudal vertebrae, metatarsals, and sternebrae. No dose-related trend was observed for the
decreases in ossification or in maternal or body weight between the low- and high-cyanide dose
groups.
A Japanese study (Amo, 1973) reported that 0.05 mg/kg-day of cyanide administered in
drinking water decreased the fertility and survival rate in the Fl generation and produced 100%
mortality in the F2 generation in mice. Although no other studies exist on F2 animals treated
with cyanide, the data presented by Amo (1973) on decreased survival of the Fl generation are
not consistent with the body of available literature for cyanide, which indicates no decrease in
survival of the Fl generation of goats treated gestationally with doses twice as high (Soto-Blanco
and Gorniak, 2004) or in rats treated gestationally with doses >20 mg/kg-day (Imosemi et al.,
2005; Tewe and Maner, 1981). Additionally, studies in rats with ACH, which breaks down into
cyanide and acetone following inhalation or oral exposure, have not observed decreases in
reproductive parameters or Fl survival at inhalation exposures equivalent to 66 mg/m3 HCN
(Monsanto Co., 1985a, b). Furthermore, gavage dosing of pregnant Sprague-Dawley rats with
doses of ACH equivalent to 3 mg/kg-day during GDs 6-15 did not decrease survival in offspring
compared with controls, nor were there any difference in number of viable fetuses,
postimplantation losses, mean fetal body weight, fetal sex distribution, or fetal malformations
between treated animals and controls (IRDC, 1984).
4.3.2. Inhalation Studies
No studies exist on the potential reproductive or developmental toxicity of inhaled
cyanide. However, male and female fertility indices were investigated in rats exposed via
inhalation to the cyanide precursor ACH, which decomposes to acetone and cyanide (IPCS,
2005). At room temperature, ACH is primarily a liquid (boiling point: 95°C); however, in these
inhalational studies, the ratio of target and analytical air concentrations were close to unity,
indicating that ACH was primarily present as a vapor.
In a male fertility study (Monsanto Co., 1985a), Sprague-Dawley rats (15/group) were
exposed by inhalation to ACH at 0, 35, 104, or 209 mg/m3 for 6 hours/day, 5 days/week over a
period of 69 days. These doses are equivalent to 0, 11, 32, or 65 mg/m3 HCN. Following the
exposure period, males were mated with three nonexposed females each. Pregnant females were
sacrificed at mid-gestation (GDs 13-15), and pre- and postimplantation losses were determined.
Males were sacrificed 3 weeks following cessation of exposure. Histologic analysis of
reproductive organs, including the testis, epididymis, prostate gland, and seminal vesicle, was
conducted; reproductive organ weight and sperm parameters were not evaluated. No treatment-
related differences were seen in mean body weight, clinical chemistry, or histology of treated
45
-------
males. The mating efficiency, number of live implants, and pre- and postimplantation losses
were not different between treated and control groups.
Female Sprague-Dawley rats (24/group) were exposed by inhalation 6 hours/day,
7 days/week for 21 days to 0, 38, 108, or 207 mg/m3 ACH (0, 12, 33, or 64 mg/m3 HCN) and
then mated with untreated males (Monsanto Co., 1985b). Exposure of the females was
continued until the day of mating, and the females were sacrificed at mid-gestation (GDs 13-15)
to determine pregnancy status, nidations, pre- and postimplantation loss, and histology of the
ovaries and uteri. No clinical signs of toxicity were observed in treated animals except for dose-
related observations of red nasal discharge or encrustation in some animals. No treatment-
related differences were seen in mean body weight, clinical chemistry, or histology. Mating
efficiency, pregnancy rates, number of live implants, and pre- and postimplantation losses in
treated animals were comparable to control values.
4.4. OTHER DURATION- OR ENDPOINT-SPECIFIC STUDIES
4.4.1. Acute Oral Studies
In evaluating the oral toxicity of cyanide, both the total amount administered and the rate
of absorption are important (U.S. EPA, 1992), because toxicity results from exceeding the
body's capacity for detoxification of cyanide, which occurs mainly in the liver. At high doses of
cyanide, the availability of the sulfur donor needed for detoxification by the enzyme rhodanese
can become rate limiting. If absorption of ingested cyanide proceeds too quickly, then the
capacity of the liver to form thiocyanate via first-pass metabolism may be exceeded. In contrast,
slow absorption of the same total oral load of cyanide may allow complete metabolism by the
liver. Similarly, an acute cyanide dose is more toxic when administered by inhalation compared
with the same dose administered by ingestion, because the inhalation route bypasses first-pass
metabolism in the liver and directly enters systemic circulation.
The significant impact of absorption on the rate of detoxification of cyanide is
responsible for the observation that median lethal dose (LDso) values for NaCN (presented
below) are lower than the acute and chronic LOAEL values. The LD50 values are based on bolus
doses that result in rapid absorption of a large amount of cyanide that overwhelms the
detoxification capacity of the body. In contrast, the acute and chronic LOAEL values are based
on cyanide administration at a lower dose rate over the course of a day. This slower dose rate
means that the body is able to detoxify higher doses of cyanide (on the basis of administered
mg/kg) without being overwhelmed, and thus, it can handle a higher total dose load.
Acute oral LDso values for cyanide in rats range from 3 mg/kg (Ballantyne, 1988) to
8 mg/kg (Smyth et al., 1969) for cyanide administered as NaCN. Single daily doses of 4 mg/kg
in rats and 6 mg/kg in mice as KCN resulted in 95% mortality (Ferguson, 1962). Dermal LD50
values in rabbits range from 4.1 to 8.9 mg/kg. Clinical signs observed following single dermal
46
-------
doses ranging from 0.9 to 2.5 mg/kg include rapid breathing, dizziness, weakness, convulsions,
and loss of consciousness (ATSDR, 2006).
Palmer and Olson (1979) administered KCN to groups of seven male Sprague-Dawley
rats at 0 or 200 ppm in drinking water or 0 or 200 ppm in feed for 21 days. Using an average
body weight of 0.12 kg and a water consumption rate of 0.17 L/kg-day as per U.S. EPA (1988),
this assessment estimated the daily CN~ intake to be approximately 14 mg/kg-day. The only
endpoints evaluated were body weight gain and liver weight. A statistically significant 17%
increase in absolute liver weight was observed relative to controls; thus, the LOAEL was the
single dose tested, 14 mg/kg-day.
The dietary part of this study was inadequate for evaluation of toxicity due to instability
of cyanide concentrations in feed. Although CN~ ingestion was estimated at 9 mg/kg-day by
using default assumptions for body weight and food consumption (U.S. EPA, 1988), the study
authors noted that subsequent analysis of cyanide in feed resulted in <20% recovery of the
predicted value compared to 95% recovery for cyanide added to feed immediately prior to
analysis. Therefore, due to uncertainties regarding actual animal dosage, a LOAEL could not be
identified from the dietary part of this study.
Sousa et al. (2002) administered KCN to adult male Wistar rats in drinking water at target
doses of 0, 0.3, 0.9, 3.0, or 9.0 mg/kg-day for 15 days, equivalent to CN~ doses of 0, 0.12, 0.36,
1.2, and 3.6 mg/kg-day. There were 10 rats/group, except for the high-dose group, which
included 6 rats. Weight gain was significantly decreased at the high dose to about a third that of
the controls; however, weight gain was normal in the next lower dose group. There were no
effects on serum levels of TS, T/i, or serum levels of ALT, while AST exhibited sporadic
statistically significant changes that were determined not to be dose related. Serum levels of urea
or creatinine were unaffected by treatment. "Moderate" to "severe" congestion and cytoplasmic
vacuolization of the epithelial cells of the proximal tubules were observed in the kidneys of rats
at the two highest doses. Hydropic degeneration of hepatocytes was also noted at the highest
dose. Reabsorption vacuoles were observed in the thyroid gland of animals in all groups,
including controls, and increased in severity with increasing dose. However, quantitative
incidence or severity data for the histopathologic observations were not reported. Based on
moderate kidney vacuolization and congestion, aNOAEL of 0.36 mg/kg-day and a LOAEL of
1.2 mg/kg-day CIST were identified from this study.
Kreutler et al. (1978) evaluated the short-term effects on the thyroid of oral exposure to
cyanide. Male weanling albino rats (strain not specified; 10-24 animals/group) were fed diets
containing either low protein (2% casein) or normal protein (20% casein) for 2 weeks; treated
rats received the same diets supplemented with 0.2% KCN (equivalent to 99 mg/kg-day CIST,
using an average body weight of 95 g and average food consumption rates [U.S. EPA, 1988]).
Additional groups were administered the low-protein diet with or without KCN and with or
without iodide supplementation. Body weights and food consumption were recorded. Blood
47
-------
was collected and evaluated for serum TSH levels. Thyroids were removed and weighed. No
difference in body weight was observed among cyanide-treated rats and their respective control
groups. Rats treated with cyanide on the low-protein diet had significantly elevated serum TSH
levels and increased thyroid weights compared to the low-protein control group; supplementation
with iodide in addition to cyanide eliminated these effects. There was no effect on serum TSH or
thyroid weight in the cyanide-treated rats on the normal-protein diet. This study suggests that
severely protein- and iodine-deficient diets are likely to increase the sensitivity of the thyroid
gland to cyanide ingestion.
4.4.2. Acute Inhalation Studies
Relatively few inhalation studies providing quantitative data are available in animals
exposed repeatedly to HCN. Some studies of acute exposure are available in rats, rabbits, and
monkeys (Bhattacharya et al., 1994; Purser et al., 1984; Hugod, 1981). Inhalation LCso values
reported in animals range from 151 to 579 mg/m3 HCN in various species (ATSDR, 2006).
These studies provide limited information because sample sizes were either small (Purser et al.,
1984) or only a single organ or endpoint was assessed (Bhattacharya et al., 1994; Hugod, 1981;
Valade, 1952).
Purser et al. (1984) exposed cynomolgus monkeys individually to 100, 102, 123, 147, or
156 ppm HCN for up to 30 minutes. These concentrations correspond to 111, 113, 136, 163, or
172 mg/m3 HCN. A single monkey was exposed per concentration, with one monkey exposed to
both 100 and 147 ppm in separate experiments. There was no control group. The time to
incapacitation decreased with increasing exposure levels and ranged from 8 to 19 minutes. The
authors noted that three of the exposures (exposure levels not reported, presumably the three
highest concentrations) were terminated within 30 minutes due to the severity of the symptoms.
The observed symptoms included hyperventilation, decreased and arrhythmic heart rate, loss of
muscle tone and reflexes, and convulsions. Blood cyanide levels reached steady state within
10 minutes. There was no correlation between air concentration and blood cyanide levels.
Bhattacharya et al. (1994) investigated the effects of inhalation of 55 ppm HCN
(61 mg/m3 HCN) for 30 minutes on the pulmonary mechanics of six male Wistar rats.3 In treated
animals, the airflow was increased (20%), accompanied by increased transthoracic pressure
(40%) and tidal volume (50%). The respiratory rate, compliance, and minute volume decreased
50, 60, and 25%, respectively, accompanied by a decrease in pulmonary phospholipids (i.e.,
surfactant) of about 10-30%. Other effects of cyanide were not evaluated.
Extensive involvement of the CNS in cyanide toxicity was demonstrated by Valade
(1952), who exposed groups of four dogs to 50 mg/m3 (45 ppm) HCN for a varying number of
30-minute exposure periods conducted at 2-day intervals. Clinical signs included tremors,
3Data for the investigated parameters were presented graphically and thus, the magnitudes of change were estimated
for this review.
48
-------
stiffness, ataxia, dyspnea, vomiting, and diarrhea. In the longest-term exposure of 36 days
(consisting of 19 exposure periods), two of the four dogs died. Necropsies of the dead and
surviving dogs all showed histopathology in the brain consisting of vasodilation, hemorrhages,
and various cellular lesions.
Myocardial morphology in rabbits was investigated following inhalation of HCN as part
of a study attempting to identify constituents of tobacco smoke responsible for the increased risk
of cardiovascular disease observed in smokers (Hugod, 1981). Male rabbits (22/group) were
exposed to 0 or 0.5 ppm HCN for a period of 4 weeks, after which the animals were sacrificed
and examined for myocardial abnormalities. Following blinded morphologic examination, no
significant effects of cyanide were detected on myocardial ultrastructure.
4.4.3. Neurotoxicity Studies
Crampton et al. (1979) reported on a study in which baboons (7-10/group) were fed a
low cobalamin (vitamin 812) diet that was either supplemented with hydroxocobalamin (control)
or unsupplemented. Treated animals received 1 mg/kg-day KCN subcutaneously for 5 days.
The body weight of treated animals (with and without hydroxycobalamin supplementation) did
not differ from that of untreated animals. No neurological effects were evident from nerve
conduction measurements or in extensive histopathologic examination of the nervous system,
apparently the only organ system examined.
As described in Section 4.1.2, neurological symptoms have been reported in populations
that traditionally consume foods with high concentrations of cyanogenic glycosides, such as
cassava (ATSDR, 2006; Banea-Mayambu et al., 1997; Ministry of Health Mozambique, 1984;
Osuntokun, 1973). Effects include spastic paraparesis, ataxic tropic neuropathy, optic atrophy,
and decreased nerve conduction velocity. Osuntokun (1973) reported that the neurological
effects correlated with blood thiocyanate levels, but other reports found no correlation between
disease severity and thiocyanate level (Ministry of Health, Mozambique, 1984). Several studies
(Oluwole et al., 2003; Banea-Mayambu et al., 1997; Kamalu, 1993; Olusi et al., 1979) indicated
that constituents of cassava other than cyanide, such as the parental cyanogenic glycoside,
linamarin, may directly contribute to the characteristic endemic neurotoxicity observed in these
populations. Specifically, an ecological epidemiologic study conducted in Zaire (Banea-
Mayambu et al., 1997) indicated that prevalence of this endemic neuropathy was more closely
correlated with urinary linamarin than urinary thiocyanate.
Fechter et al. (2002) evaluated the effect of HCN exposure on hearing loss and its
interaction with noise-induced hearing loss. Male Long-Evans rats were exposed to HCN for
3.5 hours/day at concentrations of 0, 10, 30, or 50 ppm (equivalent to 0, 11, 33, or 55 mg/m3
HCN, respectively), to noise alone (i.e., 100 dB volume octave band noise for 2 hours/day
unaccompanied by cyanide exposure), or to noise plus 0, 10, 30, or 50 ppm HCN. Groups of
16 animals were exposed to air alone (0 ppm HCN) or noise alone, and groups of 6-12 animals
49
-------
were exposed to HCN or HCN plus noise. Hearing loss was assessed 4 weeks after exposure by
evaluating pure tone compound action potential (CAP) thresholds at frequencies ranging from
2 to 64 kHz (i.e., measuring the response at low through high pitches). The CAP threshold is a
measure of change in the electrochemical response of nerve cells in response to auditory
stimulation, a response that is considered to be a measure of cochlear function. This approach
was used in order to evaluate permanent hearing loss rather than the transient loss that occurs
immediately after exposure. Histologic analysis was also conducted on unexposed rats and on
rats exposed to noise alone or in combination with 10 or 30 ppm HCN (three to four rats per
group).
CAP thresholds were not affected by exposure to 10 or 30 ppm HCN. At 50 ppm HCN
(in the absence of noise), CAP thresholds were slightly elevated, but significant differences
among treated groups relative to control were not observed (using ANOVA for repeated
measures). As expected, noise alone did increase the CAP threshold, indicating hearing loss. In
the groups exposed to noise and HCN, there was a concentration-related increase in the CAP
threshold at frequencies of 12-40 kHz, with statistically significant differences at 30 and 50 ppm
as compared with controls. These data indicate that HCN can potentiate noise-induced hearing
loss, but they do not indicate an effect of HCN alone on hearing loss.
In a related study from the same laboratory, i.p. injection of rats with 7 mg/kg KCN
(2.8 mg/kg CIST) caused significant transient hearing loss (Tawackoli et al., 2001). The authors
also found that, in the absence of noise, auditory function recovered as cyanide was eliminated
from the blood. These studies together suggest that hearing loss from cyanide exposure is a
potentially sensitive neurological marker of toxicity. The return of function with the elimination
of cyanide from the blood raises the question of whether a permanent effect would occur under
conditions of high noise exposure and prolonged elevation of blood cyanide levels.
4.4.4. Immune Endpoints
Studies specifically designed to evaluate immune endpoints have not been located in the
HCN database. Additionally, no functional immune measures were identified in the database.
Limited information on immune endpoints exists from human occupational studies and animal
studies. El Ghawabi et al. (1975) found that the percentage of lymphocytes in peripheral blood
was statistically significantly elevated over controls in workers occupationally exposed to HCN
(7-12 mg/m3) for 5-15 years. The percentage of lymphocytes in exposed workers was 42%
(range 32-50%) compared to 30% in controls (range 26-40%). The total number of leucocytes
did not differ between groups. The biological significance of this magnitude of change in the
relative percentage of lymphocytes is unclear, as is the impact of other chemicals to which the
workers were concomitantly exposed. Another occupational study (Blanc et al., 1985) examined
workers an average of 11 months after cessation of exposure to average concentrations of
50
-------
17 mg/m3 (15 ppm) for a mean duration of 11 months. Analyses included a complete blood
count and differential with no significant findings reported for these endpoints.
There are no animal inhalation studies that evaluate the immunotoxicity of HCN, but
inhalation studies on related compounds are available. These studies have evaluated limited
immune-relevant endpoints and are mostly negative. A 3-month inhalation study of ACH (HCN
exposure equivalent of up to 18 mg/m3 or 16 ppm) in rats examined spleen weight and gross and
microscopic histopathology of the spleen, lymph nodes, and thymus. In addition, hematology
was examined, including white blood cell (WBC) and differential WBC counts. No changes
were seen in these endpoints (Monsanto Co., 1985a, b). Six-month inhalation studies of (CN>2
inhalation in male rats and monkeys exist (HCN exposure equivalent of 28 mg/m3 or 25 ppm).
Gross necropsy was performed on the spleen and bone marrow. No changes were seen in these
endpoints in either species (Lewis et al., 1984).
Oral studies of cyanide have examined limited immune endpoints. Three-month drinking
water studies in rats and mice (NTP, 1993) with doses up to 12.5 mg/kg-day in rats and
24 mg/kg-day in mice examined immune organs (spleen, thymus, bone marrow, and lymph
nodes) and conducted hematology, including WBC and differential WBC counts. NTP (1993)
did not demonstrate significant changes in any of these endpoints. Additionally, a 2-year oral
study in rats with doses up to 10.8 mg/kg-day examined spleen, thymus, and hematology
endpoints and did not note immunological effects (Howard and Hanzal, 1955).
4.5. MECHANISTIC DATA AND OTHER STUDIES IN SUPPORT OF THE MODE OF
ACTION
4.5.1. Genotoxicity
KCN was not mutagenic in Salmonella typhimurium strains TA82, TA97, TA102, TA98,
TA100, TA1535, TA1537, or TA1538 in the reverse mutation assay with or without metabolic
activation (De Flora et al., 1984; De Flora, 1981). NaCN was not mutagenic in S. typhimurium
strains TA97, TA98, TA100, or TA1535 with or without metabolic activation (NTP, 1993). A
positive response was reported, however, for HCN in S. typhimurium strain TA100 without
metabolic activation; adding metabolic activation reduced the magnitude of the positive response
to 40% of what it had been without metabolic activation (Kushi et al., 1983). Negative results
were obtained in the DNA-repair test in Escherichia coli strains WP67, CM871, and WP2 (De
Flora et al., 1984) and in a test for inhibition of DNA synthesis in HeLa cells (Painter and
Howard, 1982).
Overall, cyanide has tested negative in bacterial mutagenicity studies with and without S9
activation (NTP, 1993; De Flora et al., 1984; De Flora, 1981), although a positive result was
obtained in S. typhimurium strain TA100 with and without S9 activation (Kushi et al., 1983).
Neither standard chromosome aberration assays nor mammalian gene mutation studies of
51
-------
cyanide are available. Cyanide was negative in assays for the production of DNA damage and
repair (De Flora et al., 1984; Painter and Howard, 1982).
4.5.2. Acute Neurotoxicity
The mode of action for the acute toxicity of cyanide is well understood (Klaassen, 2001;
Hall and Rumack, 1990). Cyanide is considered a chemical asphyxiant because it impairs
aerobic metabolism without affecting oxygen delivery to the tissues. It has a high affinity for
iron in the ferric state, resulting in binding to and inactivation of tissue cytochrome c oxidase.
Since cytochrome c oxidase normally accepts oxygen from the blood and functions as an
electron acceptor in cellular energy production, this inactivation inhibits cellular respiration. As
anaerobic metabolism proceeds, blood levels of pyruvic acid, lactic acid, and NADPH rise; the
ATP/adenosine diphosphate (ADP) ratio decreases. The earliest effects of acute cyanide toxicity
occur in organs with high aerobic energy demands, particularly the brain and heart. The
inhibition of oxygen use by cells causes oxygen tension to rise in the peripheral tissues, which
results in a decrease in the unloading gradient for oxyhemoglobin. Thus, oxyhemoglobin is
present in the venous blood. In addition to cytochrome c oxidase, cyanide binds to other
metalloproteins and other cellular molecules, including catalase, peroxidase, methemoglobin,
and hydroxycobalamin; this binding also contributes to the symptoms of acute cyanide toxicity.
Cyanide also stimulates the release of secondary neurotransmitters and catecholamines
from the adrenal glands and adrenergic nerves (Kiuchi et al., 1992; Kanthasamy et al., 1991).
Thus, the cardiac effects and the peripheral autonomic responses observed following cyanide
exposure appear to be due to the increase of plasma catecholamine levels. CNS necrosis and
demyelination caused by cyanide may be due to vasoconstriction and low blood flow in the
brain, resulting from low carbon dioxide levels (Brierley et al., 1976). Alternatively, the
decreased ATP/ADP ratio may alter energy-dependent calcium homeostasis in nerve cells
(Johnson et al., 1986). Thus, the acute effects of cyanide result primarily from the interruption of
aerobic metabolism and from the release of secondary neurotransmitters and catecholamines;
these effects include altered respiration, vomiting, nausea, and weakness and ultimately
convulsions, coma, and death.
4.5.3. Thyroid Disruption
The primary cyanide metabolite, SCIST, has the same ionic charge and is of similar size as
iodide. SOT competitively inhibits iodide uptake in the thyroid by the sodium-iodide (Na+/F)
symporter (NIS). Iodine is essential for the normal production of the thyroid hormones TS and
14. The NIS is a transmembrane protein that actively transports iodide from the bloodstream,
against an electrical and chemical gradient, and concentrates it in the thyroid gland. The human
NIS has been reported to have greater affinity for thiocyanate than for iodide (De Groef et al.,
2006; Tonacchera et al., 2004; Wolff, 1998). In addition to reducing iodide uptake by the
52
-------
thyroid, thiocyanate may also cause iodide already accumulated in the thyroid to be discharged
(Wolff, 1998). Additional compounds that have a similar mode of action (i.e., competitive
inhibition of the NTS) include perchlorate, nitrate, chlorate, and fluoroborate; however, each of
these compounds has differing affinities for the NTS and thus different potencies of iodine uptake
inhibition (Tonacchera et al., 2004; Van Sande et al., 2003; Greer et al., 1966). For example,
perchl orate has been estimated to be 15-20 times more potent than thiocyanate in terms of iodine
uptake inhibition (Tonacchera et al., 2004; Greer et al., 1966).
The effect of thiocyanate on the thyroid gland is dose dependent and controlled by
homeostatic processes that tightly regulate and control the synthesis of essential thyroid
hormones in order to ensure a constant systemic supply to meet physiological needs (NRC,
2005). If thiocyanate interference with iodide uptake is of sufficient magnitude to decrease the
production and secretion rate of thyroid hormones (T4 and T3), then circulating levels of these
hormones decrease. Homeostatic mechanisms mediated mainly via the hypothalamo-pituitary-
thyroid feedback axis are rapidly activated to modulate thyroid hormone synthesis (NRC, 2005;
Hill et al., 1989). As the blood levels of these hormones drop, the hypothalamus, through the
release of thyrotropin-releasing hormone, stimulates the pituitary gland to produce TSH. TSH
stimulates the thyroid gland to increase the rate at which it produces and secretes thyroid
hormones. Elevated TSH levels stimulate histologic changes meant to increase thyroid secretion,
such as increased size and number of thyroid cells (Guyton and Hall, 2000). Clinically, this
increased size and number of thyroid cells manifests as an enlarged thyroid gland (goiter). It is
only when thiocyanate intake levels are sustained and high enough to overwhelm homeostatic
processes that decreased synthesis and secretion of thyroid hormones would be expected to occur
and thus result in hypothyroidism and effects secondary to hypothyroidism. This mode of action
may be relevant to the thyroid effects observed in both humans and animals, including thyroid
gland enlargement, decreased thyroid hormones, and increased TSH (Manzano et al., 2007;
Banerjee et al., 1997; Kamalu and Agharanya, 1991; Jackson, 1988; Blanc et al., 1985; Philbrick
et al., 1979; El Ghawabi et al., 1975).
4.5.4. Reproductive Effects
The NTP (1993) observed a suite of reproductive effects in rats and mice, including
decreased epididymis weight, testis weight, and testicular spermatid count in rats and mice
treated for 3 months with NaCN in drinking water. The mode of action of these reproductive
effects is not well established. However, some data exist in hypothyroid animals, suggesting that
disruptions in thyroid hormone levels may affect the male reproductive system. Studies in
humans and animals have demonstrated that cyanide exposure can result in decreased thyroid
hormone levels (Jackson, 1988; Philbrick et al., 1979; El Ghawabi et al., 1975). Therefore, it is
possible that the observed reproductive effects following exposure to cyanide may be mediated
through decreases in thyroid hormones mediated through the cyanide metabolite thiocyanate.
53
-------
Thyroid hormones are important in the growth and development of a wide range of
tissues, including the male reproductive tract (Kobayashi et al., 2007; Wistuba et al., 2007).
Some reports investigating the developmental effects of hypothyroidism in animal models have
indicated male reproductive effects, including altered maturation of male reproductive organs
and impaired spermatogenesis (Wistuba et al., 2007; Del Rio et al., 2003; Maran and Aruldhas,
2002). Persistent neonatal hypothyroidism in animal models has been shown to result in reduced
reproductive organ weight and decreased sperm count and motility (Sahoo et al., 2008; Hamouli-
Said et al., 2007; Del Rio et al., 1998; Kumar et al., 1994). Conversely, transient neonatal
hypothyroidism has been shown to cause increased testis size and increased sperm production
(Sahoo et al., 2008; Joyce et al., 1993; Cooke, 1991). These experimental observations indicate
that reproductive tissues are sensitive to thyroid hormone levels during development.
In addition to thyroid effects on the growth and development of the reproductive tissues,
some research has suggested that the adult reproductive system is also modulated by thyroid
hormones. In adults, proper thyroid function has also been shown to be important for
maintenance of fertility in adult males and females (Trokoudes et al., 2006; Poppe and
Velkeniers, 2004). Hypothyroidism in adult males has been noted to cause alterations in sex
steroid hormone metabolism, spermatogenesis, and fertility (Krassas and Pontikides, 2004).
Mechanistic studies in adult animals have indicated that reproductive organs, including
the testis and epididymis, may be sensitive to alterations in thyroid hormone levels. A study by
De Paul et al. (2008) demonstrated staining for thyroid receptor protein and mRNA in the adult
rat epididymis, which was shown to be increased in hypothyroid rats showing responsiveness of
adult epididymis tissue in response to decreased thyroid hormone. Additional studies have found
cellular and ultrastructural changes in the adult rat epididymis following induced hypothyroidism
(following thyroidectomy) (Del Rio et al., 2003, 2001, 1979), some of which were reversible
following supplementation with thyroid hormone (Del Rio et al., 1979). Similarly, another study
found reduced epididymis weight in thyroidectomized rats, which was reversible following T4
supplementation (Kala et al., 2002).
In summary, research exists to suggest that reproductive tissues in developing and adult
animals are responsive to alterations in thyroid hormone levels. Additional evidence also exists
to suggest that specific structural changes and decreased epididymis weight can be mediated
through hypothyroidism in the adult animal. Though some information supports this
hypothetical mode of action that reproductive effects observed in the NTP (1993) study may be
due to alterations in thyroid function due to exposure to cyanide, specifically the cyanide
metabolite thiocyanate, uncertainty exists due to the lack of any measurement of indicators of
thyroid function, such as thyroid hormones (TSH, TS, T/i) or thyroid weight (NTP, 1993).
4.6. SYNTHESIS OF MAJOR NONCANCER EFFECTS AND MODE OF ACTION
54
-------
Tables 4-6 and 4-7 present summaries of noncancer effects from repeated oral and
inhalation exposure to cyanide. Chronic and subchronic cyanide oral exposure studies in
experimental animals indicate that the thyroid, CNS, and male reproductive organs are sensitive
targets of toxicity (Manzano et al., 2007; Soto-Blanco et al., 2002a, b; NTP, 1993; Jackson,
1988). Information from human occupational studies suggests that subchronic and chronic
inhalation exposure to cyanide may be associated with CNS symptoms (including headache,
weakness, and changes in taste and smell) and thyroid alterations (enlargement, altered iodine
uptake, increased TSH, and decreased TS and 14) (Banerjee et al., 1997; Leeser et al., 1990;
Blanc et al., 1985; El Ghawabi et al., 1975). Another study also suggests that chronic exposure
to HCN in a metal-tempering plant may reduce pulmonary function in chronically exposed
workers (Chatgtopadhyay et al., 2000).
55
-------
Table 4-6. Summary of subchronic and chronic oral toxicity studies for cyanide in animals
Species,
strain, sex
Reference
Dose
(mg/kg-d CNl
Route
Duration
Response at LOAEL
NOAEL
LOAEL
Comments
Subchronic and chronic
Pig
(6 or 10/group;
sex not
specified)
Rat, Wistar
(6-7 males/
group)
Rat, SD
(26^0/group)
Rat, F344
(10/sex/
group)
Mouse,
B6C3FJ
(10/sex/
group)
Dog, mongrel
(6 males/
group)
Manzano et al.
(2007)
Soto-Blanco et
al. (2002a)
Leuschner and
Neumann
(1989)
NTP (1993)
NTP (1993)
Kamalu
(1993);
Kamalu and
Agharanya
(1991)
0, 0.8, 1.6, 2.4
0,0.06,0.12,0.24
0, 16, 32, 64
Males:
0,0.16,0.48, 1.4,
4.5, 12.5
Females:
0,0.16,0.53, 1.7,
4.9, 12.5
Males:
0, 0.26, 0.96, 2.7,
8.6, 24.4
Females:
0,0.32, 1.1,3.3,
10.1,28.8
0, 1.04
Diet; KCN
Gavage
Drinking
water;
KCN
Drinking
water;
NaCN
Drinking
water;
NaCN
Diet; NaCN
10 wk
12 wk
13 wk
13 wk
13 wk
14 wk
Increased thyroid weight
Histopathologic changes in
CNS
Decreased body weight
Decrease in cauda epididymis
weight and sperm motility
Decrease in cauda epididymis
and epididymis weight
Casts in renal tubules, adrenal
gland hypertrophy and
hyperplasia, and decreased
spermatids in stage VIII of the
spermatogenic cycle;
decreased T3; increased
thyroid weight
1.6
Not
determined
16
Not
determined
8.6
None
2.4
Not
determined
32
1.4
24.3
1.04
All doses showed altered
histology in thyroid, liver,
kidney, and CNS (no
incidences given for histologic
lesions).
No incidences given; no
changes in T3 and T4.
Unpublished study.
Thyroid hormones and thyroid
weight not measured.
Thyroid hormones and thyroid
weight not measured.
Dogs suffered from parasitic
infections.
56
-------
Table 4-6. Summary of subchronic and chronic oral toxicity studies for cyanide in animals
Species,
strain, sex
Goat (6-
8 males/
group)
Pig
(3/group,
mixed sexes)
Rabbit
(6 males/
group)
Rat, strain not
specified
(10 males/
group)
Rat (10/sex/
group)
Reference
Soto-Blanco et
al. (2002b)
Jackson
(1988)
Okolie and
Osagie (2000,
1999)
Philbrick et al.
(1979)
Howard and
Hanzal (1955)
Dose
(mg/kg-d CNl
0,0.12,0.24,0.48,
1.2
0, 0.4, 0.7, 1.2
0.2, 20
0,44
0,4.3, 10.8
Route
Milk,
drinking
water
Gavage in
water;
KCN
Diet;
KCN
Diet
Diet
Duration
5 mo
6 mo
10 mo
11.5 mo
2yr
Response at LOAEL
Histopathologic changes in
CNS
Decreased T3 and T4;
behavioral changes
Decreased body weight, focal
liver necrosis, tubular and
glomerular necrosis of
kidneys, pulmonary edema and
necrosis
Vacuolation of spinal cord
white matter; decreased T4 at 4
mo and increased thyroid
weight at 11.5 mo
None
NOAEL
Not
determined
0.7
0.2
None
10.8
LOAEL
Not
determined
1.2
20
44
None
Comments
Inadequate dose-response
characterization to identify
NOAEL/LOAEL.
Single daily bolus dose; no
other endpoints evaluated.
Cyanide in control group from
determination of basal amount
in feed.
Reproductive and developmental
Rat, Wistar
(20 dams,
5 pups/ group)
Rat, Wistar
(20 dams,
5 pups/ group)
Goat
(5-8/group)
Goat (7/group)
Imosemi et al.
(2005)
Malomo et al.
(2004)
Soto-Blanco
and Gorniak
(2004)
Soto-Blanco
and Gorniak
(2003)
0,20
0,20
0, 0.4, 0.8, 1.2
0, 0.4, 0.8, 1.2
Diet
Diet
Gavage in
water;
KCN
Gavage in
water
Gestation, up
to PND 50
Gestation, up
to PND 50
D 24-150
(birth)
Lactation d
0-90
Decreased body weight, brain,
and cerebellar weight; altered
cerebellar dimensions
Reduced thickness of ML and
increased thickness of EGL of
cerebellum
Increased T3 in dams and
offspring at birth
Vacuolation of thyroid, kidney
epithelial cells, and
hepatocytes in offspring and
dams
None
None
0.8
Not
determined
20
20
1.2
Not
determined
Same study as Malomo et al.
(2004); aggressive and restless
behavior noted in treated dams.
Indicative of delayed
maturation and migration of
cerebellar cells.
Some of the dams in the highest
dose group experienced tremors
and ataxia.
Histological lesions reported in
treated dams and offspring but
incidence not given.
57
-------
Table 4-6. Summary of subchronic and chronic oral toxicity studies for cyanide in animals
Species,
strain, sex
Rat, strain not
specified
(20 females/
group)
Reference
Tewe and
Maner(1981)
Dose
(mg/kg-d CNl
1.2,21.6 (adults)
1.9,34.3
(weanlings)
Route
Diet;
cassava and
KCN
Duration
Throughout
mating,
gestation,
and lactation
Response at LOAEL
Decreased food consumption,
growth rate, liver weight in
weanlings
NOAEL
1.9
LOAEL
34.3
Comments
Control animals fed basal diet
containing low-HCN cassava;
treated animals fed basal diet
supplemented with KCN.
58
-------
Table 4-7. Summary of subchronic and chronic inhalation toxicity studies for cyanide in humans
Study population
Males
n = 36 exposed workers
n = 20 unexposed controls
Electroplating workers across
3 factories
Males
n = 63 HCN exposed workers
n = 100 DPO plant workers
Males, n = 36 exposed workers,
divided into
low(n= 13)
medium (n = 14) or
high (n = 9) exposure
Silver reclaiming facility
Reference
El Ghawabi et al.
(1975)
Leeseretal. (1990)
Blanc etal. (1985)
Exposure
7.07,8.9, 11.5mg/m3HCN
(6.4,8.1, 10.4 ppm)
0.03-1. 03 mg/m3 HCN
16.6 mg/m3 HCN (15 ppm)
24-hr TWA taken 1 d after
plant closed
Duration
5-15 yr
1-32 yr
mean: 12.6 yr
0.5-21 mo;
median:
8.5 mo;
mean:
11 mo
Response
Thyroid
enlargement,
altered iodide
uptake, and CNS
symptoms;
increased
lymphocytes
Increased self-
reported symptoms,
increased
lymphocytes,
decreased
hemoglobin
Increased TSH,
increased T3
uptake, CNS
symptoms
NOAEL
mg HCN/m3
None
None
None
LOAEL
mg HCN/m3
7.07
1.03
16.6
Comments
Urinary
thiocyanate
correlated with
HCN air
concentration.
Unpublished
study; no change
inT4.
No thyroid
enlargement seen;
study conducted
1 1 months
postexposure.
59
-------
The CNS is a target of both acute and chronic cyanide exposure. Symptoms of severe
CNS toxicity following acute cyanide exposure include respiratory depression, convulsions,
coma, and death. Chronic and subchronic inhalation exposure in workers has been reported to
result in symptoms, including headaches, weakness, nausea, and changes in taste and smell, at
doses ranging from 1 to 17 mg/m3 (Leeser et al., 1990; Blanc et al., 1985; El Ghawabi et al.,
1975). Behavioral changes and CNS lesions have been reported in animals exposed orally to
cyanide. Histopathologic effects on various CNS structures also have been observed following
subchronic exposure in rats (Soto-Blanco et al., 2002a; Philbrick et al., 1979) and goats (Soto-
Blanco et al., 2002b). Philbrick et al. (1979) reported that vacuolation was observed in the spinal
cord white matter of rats treated with cyanide or thiocyanate for 1 year with 44 mg/kg-day.
Studies by Soto-Blanco et al. (2002a, b) in rats and goats at doses of 0.24-1.2 mg/kg-day
reported effects, including neuron loss in the hippocampus, spheroids on white and gray matter
of the spinal cord, damaged Purkinje cells, and loss of white matter in the cerebellum; however,
no quantitative data or statistical analyses were presented for these effects. Behavioral changes
in pigs and rats have also been observed with oral exposure to cyanide. Jackson (1988) observed
that pigs administered KCN equivalent to 0.4-1.2 mg/kg-day cyanide in the drinking water for 6
months exhibited increased flight response, a decrease in fighting, and a decrease in exploratory
behavior. Additionally, pregnant rats treated with 20 mg/kg-day cyanide by gavage
demonstrated aggressive and restless behavior (Imosemi et al., 2005). Though the mode of
action of acute cyanide toxicity is well understood (Klaassen, 2001; Hall and Rumack, 1990), the
mode of action of CNS changes observed with chronic cyanide exposure is unclear. It is
plausible, due to the mode of action of cyanide of inhibition of ATP synthesis, that CNS changes
upon chronic CIST exposure may also be due to energy deprivation in areas of high metabolic
activity in the brain. Conversely, a chronic study in rats found similarly increased vacuolation of
spinal cord white matter, astrogliosis, and fluid accumulation compared to controls regardless of
whether animals were treated with KCN or KSCN, indicating that these histologic lesions may
be due to SCN~ (Philbrick et al., 1979).
The thyroid is also an organ sensitive to cyanide, particularly following long-term
exposure. Cyanide's effects on the thyroid are mediated by its metabolite, thiocyanate. Thyroid
enlargement and altered iodine uptake have been seen in workers exposed for 5-15 years to
HCN at concentrations ranging from 7 to 12 mg/m3 (El Ghawabi et al., 1975). A retrospective
study of a group of 36 male former workers who had been exposed to HCN fumes in a silver-
reclaiming facility for a mean duration of 11 months found that TSH levels, although still within
normal levels, were statistically significantly elevated compared with those of controls (Blanc et
al., 1985). Thyroid effects, including enlargement, decreased hormone levels, and altered
histology have also been seen in experimental animals orally treated with cyanide (Manzano et
al., 2007; Soto-Blanco and Gorniak, 2003; Jackson, 1988; Philbrick et al., 1979). Rats treated
for 4 months at 44 mg/kg-day had significantly decreased plasma T4 levels (53%) and decreased
60
-------
T4 secretion rates (68%) compared with those of controls; however, after 11 months of cyanide
treatment, 14 levels no longer differed from controls, although 14 secretion rates were depressed
27%. At the termination of the study, thyroid weights were significantly increased by 43% in
treated animals (Philbrick et al., 1979). A study in pigs found a dose-related decrease in T3 and
14 (23 and 13%, respectively) in animals administered 1.2 mg/kg-day CIST (Jackson, 1988).
Another study in pigs found increased thyroid weight (24%) in animals administered 4.3 mg/kg-
day CIST for 10 weeks, although significant alterations in thyroid hormones were not observed
(Manzano et al., 2007). Histologic changes in the thyroid, characterized by an increased number
of reabsorption vacuoles on the colloid of the thyroidal follicles, were observed in dams and
offspring treated for 90 days with 0.4-1.2 mg/kg-day CIST (Soto-Blanco and Gorniak, 2003).
However, this study found significantly increasedT4 levels in dams treated with 1.2 mg/kg-day,
which adds uncertainty to the interpretation of effects observed in this study. It is well
established that the anti-thyroid effects of cyanide are due to its primary metabolite, SCIST.
Thiocyanate competitively inhibits the active transport of iodide into the thyroid gland, and thus,
if homeostatic mechanisms of the thyroid are overwhelmed, the available concentration of the
iodine-based thyroid hormones T3 and 14 can be decreased (Crump and Gibbs, 2005; NRC,
2005; Guyton and Hall, 2000).
Alterations in the male reproductive system have also been observed in some studies.
Male reproductive effects were reported in rats and mice following exposure in drinking water
(NTP, 1993) and in dogs following exposure in feed (Kamalu, 1993). Decreased epididymis,
cauda epididymis, and testis weight, and decreased sperm counts were seen in rats treated with
>12.5 mg/kg-day for 13 weeks (NTP, 1993). Decreased relative epididymis and cauda
epididymis weights were also observed in mice treated with>24.3 mg/kg-day for 13 weeks
(NTP, 1993). In the Kamalu (1993) study, dogs treated with 1.04 mg/kg-day for 14 weeks
exhibited histologic changes in the testis, including a significantly decreased percentage of
tubules in stage VIII of the spermatogenic cycle. The NTP (1993) authors suggested that the
male reproductive effects may be related to perturbations in hormonal balance, although thyroid
hormones or thyroid weight were not evaluated as part of the study. However, some data from
hypothyroid animals suggest that thyroid disruption can result in reproductive changes, including
sperm decrements and reproductive organ weight decreases (see Section 4.5.4). It is not known
whether workers occupationally exposed to cyanide would be expected to suffer reproductive
effects, since it does not appear that sperm or other reproductive parameters were investigated in
the available occupational studies (Blanc et al., 1985; El Ghawabi et al., 1975).
Cyanide affects the kidneys and liver at doses similar to those that result in male
reproductive toxicity, depending on the species tested. Kidney effects, characterized by casts in
renal tubules, were reported in dogs following subchronic oral dosing at 1.04 mg/kg-day for
14 weeks (Kamalu, 1993). Rabbits dosed at 20 mg/kg-day for 40 weeks had evidence of tubular
and glomerular necrosis (Okolie and Osagie, 1999). Additionally, goat dams and offspring
61
-------
treated with 0.4-1.2 mg/kg-day throughout lactation were reported to have vacuolation of the
kidney epithelial cells. Similarly, liver effects with cyanide exposure have been seen in rats,
mice, rabbits, and goats. Goats treated with 0.4-1.2 mg/kg-day for 90 days were reported to
have vacuolation of hepatocytes (Soto-Blanco and Gorniak, 2003). Focal necrosis was seen in
rabbits treated with 20 mg/kg-day in the diet for 10 months (Okolie and Osagie, 1999), and
increased liver weight was observed in female rats and mice administered 12.5 or 23.4 mg/kg-
day, respectively, in drinking water for 13 weeks (NTP, 1993). The mode of action for the
effects of cyanide on the liver and kidneys has not been identified. One hypothesis suggested by
Okolie and Osagie (1999) is that interference with thyroid function by thiocyanate can lead to
decreased tissue protein turnover, resulting in depressed growth, as well as liver and kidney
effects. Additionally, high-dose exposure of cyanide in heavily oxygen-dependant tissues, such
as the liver, may result in ATP deprivation of the tissues.
4.7. EVALUATION OF CARCINOGENICITY
Under the U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment, there is
"inadequate information to assess the carcinogenic potential" of cyanide. Studies examining
cancer incidence in occupationally exposed cyanide workers are not available. Studies of cancer
in populations exposed to thiocyanate via diet were limited to examinations of thyroid cancer,
and the results are generally not positive (Bosetti et al., 2002; Kolonel et al., 1990), although one
recent case-control study has associated high consumption of goitrogenic food and low iodine
intake with increased incidence of thyroid cancer in women (Truong et al., 2010). The currently
available data indicate that cyanide is not genotoxic. Bacterial mutagenicity assays, with and
without activation, are predominantly negative (NTP, 1993; De Flora et al., 1984; De Flora,
1981). The only available chronic animal study of cyanide that analyzed a wide variety of
tissues is an oral study in rats (Howard and Hanzal, 1955), in which tumors or lesions were not
associated with either dose group following dietary administration of cyanide at doses up to
10.8 mg/kg-day for 2 years. This study is limited by small sample sizes (n = 10), histopathologic
assessment of only a subset of potential target organs of carcinogenicity, and uncertainty
regarding dose due to HCN volatility issues. Therefore, the available data for cyanide are
inadequate to assess the carcinogenic potential of cyanide.
4.8. SUSCEPTIBLE POPULATIONS AND LIFE STAGES
4.8.1. Possible Childhood Susceptibility
Due to the mode of action of the primary cyanide metabolite, thiocyanate, fetuses
exposed in utero may be most susceptible to the effects of cyanide exposure, leading to
potentially reduced thyroid hormone production during critical periods of brain development.
The effects of significantly reduced thyroid hormone levels that result from untreated subclinical
or clinical hypothyroidism are much more severe in the developing young than in adults (NRC,
62
-------
2005; Guyton and Hall, 2000). While hypothyroidism in adults typically results in goiter (an
enlarged thyroid), congenital hypothyroidism has been associated with stunted bodily growth and
impaired mental development in fetuses, infants, and young children. Hypothyroidism has been
associated with neurodevelopmental delay and functional and structural neurological deficits in
young humans and/or rats (NRC, 2005). In human fetuses and neonates, the effects of severe
congenital hypothyroidism associated with severe maternal hypothyroidism during pregnancy
are irreversible and are characterized by long-term impairment of behavior, locomotor ability,
speech, hearing, and cognition (Chan and Kilby, 2000). In moderately hypothyroid neonates
born to mothers who were hypothyroid during pregnancy, prompt supplementation with thyroid
hormone may restore neurodevelopmental function to some degree (NRC, 2005). A
retrospective study of over 25,000 pregnant women indicated that mild subclinical maternal
hypothyroidism can adversely affect neurological development. Children (aged 7-9 years)
whose mothers had subclinical hypothyroidism during pregnancy were found to have IQ scores
7 points lower than the matched control children (Haddow et al., 1999). Another study observed
that pregnant women with subclinical hypothyroidism were 3 times more likely to have a
placental abruption (relative risk [RR] 3.0, 95% CI 1.1-8.2), a serious pregnancy complication
associated with high perinatal mortality rates. This study also observed that pregnant women
with subclinical hypothyroidism were twice as likely to have a preterm birth (RR 1.8, 95% CI
1.1-2.9) than controls (Casey et al., 2005). These authors speculated that the reduced IQ of
children born to mothers with subclinical hypothyroidism may be related to the effects of
prematurity; however, the mean gestational week at delivery in hypothyroid women in Haddow
et al. (1999) was no different from controls. Additional studies support the indication of lower
neurodevelopmental scores in offspring of women with free T4 levels in the lowest 10
percentile and normal TSH values during early gestation (Kooistra et al. 2006; Pop et al., 2003).
In rats, hypothyroidism during development has been associated with anatomical
alterations in the cerebellum, including reduction of growth and branching of Purkinje cells,
delayed proliferation and migration of granule cells, delayed myelination, and changes in
synaptic connections among cerebellar neurons (Koibuchi and Chin, 2000). Although animal
models may provide information on the potential neuroanatomical and neurophysiological
effects of highly reduced maternal thyroid hormone levels during gestation, they are limited in
the ability to assess subtle changes in neurodevelopment, cognition, and behavior that may occur
in humans. Further, the homeostatic system in humans is regarded as more robust and elastic
than that in rodents (NRC, 2005). Thus, animal models provide qualitative, but not quantitative,
information on the effects of low human fetal availability of thyroid hormones during gestation
and early development (Jahnke et al., 2004). The magnitude of decrease in serum T4 levels that
might result in neurodevelopmental effects during gestation and early childhood is not well
characterized and would depend on many factors, including whether T4 levels are reduced during
critical windows of development, the extent of the ability of the fetus to produce its own thyroid
63
-------
hormones to compensate for decreased maternal thyroid hormone availability, and the nature and
extent of other nutritional deficiencies associated with thyroid hormone production (e.g., iodine,
selenium).
Although cyanide is a known neurotoxicant, a dose-response characterization of
neurodevelopmental toxicity resulting from competitive inhibition of iodide uptake in the thyroid
gland by its thiocyanate metabolite has not been demonstrated. This relationship is complicated
by the interdependence of several factors including the intake of iodine, protein, and selenium
(and likely additional micronutrients), exposure to other chemicals that modify thyroid function,
and pre-existing thyroid conditions (Pearce and Braveman, 2009; Triggiani et al., 2009).
However, based upon the mode of action of thyroid disruption and studies following offspring
neurodevelopment and pregnancy outcomes of hypothyroid mothers, it is clear that the
developing fetus and infant are at a disproportionately high risk from chemicals such as cyanide,
which antagonize thyroid function.
4.8.2. Possible Gender Differences
Experimental animal studies have indicated that male reproductive toxicity is a target of
chronic cyanide toxicity in rats, mice, and dogs (Kamalu, 1993; NTP, 1993). NTP (1993) found
reduced testicular spermatid count in male rats and decreased reproductive organ weights,
including the epididymis and testes, in both rats and mice. However, effects on female
reproductive organs were not reported in rats or mice at any dose tested (NTP, 1993). Few
studies identifying effects from cyanide exposure studied both sexes; therefore, information
suitable to assess gender differences is lacking.
Population-based studies of thyroid disorders indicate gender-related trends, with women
being more likely to develop goiter and hypothyroidism (NLM, 2008b; Knudsen et al., 2002).
Additionally, analysis of a large data set from the 2001-2002 National Health and Nutrition
Examination Survey (NHANES) indicated statistically significant associations between urinary
levels of perchlorate (which shares a common mode of action of competitive iodine inhibition
with SCIST) and changes in TSH and T4 levels in women but not in men, with the association
strongest in women with low iodine intake (Steinmaus et al., 2007; Blount et al., 2006).
Therefore, women, especially those with low iodine intake, may be more susceptible to thyroid
disruption compared to men.
4.8.3. Other Susceptible Populations
Due to the ability of thiocyanate to competitively inhibit iodine uptake, people with
preexisting hypothyroidism or low iodine intake, especially pregnant and lactating women, may
represent a susceptible population due to an increased need for iodine during these periods
(WHO, 1994). Kreutler et al. (1978) observed that thyroid effects in rats induced by exposure to
KCN could be attenuated by administering iodine concurrently with cyanide. Additionally, an
64
-------
epidemiologic study by Brauer et al. (2006) demonstrates that populations with low ratios of
urinary iodine to urinary thiocyanate are at increased risk of developing enlarged thyroid.
Populations with low iodine intake exposed to additional chemicals that operate by a
similar mode of action as thiocyanate also represent an additional sensitive population because of
expected additive effects. In addition to thiocyanate, other chemicals, such as perchlorate and
nitrate, can act as competitive inhibitors of NTS, the membrane protein that actively transports
iodine into the thyroid follicular cells (De Groef et al., 2006; Van Sande et al., 2003). A recent
study by Steinmaus et al. (2007) investigated the relationship between urinary levels of
thiocyanate and perchlorate and thyroid hormone levels in women with low iodine intake
(categorized as < 100 |ig/L urinary iodine) by using cross-sectional human data gathered as part
of NHANES. The authors found no association between urinary thiocyanate levels and 14 or
TSH levels in women with low iodine intake. However, an association was seen between
exposure to perchlorate and decreased serum 14 levels. The authors found that this observed
association was strengthened when thiocyanate exposure was considered together with
perchlorate.
The effect of co-exposures of cyanide with other chemicals is not completely understood.
For example, carbon monoxide co-exposure with hydrogen cyanide in acute exposure scenarios,
such as structural fires, is thought to potentiate effects including hypoxia and mortality (Eckstein
and Maniscalco 2006; Moore et al., 1991; Norris et al., 1986); however, the impact of chronic,
low level co-exposure of these two chemicals, such as through vehicle exhaust and tobacco
smoke, is not known.
People with protein deficiency may also be more sensitive to the cyanide-induced thyroid
effects. Kreutler et al. (1978) observed that rats on a low-protein diet (2% casein) demonstrated
increased plasma TSH and thyroid weights following KCN administration, whereas rats on a
normal-protein diet (20% casein) exposed to the same concentrations of cyanide did not develop
these effects. Reduced availability of sulfur-containing amino acids in protein-deficient diets
may impact the concentration of sulfur donors available for the detoxification of cyanide (Frakes
et al., 1986). Studies in human populations ingesting large quantities of cyanide-containing
foods, such as those made from cassava flour, also suggest that increased susceptibility to
thyrotoxic effects are also associated with dietary deficiencies of protein, iodide, vitamin 812, or
other vitamins (Makene and Wilson, 1972; Osuntokun et al., 1969).
65
-------
5. DOSE RESPONSE ASSESSMENTS
5.1. ORAL REFERENCE DOSE (RfD)
5.1.1. Choice of Principal Study and Critical Effect
The dose-response data available on subchronic or chronic oral exposure to cyanide are
limited to experimental studies in animals. Though clinical data from several acute human
exposures are available, no chronic or subchronic studies which provide dose-response
information following oral exposure to cyanide in humans exist. Two chronic oral exposure
studies exist for cyanide, Philbrick et al. (1979) and Howard and Hanzal (1955), both in rats,
though they are limited by the use of only one high dose (Philbrick et al., 1979) or the failure to
detect effects (Howard and Hanzal, 1955). Additionally, there are two well-designed subchronic
studies in rats and mice that tested multiple dose levels and examined an array of endpoints and
tissues (NTP, 1993), and other subchronic studies in a variety of animal models assessing more
limited endpoints and tissues (Manzano et al., 2007; Soto-Blanco et al., 2002a, b; Okolie and
Osagie, 2000, 1999; Kamalu, 1993; Jackson, 1988). Furthermore, several developmental studies
on oral cyanide exposure in rats and goats exist (Imosemi et al., 2005; Malomo et al., 2004;
Soto-Blanco and Gorniak, 2004; Tewe and Maner, 1981).
Manzano et al. (2007) treated 6 or 10 pigs per group (sex not specified and number of
animals unclear due to inconsistencies in reporting) with potassium cyanide (KCN) administered
in the diet at 0, 1.4, 2.8, or 4.3 mg/kg-day for 10 weeks. An increase of 24% in thyroid weight
was seen at 4.3 mg/kg-day. Histologic alterations were reported in the thyroid, liver, kidney, and
CNS in all dosed animals. However, no incidence data or statistical analyses were provided for
these histologic findings, precluding a characterization of the dose response for these effects.
This study identified a LOAEL of 4.3 mg/kg-day and a NOAEL of 2.8 mg/kg-day for a
statistically significant increase in thyroid weight. This study is limited by poor reporting of
study design and observed histologic effects. Due to these limitations and the availability of
studies demonstrating effects at lower levels, this study was not selected as the principal study.
Jackson (1988) evaluated the effects of gavage administration of KCN on behavior and
thyroid function in miniature pigs. Doses of 0, 0.4, 0.7, or 1.2 mg/kg-day KCN were
administered by gavage to three pigs per group (mixed sexes) for 24 weeks. Both TS and T4
demonstrated a dose-related decrease (23 and 13%, respectively) that was statistically significant
by week 18 of the study. Changes in thyroid hormones were portrayed graphically as means,
without reporting variance or data for individual animals. The author concluded that the overall
pattern of behavioral changes, characterized as an increased ambivalence and slower response to
stimuli, was different in the highest dose group compared to control animals. Based on
behavioral changes and decreased thyroid hormones, the LOAEL and NOAEL values are 1.2 and
0.7 mg/kg-day cyanide ion (CIST), respectively. The biological significance of the behavioral
66
-------
changes observed in this study is unclear. In addition, the utility of this study is limited by the
use of bolus dosing. In comparison to relatively steady intake throughout the day via dietary
administration, bolus dosing produces higher peak blood levels as the entire daily dose is rapidly
absorbed. This difference is especially important considering that the toxicity of cyanide is
highly rate dependent. Thus, the findings from bolus exposure to cyanide are considered less
relevant to subchronic or chronic exposure conditions in humans. Due to the use of a bolus
regimen, this study was not considered appropriate for selection as the principal study.
Soto-Blanco et al. (2002a) treated Wistar rats (six to seven per group) with 0, 0.06, 0.12,
or 0.24 mg/kg-day by gavage for 12 weeks and reported histopathologic changes in the CNS.
The same authors also conducted a 5-month drinking water study in female goats (six to eight
per group) with concentrations ranging from 0 to 1.2 mg/kg-day (Soto-Blanco et al., 2002b). In
these studies, the authors reported a variety of histopathologic changes in the CNS described as
spheroids on the spinal cord, neuronal loss in the hippocampus, damaged Purkinje cells, gliosis,
and loss of cerebellar white matter. However, the authors provided no quantitative data,
precluding a dose-response characterization of the reported effects. The lack of quantitation of
the observed histologic effects and the use of bolus dosing precluded further consideration as the
principal study.
Kamalu (1993) evaluated the toxicity of sodium cyanide (NaCN) administered to
22-week-old mongrel dogs (six males per group) for 14 weeks. The diet was supplemented with
NaCN corresponding to 0 or 1.04 mg/kg-day CIST. Lesions in the kidneys and adrenal gland
were reported at the only dose tested; however, no quantitation of these lesions was provided. In
the testes, a specialized morphologic analysis indicated that the treated dogs had a significantly
decreased percentage of tubules in stage VIII of the spermatogenic cycle, as compared with
controls (p < 0.01). An evaluation of the thyroid of animals in this study was published by
Kamalu and Agharanya (1991). Serum T3 was significantly decreased (55%) and thyroid weight
was significantly increased (23%) in the cyanide-exposed group. Based on thyroid enlargement
and histopathologic changes in the kidneys, testes, and adrenal glands, the only dose tested,
1.04 mg/kg-day, was considered to be the LOAEL. The authors reported that the animals
suffered from recurring parasitic infestations that required treatment with pharmaceuticals
throughout the study. Therefore, the use of the data from the Kamalu (1993) and Kamalu and
Agharanya (1991) studies are limited by the use of dogs of compromised health status and were
not selected as the principal study for the derivation of the RfD.
An unpublished study by Leuschner and Neumann (1989) administered KCN to male
Sprague-Dawley rats (26-40/group) in drinking water for 13 weeks. Administered doses were 0,
40, 80, or 160 mg KCN/kg-day or 0, 16, 32, or 64 mg CN7kg-day. Early mortality was observed
at the high dose with 11 animals dying prematurely. Body weight was statistically significantly
decreased 42% at the high dose and 15% in the mid dose. Organ weight changes were not
observed at the lowest dose level tested (16 mg CNVday). Absolute thymus weight was
67
-------
statistically significantly decreased (20%) at the mid-dose level (32 mg GST/day). Statistically
significant relative and absolute organ weight changes were seen at the highest dose level (64 mg
CN7kg-day), although these changes were inconsistent. At the highest dose, absolute heart,
liver, spleen, kidney, and brain weights were statistically significantly decreased compared to the
controls. However, when relative weights were calculated, all organs showed increased weight
compared to controls (except for the thymus, which was decreased). A LOAEL of 32 mg/kg-day
was identified based on decreased body weight. Due to the availability of studies demonstrating
effects at lower doses, this study was not selected as the principal study.
Studies observing low-level developmental effects were also considered in the selection
of the principal study and critical effect (Soto-Blanco and Gorniak, 2004, 2003). Soto-Blanco
and Gorniak (2004) administered gavage doses of CW equivalent to 0, 0.4, 0.8, or 1.2 mg/kg-
day throughout gestation (GD 24 to birth) to pregnant goats (six per group) and observed
elevated TS (but not 14) levels in dams and offspring tested at birth in the highest dose group.
Another publication by the same authors (Soto-Blanco and Gorniak, 2003) treated goats with 0,
0.4, 0.8, or 1.2 mg/kg-day during lactation (PNDs 0-90) and identified vacuolation of kidney
epithelial cells and hepatocytes in offspring and dams at unspecified doses. Incidence or severity
of the reported histologic lesions was not provided, precluding any characterization of dose
response. These studies are limited by the use of bolus doses and the lack of dose-response
characterization, which precludes their selection as principal studies.
A study by NTP (1993) examined the toxicity of cyanide over a wide range of doses.
NTP administered NaCN in drinking water to rats and mice (10/sex/group) at concentrations of
0, 0.16, 0.48, 1.4, 4.5, and 12.5 mg/kg-day CW in male rats; 0, 0.16, 0.53, 1.7, 4.9, and
12.5 mg/kg-day in female rats; 0, 0.26, 0.96, 2.7, 8.6, and 24.4 mg/kg-day CIST in male mice; and
0, 0.32, 1.1, 3.3, 10.1, and 28.8 mg/kg-day in female mice. Reproductive effects were observed
in male animals of both species, although rats appeared to be the more sensitive species. In rats,
a statistically significant decrease in cauda epididymis weight (7%) was seen at doses
>1.4 mg/kg-day. A 7% decrease in whole epididymis weight (as compared to cauda epididymis
weight) was seen at 12.5 mg/kg-day. At the highest dose tested, 12.5 mg/kg-day, epididymis and
cauda epididymis weights were decreased by 7 and 13%, respectively. Dose-related decreases in
testis weight (8%), number of spermatid heads (14%), and spermatid concentration (14%) were
also found to be significant at doses >12.5 mg/kg-day. No change in epididymal sperm count
was observed at any dose; however, a statistically significant decrease in epididymal sperm
motility was observed at doses >1.4 mg/kg-day CIST, although it did not appear to increase in
severity with dose.
In consideration of the available studies reporting low-dose effects of chronic and
subchronic oral exposure to cyanide in animals, the NTP (1993) study was chosen as the
principal study. This study was well designed, with five dose groups of 10 animals per group per
sex and species. Numerous tissues and endpoints were assessed, and methods and observed
68
-------
effects were thoroughly reported. This study identified statistically significant male reproductive
effects in rats and mice that increased in severity in a dose-dependent manner. The observed
effects included decreased cauda and whole epididymis weights, decreased testes weight, and
altered sperm parameters.
The reproductive effects observed by NTP (1993) are consistent with an effect on male
reproductive endpoints, including organ weights and sperm parameters, although the magnitude
of the effects alone may be insufficient to decrease fertility in rats. However, human males have
markedly lower rates of sperm production and sperm counts compared with rats; thus, the
potential impact of decrements in sperm quality in humans is considered to be greater than that
of rats (U.S. EPA, 1996; Working, 1988). Furthermore, the cyanide database contains limited
additional support for the specific endpoint of reproductive toxicity (Kamalu, 1993). Therefore,
for the above reasons, NTP (1993) was chosen as the principal study, and all statistically
significantly altered reproductive endpoints in rats and mice were benchmark dose (BMD)
modeled and are presented in Section 5.1.2 and Appendix B.
EPA has selected decreased cauda epididymis weight as the critical effect because it was
determined that this effect represents the most sensitive endpoint indicative of male reproductive
toxicity. The cauda epididymis is one of the three primary subsections of the epididymis (along
with the caput and corpus) and functions as the site of sperm storage and maturation. Because
the cauda is part of the epididymis, these weights are not independent endpoints. BMD analysis
of the observed reproductive data from rats and mice indicated decreased (absolute) whole
epididymis and cauda epididymis weights to be the most sensitive reproductive effects observed.
Points of departure (PODs) for these endpoints identified through BMD modeling were virtually
identical. However, examination of the organ weight data from the principal study (NTP, 1993)
indicated that data for decreased cauda epididymis weight was statistically significantly
decreased (7%) at the lowest dose tested, whereas the decrease in whole epididymis weight (2%)
was not. It is possible that use of whole epididymis weight may mask the effect first observed in
the cauda region of the epididymis. Thus, decreased cauda epididymis weight was considered a
more sensitive effect.
Altered sperm parameters support the observed decreases in reproductive organ weight
seen in the NTP (1993) study. At the lowest dose examined, 1.4 mg/kg-day CIST, a modest, but
statistically significant decrease in epididymal sperm motility (4%) was observed, although its
severity did not increase with dose. Additionally, at the highest dose tested, testicular spermatid
count was statistically significantly decreased (14%). Epididymal sperm count was not affected
at any dose tested.
Human male fertility is established to be lower than that of rodent test species; thus,
human fertility may be more susceptible to damage from toxic agents (Working, 1998; U.S.
EPA, 1996). Therefore, according to the U.S. EPA Guidelines for Reproductive Toxicity Risk
Assessment (U.S. EPA, 1996), statistically significant changes to measures in sperm parameters
69
-------
including sperm count, morphology, or motility are considered adverse. However, no decrease
in epididymal sperm count and only a modest decrease in sperm motility were observed at doses
at which the cauda epididymis weight was statistically significantly decreased. The data from
NTP (1993) suggest that cauda epididymis weight is an effect that precedes more severe
decrements in sperm parameters, such as decreased testicular spermatid count, seen at the highest
dose. Therefore, decreased cauda epididymis weight, the most sensitive effect observed in this
study, was chosen as the critical effect.
Several studies in a variety of experimental models described above have reported
LOAELs in the same range or lower than the reproductive effects identified by NTP (1993).
However, interpretations of these studies are complicated by various issues, including limited
reporting of methods, incidences, severity, and statistical significance of observed effects in
addition to the use of bolus dosing regimens or animals of compromised health status (Manzano
et al., 2007; Soto-Blanco and Gorniak, 2004, 2003; Soto-Blanco et al., 2002a, b; Kamalu, 1993;
Jackson, 1988). Nevertheless, possible reference values based on the observed effects from
Jackson (1988), Manzano et al. (2007), and Kamalu (1993) are presented for comparison in
Section 5.1.4.
5.1.2. Method of Analysis
Statistically significantly altered reproductive endpoints in rats and mice observed in the
NTP (1993) study were BMD modeled, including decreased cauda and whole epididymis
weights, decreased testes weight, and altered sperm parameters (Table 5-1). Epididymal sperm
motility, although statistically significantly decreased in all treated groups, did not exhibit a
dose-response relationship, and thus, was not amenable to BMD modeling. For reproductive
organ weight changes, absolute reproductive organ weights, as opposed to relative organ
weights, were modeled. The absolute reproductive organ weight data presented by NTP (1993)
showed dose-related decreases in rats and mice. Relative organ weights did not show a stronger
dose-response than absolute organ weights. The study found body weight decreases in the
highest dose group of male rats (6%,/> < 0.05) and in the highest dose of male mice (4%, not
statistically significant). Given the lack of substantive body weight changes in rats and mice,
especially at lower doses that showed organ weight changes, relative organ weights were not
analyzed further.
70
-------
Table 5-1. Reproductive endpoints in male rats and mice observed following
administration of NaCN in drinking water for 13 weeks
Concentration (ppm)
0
30
100
300
Rats
Dose (mg/kg-d)
0
1.4
4.5
12.5
Weights (g)a
Testis, absolute
Epididymis, absolute
Cauda epididymis, absolute
1.58 ±0.094
0.448 ±0.019
0.162 ±0.009
1.56 ±0.063
0.437 ±0.016
0.150 ±0.013b
1.52 ±0.063
0.425 ± 0.022b
0.148 ±0.013b
1.46±0.063C
0.417 ±0.016C
0.141 ±0.009C
Spermatid measurements3
Spermatid count (/10~4 mL)
89.28 ±9.64
84.68 ±12.74
82.90 ±9.99
77.10±6.96b
Mice
Dose (mg/kg-d)
0
2.7
8.6
24.3
Weights (g)a
Epididymis, absolute
Cauda epididymis, absolute
0.049 ± 0.003
0.017 ±0.003
0.047 ± 0.006
0.016 ±0.000
0.047 ± 0.003
0.015 ±0.003b
0.044 ± 0.003b
0.014 ±0.003b
aValues are mean ± SD.
bSignificantly different from control atp < 0.05 using Shirley's test.
Significantly different from control atp < 0.01 using Shirley's test.
Source: NTP(1993).
Continuous models (i.e., linear, polynomial, and power) with constant variance were fit
to the data by using U.S. EPA BMD software (BMDS) (version 1.4.1). The other continuous
model available in BMDS, the Hill model, was not fit to these data because fitting of the Hill
model requires the estimation of four parameters (i.e., intercept, v, n, and k), which necessitates
having a minimum of five dose groups in order to have adequate degrees of freedom for testing
model fit. The NTP (1993) study employed only four dose groups, and thus, the Hill model
could not be fit to these data.
A benchmark response (BMR) level was selected corresponding to a change in the mean
response equal to 1 SD from the control mean for cauda epididymis weight. Information
regarding the degree of change in this endpoint that is considered biologically significant was not
available in the literature. Therefore, the BMR for continuous data of 1 SD change in the control
mean was selected under the assumption that it represents a minimally biologically significant
response level. By using the best fitting model for this data set, a 1 SD change was equivalent to
a 7% decrease in cauda epididymis weight. The BMD modeling reports generated from
modeling the reproductive endpoints from the NTP (1993) study are summarized below in Table
5-2.
71
-------
Table 5-2. BMD modeling results for observed reproductive endpoints
Endpoint
Fitted model
Goodness-of-fit
/7-value
AIC
BMD
BMDL
Rats
Cauda epididymis weight
(absolute)
Epididymis weight (absolute)
Testis weight (absolute)
Spermatid concentration
Linear
Polynomial
Power
Linear
Polynomial
Power
Linear
Polynomial
Power
Linear
Polynomial
Power
0.08
0.11
0.08
0.22
0.73
0.22
0.82
0.98
0.82
0.70
0.53
0.70
-312.75
-313.10
-312.75
-274.73
-275.64
-274.73
-167.94
-166.32
-167.94
227.04
228.73
227.04
8.4
3.5
8.4
8.2
3.2
8.2
7.4
5.3
7.4
11.2
8.5
11.2
5.6
1.9
5.6
5.6
1.8
5.6
5.1
2.4
5.1
6.9
2.9
6.9
Mice
Epididymis weight (absolute)
Cauda epididymis weight
(absolute)
Linear
Polynomial
Power
Linear
Polynomial
Power
0.67
0.38
0.67
0.87
0.82
0.69
-378.71
-376.73
-378.71
-402.99
-399.67
-400.99
21.5
20.5
21.5
25.9
16.3
16.3
13.0
6.6
13.0
14.6
5.2
14.6
AIC = Akaike's Information Criterion; BMDL = 95% lower confidence limit on the BMD
Data source: NTP (1993).
All three models provided an adequate fit to this data set based on the goodness-of-fit
statistic (p > 0.1). Of these three models, the polynomial model provided the best fit to the data
based on this model's exhibiting the lowest Akaike's Information Criterion (AIC) and visual
inspection of the plot of observed versus expected values across the three models. The detailed
BMD modeling output for the selected polynomial model is presented in Appendix B. The BMD
associated with a 1 SD decrease in cauda epididymis weight in rats is 3.5 mg/kg-day, and its
95% lower confidence limit (BMDL) is 1.9 mg/kg-day.
5.1.3. RfD Derivation—Including Application of Uncertainty Factors (UFs)
The BMDL of 1.9 mg/kg-day based on decreased cauda epididymis weight in rats was
used as the POD for the derivation of the RfD. A total UF of 3,000 was applied to the POD of
1.9 mg/kg-day: 10 for interspecies extrapolation from animals to humans (UFA), 10 for human
72
-------
intraspecies variability (UFH), 10 to account for the use of a sub chronic study (UFS), and 3 to
account for database deficiencies (UFo).
A default 10-fold UF was applied to account for uncertainties in extrapolating from
laboratory animals to humans. Humans and laboratory animals have qualitatively similar
absorption, distribution, metabolism, and excretion of cyanide. However, quantitative
comparisons of toxicokinetic parameters are lacking. Additionally, a wide range of sensitivity to
effects of cyanide has been observed between different species of experimental animals. The
available data do not provide quantitative information on the difference in susceptibility to
cyanide between rats and humans.
A default 10-fold UF was applied to account for variation in susceptibility among
members of the human population (i.e., interindividual variability). Insufficient information is
available to quantitatively estimate variability in human susceptibility to cyanide.
A 10-fold UF was applied for the extrapolation of subchronic-to-chronic exposure
duration. The 91-day study by NTP (1993) falls well short of a lifetime duration. In addition,
there is a lack of data on male reproductive parameters following chronic administration of
cyanide, and the mode of action of the reproductive effects observed in this study is unclear.
Therefore, it is unknown whether effects would be more severe or would be observed at lower
doses with a longer exposure duration. For these reasons, the UF of 10 to extrapolate from a
study with a subchronic duration was applied.
A 3-fold UF was applied to account for deficiencies in the cyanide toxicity database,
including the lack of a multigenerational reproductive toxicity study and a sensitive
neurodevelopmental toxicity study. The database includes limited human data from
epidemiological studies of workers exposed by inhalation to hydrogen cyanide (HCN)
(Chatgtopadhyay et al., 2000; Banerjee et al., 1997; Blanc et al., 1985; El Ghawabi et al., 1975).
The database also includes studies in laboratory animals, including chronic and subchronic
dietary exposure studies and developmental studies. The database includes oral toxicity studies
in various animal species, including rats, mice, rabbits, dogs, pigs, and goats. A developmental
study with skeletal and visceral examination has not been conducted for cyanide; however,
developmental studies exist in rats and goats evaluating the thyroid, kidney, liver, pancreas,
brain, and CNS of gestationally and/or lactationally exposed offspring (Imosemi et al., 2005;
Malomo et al., 2004; Soto-Blanco and Gorniak, 2004, 2003; Tewe and Maner, 1981). External
or overt developmental effects with cyanide exposure have not been noted at doses up to
1.2 mg/kg-day in goats (Soto-Blanco and Gorniak, 2004, 2003) and 21.6 mg/kg-day in rats
(Imosemi et al., 2005; Malomo et al., 2004; Tewe and Maner, 1981). However, due to the mode
of action of thiocyanate involving competitive iodine uptake inhibition and implications for
neurotoxicity in the developing animal, the lack of a sensitive neurodevelopmental toxicity study
to assess endpoints sensitive to thyroid disruption is an additional weakness in this database. The
cyanide database is also lacking an appropriately designed multigenerational reproductive
73
-------
toxicity study, although an assessment of reproductive organs was included as a component of
the 13-week NTP (1993) studies in rats and mice, and testicular histology was also assessed in
dogs (Kamalu, 1993). These studies in adult animals demonstrated low-dose reproductive
effects. The observance of these effects reinforces the need for a multigenerational assessment
of reproductive endpoints. Therefore, in consideration of the above data gaps, a 3-fold UF to
account for deficiencies in the database was applied.
The UF for LOAEL-to-NOAEL extrapolation was not used because the current approach
is to address this factor as one of the considerations in selecting a BMR for BMD modeling. In
this case, a BMR of a 1 SD change from the control mean in epididymis weight was selected
under an assumption that it represents a minimal biologically significant response level.
The oral RfD for CW was calculated as follows:
RfD = BMDLio - UF
= 1.9 mg/kg-day-3,000
= 6.3 x 10~4 mg/kg-day (rounded to 6 x 10~4 mg/kg-day)
The RfDs for simple cyanide salts like NaCN and KCN, which freely dissociate into
cyanide, are calculated from the RfD for CIST by adjusting for molecular weight (i.e., the RfD is
multiplied by the ratio of the total molecular weight of the compound to the molecular weight of
the CN~):
RfD for aqueous HCN [HCN(aq)] = 6.3 x 10'4 x 27/26 = 7 x 10'4 mg/kg-day
RfD for NaCN = 6.3 x 10'4 x 49/26 = 1 x lO'3 mg/kg-day
RfD for KCN = 6.3 x 10'4 x 65/26 = 2 x 10'3 mg/kg-day
RfD for calcium cyanide4 [Ca(CN)2] = 6.3 x 10'4 x 92/(2 x 26) = 1 x lO'3 mg/kg-day
RfD for potassium silver cyanide5 [KAg(CN)2] = 6.3 x 10~4 x 199/26 = 5 x 10~3 mg/kg-day
RfD for cyanogen5 (CN)2 = 6.3 x 10'4 x 52/26 = 1 x lO'3 mg/kg-day
Use of the RfD for free cyanide to calculate RfDs of other cyanide compounds may be
merited, but the ability of the individual cyanogenic species to dissociate and release free
cyanide in aqueous solution (and at physiological pHs) should be taken into consideration. If
dissociation of the compound is expected, then liberated cations should be considered for
potential toxicity independent of CW. Also, some metallocyanides, such as copper cyanide,
have chemical-specific data and are not included in this analysis.
4Two molar equivalents of free CN released in water.
5One molar equivalent of free CN" released in water.
74
-------
5.1.4. RfD Comparison Information
Reduced reproductive organ weight, altered sperm parameters, increased thyroid weight,
altered thyroid hormones, and altered testicular, kidney, and adrenal histopathology are observed
low-level effects following subchronic oral exposure to cyanide (Manzano et al, 2007; Kamalu,
1993; NTP, 1993; Jackson, 1988). Table 5-3 provides a summary of potential PODs and
resulting potential reference values derived from these endpoints. Additionally, Figure 5-1
provides a graphical representation of this information. This figure should be interpreted with
caution since the PODs across studies are not necessarily comparable, nor is the confidence the
same in the data sets from which the PODs were derived. The PODs presented in this figure are
based on either a BMDLiso, NOAEL, or LOAEL (if no NOAEL was available).
Table 5-3. Potential PODs with applied UFs and resulting potential
reference values
Effect
Decreased testis weight
Decreased epididymis
weight
Decreased cauda
epididymis weight
Decreased testicular
spermatid concentration
Decreased epididymis
weight
Decreased cauda
epididymis weight
Altered thyroid
hormones, behavioral
changes
Increased thyroid
weight
Kidney, testes, and
adrenal gland effects
Potential
POD
5.1b
1.8b
1.9b
6.9b
13b
14.6b
0.7C
1.6d
1.04e
Species
Rat
Rat
Rat
Rat
Mouse
Mouse
Pig
Pig
Dog
UFa
Total
3,000
3,000
3,000
3,000
3,000
3,000
3,000
3,000
30,000
A
10
10
10
10
10
10
10
10
10
H
10
10
10
10
10
10
10
10
10
L
NA
NA
NA
NA
NA
NA
1
1
10
s
10
10
10
10
10
10
10
10
10
D
3
3
3
3
3
3
3
3
3
Potential
reference
value
2 x 1Q-3
6 x 1Q-4
6 x 1Q-4
2 x 1Q-3
4 x 1Q-3
5 x 1Q-3
2 x 1Q-4
5 x 1Q-4
4 x 1Q-5
aUFs: A = animal to human (interspecies); H = interindividual (intraspecies); L = LOAEL-to-NOAEL;
S = subchronic to chronic duration; D = database deficiency.
bBMDL based on BMD modeling of a 1 SD change. Source: NTP (1993); NA, not applicable.
CPOD based on NOAEL. Sources: Jackson (1988);
dPOD based on NOAEL. Source: Manzano et al. (2007).
ePOD based on LOAEL. Source: Kamalu (1993).
75
-------
100
10
1
t
} 0.1
M
a
0.01
0.001
0.0001
0.00001
f
Deere
testis
weigl
(NTP
1993;
^
as
it,
in
i
\
ed "
rats Decr(
epidi
weig
(NTF
1993
n in
r f
|f
tl
r
5 i
It
?
^
? ^
^ ^
! J
A ft \
\ t • o*
I H Decreased epic
testicular wei
j *T-> j spermatid mic
:ased *Decreased ^ . /vrT
, . , concentration (Ml
dymis cauda
•1-1 • rats
it, rats epididymis ^ j
, weight, rats '
) (NTP, 1993)
IT
P. ?
1 i
! — !
Dec
reased Cauc
idymis epic
ght, wei£
e (NT
P, 1993)
n
if
A
?
/ u
/ ;
/ ;
t 'l
1 >
reased f
la _(
idymis ^
'ht, mice H
P, 1993) |
Thy
hon
beh
chai
(Jac
•j
, ^
\ \
'
P In
roid th
none and w
avioral n
ages, pigs 2(
kson, 1988)
1
rn
r
/ r
t f
t f
A t
1 \
m
m
creased !
yroid -|
eight, pigs I
/lanzano, 1
)07)
t
t
i
•
•
•
•
•
•
•
=
P Changes in
kidney, testes
and adrenal
gland, dogs
(Kamalu, 199
Figure 5-1. Potential reference value comparison array
* Point of departure
• Potential reference value
* Critical effect and RfD
UF, animal to human
UF, human variability
UF, subchronic to chronic
UF, LOAEL to NOAEL
UF, database
76
-------
A composite UF of 30,000 was applied to the LOAEL for the endpoints identified in the
Kamalu (1993) study, which is generally considered too large to support derivation of a
reference value. In the report, A Review of the Reference Dose and Reference Concentration
Processes (U.S. EPA, 2002), the RfD/RfC Technical Panel concluded that, in cases where the
total UF is more than 3,000, it is unlikely that the study or database is sufficient to derive a
reference value. Thus, the magnitude of the uncertainty associated with this specific study
indicates it is insufficient to support derivation of a potential reference value.
Some indication of the confidence associated with the resulting potential reference values
is reflected in the magnitude of the total UF applied to the POD (i.e., the size of the bar);
however, the text of Sections 5.1.1 and 5.1.2 should be consulted for a more complete
understanding of the issues associated with each data set, the rationale for the selection of the
principal study, and the critical effect used to derive the RfD. As discussed in Section 5.1.1,
among the studies considered, the subchronic study by NTP (1993) provided the data set for the
derivation of the RfD.
5.1.5. Previous RfD Assessment
An RfD for cyanide of 2 x 10"2 mg/kg-day was posted on the IRIS database in 1985 and
was based on co-principal studies (Philbrick et al., 1979; Howard and Hanzal, 1955). Howard
and Hanzal (1955) fed food fumigated with HCN to rats for 2 years and identified the high dose
of 10.8 mg/kg-day as the NOAEL. Philbrick et al. (1979) evaluated the thyroid and nervous
system in rats fed KCN for 11.5 months. The single dose tested, 44 mg/kg-day, was identified as
a LOAEL, based on myelin degeneration in the CNS and increased thyroid gland weight. These
two studies were considered together to identify the critical effect and the POD. The previous
RfD was based on a NOAEL of 10.8 mg/kg-day. The NOAEL was divided by a total UF of 500,
including a factor of 10 each for extrapolation from animals to humans and intraspecies
variability. A modifying factor of 5 was used to account for the apparent tolerance to cyanide
when it is ingested with food compared with administration by gavage or by drinking water (U.S.
EPA, 1992). However the apparent difference may have been due to instability of the cyanide
concentrations in the feed rather than differences in bioavailability (Palmer and Olson, 1979).
Since the posting of the 1985 IRIS RfD for cyanide, several new subchronic studies by
the oral route have been published. In addition, new data are available that evaluate potential
health effects following perturbations of thyroid function, in general, in pregnant women and
their offspring (see Section 4.8.1). Specifically, these studies show increased pregnancy
complications and decrements in learning and memory in offspring of women with subclinical
hypothyroidism (Kooistra et al., 2006; Casey et al., 2005; Pop et al., 2003; Haddow et al., 1999).
The revised RfD was derived based on a subchronic study by NTP (1993), which indicated
sensitive reproductive effects in rats and mice following a 91-day exposure to NaCN. In
77
-------
summary, the current assessment includes new chemical-specific data for cyanide and new
information regarding the severity of low level thyroid perturbations in sensitive populations.
5.2. INHALATION REFERENCE CONCENTRATION (RfC)
5.2.1. Choice of Principal Study and Critical Effect
Limited data are available on the effects of long-term inhalation exposure to cyanide.
Several occupational studies investigated the effects of inhalation exposure to HCN, and three of
these studies provide evidence of effects on the thyroid. One of these studies included
environmental exposure data based on breathing zone samples for the individual study
participants. There are no subchronic or chronic inhalation exposure studies of HCN in animals.
El Ghawabi et al. (1975) reported statistically significantly altered rates of iodide uptake
by the thyroid, thyroid enlargement, and CNS symptoms (e.g., self-reported increased incidence
of headache, weakness, and sensory changes for taste and smell) in workers (n = 36) exposed to
HCN for 5-15 years in three electroplating factories. Individual breathing zone measurements of
HCN were collected from each worker. The mean concentrations across factories ranged from
7.07 to 11.5 mg/m3 HCN and the values for individual workers ranged from 4.6 to 13.7 mg/m3
HCN. Urine SCIST levels, an indicator of internal dose, collected from workers were highly
correlated with individual HCN exposure concentrations (see Figure 4-1). Twenty of the
exposed workers (56%) were identified with mild to moderate thyroid enlargement. Radioactive
iodine uptake measured following a 2-day break in HCN exposure indicated statistically
significantly elevated iodide uptake after 4 hours (38.7 compared to 22.4%) and 24 hours
(49.3 compared to 39.9%) as compared to controls.
Increased 24-hour uptake of radioactive iodide by the thyroid has been reported to occur
in hyperthyroidism, iodine deficiency, and goiter (NLM, 2008a; Ravel, 1995). The study authors
concluded that the increased iodine uptake observed in the workers following the 2-day cessation
in exposure was a postexposure response to depletion of iodine in the thyroid. A similar increase
in iodide uptake has been seen with perchlorate (ClO/f), another competitive inhibitor of iodide
uptake, following cessation of exposure. Lawrence et al. (2000) measured iodine uptake in
volunteers administered doses of ClO/f at baseline, at 2 weeks of dosing, and then 2 weeks
postexposure cessation. The authors reported that iodide uptake decreased 10-38% in the low-
and high-dose groups (compared to baseline) at 2 weeks of dosing. Two weeks after exposure
was discontinued, iodide uptake was statistically significantly increased 22 and 25%, indicating a
rebound effect in iodide accumulation postexposure.
The lowest mean concentration of HCN recorded in the three factories, 7.07 mg/m3, is
designated as a LOAEL for thyroid enlargement and altered iodide uptake. The study authors
also indicated some coexposure of the workers to gasoline, alkali, and acid during the
electroplating process, although the magnitudes of these exposures were not quantified and it is
unclear if these exposures would impact the observed thyroid effects.
78
-------
Blanc et al. (1985) conducted a study of silver-reclamation workers (n = 36) examined an
average of 11 months following exposure. The median length of employment was 8.5 months
and mean exposure duration was 11 months. Workers were categorized into low-, moderate-, or
high-exposure groups based on their primary job activities. Information on exposure was
limited, as the plant had been shut down following the death of one worker from cyanide
overexposure. Environmental monitoring conducted the day after the plant was shut down found
that the 24-hour TWA HCN exposure was 16.6 mg/m3. Serum TSH levels in workers were
significantly elevated relative to laboratory controls. The authors noted a significant positive
trend with increasing exposure level for self-reported weight loss and several symptoms,
including dizziness, syncope, and nausea and vomiting. Serum TSH levels in workers were
reported as being significantly elevated in workers relative to laboratory controls. TS uptake in
the highest exposed workers (n = 9) was statistically significantly elevated compared to
laboratory controls. The authors reported that this elevation may reflect a post-inhibitory
response. Because there were multiple possible routes of cyanide exposure, including dermal
exposure and contamination of food, and because earlier air levels were likely higher than the
measured TWA concentration, the environmental monitoring data do not allow for the selection
of a LOAEL. Additionally, this study examined workers an average of 11 months post-
occupational HCN exposure and may therefore have missed effects that have the potential to
regress following cessation of exposure. The observation of significant effects on the thyroid
almost 1 year after cessation of exposure indicates that these observed thyroid effects are not
transient. Due to limitations of this study based on its retrospective design and because El
Ghawabi et al. (1975) reported significant effects at lower levels, this study was not selected for
the derivation of the RfC.
An unpublished study by Leeser et al. (1990) compared the health of 63 male cyanide salt
production workers with a control group of 100 British workers in a cross-sectional study.
Cyanide workers were exposed for periods ranging from 1 to 32 years with a mean exposure
duration of 12.6 years and mean breathing zone concentrations of cyanide up to 1 mg/m3.
Several hematological parameters in cyanide workers were statistically significantly elevated
compared to controls, including hemoglobin (15.57 vs. 15.08 g/dL), ratios associated with
hemoglobin, such as MCH and MCHC, and lymphocyte count (2.87 compared to 2.55 x 109/L).
However, the biological significance of these slight elevations in hematological parameters is
unclear. Serum T4 levels in cyanide-exposed workers were decreased in controls, but the
difference was not reported as statistically significant by the study authors (85.13 ± 2.51 vs.
89.04 ±1.81 nmol/L). Additionally, serum T4 was below the clinical reference range (60-160
nmol/L) in 3 cyanide-exposed workers. Other commonly administered, and more sensitive, tests
for thyroid function, including TSH, free T4, or iodide uptake, were not measured. It is unclear
whether a NOAEL for thyroid effects can be established by this study as only one, relatively
insensitive indicator of thyroid function was measured. A LOAEL of 1 mg/m3 cyanide for
79
-------
increased lymphocyte count and increased hemoglobin concentration was identified. A NOAEL
for thyroid effects was not identified from this study based on the lack of measurement of
sensitive thyroid parameters, although overt hypothyroidism was not observed. Due to the
failure of the study authors to include additional, more sensitive thyroid function tests, as well as
the unclear biological significance of the hematological effects observed, this study was not
selected as the principal study.
In another occupational study of electroplating workers exposed to HCN, workers
(n = 35) exposed for 5 years had significantly decreased T^ (48%) and 14 (37%) and significantly
increased TSH (142%) as compared to controls (Banerjee et al., 1997). Serum SOT was
elevated in workers compared to controls. A significant negative correlation between serum 14
and SOT concentrations and a significant positive correlation between TSH and SGST
concentrations were observed. However, no information was provided on exposure levels;
therefore, no NOAEL or LOAEL could be identified from this study.
Chandra et al. (1980) reported on a group of 23 electroplating workers chronically
exposed to average breathing zone concentrations of 0.15 mg/m3 HCN. The authors noted that
the workers complained of symptoms typical of cyanide poisoning, but provided no additional
information on specific symptoms or further analysis. In the absence of information on
measured effects, no NOAEL or LOAEL could be identified from this study, precluding its use
in a quantitative risk assessment.
Chatgtopadhyay et al. (2000) found some indication of decreased pulmonary function in
workers at a metal-tempering plant. Specifically, the authors observed decreased pulmonary
function in 24 workers exposed for a mean duration of 24 years. This study provided no
information regarding the environmental exposure levels of the workers, and thus, no NOAEL or
LOAEL could be identified, limiting this study's utility for risk assessment.
Considering the availability of studies in the HCN database, El Ghawabi et al. (1975) was
chosen as the principal study. The results of this study indicate that chronic, low-level exposure
to cyanide was associated with thyroid enlargement and altered iodine uptake in humans. This
study examined workers exposed to HCN for extended durations (5-15 years). Although this
study is limited by small sample size, it used matched controls and is not confounded by
smoking since all workers and controls were nonsmokers. The authors collected individual
breathing zone measurements of HCN exposure, which were strongly correlated with urinary
SCIST, a measure of internal exposure. The range of mean individual HCN concentrations
reported from all three plants was 6.4-10.4 ppm and the range among the 36 individuals was
4.2-12.4 ppm, indicating a similar magnitude of exposure for exposed workers. Thyroid
enlargement was strongly associated with HCN exposure, with 56% of the exposed workers
diagnosed with mild to moderate thyroid enlargement. This observation is supported by an
increased radioactive iodide uptake in workers (p < 0.001). Increased uptake of radioactive
iodide has been reported to occur in hyperthyroidism, iodine deficiency, recovery from thyroid
80
-------
suppression, and goiter (NLM, 2008a; Spencer, 2008; Ravel, 1995). The increase in iodide
uptake may have resulted from temporary weekend cessation of exposure. A similar
phenomenon of post-inhibitory response was also seen in the occupational study by Blanc et al.
(1985), which noted significantly increased T3 uptake observed in workers several months
following HCN exposure.
The thyroid alterations reported in El Ghawabi et al. (1975) are believed to be
biologically significant effects. These effects, particularly thyroid enlargement, are consistent
with those observed in oral exposure animal studies (Manzano et al., 2007; Jackson, 1988;
Philbrick et al., 1979). Additionally, other human inhalation studies have indicated thyroid
effects in exposed workers (Banerjee et al., 1997; Blanc et al., 1985). The thyroid effects
observed in El Ghawabi et al. (1975) are also supported by mode-of-action data for cyanide,
indicating competitive iodide uptake (see Section 4.6). The thyroid enlargement observed in the
HCN-exposed workers likely indicates antagonism of iodine uptake by the cyanide metabolite,
SGST. This biological response indicates a stress on the homeostatic mechanisms of the thyroid,
which is of special concern to populations that include individuals with iodine deficiency,
individuals with clinical or subclinical hypothyroidism, and the developing fetus.
The HCN inhalation database contains limited exposure-response data. Information on
acute human occupational inhalation exposure to HCN does exist, but is limited to case reports
of accidental overexposures with unclear exposure concentrations and/or durations.
Additionally, no chronic or subchronic HCN inhalation studies exist in animals. The only
available studies that report exposure data and potential health effects of inhalation of HCN are
occupational exposure studies. In consideration of this limited inhalation database, the El
Ghawabi et al. (1975) study was chosen as the principal study. This study of electroplating
workers at three factories in Egypt included individual breathing zone measurements from the
study participants and reported a strong correlation between these measurements and urinary
levels of SCIST. However, the exposure assessment was based on a single 15-minute breathing
zone sample for each worker and the potential for dermal exposure was not explicitly discussed.
Despite the weakness in the available exposure information, the El Ghawabi et al. (1975) study
was selected as the principal study, and thyroid enlargement and altered iodide uptake were
designated as the critical effects. The lowest mean concentration of HCN reported, 6.4 ppm
HCN, is designated as the LOAEL. The choice of thyroid dysfunction as a critical effect is
supported by other epidemiologic studies in a silver-reclaiming factory (Blanc et al., 1985) and
in electroplating workers (Banerjee et al., 1997).
5.2.2. Method of Analysis
A LOAEL was available from the principal study. Because quantitative concentration-
response data were available only for one concentration, benchmark concentration modeling
could not be conducted.
81
-------
The lowest mean concentration recorded among the three factories evaluated by
El Ghawabi et al. (1975) was 6.4 ppm. Assuming a temperature and pressure of 25°C and
760 mm Hg, this LOAEL in ppm was multiplied by the molecular weight of HCN and divided
by 24.45 to determine the LOAEL in mg/m3 HCN.
LOAEL (ppm) x 27/24.45 = 6.4 ppm x 27/24.45 = 7.07 mg/m3 HCN
The standard method for adjustment of LOAEL or NOAEL values from occupational
studies was employed as described in U.S. EPA (2002). Because El Ghawabi et al. (1975) did
not report daily exposure durations for exposed workers, an 8-hour/day, 5-day/week exposure
scenario was assumed. A default occupational ventilation rate of 10 m3/8-hour day and a default
ventilation rate for continuous ambient exposure of 20 m3/24-hour day were used. The exposure
was also adjusted to account for the difference between occupational exposure for 5 days/week
versus continuous ambient exposure for 7 days/week:
LOAEL(ADJ) = 7.07 mg/m3 HCN x 10/20 x 5 days/7 days = 2.5 mg/m3 HCN
5.2.3. RfC Derivation—Including Application of Uncertainty Factors (UFs)
The RfC is based on thyroid enlargement and altered iodide uptake reported in an
occupational study of electroplating workers and is supported by other occupational studies and
oral animal studies reporting similar thyroid effects. The LOAEL from this study (adjusted for
continuous exposure) of 2.5 mg/m3 HCN was used as the POD.
A total UF of 3,000 was applied to the POD: 10 for the extrapolation of a LOAEL to a
NOAEL (UFL), 3 for the extrapolation from a subchronic to chronic exposure duration (UFs),
10 for human intraspecies variability (UFH), and 10 to account for database deficiencies (UFD).
A 1-fold UF for extrapolation across species was applied because the RfC is based on
thyroid enlargement and altered iodide uptake reported in an occupational study.
A 10-fold UF was used to account for variation in susceptibility to cyanide among
members of the human population. Although some information is available on potential
sensitive populations, as described in Section 4.8, there are insufficient quantitative data to
inform the UF for human variability with chemical-specific data.
A 10-fold UF was used for extrapolating from a LOAEL to a NOAEL (UF^ because the
POD was a LOAEL.
A 3-fold UF was applied to account for extrapolation from what is assumed to be a
largely subchronic exposure to chronic exposure duration. The workers in the principal study
were exposed to cyanide for 5-15 years. Of the 36 workers, 14 had been exposed for 5 years, 14
for 5-10 years, 7 for 10-15 years, and 1 for >15 years. The mean and median exposure times for
the worker population were not reported. Twenty of the 36 exposed workers had thyroid
82
-------
enlargement; however, the authors found no correlation between duration of exposure and either
incidence or magnitude of thyroid enlargement in the workers. A lack of an association could be
related to the low sample size and/or the failure of the authors to consider iodine status of the
workers. In addition, following continued administration of cyanide in rats, thyroid effects were
less prominent at 11 months of exposure compared to 4 months of exposure (Philbrick et al.,
1979), which provides some indication (although limited), that increased duration of exposure
may not lead to an increase in thyroid effects. Therefore, it is unknown whether greater
alteration in thyroid function or increased incidence of the effect would be observed with longer
exposure duration. In the absence of information indicating the effects of HCN would not
progress in incidence or severity, a subchronic to chronic UF of 3 was applied.
A 10-fold UF was applied to account for deficiencies in the cyanide inhalation database.
The database includes limited human data from epidemiologic studies of inhalationally exposed
workers (Blanc et al., 1985; El Ghawabi et al., 1975). Inhalation studies on ACH evaluated
limited male and female reproductive endpoints and were negative for impacts on fertility
(Monsanto Co., 1985a, b). Oral studies of cyanide exposure in rodents have suggested that the
male reproductive tract is a sensitive target of cyanide toxicity following subchronic exposure
(NTP, 1993). However, the database lacks developmental and multigenerational reproductive
toxicity studies. Data are available which evaluate potential health effects following
perturbations of thyroid function in pregnant women and their offspring (see Section 4.8.1).
These studies indicate increased pregnancy complications and decrements in learning and
memory in offspring of women with subclinical hypothyroidism (Kooistra et al., 2006; Casey et
al., 2005; Pop et al., 2003; Haddow et al., 1999). Due to the proposed cyanide mode of action of
thyroid disruption (through the metabolite thiocyanate), developmental neurotoxicity studies or
developmental studies assessing maternal and fetal thyroid function are also considered data
insufficiencies. Thus, a database UF of 10 was applied in this assessment to account for the lack
of developmental and multigenerational reproductive toxicity studies.
The RfC for HCN was calculated as follows:
RfC = LOAEL(ADJ) - UF
= 2.5 mg/m3 - 3,000
= 0.00083 mg/m3 (rounded to 8 x 10"4 mg/m3)
It is recommended that the RfC for HCN should not be used to estimate an RfC for
cyanide salts due to inhalation considerations. Specifically, exposure to HCN occurs as a gas,
whereas the extremely high boiling points and vapor pressures of cyanide salts predict that
inhalation exposure would occur as aerosols. Different dosimetric approaches would apply to
83
-------
the aerosol (or particle) exposures that would result from exposure to cyanide salts, compared
with exposure to HCN gas.
5.2.4. Previous RfC Assessment
An RfC for HCN of 3 x 10~3 mg/m3 was posted on the IRIS database in 1994. This RfC
was based on findings of thyroid effects and neurological symptoms in workers from the study
by El Ghawabi et al. (1975). The POD for this RfC was based on an adjusted LOAEL of
7.07 mg/m3. The LOAEL was divided by a total UF of 1,000 comprised of UFs of 10 each to
account for the lack of a NOAEL and intrahuman variability. UFs of 3 each were applied to
account for the use of a study of less-than-chronic duration and deficiencies in the database (lack
of chronic and multigenerational reproduction studies). Since the posting of the previous IRIS
RfC for cyanide, no new chronic or subchronic inhalation studies for HCN with quantitative
information are available in the literature. The revised RfC for cyanide was derived from the
same study as the previous assessment based on the observation of altered thyroid function (as
indicated by iodine uptake) in occupationally exposed male workers. In the current assessment,
a database UF of 10 was applied to account for the lack of developmental and multigenerational
reproductive toxicity studies.
5.3. UNCERTAINTIES IN THE ORAL REFERENCE DOSE AND INHALATION
REFERENCE CONCENTRATION
The following discussion identifies uncertainties associated with the quantification of the
RfD and RfC for cyanide. Following EPA practices and guidance (U.S. EPA, 1994b, 1993), the
UF approach was applied to the chosen PODs to derive an RfD and RfC (see Sections 5.1.3 and
5.2.3). Factors accounting for uncertainties associated with a number of steps in the analyses
were adopted to account for extrapolating from an animal study to human exposure, a diverse
human population of varying susceptibilities, and database deficiencies.
The database for cyanide includes limited human data from studies of occupationally
exposed workers. Endpoints observed in inhalationally exposed workers include altered thyroid
function and CNS symptoms (including headache, weakness, nausea, and vomiting). The
database also includes oral exposure studies in laboratory animals, including limited chronic
studies, subchronic dietary exposure studies, and several developmental studies, including one
specifically assessing gross and microscopic brain morphology in rats (as discussed in
Section 4.3). Effects seen with low-dose oral exposure to cyanide include decreased
reproductive organ weight, decreased spermatid concentration, increased thyroid weight, and
histologic alterations in the CNS, kidney, testis, and adrenal glands. In addition to oral and
inhalation data, the database for cyanide includes studies on absorption, distribution, metabolism,
and excretion.
84
-------
Uncertainty exists in the selection of the most relevant animal species for human health
assessment. Studies in several species, including rodents, pigs, goats, and dogs, were considered
in the development of the RfD; however, limited data exist on differential species' sensitivity to
cyanide, especially in the context of long-term exposure.
The RfD was derived from a BMDL of 1.8 mg/kg-day, which was based on the
observation of decreased epididymis weight in male F344 rats exposed to NaCN in drinking
water for 13 weeks (NTP, 1993). This study treated male and female rats with doses of CIST
ranging from 0.16 to 12.5 mg/kg-day. Other reproductive effects observed at higher doses
included decreased caudal epididymis and testis weights and decreased spermatid count. After
consideration of all potential PODs, the RfD of 6 x 10~4 mg/kg-day was based on the observation
of decreased cauda epididymis weight in male F344 rats following subchronic dietary
administration of cyanide (NTP, 1993).
The mode of action of the decrease in cauda epididymis weight in rats is uncertain,
although limited information from other model systems of hypothyroid animals suggests that it
may be related to thyroid disruption from the primary metabolite, thiocyanate (see Section 4.5.4).
However, the critical study used as the basis for the RfD (NTP, 1993) did not assess thyroid-
related parameters, such as T4, T3, TSH, or thyroid weight. Therefore, it is not known whether
direct indicators of disruption of thyroid homeostasis accompanied the observed reproductive
effects.
Additional studies exist that determined different effects at lower doses, as discussed in
Section 5.1.1, including behavioral changes and decreased serum T4 in pigs (Jackson, 1988) and
kidney, adrenal, and testicular effects in dogs (Kamalu, 1993). Selection of either of these
studies would result in a lower RfD as portrayed in Table 5-3 and Figure 5-1. Ultimately, these
studies were deemed of lower confidence, due to issues concerning study design and reporting
(see Section 5.1.1), than the 3-month dietary study in rats and mice conducted by NTP (1993)
and thus, were not chosen as the principal study. To derive the RfD, UFs were applied to the
POD determined through BMD modeling of the critical effect of reduced cauda epididymis
weight in male rats. This study was well designed and conducted with several dose levels,
sufficient numbers of animals, and a wide range of tissues and endpoints assessed; however,
significant areas of uncertainty exist in the animal data relied upon for the RfD. UFs associated
with the extrapolation from the POD derived from an animal study to a diverse human
population of varying susceptibilities were applied.
Uncertainty exists in the selection of the BMR level utilized in the BMD modeling of the
critical effect (decreased cauda epididymis weight) to determine the POD. At this response level
in the cauda epididymis, no alteration in epididymal sperm count was detected, and thus, fertility
in these animals would not likely be affected. However, human males have markedly lower
fertility than do rats (U.S. EPA, 1996; Working et al., 1988) and thus, changes in male
reproductive endpoints may therefore be expected to have greater impact in humans as compared
85
-------
to rodents. In the absence of clear information to determine the level of change in cauda
epididymis weight related to a biologically significant change, a decrease of 1 SD in organ
weight was selected to represent a minimally adverse change.
The choice of BMD model is not expected to introduce a considerable amount of
uncertainty since 1 SD in the reduction of cauda epididymis weight is within the observable
range of the data. Other available continuous, constant variance models available in the BMD
software (i.e., the power and linear models) also showed acceptable fits (p > 0.1) and had AIC
values within 1 unit of the selected polynomial model. However, the polynomial model had
superior visual fit to the data, especially in the low-dose range. The BMDL estimates for various
models are not within a factor of 3, indicating some model dependence; therefore, in accordance
with the Benchmark Dose Technical Guidance Document (External Review Draft, U.S. EPA,
2000b), the model with the lowest BMDL estimate was selected.
Additional BMD modeling for other data sets, including additional reproductive
endpoints from the NTP (1993) study, was also conducted to provide other PODs for comparison
purposes (see Appendix B). A graphical representation of these and other potential PODs and
resulting reference values is shown in Figure 5-1 (see Section 5.1.4).
The default UF of 10 for the extrapolation from animals to humans accounts for
toxicokinetic differences and toxicodynamic differences. Physiologically based toxicokinetic
models can be useful for the evaluation of interspecies toxicokinetics; however, the cyanide
database lacks an adequate model that would inform potential differences. Data from workers
occupationally exposed to cyanide provide some information on the absorption, metabolism, and
elimination of cyanide in humans and indicate qualitatively that the toxicokinetics of cyanide are
similar between humans and animals. Additionally, some biological effects, including thyroid
enlargement and neurological symptoms observed in animals and humans (such as ataxia,
weakness, and behavioral changes), are similar in nature, indicating similar toxicodynamics.
However, the magnitude of the similarities or differences in toxicokinetic and toxicodynamic
parameters cannot be calculated due to uncertainties regarding routes of exposure and doses for
the occupationally exposed workers. Therefore, a 10-fold UF to account for interspecies
differences was used.
A 3-fold UF was applied to account for less-than-chronic exposure duration in the
occupationally exposed workers. Uncertainty exists as to whether additional or more severe
effects would be expected in these workers over longer durations.
Limited data exist on effects of cyanide in populations of occupationally exposed
workers. However, since potential variability in responses to cyanide in the greater human
population is unknown, the default UF of 10 for intrahuman variability was used. Human
variation may be larger or smaller; however, chronic cyanide-specific data to examine the
potential magnitude of human variability of response were not found.
86
-------
Uncertainties associated with data gaps in the cyanide database have been identified.
Effects on reproductive organ weight and sperm parameters have been identified in rats and mice
subchronically exposed to cyanide in the diet. However, data more fully characterizing potential
multigenerational reproductive effects are lacking. Gross developmental effects have not been
observed in the few, limited developmental studies available (Imosemi et al., 2005; Malomo et
al., 2004; Tewe and Maner, 1981). Due to thiocyanate's proposed mode of action of competitive
inhibition of iodide uptake in the thyroid, the lack of studies evaluating subtle
neurodevelopmental and behavioral outcomes adds uncertainty to this assessment. It is unclear
from the available database whether perturbation of thyroid function sufficient for the induction
of subclinical or clinical hypothyroidism would be expected to occur below the POD for reduced
epididymis weight in rats identified in NTP (1993), since thyroid hormone levels were not
measured in this study (though examination of the thyroid showed no increase in weight or
histologic lesions). Therefore, a 3-fold UF for database deficiencies was applied to the POD to
account for uncertainty regarding potential neurological effects during development and for the
lack of data on multigenerational reproductive toxicity.
The lack of specific immune-related data on cyanide represents a data gap. Subchronic
and chronic studies on cyanide and cyanide-related compounds have evaluated limited immune
endpoints, such as organ weights, histopathology (NTP, 1993; Monsanto Co., 1985a, b; Lewis et
al., 1984; Howard and Hanzal, 1955), and hematological parameters (NTP, 1993; Blanc et al.,
1985; Monsanto Co., 1985a, b; Howard and Hanzal, 1955), and are generally negative (see
Section 4.4.4). Two studies have found an elevation in percent lymphocytes in exposed workers
as compared to controls (Leeser et al., 1990; El Ghawabi et al., 1975). This finding is of unclear
significance, considering the nonspecific nature of this hematological parameter. Overall,
although the examined immune endpoints in the cyanide database appear normal, the lack of
functional immune assays precludes a confident conclusion regarding potential immune toxicity
of cyanide.
The HCN inhalation database contains limited exposure-response data. Additionally, no
chronic or subchronic HCN inhalation studies exist in animals. Several of the available studies
may not meet current standards for study design, reporting, and peer review due to technological
and experimental developments that may have occurred from the time of reference publication.
The only available studies reporting exposure data and potential health effects of inhalation of
HCN are occupational exposure studies. In consideration of this limited inhalation database, El
Ghawabi et al. (1975) was selected as the most appropriate study for the derivation of the RfC.
The RfC was derived from a LOAEL(ADJ) of 2.5 mg/m3 HCN, which was based on
thyroid enlargement and altered iodide uptake in a cohort of workers in three electroplating
facilities who had been exposed to HCN for 5-15 years (El Ghawabi et al., 1975). This study is
the only extended duration epidemiologic study in which concurrent exposure concentrations
were measured. The mean cyanide air concentrations in the breathing zone of workers at the
87
-------
three plants were 7.07-11.5 mg/m3 HCN. The lowest mean concentration recorded in the three
factories, 7.07 mg/m3 HCN, is designated as a LOAEL. Twenty male volunteers of the same age
group and socioeconomic status who had no occupational exposure to cyanide were chosen as
controls. Effects observed in exposed workers included thyroid enlargement and increased
iodide uptake in the thyroid. The effects observed are consistent with known effects of cyanide
and reported effects in other studies of cyanide-exposed workers and are also supported by
similar effects observed in animals orally exposed to cyanide. However, significant areas of
uncertainty exist in the human data relied upon for the RfC.
Some uncertainties exist in the exposure doses measured. The individual breathing zone
concentrations of HCN were measured in the three different factories over a period of 2 months.
No information was given regarding how the current cyanide environmental monitoring data
may compare to conditions over the last several years. Furthermore, the authors acknowledged
that workers were co-exposed to other chemicals during the electroplating process, including
gasoline, alkali, and acid, but did not quantitate the magnitude of these exposures. It is unclear
how these exposures would be expected to impact the effects observed in El Ghawabi et al.
(1975). However, the thyroid effects, specifically the thyroid enlargement, are supported by
human and animal data and are consistent with the mode of action of the cyanide metabolite,
SCIST (see Section 4.6). Additionally, as with most occupational exposure scenarios, the
possibility exists for exposure through the dermal route. However, Table 4-1 provides evidence
indicating that the individual breathing zone measurements of HCN closely correlated with an
internal measure of exposure (urinary SCIST).
No NOAEL was identified in the El Ghawabi et al. (1975) study. Effects in this study
were found at the lowest exposure concentrations measured; therefore, the LOAEL identified for
this study does not indicate where a threshold of effects would lie and the data provided in the
study are not sufficient for a dose-response analysis. To account for the uncertainty in the use of
a LOAEL for the POD, a factor of 10 was applied in the derivation of the RfC.
Several assumptions were made in the conversion of the LOAEL observed in El Ghawabi
et al. (1975) to an adjusted dose. The average temperatures in the factories were not included in
the study; therefore, conversions from ppm to mg/m3 exposure concentrations were based on
standard temperature and pressure. Additionally, the daily and weekly cyanide exposure
durations were not explicitly stated in the study; therefore, an 8-hour/day, 5-day/week exposure
was assumed. Other uncertainties in the exposure assessment of the workers include potential
variability in exposures among workers based on specific duties or locations in the factories;
however, this uncertainty is limited since breathing zone samples from individual workers were
averaged to determine mean factory exposure.
Although significant areas of uncertainty remain in the human epidemiologic data relied
upon for the RfC, the use of human exposure data eliminates the substantial uncertainty inherent
in the extrapolation of an animal study to humans. The study by El Ghawabi et al. (1975) was
-------
conducted by using a small population of male workers (n = 36) and cannot be expected to
capture the human variability of response to cyanide exposure. Additionally, the workers
included in the study may represent a low-sensitivity group with other more affected workers not
continuing employment (i.e., the healthy worker effect). Therefore, in the absence of cyanide-
specific data to account for the heterogeneity of human sensitivity, a factor of 10 was used to
account for uncertainty associated with human variation in the derivation of the RfC. Human
variation in response to cyanide exposure may be larger or smaller; however, chemical-specific
data to assess the potential magnitude of variability are unavailable. Of the 36 workers, average
exposure duration was about 7 years, with the minimum exposure duration being 5 years and the
maximum being 15 years.
Uncertainty exists regarding whether progression of effects would be expected with
longer exposure time. Twenty of the 36 exposed workers had thyroid enlargement rated as being
mild to moderate; however, the authors found no correlation between duration of employment at
the factory and either incidence or magnitude of enlargement. Additionally, it is not known
whether CNS symptom incidence or severity would be expected to increase with increasing
exposure duration or whether CNS effects would be detected in chronically exposed workers at
lower concentrations. It is also possible that different sensitive endpoints may be detected in
studies of longer duration. For instance, chronic inhalation exposure in workers at a metal
tempering plant indicated some deterioration in pulmonary function (Chatgtopadhyay et al.,
2000). However, no exposure monitoring from this study was available for dose-response
comparison to the El Ghawabi et al. (1975) study, identified as the principal study. Therefore, to
account for uncertainties regarding the exposure duration of the POD, a 3 fold UF was applied.
Uncertainties associated with data gaps in the cyanide database have been identified. The
database includes limited human data from epidemiologic studies of occupationally exposed
workers. No animal studies exist that employed extended duration inhalation exposure to HCN,
although inhalation studies on the related compounds ACH and (CNh exist. Studies to assess
developmental or multigenerational reproductive toxicity of HCN for the inhalational route of
exposure are not available. Thus, a 10-fold UF was applied to account for limitations in the
inhalation database.
5.4. CANCER ASSESSMENT
The only available chronic study of cyanide that analyzed a wide variety of tissues
following near lifetime exposure is an oral rat study (Howard and Hanzal, 1955); no tumors or
lesions were associated with either dose group following dietary administration of cyanide at
doses up to 10.8 mg/kg-day for 2 years. This study is limited by small sample sizes (10/group),
histopathologic assessment of only a subset of potential target organs of carcinogenicity, and
uncertainty regarding dose due to volatility. Overall, the data are inadequate for an assessment
of the human carcinogenic potential of cyanide, based on EPA's Guidelines for Carcinogenic
89
-------
Risk Assessment (U.S. EPA, 2005a). Therefore, no quantitative cancer assessment was
conducted.
Uncertainty exists as to the potential of cyanide to induce thyroid tumors. Treatment of
rodents with chemicals that cause decreased levels of circulating thyroid hormones may result in
compensatory increased TSH levels, increased cellularity and size of the thyroid gland, and
finally tumors of the thyroid (U.S. EPA, 1998b). In the case of cyanide, there is evidence of
decreases of circulating thyroid hormones and compensatory increases in TSH levels followed
by the formation of goiter in rodents and humans. However, the rat study by Howard and Hanzal
(1955) did not identify thyroid tumors following oral exposure to HCN for 2 years. Additionally,
studies examining cancer incidence in occupationally exposed cyanide workers are not available.
Several case-control studies have indicated that development of goiter is a significant risk factor
for the development of thyroid cancer (Truong et al., 2005; Ron et al., 1987). Smokers, a
subpopulation with elevated exposure to HCN, have consistently been shown to have a decreased
risk of thyroid cancer in case-control studies (Kreiger and Parkes, 2000; Galanti et al., 1996).
Additionally, two case-control studies did not find an association between thyroid cancer and
intake of food high in cyanogenic compounds (Bosetti et al., 2002; Kolonel et al., 1990).
However, a recent case-control study in a population with a high background risk of thyroid
cancer has associated high consumption of goitrogenic food and low iodine intake with increased
incidence of thyroid cancer in women (Truong et al., 2010). Therefore, the potential for cyanide
to influence the development of thyroid tumors is unclear, but merits further investigation.
90
-------
6. MAJOR CONCLUSIONS IN CHARACTERIZATION OF
HAZARD AND DOSE RESPONSE
6.1. HUMAN HAZARD POTENTIAL
Cyanide compounds are used in a number of industrial processes, including mining,
metallurgy, manufacturing, and photography, due to their ability to form stable complexes with a
range of metals. Cyanide has been employed extensively in electroplating, in which a solid
metal object is immersed in a plating bath containing a solution of another metal with which it is
to be coated in order to improve the durability, electrical resistance, and/or conductivity of the
object. Hydrogen cyanide (HCN) has also been used in gas chamber executions and in chemical
warfare. The cyanide salts, sodium cyanide (NaCN) and potassium cyanide (KCN), have also
been used as rodenticides. Use in industrial processes is the main origin of cyanide in the
environment, but cyanide is also released from biomass burning, volcanoes, and natural biogenic
processes from higher plants, bacteria, and fungi (ATSDR, 2006). Additionally, cyanogenic
compounds, which are converted to cyanide in the body, naturally occur in many plant foods,
including cassava root, almonds, millet sprouts, lima beans, soy, spinach, bamboo shoots, and
sorghum. Exposure to cyanide also occurs from smoking. Thiocyanate (SGST), the primary
metabolite of cyanide, is found in plasma or blood at approximately 0.5-4 |ig/L in nonsmokers
and approximately 6-22 |ig/L in smokers (Chandra et al., 1980).
The available data show that cyanide is rapidly and extensively absorbed via the oral,
inhalation, and dermal routes, although quantitative data on the percent or extent of absorption
are limited. At physiological pH, cyanide is distributed in the body as HCN, and thus, the
toxicokinetics for freely dissociating cyanide compounds are the same. Cyanide distributes
rapidly and fairly uniformly throughout the body following absorption. Inhaled or dermally
absorbed HCN enters the systemic circulation immediately. In contrast, ingested cyanide is
primarily converted to thiocyanate via first-pass metabolism in the liver. Immediately following
oral exposure in humans, tissues containing cyanide included the liver, brain, spleen, blood,
kidneys, and lungs (Ansell and Lewis, 1970; Gettler and Baine, 1938). Following acute
inhalation exposure in humans and animals, cyanide is found in the lung, heart, blood, kidneys,
and brain (Ballantyne, 1983, as cited in ATSDR, 2006; Gettler and Baine, 1938). The major
metabolic pathway for cyanide is conversion to thiocyanate, primarily by rhodanese.
Detoxification of cyanide by rhodanese is rapid with the concentration of sulfur-containing donor
molecules as the rate-limiting factor. Rhodanese is widely distributed throughout the body, but
is located at the highest concentration in the liver. Toxicokinetic studies in animals indicate
rapid decreases in cyanide blood concentration within 3 hours following dosing, with the half-
life of elimination for thiocyanate for all species about 10 times longer (Sousa et al., 2003).
91
-------
Cyanide is primarily excreted in the urine as thiocyanate following both inhalation and oral
exposures and is not thought to accumulate in the blood and tissues.
Several reports on occupationally exposed workers indicate that chronic inhalational
exposure to low concentrations of cyanide can cause thyroid effects and CNS symptoms
(Banerjee et al., 1997; Blanc et al., 1985; El Ghawabi et al., 1975). The results of these
occupational studies suggest that chronic exposure to cyanide may be associated with alterations
in thyroid gland function, including enlargement, altered iodine uptake, and decreased thyroid
hormones, and subjective CNS symptoms. Another study also suggests that chronic exposure to
cyanide fumes in a metal-tempering plant may reduce pulmonary function in chronically exposed
workers (Chatgtopadhyay et al., 2000). Chronic or subchronic inhalation studies of HCN in
experimental animals were not found.
Epidemiologic studies of populations in developing countries consuming cyanogenic
compounds in food have been conducted (Madhusudanan et al., 2008; Oluwole et al., 2003;
Osuntokun, 1973; Makene and Wilson, 1972). These studies are confounded by the presence of
other potentially toxic dietary components associated with cyanogenic foods, such as the
cyanogenic glycoside linamarin, and the high prevalence of iodine, protein, and vitamin
deficiencies in the studied populations. Because of the aforementioned challenges, use of
epidemiologic studies of human dietary cyanogenic exposure is limited for the purposes of this
hazard assessment for cyanide.
No epidemiologic studies exist of long-term human exposure to cyanide by the oral route.
Information on human oral exposure to cyanide is limited to acute effects following suicide
attempts or accidental poisoning. Acute oral exposure to cyanide has been observed to result in
typical signs of cyanide poisoning, including CNS depression, convulsions, coma, and death.
Chronic and subchronic oral studies in experimental animals indicate that the thyroid, CNS, and
male reproductive organs are sensitive targets of cyanide toxicity (Manzano et al., 2007; Soto-
Blanco et al., 2002a, b; NTP, 1993; Jackson, 1988).
Histologic changes in the CNS have been observed following longer-term exposure to
cyanide in some animal models. In rats exposed to cyanide in the diet for 1 year, increased
vacuolation in the spinal cord white matter and exacerbation of methionine deficiency-induced
spinal cord demyelination were observed (Philbrick et al., 1979). In addition, histopathologic
effects, including neuronal loss, spheroids, damaged Purkinje cells, and loss of white matter in
various CNS structures, were observed in rats following a 12-week oral exposure period (Soto-
Blanco et al., 2002a) and in goats following 5 months of oral exposure (Soto-Blanco et al.,
2002b). However, histopathology of the nervous system has not been identified in other chronic
or subchronic studies (NTP, 1993; Howard and Hanzal, 1955) conducted at doses lower than that
administered by Philbrick et al. (1979). In addition to histologic changes observed in some
studies, subtle behavioral changes were noted in pigs orally exposed to 1.2 mg/kg-day cyanide in
drinking water for 6 months.
92
-------
Chronic and subchronic exposure to cyanide is known to induce thyroid effects due to the
cyanide metabolite, thiocyanate. Thiocyanate adversely affects the thyroid gland via competitive
inhibition of iodide uptake and perturbation of the homeostatic feedback mechanisms that
regulate the synthesis and secretion of essential thyroid hormones. Philbrick et al. (1979)
reported decreased serum 14 levels and increased thyroid weights in rats but no histopathologic
changes in the thyroid gland. Subchronic studies in rats and mice (NTP, 1993) conducted with a
range of doses lower than the single dose tested by Philbrick et al. (1979) did not observe
adverse histopathology or increased weight of the thyroid gland, although thyroid hormone
levels were not evaluated. Studies in pigs have noted increased thyroid weights, altered thyroid
histology, and decreased thyroid hormones (Manzano et al., 2007; Jackson, 1988) at doses in the
range of the NTP (1993) study and several times lower than the Philbrick et al. (1979) study. It
is apparent through comparisons of thyroid effects in animal models that sensitivity of the
thyroid to the effects of cyanide appears to vary widely among species.
Reproductive effects, including decreased epididymis, cauda epididymis, and testis
weights and decreased sperm parameters (epididymal sperm motility and testicular spermatid
counts), have been observed in rats in a subchronic dietary study by NTP (1993). Decreases in
the cauda epididymis and epididymis weights were also seen in mice (NTP, 1993). Histologic
examination of reproductive organs did not reveal any lesions. Additionally, reproductive
effects, specifically, alterations in testicular histology, have also been observed in a 14-week
study in dogs (Kamalu, 1993). The mode of action of the reproductive effects following
subchronic cyanide exposure in rodents is unclear, although some data in hypothyroid animal
models suggest that these effects may be secondary to thyroid perturbation.
Cyanide exerts its acute effects, including CNS depression, convulsions, coma, and death,
by binding with cytochrome c oxidase, a key enzyme in the production of ATP by way of
oxidative phosphorylation. The steep dose response occurring with acute high-dose exposures is
thought to be due to cyanide overload, resulting in saturation of detoxification pathways that
metabolize cyanide to less acutely toxic intermediate compounds. At lower dose rates, an
efficient detoxification system (primarily via rhodanese with sulfur donors as the rate-limiting
factor) catalyzes the transformation of cyanide to thiocyanate, its primarily metabolite.
Thiocyanate is not known to be acutely toxic, although long-term exposures can adversely affect
the thyroid gland via iodide uptake inhibition and decreased thyroid hormone synthesis. The
chronic effects of cyanide and thiocyanate on other organ groups are not clear.
6.2. DOSE RESPONSE
6.2.1. Noncancer—Oral
The male reproductive effects observed by NTP (1993) have been identified as the
critical effects for the development of the cyanide RfD. The NTP (1993) study was well
designed with five treatment groups with doses spanning two orders of magnitude. Numerous
93
-------
tissues and endpoints were assessed in both rats and mice. This study identified a suite of
statistically significant reproductive effects in both species, including decreased epididymis
weights (cauda and whole), decreased testes weight, and altered sperm parameters. BMD
analysis of the observed reproductive data from rats and mice indicated decreased cauda
epididymis weight to be the most sensitive reproductive effect observed. Thus, decreased cauda
epididymis weight was chosen as the critical effect. This effect is believed to be one that would
likely precede substantial decrements in sperm parameters and fertility in this test species. The
cyanide database contains additional, limited support for the specific endpoint of reproductive
toxicity. Altered testicular histopathology, including a significantly decreased percentage of
tubules in stage VIII of the spermatogenic cycle, was observed in dogs ingesting 1 mg/kg-day
cyanide for 14 weeks (Kamalu, 1993).
In addition to the reproductive effects observed in NTP (1993), other sensitive effects
observed in animals, including increased thyroid weight, altered thyroid hormones, and altered
testicular, kidney, and adrenal histopathology, were also considered as potential critical effects
(see discussion in Section 5.1.1). Though these effects were not ultimately selected for the
derivation of the RfD, RfDs for these endpoints were quantified for comparison purposes.
BMD modeling was conducted to calculate potential PODs for deriving the RfD by
estimating the effective dose at a specified level of response (BMDX) and its 95% lower bound
confidence limit (BMDLX). A BMR level was selected corresponding to a change in the mean
response equal to 1 control SD from the control mean for cauda epididymis weight. In this case,
a 1 SD change in cauda epididymis weight was selected under an assumption that it represents a
minimal biologically significant change. Using the best fitting model for this data set, a 1 SD
change was equivalent to a 7% decrease in cauda epididymis weight. Additional BMD modeling
for other amenable data sets was also conducted to provide other PODs for comparison purposes
(see Appendix B). PODs for these endpoints and other PODs determined through a
NOAEL/LOAEL approach were considered for the derivation of the RfD. Tabular and graphical
representations of these potential PODs and resulting potential reference values are shown in
Table 5-2 and Figure 5-1, respectively.
The RfD of 6 x 10~4 mg/kg-day was calculated from a BMDLiso of 1.9 mg/kg-day based
on decreased cauda epididymis weight in rats in the subchronic oral study conducted by NTP
(1993). A total UF of 3,000 was applied to the POD: 10 for the extrapolation from animals to
humans (UFA), 10 for the extrapolation from a subchronic to chronic exposure duration (UFS);
10 for human intraspecies variability (UFn); and 3 to account for database deficiencies (UFo).
Information was unavailable to quantitatively assess toxicokinetic or toxicodynamic differences
between animals and humans and the potential variability in human susceptibility; thus, the
interspecies and intraspecies UFs of 10 were applied. To address the subchronic, 13-week study
duration of the principal study, a 10-fold UF (UFs) was also applied. Additionally, a 3-fold
94
-------
database UF was considered necessary due to the lack of information regarding potential
multigenerational reproductive effects and the lack of a sensitive neurodevelopmental study.
The overall confidence in the RfD is low to medium. Confidence in the principal study
(NTP, 1993) is medium. This study was well conducted, involved a sufficient number of
animals per group (including both sexes of two species), used several dose levels, and assessed a
wide range of tissues and endpoints. However, this study did not evaluate thyroid endpoints and
was only 90 days in duration. Confidence in the database is low to medium. The cyanide
database includes occupational inhalation exposure studies in humans, chronic and subchronic
dietary exposure studies in laboratory animals, and several developmental studies in laboratory
animals. However, the database is lacking a multigenerational reproductive toxicity study, a
sensitive neurodevelopmental study, and a chronic study evaluating noncancer endpoints.
Therefore, reflecting low to medium confidence in the database and medium confidence in the
principal study, the overall confidence in the RfD is low to medium.
6.2.2. Noncancer—Inhalation
No new chronic or subchronic inhalation studies for HCN (with quantitative information)
have been published in the literature since the development of the previous HCN RfC.
Therefore, the inhalation database was reevaluated, but the principal study and critical effect
selected for the RfC were unchanged. The RfC is based on an occupational study reporting
thyroid enlargement and altered iodide uptake. The principal study (El Ghawabi et al., 1975)
identified effects on the thyroid that were consistent with the proposed mode of action of
cyanide. Only a LOAEL could be identified from this study. In addition, there was potential
coexposure of the workers to other substances, including gasoline, alkali, and hydrochloric acid,
through the electroplating process. Nonetheless, the reported thyroid alterations are consistent
with the reported effects of cyanide exposure in other occupational studies and animal studies.
The RfC of 8 x 10'4 mg/m3 was derived from a LOAEL(ADJ) of 2.5 mg/m3 HCN, which
was based on thyroid enlargement and altered iodide uptake in a cohort of workers in three
electroplating facilities who had been exposed to HCN for 5-15 years (El Ghawabi et al., 1975).
The authors recorded multiple individual breathing zone samples and reported the mean HCN
exposure level for each factory. The lowest mean HCN concentration was designated as the
LOAEL. The study authors did not report the data in a manner that allowed evaluation of an
exposure duration response or a concentration response. Other studies of occupationally exposed
workers either did not provide exposure data or included higher exposure levels and did not
control for confounding variables.
A total UF of 3,000 was applied to the POD: 10 for the extrapolation of a LOAEL to a
NOAEL (UFL), 3 for the extrapolation from a subchronic study (UFS), 10 for human intraspecies
variability (UFn), and 10 to account for database deficiencies (UFo). A 1-fold UF for
extrapolation across species was applied because the RfC is based on thyroid enlargement and
95
-------
altered iodide uptake reported in an occupational study. The occupational study by El Ghawabi
et al. (1975) identified a LOAEL and a 10-fold UF was applied (UFi) to extrapolate to a
NOAEL. Information was unavailable to quantitatively assess the variability in susceptibility to
cyanide in the human population; therefore, a 10-fold UF for intraspecies variability was applied
(UFn). Additionally, a subchronic-to-chronic UF of 3 was applied to extrapolate from what is
assumed to be largely subchronic exposure in exposed workers to chronic exposure. A 10-fold
UF was applied for uncertainties in the database, specifically the lack of a multigenerational
toxicity study and a sensitive neurodevelopmental study. Several limited occupational inhalation
studies are available in the cyanide database.
Reflecting medium confidence in the principal study (El Ghawabi et al., 1975) and low to
medium confidence in the inhalation database, the overall confidence in the cyanide RfC is low
to medium.
6.2.3. Cancer
Cyanide has not been subjected to a complete standard battery of genotoxicity assays,
although, overall, the available data indicate that cyanide is not genotoxic. No adequate
carcinogenicity studies of cyanide are available in animals or humans. In a 2-year chronic study
in rats, no evidence of tumorigenicity was observed (Howard and Hanzal, 1955). However, the
number of animals per dose group limited the power of the study and only a limited set of target
tissues was evaluated histopathologically. Based on these considerations and in accordance with
the U.S. EPA (2005a) Guidelines for Carcinogen Risk Assessment, there is "inadequate
information to assess the carcinogenic potential" of cyanide.
96
-------
7. REFERENCES
Abuye, C; Kelbessa, U; Wolde-Gebriel, S. (1998) Health effects of cassava consumption in south Ethiopia. East Afr
MedJ 75(3): 166-170.
Ahmed, AE; Farooqui, MY. (1982) Comparative toxicities of aliphatic nitriles. Toxicol Lett 12(2-3): 157-163.
Aminlari, M; Vaseghi, T; Kargar, MA. (1994) The cyanide-metabolizing enzyme rhodanese in different parts of the
respiratory systems of sheep and dog. Toxicol Appl Pharmacol 124(1):67-71.
Amo, H. (1973). Effects of long term trace amount oral administration of CN and heavy metals on breeding and
genetic factors of mice. Nagoya City Medical School Journal 24(1). (translation)
Ansell, M; Lewis, FA. (1970) A review of cyanide concentrations found in human organs. A survey of literature
concerning cyanide metabolism, 'normal', non-fatal, and fatal body cyanide levels. J Forensic Med 17(4): 148-155.
ATSDR (Agency for Toxic Substances and Disease Registry). (2006) Toxicological profile for cyanide. Public
Health Service, U.S. Department of Health and Human Services, Atlanta, GA. Available online at
http://www.atsdr.cdc.gov/toxprofiles/tp8.html
Ballantyne, B. (1988) Toxicology and hazard evaluation of cyanide fumigation powders. J Toxicol Clin Toxicol
26(5-6):325-335.
Banea-Mayambu, JP; Tylleskar, T; Gitebo, N; et al. (1997) Geographical and seasonal association between
linamarin and cyanide exposure from cassava and the upper motor neurone disease konzo in former Zaire. Trop
Med Int Health 2(12):1143-1151.
Banerjee, KK; Bishayee, A; Marimuthu, P. (1997) Evaluation of cyanide exposure and its effect on thyroid function
of workers in a cable industry. J Occup Environ Med 39(3):258-260.
Barrere, X; Valeix, P; Preziosi, P et al. (2000) Determinants of thyroid volume in healthy French adults participating
in the SU.VI.MAX cohort. Clin Endocrinol 52(3):273-278.
Beck, MT. (1987) Critical survey of stability constants of cyano complexes. Pure Appl Chem 59(12): 1703-1720.
Bhattacharya, R; Kumar, P; Sachan, AS. (1994) Cyanide induced changes in dynamic pulmonary mechanics in rats.
Indian J Physiol Pharmacol 38(4):281-284.
Billaut-Laden, I; Allorge, D; Crunelle-Thibaut, A; et al. (2006) Evidence for a functional genetic polymorphism of
the human thiosulfate sulfurtransferase (Rhodanese), a cyanide and H2S detoxification enzyme. Toxicology
Blanc, P; Hogan, M; Mallin, K; et al. (1985) Cyanide intoxication among silver-reclaiming workers. JAMA
253(3):367-371.
Blount, BC; Pirkle, JL; Osterloh, JD; et al. (2006) Urinary perchlorate and thyroid hormone levels in adolescent and
adult men and women living in the United States. Environ Health Perspect 1 14(12): 1865-1871.
Bonmarin, I; Nunga, M; Perea, WA. (2002) Konzo outbreak, in the south-west of the Democratic Republic of
Congo, 1996. JTropPediatr48(4):234-238.
Bosetti, C; Negri, E; Kolonel, L; et al. (2002) A pooled analysis of case-control studies of thyroid cancer. VII.
Cruciferous and other vegetables (international). Cancer Causes Control 13(8):765-775.
Boxer, GE; Rickards, JC. ( 1 952) Studies on the metabolism of the carbon of cyanide and thiocyanate. Arch
Biochem Biophys 39(l):7-26.
97
-------
Brandt-Rauf, PW; Fallen, LF Jr; Tarantini, T; et al. (1988) Health hazards of fire fighters: exposure assessment. Br
JIndMed45(9):606-612.
Brauer, VF; Below, H; Kramer, A; et al. (2006) The role of thiocyanate in the etiology of goiter in an industrial
metropolitan area. Eur J Endocrinol 154(2):229-235.
Brierley, JB; Brown, AW; Calverley, J. (1976) Cyanide intoxication in the rat: physiological and neuropathological
aspects. J Neurol Neurosurg Psychiatr 39(2): 129-140.
Campbell, MA; Golub, HS; Iyer, P; et al. (2009) Reduced water intake: implications for rodent developmental and
reproductive toxicity studies. Birth Defects Res B Dev Reprod Toxicol 86(3): 157-175.
Carella, F; Grassi, MP; Savoiardo, M; et al. (1988) Dystonic-Parkinsonian syndrome after cyanide poisoning:
clinical and MRI findings. J Neurol Neurosurg Psychiatr 51(10): 1345-1348.
Casey, BM; Dashe, JS; Wells, CE; et al. (2005) Subclinical hypothyroidism and pregnancy outcomes. Obstet
Gynecol 105(2):239-245.
Chan, S; Kilby, MD. (2000) Thyroid hormone and central nervous system development. J Endocrinol 165(1): 1-8.
Chandra, H; Gupta, BN; Bhargava, SK; et al. (1980) Chronic cyanide exposure~A biochemical and industrial
hygiene study. J Anal Toxicol 4(4):161-165.
Chatgtopadhyay, BP; Gangopadhyay, PK; Alam, JSK. (2000) Long term effect of cyanide fumes exposure on
ventilatory pulmonary function among the workers of a metal tempering plant. Biomedicine 20(3):207-218.
Chen, KK; Rose, CL. (1952) Nitrite and thiosulfate therapy in cyanide poisoning. JAMA 149:113-119.
Cliff, J; Lundquist, P; Rosling, H; et al. (1986) Thyroid function in a cassava-eating population affected by epidemic
spastic paraparesis. Acta Endocrinol 113(4):523-528.
Cooke, PS. (1991) Thyroid hormones and testis development: a model system for increasing testis growth and sperm
production. AnnN Y Acad Sci 637:122-132.
Cotton, FA; Wilkinson, G. (1980) Advanced inorganic chemistry: a comprehensive text. 4th edition. New York,
NY: John Wiley & Sons.
Crampton, RF; Gaunt, IF; Harris, R; et al. (1979) Effects of low cobalamin diet and chronic cyanide toxicity in
baboons. Toxicology 12(3):221-234.
Crump, KS; Gibbs, JP. (2005) Benchmark calculations for perchlorate from three human cohorts. Environ Health
Perspect 113(8):1001-1008.
Dahl, AR. (1989) The cyanide-metabolizing enzyme rhodanese in rat nasal respiratory and olfactory mucosa.
Toxicol Lett 45(2-3): 199-205.
De Flora, S. (1981) Study of 106 organic and inorganic compounds in the Salmonella/microsome test.
Carcinogenesis 2(4):283-298.
De Flora, S; Camoirano, A; Zanacchi, P; et al. (1984) Mutagenicity testing with TA97 and TA102 of 30 DNA-
damaging compounds, negative with other Salmonella strains. Mutat Res 134(2-3): 159-165.
De Groef, B; Decallonne, BR; Van der Geyten, S; et al. (2006) Perchlorate versus other environmental
sodium/iodide symporter inhibitors: potential thyroid-related health effects. Eur J Endocrinol 155(1): 17-25.
De Paul, AL; Mukdsi, JH; Pellizas, CG; et al. (2008) Thyroid hormone receptor alpha 1-beta 1 expression in
epididymal epithelium from euthyroid and hypothyroid rats. Histochem Cell Biol 129(5):631-642.
98
-------
Del Rio, AG; Valdez Toledo, CL; Quiros, MC. (1979) Thyroid gland and epididymal function in rats~histological
study. Arch Androl 3(1): 19-22.
Del Rio, AG; Blanco, AM; Niepomniscze, H; et al. (1998) Thyroid gland and epididymal sperm motility in rats.
Arch Androl 41(l):23-26.
Del Rio, AG; Palaoro, LA; Blanco, AM; et al. (2001) Epididymal scanning electron microscopy study in
hypothyroid rats. Arch Androl 46(l):73-77.
Del Rio, AG; Palaoro, LA; Canessa, OE; et al. (2003) Epididymal cytology changes in hypothyroid rats. Arch
Androl 49(4):247-255.
Devlin, DJ; Mills, JW; Smith, RP. (1989a) Histochemical localization of rhodanese activity in rat liver and skeletal
muscle. Toxicol Appl Pharmacol 97(2):247-255.
Devlin, DJ; Smith, RP; Thron, CD. (1989b) Cyanide metabolism in the isolated, perfused, bloodless hindlimbs or
liver of the rat. Toxicol Appl Pharmacol 98(2):338-349.
Doherty, PA; Perm, VH; Smith, RP. (1982) Congenital malformations induced by infusion of sodium cyanide in the
golden hamster. Toxicol Appl Pharmacol 64(3):456-464.
Drawbaugh, RB; Marrs, TC. (1987) Interspecies differences in rhodanese (thiosulfate sulfurtransferase, EC 2.8.1.1)
activity in liver, kidney and plasma. Comp Biochem Physiol B 86(2):307-310.
Drinker, P. (1932) Hydrocyanic acid gas poisoning by absorption through the skin. J Ind Hyg 14:1-2.
Eckstein, M; Maniscalco, PM. (2006) Focus on smoke inhalation-the most common cause of acute cyanide
poisoning. Prehosp Disaster Med 21(2):s49-55.
El Ghawabi, SH; Gaafar, MA; El-Saharti, AA; et al. (1975) Chronic cyanide exposure: a clinical, radioisotope, and
laboratory study. Br J Ind Med 32(3):215-219.
Farooqui, MY; Ahmed, AE. (1982) Molecular interaction of acrylonitrile and potassium cyanide with rat blood.
Chem Biol Interact 38(2): 145-159.
Fechter, LD; Chen, GD; Johnson, DL. (2002) Potentiation of noise-induced hearing loss by low concentrations of
hydrogen cyanide in rats. Toxicol Sci 66(1):131-138.
Feldstein, M; Klendshoj, NC. (1954) The determination of cyanide in biologic fluids by microdiffusion analysis. J
Lab din Med 44(1): 166-170.
Ferguson, HC. (1962) Dilution of dose and acute oral toxicity. Toxicol Appl Pharmacol 4:759-762.
Fiksel, J; Cooper, C; Eschenroeder, A; et al. (1981) Exposure and risk assessment for cyanide. Monitoring and Data
Support Division, Office of Water Regulations and Standards, U.S. Environmental Protection Agency, Washington,
DC; EPA/440/4-85/008. Available from the National Technical Information Service, Springfield, VA; PB85-220572.
Frakes, RA; Sharma, RP; Willhite, CC; et al. (1986) Effect of cyanogenic glycosides and protein content in cassava
diets on hamster prenatal development. Fundam Appl Toxicol 7(2): 191-198.
Galanti, MR; Sparen, P; Karlsson, A; et al. (1995) Is residence in areas of endemic goiter a risk factor for thyroid
cancer? IntJ Cancer 61(5):615-621.
Galanti, MR; Hansson, L; Lund, E; et al. (1996) Reproductive history and cigarette smoking as risk factors for
thyroid cancer in women: a population-based case-control study. Epidemiol Biormarkers Prev 5(6):425-431.
Gettler, AO; Baine, JO. (1938) The toxicology of cyanide. Am J Med Sci 195:182-198.
99
-------
Goldstein, F; Rieders, F. (1953) Conversion of thiocyanate to cyanide by an erythrocytic enzyme. Am J Physiol
173(2):287-290.
Grandas, F; Artieda, J; Obeso, JA. (1989) Clinical and CT scan findings in a case of cyanide intoxication. Mov
Disord4(2):188-193.
Greer, MA; Stott, AK; Milne, KA. (1966) Effects of thiocyanate, perchlorate and other anions onthyroidal iodine
metabolism. Endocrinology 79(2):237-247.
Guyton, AC; Hall, JE. (2000) The thyroid metabolic hormones. In: Textbook of medical physiology, 10th edition.
Philadelphia, PA: W.B. Saunders Company; pp. 858-868.
Haddow, JE; Palomaki, GE; Allan, WC; et al. (1999) Maternal thyroid deficiency during pregnancy and subsequent
neuropsychological development of the child. N Engl J Med 341(8):549-555.
Hall, AH; Rumack, BH. (1990) Cyanide. In: Haddad, LM; Winchester, JF; eds. Clinical management of poisoning
and drug overdose. 2nd edition. Philadelphia, PA: W.B. Saunders Company; pp. 1103-1111.
Hamouli-Said, Z; Tahari, F; Hamoudi, F; et al. (2007) Comparative study of the effects of pre and post natal
administration of a thyroid drug on testicular activity in adult rat. Folia Histochem Cytobiol 45 Suppl 1: S51-S57.
Han, H; Kwon, H. (2009) Estimated dietary intake of thiocyanate from Brassicaceae family in Korean diet. J
Toxicol Environ Health A 72(21-22): 1380-1387.
Haque, MR; Bradbury, JH. (1999) Simple method for determination of thiocyanate in urine. Clin Chem
45(9):1459-1464.
Hill, RN; Erdreich, LS; Paynter, OE; et al. (1989) Thyroid follicular cell carcinogenesis. Fundam Appl Toxicol
12(4):629-697.
Himwich, WA; Saunders, JP. (1948) Enzymatic conversion of cyanide to thiocyanate. Am J Physiol 153(2):348-
354.
Howard, JW; Hanzal, RF. (1955) Chronic toxicity for rats of food treated with hydrogen cyanide. J Agric Food
Chem3(4):325-329.
Huang, J; Niknahad, H; Khan, S; et al. (1998) Hepatocyte-catalysed detoxification of cyanide by L- and D-cysteine.
Biochem Pharmacol 55(12): 1983-1990.
Hugod, C. (1981) Myocardial morphology in rabbits exposed to various gas-phase constituents of tobacco smoke--
an ultrastructural study. Atherosclerosis 40(2):181-190.
Imosemi, IO; Malomo, AO; Oladejo, OW; et al. (2005) Gross morphological studies on the effect of cyanide on the
developing cerebellum of Wistar rat (Rattus novegicus). Afr J Med Sci 34(l):59-63.
IPCS (International Programme on Chemical Safety). (1992) Cyanides. Poison information monograph. Vol. 3.
World Health Organization, Geneva, Switzerland. Available online at http://www.inchem.org/documents/pims
/chemical/pimgOOS .htm
IPCS (International Programme on Chemical Safety). (2004) Hydrogen cyanide and cyanides: human health aspects.
Concise International Chemical Assessment Document. Vol. 61. World Health Organization, Geneva, Switzerland.
Available online at http://www.inchem.org/documents/cicads/cicads/cicad61.htm
IPCS (International Programme on Chemical Safety). (2005) Acetone cyanohydrin. International chemical safety
card. Prepared by the International Programme on Chemical Safety, World Health Organization, Geneva,
Switzerland and the Commission of the European Communities (now the European Commission), Brussels,
Belgium. Available online at http://www.inchem.org/documents/icsc/icsc/eics0611.htm
100
-------
IRDC (International Research and Development Corporation). (1984) Teratology study in rats with test article
acetone cyanohydrin with cover letter dates 042586. Conducted by the International Research and Development
Corporation, Mattawan, MI for Monsanto Co., St. Louis, MO; Report IL-83-105; Submitted under TSCA Section
8D: EPA Document No. 878216401; NTIS No. OTS0510329.
Jackson, LC. (1988) Behavioral effects of chronic sublethal dietary cyanide in an animal model: implications for
humans consuming cassava (Manihot esculenta). Hum Biol 60(4):597-614.
Jahnke, GD; Choksi, NY; Moore, JA; et al. (2004) Thyroid toxicants: assessing reproductive health effects. Environ
Health Perspect 112(3):363-368.
Johnson, JD; Meisenheimer, TL; Isom, GE. (1986) Cyanide-induced neurotoxicity: role of neuronal calcium.
Toxicol Appl Pharmacol 84(3):464-469.
Joyce, KL; Porcelli, J; Cooke, PS. (1993) Neonatal goitrogen treatment increases adult testis size and sperm
production in the mouse. J Androl 14(6):448-455.
Kala, N; Ravisankar, B; Govindarajulu, P; et al. (2002) Impact of foetal-onset hypothyroidism on the epididymis of
mature rats. Int J Androl 25(3): 139-148.
Kamalu, BP. (1991) The effect of a nutritionally-balanced cassava (Manihot esculenta Crantz) diet on endocrine
function using the dog as a model. 1. Pancreas. Br J Nutr 65(3):365-372.
Kamalu, BP. (1993) Pathological changes in growing dogs fed on a balanced cassava (Manihot esculenta Crantz)
diet. BrJNutr69(3):921-934.
Kamalu, BP; Agharanya, JC. (1991) The effect of a nutritionally-balanced cassava (Manihot esculenta Crantz) diet
on endocrine function using the dog as a model. 2. Thyroid. Br J Nutr 65(3):373-379.
Kanthasamy, AG; Borowitz, JL; Isom, GE. (1991) Cyanide-induced increases in plasma catecholamines:
relationship to acute toxicity. Neurotoxicology 12(4):777-784.
Kiuchi, Y; Inagaki, M; Izumi, J; et al. (1992) Effect of local cyanide perfusion on rat striatal extracellular dopamine
and its metabolites as studied by in vivo brain microdialysis. Neurosci Lett 147(2): 193-196.
Klaassen, CD. (2001) Nonmetallic environmental toxicants. In: Hardman, JG; Limbird, LE; Oilman, AE; eds.
Goodman and Oilman's the pharmacological basis of therapeutics. 10th edition. New York, NY: McGraw-Hill; pp.
1877-1902.
Knowles, EL; Bain, JT. (1968) Medical cover required in large scale production of cyanides and hydrocyanic acid.
Chemlnd 8:232-235.
Knudsen, N; Laurberg, P; Perrild, H; et al. (2002) Risk factors for goiter and thyroid nodules. Thyroid 12(10):879-
Kobayashi, K; Kubota, H; Saegusa, J. (2007) Testicular development in growth-retarded mice. Exp Anim
56(5):393-397.
Koibuchi, N; Chin, WW. (2000) Thyroid hormone action and brain development. Trends Endocrinol Metab
11(4):123-128.
Kolonel, LN; Hankin, JH; Wilkens, LR; et al. (1990) An epidemiologic study of thyroid cancer in Hawaii. Cancer
Causes Control l(3):223-234.
Kooistra, L; Crawford, S; van Baar, AL; et al. (2006) Neonatal effects of maternal hypothyroxinemia during early
pregnancy. Pediatrics 117(1):161-167.
Krassas, GE; Pontikides, N. (2004) Male reproductive function in relation with thyroid alterations. Best Pract Res
Clin Endocrinol Metab 18(2):183-195.
101
-------
Kreiger, N; Parkes, R. (2000) Cigarette smoking and the risk of thyroid cancer. Eur J Cancer 36(15):1969-1973.
Kreutler, PA; Varbanov, V; Goodman, W; et al. (1978) Interactions of protein deficiency, cyanide, and thiocyanate
on thyroid function in neonatal and adult rats. Am J Clin Nutr 31(2):282-289.
Kumar, PN; Aruldhas, MM; Juneja, SC. (1994) Influence of hypothyroidism induced at prepuberty on epididymal
lipids and the number and motility of spermatozoa in rats. Int J Androl 17(5):262-270.
Kushi, A; Matsumoto, T; Yoshida, D. (1983) Mutagen from the gaseous phase of protein pyrolyzate. Agric Biol
Chem 47:1979-1982. (As cited in ATSDR, 2006).
Lam, KK; Lau, FL. (2000) An incident of hydrogen cyanide poisoning. Am J Emerg Med 18(2): 172-175.
Landahl, HD; Herrmann, RG. (1950) Retention of vapors and gases in the human nose and lung. Arch Ind Hyg
OccupMedl(l):36-45.
Lawrence, WW. (1947) The toxicity of sodium cyanide at slow rates of infusion. Fed Proc 6(1):349.
Lawrence, JE; Lamm, SH; Pino, S; et al. (2000) The effect of short-term low-dose perchlorate on various aspects of
thyroid function. Thyroid 10(8):659-663.
Leeser, JE; Tomenso, JA; Bryson, DD. (1990) A cross-sectional study of the health of cyanide salt production
workers. Report No. OHS/R/2, ICI Central Toxicology Laboratory, Alderley Park, Macclesfield, Cheshire, U.K.
Unpublished manuscript.
Leuschner, F; Neumann, BW. (1989) 13-Week toxicity study of potassium cyanide administered to Sprague-Dawley
rats in the drinking water. Unpublished manuscript.
Leuschner, J; Winkler, A; Leuschner, F. (1991) Toxicokinetic aspects of chronic cyanide exposure in the rat.
Toxicol Lett 57(2): 195-201.
Lewis, JL; Rhoades, CE; Gervasi, PG; et al. (1991) The cyanide-metabolizing enzyme rhodanese in human nasal
respiratory mucosa. Toxicol Appl Pharmacol 108(1): 114-120.
Lewis, TR; Anger, WK; Te Vault, RK. (1984) Toxicity evaluation of sub-chronic exposures to cyanogen in
monkeys and rats. J Environ Pathol Toxicol Oncol 5(4-5):151-163.
Liebowitz, D; Schwartz, H. (1948) Cyanide poisoning: report of a case with recovery. Am J Clin Pathol
18(12):965-970.
Lundquist, P; Rosling, H; Sorbo, B. (1985) Determination of cyanide in whole blood, erythrocytes, and plasma.
ClinChem31(4):591-595.
Madhusudanan, M; Menon, MK; Ummer, K; et al. (2008) Clinical and etiological profile of tropical ataxic
neuropathy in Kerala, South India. Eur Neurol 60(l):21-26.
Makene, WJ; Wilson, J. (1972) Biochemical studies in Tanzanian patients with ataxic tropical neuropathy. J Neurol
Neurosurg Psychiatr 35(1):31-33.
Malomo, AO; Imosemi, IO; Osuagwu, FC; et al. (2004) Histomorphometric studies on the effect of cyanide
consumption of the developing cerebellum of Wistar rat (Rattus novegicus). West Afr J Med 23(4):323-328.
Manzano, H; de Sousa, AB; Soto-Blanco, B; et al. (2007) Effects of long-term cyanide ingestionby pigs. Vet Res
Commun31(l):93-104.
Maran, RR; Aruldhas, MM. (2002) Adverse effects of neonatal hypothyroidism on Wistar rat spermatogenesis.
Endocr Res 28(3): 141-154.
102
-------
McMillan, DE; Svoboda, AC. (1982) The role of erythrocytes in cyanide detoxification. J Pharmacol Exp Ther
221(l):37-42.
McNamara, BP. (1976) Estimates of the toxicity of hydrocyanic acid vapors in man. Edgewood Arsenal, U.S.
Department of the Army, Aberdeen Proving Ground, MD; Technical Report EB-TR-76023. Available online at
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=AD A028501&Location=U2&doc=GetTRDoc.pdf.
Ministry of Health, Mozambique. (1984) Mantakassa: an epidemic of spastic paraparesis associated with chronic
cyanide intoxication in a cassava staple area of Mozambique. 1. Epidemiology and clinical and laboratory findings
in patients. Bull WHO 62(3):477-484. Available online at http://whqlibdoc.who.int/bulletin/1984/Vol62-
No3/bulletin_1984_62(3)_477-484.pdf
Monsanto Co. (1985a) Male fertility study of Sprague-Dawley rats exposed by inhalation route to acetone
cyanohydrin with cover letter dates 042586. Monsanto Environmental Health Laboratory, Monsanto Co., St. Louis,
MO; Report ML-82-144. Submitted under TSCA Section 8D; EPA Document No. 878216404; NTIS No.
OTS0510332.
Monsanto Co. (1985b) Female fertility study of Sprague-Dawley rats exposed by inhalation route to acetone
cyanohydrin. Monsanto Co., St. Louis, MO; Report ML-82-145. Submitted under TSCA Section 8D; EPA
Document No. 878216396.
Moore, SJ; Ho, IK; Hume, AS. (1991) Severe hypoxia produced by concomitant intoxication
with sublethal doses of carbon monoxide and cyanide. Toxicol Appl Pharmacol 109(3):412-20.
NLM (National Library of Medicine). (2008a) Medline Plus medical encyclopedia: radioactive iodine uptake.
Available online at http://www.nlm.nih.gov/MEDLINEPLUS/ency/article/003689.htm
NLM (National Library of Medicine). (2008b) Medline Plus medical encyclopedia: hypothyroidism. Available
online at http://www.nlm.nih.gov/MEDLINEPLUS/ency/article/003689.htm
Norris, JC; Moore, SJ; Hume, AS. (1986) Synergistic lethality induced by the combination of carbon monoxide and
cyanide. Toxicology 40(2): 121-9.
NRC (National Research Council). (1983) Risk assessment in the federal government: managing the process.
Washington, DC: National Academy Press.
NRC (National Research Council). (2005) Health implications of perchlorate ingestion. Washington, DC: National
Academy Press.
NTP (National Toxicology Program). (1993) NTP technical report on toxicity studies of sodium cyanide (CAS No.
143-33-9) administered in drinking water to F344/N rats and B6C3Fi mice. Public Health Service, U.S. Department
of Health and Human Services; NTP TR 37; NIH Publication 94-3386. Available online at
http ://ntp. niehs. nih.gov/ntp/htdocs/ST_rpts/tox03 7 .pdf
Okafor, PN; Okorowkwo, CO; Maduagwu, EN. (2002) Occupational and dietary exposures of humans to cyanide
poisoning from large-scale cassava processing and ingestion of cassava foods. Food Chem Toxicol 40(7): 1001-
1005.
Okoh, PN. (1983) Excretion of 14C-labeled cyanide in rats exposed to chronic intake of potassium cyanide. Toxicol
Appl Pharmacol 70(2):335-339.
Okoh, PN; Pitt, GA. (1982) The metabolism of cyanide and the gastrointestinal circulation of the resulting
thiocyanate under conditions of chronic cyanide intake in the rat. Can JPhysiol Pharmacol 60(3):381-386.
Okolie, NP; Osagie, AU. (1999) Liver and kidney lesions and associated enzyme changes induced in rabbits by
chronic cyanide exposure. Food Chem Toxicol 37(7):745-750.
Okolie, NP; Osagie, AU. (2000) Differential effects of chronic cyanide intoxication on heart, lung and pancreatic
tissues. Food Chem Toxicol 38(6):543-548.
103
-------
Olusi, SO; Oke, OL; Odusote, A. (1979) Effects of cyanogenic agents on reproduction and neonatal development in
rats. BiolNeonate 36(5-6) :233-243.
Oluwole, OS; Onabolu, AO; Cotgreave, IA; et al. (2003) Incidence of endemic ataxic polyneuropathy and its
relation to exposure to cyanide in a Nigerian community. J Neurol Neurosurg Psychiatr 74(10): 1417-1422.
Osuntokun, BO. (1973) Atoxic neuropathy associated with high cassava diets in West Africa. In: Nestel, B;
Maclntyre, R; eds. Chronic cassava toxicity: Proceedings of an interdisciplinary workshop London, England, 29-30
January 1973. Ottawa: International Development Research Centre; pp. 127-138.
Osuntokun, BO; Monekosso, GL; Wilson, J. (1969) Relationship of a degenerative tropical neuropathy to diet:
report of a field survey. Br Med J 1(5643):547-550.
Painter, RB; Howard, R. (1982) The Hela DNA-synthesis inhibition test as a rapid screen for mutagenic carcinogens.
MutatRes 92(l-2):427-437.
Palmer, IS; Olson, OE. (1979) Partial prevention by cyanide of selenium poisoning in rats. Biochem Biophys Res
Commun 90(4): 1379-1386.
Pearce, EN; Braverman, LE. (2009) Environmental pollutants and the thyroid. Best Pract Res Clin Endocrinol
Metab 23(6):801-813.
Pettigrew, AR; Fell, GS. (1973) Microdiffusion method for estimation of cyanide in whole blood and its application
to the study of conversion of cyanide to thiocyanate. Clin Chem 19(5):466-471.
Pettigrew, AR; Logan, RW; Willocks, J. (1977) Smoking in pregnancy-effects on birth weight and on cyanide and
thiocyanate levels in mother and baby. Br J Obstet Gynaecol 84(1):31-34.
Philbrick, DJ; Hopkins, JB; Hill, DC; et al. (1979) Effects of prolonged cyanide and thiocyanate feeding in rats. J
Toxicol Environ Health 5(4):579-592.
Pop, VJ; Brouwers, EP; Vader, HL; et al. (2003) Maternal hypothyroxinaemia during early pregnancy and
subsequent child development a 3-year follow-up study. Clin Endocrinol 59(3):282-288.
Poppe, K; Velkeniers, B. (2004) Female infertility and the thyroid. Best Pract Res Clin Endocrinol Metab
18(2):153-165.
Potter, AL. (1950) The successful treatment of two recent cases of cyanide poisoning. Br JInd Med 7(3): 125-130.
Purser, DA; Grimshaw, P; Berrill, KR. (1984) Intoxication by cyanide in fires: a study in monkeys using
polyacrylonitrile. Arch Environ Health 39(6):394^00.
Ravel, R. (1995) Thyroid function tests. In: Clinical laboratory medicine: clinical application of laboratory data. 6th
edition. Mosby, Inc.; pp. 477^78.
Ron, E; Kleinerman, RA; Boice, JD Jr; et al. (1987) A population-based case-control study of thyroid cancer. J Natl
Cancer Inst 79(1): 1-12.
Rosenberg, NL; Myers, JA; Martin, WR. (1989) Cyanide-induced Parkinsonism: clinical, MRI, and 6-fluorodopa
PET studies. Neurology 39(1): 142-144.
Sahoo, DK; Roy, A; Bhanja, S; et al. (2008) Hypothyroidism impairs antioxidant defense system and testicular
physiology during development and maturation. Gen Comp Endocrinol 156(1):63-70.
Schulz, V. (1984) Clinical pharmacokinetics of nitroprusside, cyanide, thiosulphate and thiocyanate. Clin
Pharmacokinet 9(3):239-251.
104
-------
Schulz, V; Bonn, R; Kindler, J. (1979) Kinetics of elimination of thiocyanate in 7 healthy subjects and in 8 subjects
with renal failure. Klin Wochenschr 57(5):243-247.
Schulz, V; Gross, R; Pasch, T; et al. (1982) Cyanide toxicity of sodium nitroprusside in therapeutic use with and
without sodium thiosulphate. Klin Wochenschr 60(22): 1393-1400.
Singh, BM; Coles, N; Lewis, P; et al. (1989) The metabolic effects of fatal cyanide poisoning. Postgrad Med J
65(770):923-925.
Smyth, HF, Jr; Carpenter, CP; Weil, CS; et al. (1969) Range-finding toxicity data: List VII. Am Ind Hyg Assoc J
30(5):470-476.
Soto-Blanco, B; Gorniak, SL. (2003) Milk transfer of cyanide and thiocyanate: cyanide exposure by lactation in
goats. VetRes34(2):213-220.
Soto-Blanco, B; Gorniak, SL. (2004) Prenatal toxicity of cyanide in goats~a model for teratological studies in
ruminants. Theriogenology 62(6): 1012-1026.
Soto-Blanco, B; Marioka, PC; Gorniak, SL. (2002a) Effects of long-term low-dose cyanide administration to rats.
Ecotoxicol Environ Saf 53(1):37-41.
Soto-Blanco, B; Maiorka, PC; Gorniak, SL. (2002b) Neuropathologic study of long term cyanide administration to
goats. Food Chem Toxicol 40(11): 1693-1698.
Sousa, AB; Soto-Blanco, B; Guerra, JL; et al. (2002) Does prolonged oral exposure to cyanide promote
hepatotoxicity and nephrotoxicity? Toxicology 174(2):87-95.
Sousa, AB; Manzano, H; Soto-Blanco, B; et al. (2003) Toxicokinetics of cyanide in rats, pigs and goats after oral
dosing with potassium cyanide. Arch Toxicol 77(6):330-334.
Spencer, CA. (2008) Evaluation of thyroid function in health and disease. In: Thyroid disease manager. South
Dartmouth, MA: Endocrine Education, Inc.; pp 4-8. Available online at http://www.thyroidmanager.org
Steinmaus, C; Miller, MD; Howd, R. (2007) Impact of smoking and thiocyanate on perchlorate and thyroid hormone
associations in the 2001-2002 National Health and Nutrition Examination Survey. Environ Health Perspect
115(9):1333-1338.
Sylvester, DM; Hayton, WL; Morgan, RL; et al. (1983) Effects of thiosulfate on cyanide pharmacokinetics in dogs.
Toxicol Appl Pharmacol 69(2):265-271.
Sylvester, M; Sander, C. (1990) Immunohistochemical localization of rhodanese. Histochem J 22(4): 197-200.
Taga, I; Oumbe, VA; Johns, R; et al. (2008) Youth of west-Cameroon are at high risk of developing IDD due to low
dietary iodine and high dietary thiocyanate. Afr Health Sci 8(3): 180-185.
Tawackoli, W; Chen, GD; Fechter, LD. (2001) Disruption of cochlear potentials by chemical asphyxiants cyanide
and carbon monoxide. Neurotoxicol Teratol 23(2): 157-165.
Teles, FF. (2002) Chronic poisoning by hydrogen cyanide in cassava and its prevention in Africa and Latin America.
Food Nutr Bull 23(4):407^12.
Tewe, OO; Maner, JH. (1981) Long-term and carry-over effect of dietary inorganic cyanide (KCN) in the life cycle
performance and metabolism of rats. Toxicol Appl Pharmacol 58(1): 1-7.
Tonacchera, M; Pinchera, A; Dimida, A; et al. (2004) Relative potencies and additivity of perchlorate, thiocyanate,
nitrate, and iodide on the inhibition of radioactive iodide uptake by the human sodium iodide symporter. Thyroid
14(12):1012-1019.
105
-------
Triggiani, V; Tafaro, E; Giagulli, VA; et al. (2009) Role of iodine, selenium and other micronutrients in thyroid
function and disorders. Endocr Metab Immune Disord Drug Targets 9(3):277-294.
Trokoudes, KM; Skordis, N; Picolos, MK. (2006) Infertility and thyroid disorders. Curr Opin Obstet Gynecol
18(4):446-451.
Truong, T; Orsi, L; Dubourdieu, D; et al. (2005) Role of goiter and of menstrual and reproductive factors in thyroid
cancer: a population-based case-control study in New Caledonia (South Pacific), a very high incidence area. Am J
Epidemiol 161(11): 1056-1065.
Truong, T; Baron-Dubourdieu, D; Rougier, Y; et al. (2010) Role of dietary iodine and cruciferous vegetables in
thyroid cancer: a countrywide case-control study in New Caledonia. Cancer Causes Control Apr 2. [Epub ahead of
print]
Tsuge, K; Kataoka, M; Seto, Y. (2000) Cyanide and thiocyanate levels in blood and saliva of healthy adult
volunteers. J Health Sci 46(5):343-350.
U.S. EPA (1985) Health and environmental effects profile for acetone cyanohydrin. Environmental Criteria and
Assessment Office, Office of Health and Environmental Assessment, Cincinnati, OH; EPA/600/X-85/366. Available
from the National Technical Information Service, Springfield, VA; PB88-170816.
U.S. EPA (1986a) Guidelines for the health risk assessment of chemical mixtures. Federal Register 51(185):34014-
34025. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (1986b) Guidelines for mutagenicity risk assessment. Federal Register 51(185):34006-34012. Available
online at http://www.epa.gov/iris/backgrd.html U.S. EPA (1988) Recommendations for and documentation of
biological values for use in risk assessment. Prepared by the Environmental Criteria and Assessment Office, Office
of Health and Environmental Assessment, Cincinnati, OH for the Office of Solid Waste and Emergency Response,
Washington, DC; EPA 600/6-87/008. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (1991) Guidelines for developmental toxicity risk assessment. Federal Register 56(234):63798-63826.
Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (1992) Drinking water criteria document for cyanide. Environmental Criteria and Assessment Office,
Office of Health and Environmental Assessment, Cincinnati, OH; EPA/500/ECAO-CIN-442. Available from the
National Technical Information Service, Springfield, VA; PB92-173319.
U.S. EPA (1993) Reference dose (RID): description and use in health risk assessments. Background Document 1A,
March 15, 1993. Available online at http://www.epa.gov/NCEA/iris/rfd.htm.
U.S. EPA (1994a) Interim policy for particle size and limit concentration issues in inhalation toxicity studies.
Federal Register 59(206):53799. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (1994b) Methods for derivation of inhalation reference concentrations and application of inhalation
dosimetry. Office of Research and Development, Washington, DC; EPA/600/8-90/066F. Available online at
http://www.epa.gov/iris/backgrd.html
U.S. EPA (1995) Use of the benchmark dose approach in health risk assessment. Risk Assessment Forum,
Washington, DC; EPA/630/R-94/007. Available online at
http://cfpub.epa. gov/ncea/raf/recordisplay.cfm?deid=42601
U.S. EPA (1996) Guidelines for reproductive toxicity risk assessment. Federal Register 61(212):56274-56322.
Available online at http://www.epa.gov/iris/backgrd.html.
U.S. EPA (1998a) Guidelines for neurotoxicity risk assessment. Federal Register 63(93):26926-26954. Available
online at http://www.epa.gov/iris/backgrd.html
U.S. EPA. (1998b) Assessment of thyroid follicular cell tumors. U.S. Environmental Protection Agency,
Washington, DC, EPA/630/R-97/002.
106
-------
U.S. EPA (2000a) Science policy council handbook: risk characterization. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA 100-B-00-002. Available online at
http://www.epa.gov/iris/backgrd.html
U.S. EPA (2000b) Benchmark dose technical guidance document. External review draft. Risk Assessment Forum,
Washington, DC; EPA/630/R-00/001. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (2000c) Supplementary guidance for conducting health risk assessment of chemical mixtures. Risk
Assessment Forum, Washington, DC; EPA/630/R-00/002. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (2002) A review of the reference dose and reference concentration processes. Risk Assessment Forum,
Washington, DC; EPA/630/P-02/0002F. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (2003) Occurrence estimation methodology and occurrence findings report for the six-year review of
existing national primary drinking water regulations. Office of Water, U.S. Environmental Protection Agency;
EPA/ 815/R-03/006. Available from the National Technical Information Service, Springfield, VA; PB2003-107405.
U.S. EPA (2005a) Guidelines for carcinogen risk assessment. Risk Assessment Forum, Washington, DC;
EPA/630/P-03/001F. Available online at http://www.epa.gov/iris/backgrd.html
U.S. EPA (2005b) Supplemental guidance for assessing susceptibility from early-life exposure to carcinogens. Risk
Assessment Forum, Washington, DC; EPA/630/R-03/003F. Available online at
http://www.epa.gov/iris/backgrd.html
U.S. EPA (2006a) Science policy council handbook: peer review. Third edition. Office of Science Policy, Office of
Research and Development, Washington, DC; EPA/100/B-06/002. Available online at
http://www.epa.gov/iris/backgrd.html
U.S. EPA (2006b) A framework for assessing health risk of environmental exposures to children. National Center
for Environmental Assessment, Washington, DC, EPA/600/R-05/093F. Available online at
http://cfpub.epa. gov/ncea/cfm/recordisplay.cfm?deid=158363.
Uitti, RJ; Rajput, AH; Ashenhurst, EM; et al. (1985) Cyanide-induced Parkinsonism: a clinicopathologic report.
Neurology 35(6):921-925.
Valade, P. (1952) Lesions of the central nervous system in chronic experimental poisoning by gaseous hydrocyanic
acid. Bull Acad Natl Med 136(16-17):280-285.
Van Sande, SJ; Massart, C; Beauwens, R; et al. (2003) Anion selectivity by the sodium iodide symporter.
Endocrinology 144(l):247-252.
Vestergaard, P. (2002) Smoking and thyroid disorders- a meta-analysis. Eur J Endocrinol 146(2): 153-161.
Walton, DC; Witherspoon, MG. (1926) Skin absorption of certain gases. J Pharmacol Exp Ther 26:315-324.
Way, JL. (1984) Cyanide intoxication and its mechanism of antagonism. Annu Rev Pharmacol Toxicol 24:451^81.
Westley, JL. (1981) Thiosulfate-cyanide sulfur-transferase (rhodanese). Methods Enzymol 77:285-291.
WHO (World Health Organization). (1994) Indicators for assessing iodine deficiency disorders and their control
through salt iodization. WHO/NUT/94.6. Geneva: International Council for the Control of Iodine Deficiency
Disorders.
Wing, DA; Baskin, SI. (1992) Modifiers of mercaptopyruvate sulfurtransferase catalyzed conversion of cyanide to
thiocyanate in vitro. J Biochem Toxicol 7(2):65-72.
Wistuba, J; Mittag, J; Luetjens, CM; et al. (2007) Male congenital hypothyroid Pax8-/- mice are infertile despite
adequate treatment with thyroid hormone. J Endocrinol 192(1):99-109.
107
-------
Wolff, J. (1998) Perchlorate and the thyroid gland. Pharmacol Rev 50(1):89-105.
Wolfsie, JH; Shaffer, CB. (1959) Hydrogen cyanide. Hazards, toxicology, prevention and management of poisoning.
JOccupMed l(5):281-288.
Wood, JL. (1975) Biochemistry. In: Newman, AA; ed. Chemistry and biochemistry of thiocyanic acid and its
derivatives. New York, NY: Academic Press; pp. 156-221.
Wood, JL; Cooley, SL. (1956) Detoxication of cyanide by cystine. J Biol Chem 218(l):449-457.
Working, PK. (1988) Male reproductive toxicology: comparison of the human to animal models. Environ Health
Perspect 77:37-44.
Yamamoto, K; Yamamoto, Y; Hattori, H; et al. (1982) Effects of routes of administration on the cyanide
concentration distribution in the various organs of cyanide-intoxicated rats. Tohoku J Exp Med 137(l):73-78.
108
-------
APPENDIX A. SUMMARY OF EXTERNAL PEER REVIEW AND PUBLIC
COMMENTS AND DISPOSITION
The Toxicological Review of Hydrogen Cyanide and Cyanide Salts (dated August, 2009)
has undergone a formal external peer review performed by scientists in accordance with EPA
guidance on peer review (U.S. EPA, 2006a, 2000a). An external peer-review workshop was held
December 14, 2009. The external peer reviewers were tasked with providing written answers to
general questions on the overall assessment and on chemical-specific questions in areas of
scientific controversy or uncertainty. A summary of significant comments made by the external
reviewers and EPA's responses to these comments arranged by charge question follow. In many
cases, the comments of the individual reviewers have been synthesized and paraphrased in
development of Appendix A. EPA also received scientific comments from the public. These
comments and EPA's responses are included in a separate section of this appendix.
EXTERNAL PEER REVIEW PANEL COMMENTS
The reviewers made several editorial suggestions to clarify specific portions of
the text. These changes were incorporated in the document as appropriate and are not discussed
further.
(A) General Comments
1. Is the Toxicological Review logical, clear and concise? Has EPA accurately, clearly and
objectively represented and synthesized the scientific evidence for noncancer and cancer
hazard?
Comment: The peer reviewers found the document to be generally comprehensive, logical, well-
written, and clear. However, two reviewers found some parts of the document to be redundant.
Response: The structure of Toxicological Reviews is established by a template, and, therefore,
information may be presented in more than one place. An effort has been made to present a
streamlined assessment of the chronic toxicity of hydrogen cyanide and cyanide salts.
Suggestions for document improvement submitted by the panel were incorporated into the
Toxicological Review wherever possible and appropriate.
A-l
-------
Comment: One reviewer suggested a more concise summary of the primary chronic mode of
action for cyanide of thyroid perturbation, including expanded discussion of chemicals operating
through the same mode of action.
Response: The mode of action of thyroid perturbation following cyanide exposures is discussed
in Section 4.5.3. Several other chemicals known to act via this mode of action are included in
the text.
Comment: Another reviewer suggested that cyanide be expressed as hydrogen cyanide (HCN) in
the document and not cyanide ion (GST*.
Response: Though it is acknowledged in Section 2 that HCN and CN~ can interconvert based on
pH and temperature, for the sake of comparability, doses in this review are given as CIST unless
stated otherwise.
Comment: One reviewer commented that Section 4.8.1, Possible Childhood Susceptibility,
should unequivocally state that the developing fetus and children are more susceptible to cyanide
toxicity than adults.
Response: Text will be added in this section to highlight that the developing human is expected
to be the most susceptible population to chronic cyanide toxicity.
Comment: A reviewer expressed concern with the age of the database, and commented that the
document should acknowledge that much of the data are old and may not meet current standards
for study design, reporting, and peer review.
Response: Text will be added to Section 5.3, Uncertainties in the Oral Reference Dose and
Inhalation Reference Concentration, addressing the age of the cyanide database.
2. Please identify any additional studies that should be considered in the assessment of the
noncancer and cancer health effects of hydrogen cyanide and cyanide salts.
Comment: Several reviewers identified the following unpublished studies and requested they be
discussed in the document:
Leeser, J.E., Tomenso, J.A., andBryson, D.D. 1990. A cross-sectional study of the health of
cyanide salt production workers. Report No. OHS/R/2, ICI Central Toxicology Laboratory,
Alder ley Park, Macclesfield, Cheshire, U.K. Unpublished manuscript.
A-2
-------
Leuschner F. andB.W. Neumann. 1989. 13-Week toxicity study ofpotassium cyanide
administered to Spr ague-Daw ley rats in the drinking water. Unpublished manuscript.
One reviewer suggested additional information about perchlorate and carbon monoxide, which is
a component of smoke for most non-industrial inhalation exposures (including fires and cigarette
smoke).
Response: Additional studies were considered and added where relevant. Specifically,
descriptions of the unpublished studies by Leeser et al. (1990) and Leuschner and Neumann
(1989) were added to the document in Sections 4.1.3. and 4.2.1, respectively, and were included
in Section 4.6 and discussed in Section 5. Limited information is included in the draft
assessment regarding chemicals which operate by a similar mode of action, including perchlorate
(see Sections 4.5.3., 4.8.2., 4.8.3., and 5.2.1.), because a greater consideration of cumulative
effects of mixtures is outside the scope of this assessment. As this Toxicological Review is
intended to provide scientific hazard and dose-response information pertaining to chronic
exposure to cyanide, acute data of mixed gas exposures (such as HCN and carbon monoxide)
was not prioritized. However, some information regarding HCN and carbon monoxide co-
exposure has been added to the document in Section 4.8.3.
(B) Oral reference dose (RfD) for Hydrogen Cyanide and Cyanide Salts
1. A 13-week drinking water study (NTP, 1993) was selected as the basis for the RfD.
Please comment on whether the selection of this study as the principal study is scientifically
justified. Please identify and provide the rationale for any other studies that should be
selected as the principal study. Specifically, please comment on whether Jackson (1988) or
Kamalu et al. (1993) (which found potentially lower points of departure) should be given
greater consideration in the determination of the RfD.
Comments: Four reviewers agreed that of the studies presented in the Toxicological Review, the
study by NTP (1993) was the most appropriate study for the determination of the RfD. One
reviewer suggested greater consideration of the unpublished study by Leuschner and Neumann
(1989).
Response: Descriptions of the unpublished studies by Leeser et al. (1990) and Leuschner and
Neumann (1989) were added to the assessment in Sections 4 and 5.
A-2
-------
Comment: One reviewer disagreed with the selection of the NTP study and recommended the
study by Jackson (1988) as the principal study. The reviewer preferred the selection of a
principal study that examined thyroid related endpoints, in an effort to remain consistent with the
critical effect selected in the derivation of the RfC and the chronic cyanide mechanism of thyroid
disruption.
Response: One of the predominant effects observed following chronic cyanide exposure is the
disruption of thyroid function. However, in the case of the derivation of the RfD, the mechanism
of the decreased epididymis weight and the associated reproductive changes observed in the
study selected as the basis of the RfD (NTP, 1993) is not known. Though a known mechanism
or mode of action could help support the relevance of these reproductive findings in humans, the
lack of definitive data to inform the mode of action does not imply that this effect is not relevant
to humans, and does not constitute grounds to disregard the observed effects. Thus, the study by
NTP, which observed the most sensitive effects following extended cyanide oral exposure, was
retained as the basis of the RfD.
2. Decreased absolute cauda epididymis weight in male rats was selected as the critical
effect for the RfD. Please comment on whether the selection of this critical effect is
scientifically justified. Please identify and provide the rationale for any other endpoints
that should be considered in the selection of the critical effect.
Comment: Three reviewers agreed with the selection of the critical effect. One reviewer agreed
with the caveat that this critical effect may be overly conservative and a biomarker of exposure
rather than an adverse effect as changes in sperm parameters were not noted until higher does.
Response: Human male fertility is established to be lower than that of rodent test species,
suggesting that human fertility may be more susceptible to damage from toxic agents (Working,
1998; U.S. EPA, 1996). Decreased cauda epididymis weight in rats was the most sensitive of the
suite of reproductive effects observed in the study by NTP (1993), including decreased testis,
epididymis, and cauda epididymis weight, decreased testicular sperm concentration, and
decreased sperm motility. The data from NTP (1993) suggest that cauda epididymis weight is an
effect that precedes more severe decrements in sperm parameters, such as decreased testicular
spermatid count, seen at the highest dose. Therefore, decreased cauda epididymis weight, the
most sensitive effect observed in this study, was retained as the critical effect.
Comment: One reviewer commented that an endpoint that is secondary to thyroid perturbation
does not make scientific, biological, and public health sense. The reviewer further commented
that the critical effect selected should be consistent with the available mode of action data
A-4
-------
indicating the thyroid as the likely target of toxicity following chronic exposure, not the observed
reproductive effects observed inNTP (1993).
Response: See response to comment under Charge Question B1.
3. Benchmark dose (BMD) modeling methods were applied to continuous data on absolute
cauda epididymis weight to derive the point of departure (POD) for the RfD. Please
provide comments with regard to whether BMD modeling is the best approach for
determining the POD. Has the BMD modeling been appropriately conducted? Is the
benchmark response (BMR) selected for use in deriving the POD (specifically, a decrease in
the control mean of one standard deviation) scientifically justified? Please identify and
provide the rationale for any alternative approaches (including the selection of the BMR,
model, etc.) for the determination of the POD and discuss whether such approaches are
preferred to EPA's approach.
Comment: The reviewers generally agreed with the use of BMD modeling to determine the
POD for the RfD. Additionally, the reviewers found the BMD modeling to be appropriately
conducted.
Response: No response needed.
Comment: One reviewer requested that additional clarification be provided on the basis for the
selection of the BMR of 1 SD. Another reviewer believed that the use of the lower confidence
limit of 1 SD from the control mean will provide a minimally significant effect level.
Response: Ideally, BMR levels are determined based on the degree of change in the endpoint
considered biologically significant. In the absence of this information, EPA's Draft Benchmark
Dose Technical Guidance Document (2000b) recommends that a BMR of 1 SD change in the
control mean be used as the standardized basis for comparisons with continuous data. In this
case, the BMR level of one SD change in cauda epididymis weight was selected since
information regarding what magnitude of change in this endpoint would be considered adverse
was not available. Clarification regarding the selection of the BMR of 1 SD has been provided
in Section 5.1.2.
Comment: One reviewer agreed with the use of BMD modeling, but was uneasy with the use of
a BMD which is higher than the study LOAEL.
A-5
-------
Response: There is no inherent relationship between a NOAEL or LOAEL and a BMR dose.
PODs based on the NOAEL/LOAEL approach are highly dependent on study design including
dose selection, dose spacing, and sample size. NOAELs are generally based on the lack of
statistical significance and do not represent a biological threshold or imply that lower doses are
without risk. The BMD approach mitigates some of the limitations of the NOAEL/LOAEL
approach by using information from the entire dose-response curve, and thus, is preferable in
cases where data are amenable to BMD modeling.
4. Please comment on the selection of the uncertainty factors applied to the POD for the
derivation of the RfD. For instance, are they scientifically justified? If changes to the
selected uncertainty factors are proposed, please identify and provide a rationale(s).
Comment: Three reviewers suggested that the UF of 10 for interspecies extrapolation could be
reduced to 3. One reviewer suggested that the rat was more sensitive than the mouse; therefore,
a sensitive animal model was being used and thus, this UF could be reduced. However, this
reviewer acknowledged that this difference between rodents does not inform the difference
between rats and humans. Two reviewers commented that data may be available to inform
toxicokinetic differences and similarities between rats and humans, and thus, this UF may be
reduced.
Response: The interspecies UF, which takes into consideration toxicokinetic and toxicodynamic
differences between animal test species and humans, is applied to account for the uncertainty in
extrapolating from laboratory animal data to average healthy humans. In the case of HCN,
chemical-specific data are not available which quantitatively inform the difference in
susceptibility to cyanide between rats and humans; therefore, a 10-fold UF was retained.
Comment: One reviewer suggested reducing the UF of 10 for the extrapolation from the
duration of the subchronic principal study (91 days) to a chronic duration, citing what is known
about the thyroid disruption mode of action of cyanide and the existence of a chronic toxicity
study, though limited, that did not reveal toxicity.
Response: The 91-day study by NTP (1993) falls well short of a lifetime duration. In addition,
there is a lack of data on male reproductive parameters following chronic administration of
cyanide, and the mode of action of the reproductive effects observed in NTP (1993) is unclear.
Therefore, it is unknown whether effects would be more severe or would be observed at lower
doses with a longer exposure duration. For these reasons, the UF of 10 to extrapolate from a
study with a subchronic duration was retained. Additional text has been in added to
Section 5.1.3 to clarify the rationale for this UF.
A-6
-------
Comment: Two reviewers suggested a reduction of the database UF from 3 to 1. One reviewer
commented that the data gaps highlighted in the discussion of the database UF need to be
rethought, not necessarily reduced, given the existing reproductive and developmental toxicity
data for HCN, and questioned whether a two-generation reproductive toxicity study and a
neurodevelopment study would reveal effects not already observed, as these studies are not
generally sensitive to thyroid disrupting chemicals. Another reviewer commented that the
database UF should be reduced provided the critical effect selected for the derivation of the RfD
is a sensitive toxicological endpoint observed in a sensitive species.
Response: The UF for database deficiencies is typically applied when sensitive studies such as a
developmental toxicity study and a multigenerational reproductive toxicity study are not
available, or when data suggest that an additional study would likely reveal a lower POD. The
magnitude of the database UF reflects the database for a particular chemical and is not meant to
be a reflection of the severity of the chosen critical effect. The justification for the database UF
for HCN was modified in the document to reflect the lack of a neurodevelopmental study
evaluating endpoints sensitive to thyroid disruption, and a multigenerational reproductive study.
In consideration of these data gaps, a 3-fold UF to account for uncertainty in the database was
retained. Text has been added to Section 5.1.3 to clarify the rationale for the database UF.
Comment: A reviewer expressed concern with the age of the database, and commented that the
document should acknowledge that much of the data are old and may not meet current standards
for study design, reporting, and peer review.
Response: Text was added to Section 5.3, Uncertainties in the Oral Reference Dose and
Inhalation Reference Concentration, acknowledging the age of the cyanide database.
(C) Inhalation Reference Concentration (RfC) for Hydrogen Cyanide
1. The occupational inhalation study by El Ghawabi et al. (1975) was selected as the basis
for the RfC. Please comment on whether the selection of this study as the principal study is
scientifically justified. Specifically, are the study design, methods, and findings appropriate
to support the derivation of an RfC? Also, please comment on whether the scientific
justification and rationale for selecting the El Ghawabi et al. (1975) study as the principal
study given the potential for possible co-exposure to other chemicals is adequately
described. Please identify and provide the rationale for any other studies that should be
selected as the principal study.
A-7
-------
Comment: In consideration of the studies available in the inhalation database for cyanide, the
reviewers generally agreed with the selection of the El Ghawabi et al. (1975) as the basis for the
RfC, stating that this study was either the most appropriate, was a reasonable choice, or was the
best available. However, three reviewers also suggested greater consideration of the unpublished
occupational study conducted by Leeser et al. (1990), with one reviewer also recommending the
Leuschner and Neumann (1989) study for consideration as the principal study.
Response: The Leuschner and Neumann study was not considered as the principal study for the
RfC as animals were treated orally and not via inhalation. Additional consideration of the
unpublished occupational study by Leeser et al. (1990) was added to the document in
Sections 4.1.3 and 5.2.1. This study was not considered to the most appropriate study for the
RfC as no biologically significant effects were observed. It is unclear whether a NOAEL for
thyroid effects can be established by this study as only one, relatively insensitive indicator of
thyroid function was measured (serum total T/j). Other commonly administered (and more
sensitive) tests for thyroid function, including TSH and iodide uptake, were not measured.
Serum T4 levels in cyanide exposed workers were decreased in controls, but according to the
study authors did not reach statistical significance (85.13 ± 2.51 vs. 89.04 ± 1.81 nmol/L in
controls). Additionally, serum T4 was below the clinical reference range (60-160 nmol/L) in 3
of the cyanide-exposed workers.. The authors claimed these three workers "were otherwise
normal and other thyroid functional tests showed that there was no functional problem". The
authors did not state which additional tests were conducted to confirm normal thyroid function,
and they specifically state that no radiolabel studies were conducted (which likely refers to the
commonly used radioactive iodide uptake test). Information regarding the Leeser et al. (1990)
study and associated uncertainties has been added to the Toxicological Review in Sections 4.1.3
and 5.2.1.
2. Thyroid enlargement and altered iodide uptake were selected as the critical effects for
the RfC. Please comment on whether the selection of these critical effects is scientifically
justified. Please identify and provide the rationale for any other endpoints that should be
considered in the selection of the critical effect.
Comments: The panel agreed with the selection of functional thyroid effects as the endpoint for
the derivation of the RfC.
Response: No response needed.
3. The chronic RfC has been derived utilizing the NOAEL/LOAEL approach to derive the
POD for the RfC. Please provide comments as to whether this approach is the best
A-8
-------
approach for determining the POD. Has the approach been appropriately conducted?
Please identify and provide the rationale for any alternative approaches for the
determination of the POD and discuss whether such approaches are preferred to EPA's
approach.
Comment: The reviewers agreed that, considering the available data from the El Ghawabi et al.
(1975) study, using the NOAEL/LOAEL approach to derive the RfC is appropriate.
Response: No response needed.
4. Please comment on the rationale for the selection of the UFs applied to the POD for the
derivation of the RfC. If changes to the selected UFs are proposed, please identify and
provide a rationale(s).
Comment: Two reviewers generally agreed with the magnitude and application of the UFs.
Response: No response needed.
Comment: One reviewer stated that the UFs were well justified, but noted that intraspecies
variability could exceed an order of magnitude when considering potentially susceptible
populations, such as the developing fetus of an iodine- and/or protein-deficient mother; however,
this may be compensated for if any of the other UFs are overestimated.
Response: Populations especially sensitive to thyroid disruption, such as the developing fetus,
have been identified in the Toxicological Review; however, the magnitude of the variability
between the male workers in the El Ghawabi et al. (1975) study and pregnant women in the
general population is not known. Thus, in the absence of data, a 10-fold UF was applied for this
area of uncertainty and variability in the human population. Additionally, a 10-fold UF was
applied for database deficiencies to account for the lack of developmental and multigenerational
reproductive studies.
Comment: Two reviewers recommended reducing the LOAEL-to-NOAEL UF. One reviewer
commented that the LOAEL used for the POD for the thyroid disruption was the lowest
concentration among all of the factories monitored in the El Ghawabi et al. (1975) study;
therefore, a full UF to extrapolate from a LOAEL to a NOAEL was not needed. Another
reviewer commented that thyroid effects were not observed in the occupational study by Leeser
et al. (1990) and therefore, a NOAEL could be determined for this study, effectively eliminating
the NOAEL to LOAEL UF.
A-9
-------
Response: Thyroid effects were observed across the three factories in the study by El Ghawabi
et al. (1975) at concentration ranges that overlapped. The average air concentrations in the three
factories (7.07, 8.9, and 11.5 mg/m3) were similar, and the lowest mean air concentration of
7.07 mg/m3, at which thyroid disruption was observed, was selected as the POD. A NOAEL was
not identified in this study. Thus, the default LOAEL to NOAEL UF was applied.
The unpublished study by Leeser et al. (1990) was added to the toxicological review (see
Sections 4.1.3 and 5.2.1) but was not selected as the basis of the RfC. A LOAEL for thyroid
effects was not identified for this review based on the lack of measurement of sensitive thyroid
parameters (only serum total T4 was measured). Other commonly administered (and more
sensitive) tests for thyroid function, including TSH, free 14, or iodide uptake, were not examined
as part of the study design (see response to Charge Question Cl). Therefore, this study was not
used to reduce the LOAEL to NOAEL UF applied to the POD for the principal study.
Comment: Two reviewers disagreed with the subchronic-to-chronic UF of 3, with one reviewer
commenting that the exposures for some workers were for up to 15 years and there was no
correlation between duration of exposure and severity of effect (goiter). The other reviewer
theorized that thyroid enlargement probably requires months of exposure to alter the thyroid
gland by blocking iodide uptake and the severity may not be a function of length of exposure.
Response: The scientific support for this UF was re-evaluated and EPA has retained a 3-fold UF
to account for extrapolation from a subchronic to chronic exposure duration. The workers in the
principal study were exposed to cyanide for 5-15 years. Of the 36 workers, 14 had been exposed
for 5 years, 14 for 5-10 years, 7 for 10-15 years, and 1 for >15 years. The mean and median
exposure times for the worker population were not reported. Twenty of the 36 exposed workers
had thyroid enlargement; however, the authors found no correlation between duration of
exposure and either incidence or magnitude of thyroid enlargement in the workers. In addition,
following continued administration of cyanide in rats, thyroid effects were less prominent at
11 months of exposure compared to 4 months of exposure (Philbrick et al., 1979), which
provides some indication (although limited) that increased duration of exposure may not lead to
an increase in thyroid effects. A lack of an association could be related to the low sample size
and/or the failure of the authors to consider iodine status of the workers. Therefore, it is not clear
whether greater alteration in thyroid function would be observed with a longer exposure
duration. In the absence of information indicating that the effects of HCN would not progress in
incidence or severity, a subchronic to chronic UF of 3 was applied. The text in Section 5.2.3 was
revised to better characterize the support for this rationale.
Data are unavailable that indicate that thyroid enlargement may require longer periods of
exposure to alter the thyroid gland (e.g., by blocking iodide uptake). While severity may not be
A-10
-------
a function of length of exposure, evidence is lacking to support this hypothesis. Therefore, a
subchronic to chronic UF of 3 was applied.
Comment: One of the reviewers (who disagreed with the application of a subchronic to chronic
UF) suggested that EPA consider consulting a clinical endocrinologist who studies iodine
deficient populations to inform the subchronic to chronic UF.
Response: A clinical endocrinologist was not consulted on the proposed UF for subchronic to
chronic exposure duration; however, a reviewer with expertise in thyroid physiology was present
on the panel.
(D) Carcinogenicity of Hydrogen Cyanide and Cyanide Salts
1. Under EPA's 2005 Guidelines for Carcinogen Risk Assessment
(www.epa.gov/iris/backgrd.html), the Agency concluded that data are inadequate for an
assessment of the human carcinogenic potential of cyanide. Please comment on the cancer
weight of evidence characterization. Is the cancer weight of evidence characterization
scientifically justified?
Comment: Four reviewers supported the Agency's determination that data are inadequate for an
assessment of the carcinogenic potential of cyanide. One reviewer did not comment on the
cancer characterization.
Response: No response required.
Comment: One reviewer suggested additional discussion focusing on whether other known
thyroid toxicants with comparable mechanisms would be expected to induce thyroid tumors.
Another reviewer suggested discussing EPA's guidance document on thyroid cancer mediated by
TSH.
Response: Text has been added to Section 5.4 to discuss the uncertainty surrounding potential
thyroid carcinogenesis of HCN.
PUBLIC COMMENTS
A-ll
-------
Comment: EPA should utilize all available studies of cyanide regardless of exposure route and
form of the compound, including HCN gas, cyanide salt, cassava, cyanogenic glycosides, and
acetone cyanohydrin (ACH).
Response: The unpublished inhalation studies of ACH exposure in rats (Monsanto Co., 1985a,
b) are discussed in the Toxicological Review in Sections 4.2.2 and 4.3.2, but were not selected
for derivation of the RfC due to the absence of any observed effects in these studies and the
availability of inhalation studies in humans demonstrating sensitive effects on the thyroid.
Studies of human exposure to foods containing high amounts of cyanogenic glycosides,
primarily from cassava, provide important hazard identification and susceptibility information
for cyanide. These studies were included and considered in the document, but were ultimately
determined not to be the most appropriate basis for the RfD due to concerns, including limited
exposure documentation, confounding dietary deficiencies of the studied African populations
(particularly low dietary intake of protein, vitamin 812, and/or iodine; and overall malnutrition),
and confounding from other chemical constituents of cassava such as linamarin, which one study
found was associated with the endemic neurotoxicity (Konzo) observed in cassava-dependent,
protein-deficient populations (Banea-Mayambu et al., 1997).
The toxicological effects observed following exposure to cyanide were considered based
on the route of exposure (oral and inhalation) because a PBPK model, which would allow for
extrapolation across routes of exposure, is not available for cyanide.
Comment: Protecting against acute toxicity of cyanides will protect against chronic effects, as
acute toxicity results when the capacity for detoxification of cyanide is exceeded, whereas
repeated-dose toxicity occurs within the detoxification capacity and can be tolerated over a
longer period.
Response: The effects associated with cyanide vary depending on exposure. For example,
populations that have slow uptake of cyanide through diet or smoking are not reported to have
acute symptoms of cyanide, but demonstrate symptoms of chronic thyroid disruption through the
cyanide metabolite, thiocyanate (SOT). The mechanisms of cyanide toxicity following acute
and chronic exposures operate via different pathways. In evaluating the acute toxicity of
cyanide, both the total amount of cyanide administered and the rate of cyanide absorption are
important (U.S. EPA, 1992). Acute toxicity results from exceeding the body's capacity for the
detoxification of cyanide (to the metabolite, thiocyanate). Acute doses of cyanide inhibit the
enzyme cytochrome C oxidase of the mitochondrial electron chain, thus preventing the formation
of ATP, which can lead to respiratory arrest/cardiac arrest (see Section 4.1.1). Thus, the relevant
dose metric for the occurrence of acute cyanide toxicity is the peak dose of HCN in the blood.
Cyanide, operating within the reductive capacity as thiocyanate, reduces iodide uptake and
A-12
-------
affects thyroid hormone production and secretion following chronic exposure (see Section 4.5.3).
Thus, the relevant dose metric for chronic toxicity is based on the average amount of thiocyanate
in circulation, which has a much longer half-life in the body compared to cyanide. Therefore,
the toxicity associated with acute exposures is different from chronic exposures and protection
against the acute effects of cyanide does not protect against the chronic effects of cyanide
exposure.
Comment: The reference values proposed are 3-4 orders of magnitude below background
exposure levels and significantly below a well documented point of departure for deriving safe
cyanide exposures in humans. Background levels of cyanide for the general public should be
presented for perspective.
Response: Estimates of aggregate background cyanide exposure are not available. However,
information regarding exposure concentrations of cyanide in surface water, non-urban air, and
cigarette smoke is provided in Section 2. Estimates of the cyanide intake in the American diet
were not located in the available literature, though a study which estimates dietary cyanide intake
is available for a Korean population. Several studies exist which report cyanide concentrations
in food, though these studies do not report daily food intake which would be necessary to
calculate an average daily exposure of cyanide through food.
"Safe" chronic cyanide exposure levels have not been established in the scientific
literature. Population-based studies examining exposure of cyanide at low levels and potential
health outcomes have historically been limited to African populations with relatively high intake
of cyanide containing plants (primarily cassava). Human populations with high cyanide
exposure due to diet and/or tobacco smoking have been identified as having an elevated risk of
thyroid disruption (Vestergaard, 2002; Makene and Wilson, 1972; Osuntokun et al., 1969).
Comment: The testicular and epididymis effects reported in the 1993 NTP study were not
appropriate to use in deriving the RfD, as they are more likely the result of decreased water
consumption or stress. The commentors also stated that, in the absence of histopathology or a
proposed mode of action, the epididymal weight effects are inappropriate for use in risk
assessment.
Response: Data to support the hypothesis that decreased reproductive organ weights and sperm
parameters can be caused by slight to moderate decreases in drinking water consumption were
not found in the peer reviewed literature. At the LOAEL for decreased cauda epididymis weight,
a 7%, non-statistically significant decrease in water consumption was observed. In addition,
increases in hematocrit, blood urea nitrogen, or serum albumin, which have been identified as
objective indicators of dehydration (Campbell et al., 2009), were not observed at the LOAEL
A-13
-------
identified for decreased epididymis weight in the NTP (1993) study. EPA considered the cauda
epididymis weight changes to be biologically significant. Additionally, U.S. EPA guidelines
have identified statistically significant reproductive organ weight changes as endpoints useful for
reproductive risk assessment (U.S. EPA, 1996). The lack of a known mechanism or histological
lesion does not imply that the decrease in cauda epididymis weight is not relevant to humans and
does not constitute grounds to disregard the observed effects.
Comment: EPA should not rely on the El Ghawabi et al. (1975) study to establish the RfC given
its limitations, including limited exposure data and inconsistencies regarding the urinary
thiocyanate excretion in this study compared to a population in France. In addition, reported
symptoms in workers with exposure to cyanide, included headache, weakness, vomiting,
dyspnea, and precordial pain. It is suggested that EPA give greater consideration to the
subchronic ACH inhalation study (Monsanto Co., 1985a, b).
Response: The external peer review panel agreed with the selection of the El Ghawabi et al.
(1975) study for the derivation of the RfC. Strengths and limitations of the El Ghawabi et al.
(1975) study are discussed in Sections 4.1.3, 5.2.1, and 5.3 of the Toxicological Review. In
consideration of the limited inhalation database, the El Ghawabi et al. (1975) study was selected
as the principal study, with thyroid enlargement and altered iodine uptake selected as the critical
effects. This study of cyanide-exposed workers and control workers included individual
breathing zone measurements from the study participants and reported a strong correlation
between these cyanide measurements and urinary metabolite levels. Excretion of SOT in
workers in the El Ghawabi et al. (1975) study was similar to average urinary excretion of SCIST
measured in a French population not occupationally exposed to cyanide (Barrere et al., 2000).
This similarity could be due to the very low background SGST exposure of workers in the El
Ghawabi et al. (1975) study (only nonsmokers were included in the study and subjects were
asked to refrain from consumption of food high in cyanogenic glycosides), different methods of
detection of urinary SGST between studies, and the high background excretion of SCW
measured in the Barrere et al. (2000) study compared to studies in other populations (Steinmaus
et al., 2007; Brauer et al., 2006; Haque and Bradbury, 1999).
The unpublished inhalation studies of ACH exposure in rats (Monsanto Co., 1985a, b)
were discussed in the Toxicological Review in Sections 4.2.2 and 4.3.2, but were not selected for
derivation of the RfC due to the absence of any observed effects and the availability of inhalation
studies in humans demonstrating sensitive effects on the thyroid.
Comment: The review of cyanide conducted by European Centre for Ecotoxicology and
Toxicology of Chemicals (ECETOC) should be considered by EPA. This review established an
A-14
-------
acceptable chronic level of 15 ug SCN7mL in serum, which the authors equated to a daily
occupational inhalation exposure of 7.5 mg GST/m3.
Response: The ECETOC review for safe chronic exposure to cyanide was based on serum SGST
concentrations from two studies and a subsequent estimation of an occupational inhalation
concentration based on the serum SOT concentrations and the utilization of several assumptions
(regarding absorption, metabolism, and excretion rates). The ECETOC NOAEL is based on a
serum SGST level of 15 jig SCN7mL, which was based on the studies by Banerjee (1997) and
Cliff et al. (1986), which detected serum levels of 18 and 14.5 jig SCN7mL, respectively.
However, at these SCIST blood levels, statistically significant increased levels of TSH were
observed along with decreases in serum T4 levels, indicative of thyroid function perturbation.
The observation of thyroid function perturbations at 18 and 14.5 jig SGST/mL is considered by
EPA to be a LOAEL. Furthermore, the back-calculated occupational exposure value of 7.5 mg
CN7m3 is higher than the LOAEL identified by EPA in this document for thyroid effects in the
El Ghawabi et al. (1975) occupational study. Therefore, EPA maintains the recommendation of
an RfC based on data reported by El Ghawabi et al. (1975).
Comment: EPA's proposed derivation of the RfD and RfC applied large UFs, whereas the
application of ECETOC's approach provides a more reliable POD and requires a lower
composite factor to account for uncertainty: intraspecies UF of 10 and a factor of 3 to account for
sensitive subpopulations (developmental effects).
Response: The rationale for the selection of UFs for the RfC is detailed in Section 5.2.3 and
incorporates uncertainties associated with intraspecies extrapolation, extrapolation from a
LOAEL to a NOAEL, extrapolation from subchronic to chronic exposure, and database
deficiencies. The application of the UFs in the Toxicological Review of Hydrogen Cyanide and
Cyanide Salts is in concordance with the guidance outlined in the RfC Methodology (U.S. EPA,
1994b) and the Review of the Reference Dose and Reference Concentration Processes (U.S.
EPA, 2002). See previous comment and response with regard to ECETOC's approach to
deriving a chronic exposure value.
Comment: There is sufficient information to enable EPA to conclude that cyanide does not
present a carcinogenic risk based on the animal and human studies and negative data for
genotoxicity.
Response: EPA concluded that there was "inadequate information to assess the carcinogenic
potential" of cyanide. The descriptor of "not likely to be carcinogenic to humans" is appropriate
when there are robust data indicating that there is no basis for human hazard concern. As noted
A-15
-------
in U.S. EPA's 2005 Guidelines for Carcinogen Risk Assessment., this descriptor may be applied
when well-designed and well-conducted animal studies, in both sexes of two species, indicate the
lack of carcinogenic potential, and is applied in the absence of other animal or human data
indicating potential for cancer effects. In the case of cyanide, only one chronic animal study in
10 male and female rats is available (Howard and Hanzal, 1955). This study did not identify
exposure-related tumors.
Studies examining cancer incidence in occupationally exposed cyanide workers are not
available in the literature. Studies of cancer in populations exposed to thiocyanate via the diet
are limited to examinations of thyroid cancer and results are generally not positive (Bosetti et al.,
2002; Kolonel et al., 1990), although one recent case-control study has associated high
consumption of goitrogenic food and low iodine intake with increased incidence of thyroid
cancer in women (Truong et al., 2010). The database consisting of one animal study in a single
species does not provide robust and convincing evidence that cyanide is "not likely to be
carcinogenic to humans". Text for clarification of the available evidence for carcinogenicity was
added to the weight-of-evidence narrative for carcinogenicity in Section 4.7.
A-16
-------
APPENDIX B. BENCHMARK DOSE MODELING RESULTS
Decreased Absolute Cauda Epididymis Weight in Rats Exposed to Sodium Cyanide (NaCN) in
Drinking Water for 13 Weeks (NTP, 1993)
All models for continuous variables available in the EPA BMDS version 1.4.1c, except
the Hill model, were fit to the data in the Table B-l. The Hill model was not fit to these data
because fitting of the Hill model requires the estimation of four parameters (i.e., intercept, v, n,
and k), which necessitates having a minimum of five dose groups in order to have adequate
degrees of freedom for testing model fit. The NTP (1993) study has only four dose groups, and
thus, the Hill model could not be fit to these data. All models fit were constant variance models.
All models tested provided adequate fit to the data, based on the summary results reported by the
BMDS output and visual examination of the graphs. A summary of the goodness-of-fit statistics
for the tested models and resulting BMD and BMDL is presented in Table B-2.
Table B-l. Decreased cauda epididymis weight in F344 rats following
administration of NaCN in drinking water for 13 weeks
Male rats
Dose (mg/kg-d CN")
Weight (g) ± SD
Cauda epididymis, absolute
0 ppm
0
0.162 ±0.009
30 ppm
1.4
0.150 ±0.013
100 ppm
4.5
0.148 ±0.013
300 ppm
12.5
0.141 ±0.009
Table B-2. BMD modeling results for decreased cauda epididymis weight in
rats
Study
NTP (1993);
male rats
Endpoint
Cauda epididymis
weight (absolute)
Model
Linear (1° polynomial)
Polynomial (2°)
Power
AIC
-312.75
-313.10
-312.75
Goodness-of-fit
/7-value
0.08
0.11
0.08
BMD
8.4
3.5
8.4
BMDL
5.6
1.9
5.6
All models adequately described the data as demonstrated by goodness-of-fit ^-values
> 0.1. Of the available models, the polynomial (2°) was selected, based on having the lowest
AIC value. Visual inspection reveals that the model describes the data well.
B-l
-------
Polynomial Model with 0.95 Confidence Level
CD
CO
CO
CD
ro
CD
0.17
0.165
0.16
0.155
0.15
0.145
0.14
0.135
R
<
-
alynomial
\x T
<
BME
\
^\^
)L
~^ <
Biyp
H
> |
-- i
0 2 4 6 8 10 12
dose
16:0002/132008
Figure B-l. Observed and predicted decrease in cauda epididymis weight in
F344 rats following administration of NaCN in drinking water for 13 weeks.
The computer output for the polynomial model of decreased (absolute) cauda epididymis weight
follows:
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\BMDSl-4-lC\UNSAVEDl.(d)
Gnuplot Plotting File: C:\BMDSl-4-lC\UNSAVEDl.plt
Wed Feb 13 16:00:34 2008
HMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = MEAN
Independent variable = COLUMN1
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
B-2
-------
Default Initial Parameter Values
alpha =
rho =
beta 0 =
0.000125
0
0.159311
Specified
beta 1 = -0.00377087
beta 2 =
0.0001857
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho have been estimated at a boundary point,
or have been specified by the user, do not appear in the correlation matrix
alpha
beta 0
beta 1
beta 2
alpha
beta 0
beta 1
beta 2
1
-8.4e-010
-3.3e-011
4.3e-010
-8.4e-010
1
-0.72
0.61
-3.3e-011
-0.72
1
-0.97
4.3e-010
0.61
-0.97
1
Confidence Interval
Variable
Upper Conf. Limit
alpha
0.00017266
beta_0
0.165233
beta_l
-0.000635079
beta_2
0.000419578
Estimate
0.000120048
0.159311
-0.00377087
0.0001857
Parameter Estimates
Std. Err.
2.68434e-005
0.00302151
0.00159992
0.000119328
95.0% Wald
Lower Conf. Limit
6.74354e-005
0.153389
-0.00690665
-4.81776e-005
Table of Data and Estimated Values of Interest
Dose
Res .
-
1
4
12
0
.4
.5
.5
N
10
10
10
10
Obs Mean
0
0
0
.162
0.15
.148
.141
Est Mean
0.
0.
0.
0.
159
154
146
141
Obs
0.
0.
0.
0.
Std Dev
009
013
013
009
Est
0.
0.
0.
0.
Std Dev
Oil
Oil
Oil
Oil
Scaled
0.776
-1.27
0.548
-0.0551
Model Descriptions for likelihoods calculated
Model Al: Yij
Var{e (ij ) }
Mu(i) + e(ij;
SigmaA2
Model A2:
Yij = Mu(i)
B-3
-------
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model Log(likelihood)
Al 161.851147
A2 163.173943
A3 161.851147
fitted 160.552452
R 153.631041
# Param' s
5
8
5
4
2
AIC
-313.702293
-310.347886
-313.702293
-313.104903
-303.262082
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log(Likelihood Ratio) Test df
19.0858
2.64559
2.64559
2.59739
6
3
3
1
p-value
0.004021
0.4496
0.4496
0.107
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1.
model appears to be appropriate here
A homogeneous variance
The p-value for Test 3 is greater than .1.
to be appropriate here
The p-value for Test 4 is greater than .1.
to adequately describe the data
The modeled variance appears
The model chosen seems
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
B-4
-------
Confidence level = 0.95
BMD = 3.51354
BMDL = 1.89704
B-5
-------
Decreased Absolute Epididymis Weight in Rats Exposed to NaCN in Drinking Water for
13 Weeks (NTP, 1993)
All models for continuous variables available in the EPA BMDS version 1.4.1c, except
the Hill model, were fit to the data in the Table B-3. The Hill model was not fit to these data
because fitting of the Hill model requires the estimation of four parameters (i.e., intercept, v, n,
and k), which necessitates having a minimum of five dose groups in order to have adequate
degrees of freedom for testing model fit. The NTP (1993) study has only four dose groups, and
thus the Hill model could not be fit to these data. All models fit were constant variance models.
All models tested provided adequate fit to the data, based on the summary results reported by the
BMDS output and visual examination of the graphs. A summary of the goodness-of-fit statistics
for the tested models and resulting BMD and BMDL is presented in Table B-4.
Table B-3. Decreased epididymis weight in F344 rats following
administration of NaCN in drinking water for 13 weeks
Male rats
Dose (mg/kg-d CN")
Weight (g) ± SD
Epididymis, absolute
0 ppm
0
0.448 ±0.019
30 ppm
1.4
0.437 ±0.016
100 ppm
4.5
0.425 ± 022a
300 ppm
12.5
0.417 ±0.016b
"Not reported as significant in NTP (1993) but significant by Dunnett's test in independent analyses conducted for
this assessment, p < 0.05.
bSignificant by Shirley's test, p < 0.01.
Source: NTP (1993).
Table B-4. BMD modeling results for decreased epididymis weight in rats
Study
NTP (1993);
male rats
Endpoint
Epididymis weight
(absolute)
Model
Linear (1° polynomial)
Polynomial (2°)
Power
AIC
-274.73
-275.64
-274.73
Goodness-of-
fit /7-value
0.22
0.73
0.22
BMD
8.2
3.2
8.2
BMDL
5.6
1.8
5.6
All models adequately described the data as demonstrated by goodness-of-fit ^-values
> 0.1. Of the available models, the polynomial (2°) was selected, based on having the lowest
AIC value. Visual inspection reveals that the model describes the data well.
B-6
-------
Polynomial Model with 0.95 Confidence Level
CD
CO
c
o
Q.
CO
CD
or
05
CD
0.46
0.45
0.44
0.43
0.42
0.41
0.4
Polynomial
BMDL
BMD
6
dose
10
12
15:4002/132008
Figure B-2. Observed and predicted decrease in epididymis weight in F344
rats following administration of NaCN in drinking water for 13 weeks.
The computer output for the polynomial model of decreased (absolute) epididymis weight
follows:
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\BMDSl-4-lC\EPIDIDYMIS_WEIGHT_ABSOLUTE.(d)
Gnuplot Plotting File: C:\BMDSl-4-lC\EPIDIDYMIS_WEIGHT_ABSOLUTE.plt
Wed Feb 13 15:40:43 2008
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta_0 + beta_l*dose + beta_2*doseA2 + ...
Dependent variable = MEAN
Independent variable = COLUMN2
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
B-7
-------
alpha =
rho =
beta_0 =
beta_l =
beta 2 =
0.00033925
0
0.447054
-0.00654
0.000331286
Specified
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho have been estimated at a boundary point,
or have been specified by the user, and do not appear in the correlation
matrix )
alpha
beta 0
beta 1
beta 2
alpha
1
le-010
-3.9e-011
-l.Se-011
beta 0
le-010
1
-0.72
0.61
beta 1
-3.9e-011
-0.72
1
-0.97
beta
-1.8e-0
0.
-0.
2
11
61
97
1
Confidence Interval
Variable
Upper Conf. Limit
alpha
0.000440482
beta_0
0.456513
beta_l
-0.00153141
beta_2
0.000704843
Estimate
0.00030626
0.447054
-0.00654
0.000331286
Parameter Estimates
Std. Err.
6.84818e-005
0.00482605
0.00255545
0.000190594
95.0% Wald
Lower Conf. Limit
0.000172038
0.437595
-0.0115486
-4.22712e-005
Table of Data and Estimated Values of Interest
Dose
Res .
-
1
4
12
0
.4
.5
.5
N
10
10
10
10
Obs Mean
0
0
0
0
.448
.437
.425
.417
Est Mean
0
0
0
0
.447
.439
.424
.417
Obs
0.
0.
0.
0.
Std Dev
019
016
022
016
Est
0.
0.
0.
0.
Std Dev
0175
0175
0175
0175
Scaled
0.171
-0.28
0.121
-0.0121
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
B-S
-------
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that were specified by the
user
Model R: Yi = Mu + e(i)
Var{e (i) } = SigmaA2
Likelihoods of Interest
Model Log(likelihood)
Al 141.882675
A2 142.610833
A3 141.882675
fitted 141.821537
R 134.393157
# Param' s
5
8
5
4
2
AIC
-273.765351
-269.221665
-273.765351
-275.643073
-264.786314
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels? (A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.)
Test
Test 1
Test 2
Test 3
Test 4
Tests of Interest
-2*log(Likelihood Ratio) Test df
16.4354
1.45631
1.45631
0.122278
6
3
3
1
p-value
0.0116
0.6924
0.6924
0.7266
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1.
model appears to be appropriate here
A homogeneous variance
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
B-9
-------
BMD = 3.19201
BMDL = 1.79176
B-10
-------
Decreased Absolute Testis Weight in Rats Exposed to NaCN in Drinking Water for 13 Weeks
(NIP, 1993)
Table B-5. Decreased testis weight in F344 rats following administration of
NaCN in drinking water for 13 weeks
Male rats
Dose (mg/kg-d CN")
Weight (g) ± SD
Testis, absolute
0 ppm
0
1.58 ±0.094
30 ppm
1.4
1.56 ±0.063
100 ppm
4.5
1.52 ±0.063
300 ppm
12.5
1.46 ±0.063
Table B-6. BMD modeling results for decreased testis weight in rats
Study
NTP (1993);
male rats
Endpoint
Testis weight
(absolute)
Model
Linear (1° polynomial)
Polynomial (2°)
Power
AIC
-167.94
-166.32
-167.94
Goodness-of-fit
/7-value
0.82
0.98
0.82
BMD
7.4
5.3
7.4
BMDL
5.1
2.4
5.1
Linear Model with 0.95 Confidence Level
1.65
1.6
CD
CO
a 1.55
CO
CD
or
C
8 1.5
1.45
1.4
Li
-
-
:
-p :
^^^^-^"
:
~-^__^^
BMDL
^^^^^-^ ;
BIVP :
0 2 4 6 8 10 12
dose
15:0802/132008
Figure B-3. Observed and predicted decrease in testis weight in F344 rats
following administration of NaCN in drinking water for 13 weeks.
B-ll
-------
The computer output from the linear model of decreased (absolute) testis weight follows:
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\BMDSl-4-lC\TESTIS_WEIGHT_ABSOLUTE.(d)
Gnuplot Plotting File: C:\BMDSl-4-lC\TESTIS_WEIGHT_ABSOLUTE.plt
Wed Feb 13 15:08:00 2008
BMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = MEAN
Independent variable = COLUMN2
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 0.005233
rho = 0 Specified
beta_0 = 1.57305
beta 1 = -0.00935835
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho have been estimated at a boundary point,
or have been specified by the user, and do not appear in the correlation
matrix )
alpha beta 0 beta 1
alpha 1 -4.1e-010 3.3e-011
beta_0 -4.1e-010 1 -0.69
beta 1 3.3e-011 -0.69 1
Parameter Estimates
95.0% Wald
Confidence Interval
B-12
-------
Variable
Upper Conf. Limit
alpha
0.00683971
beta_0
1.60252
beta_l
-0.00494569
Estimate
0.00475554
1.57305
-0.00935835
Std. Err.
0.00106337
0.0150381
0.0022514
Lower Conf. Limit
0.00267137
1.54357
-0.013771
Table of Data and Estimated Values of Interest
Do
Res
-
1
4
12
se
0
.4
.5
.5
N
10
10
10
10
Obs
1.
1.
1.
1.
Mean
58
56
52
46
Est
1.
1.
1.
1.
Mean
57
56
53
46
Obs
0
0
0
0
Std Dev
.095
.063
.063
.063
Est
0.
0.
0.
0.
Std Dev
069
069
069
069
Scaled
0.319
0.00244
-0.501
0.18
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
87.162621
88.584611
87.162621
86.968887
79.788144
# Param's
5
8
5
3
2
AIC
-164.325243
-161.169222
-164.325243
-167.937775
-155.576289
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
B-13
-------
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1
Test 2
Test 3
Test 4
17.5929
2.84398
2.84398
0.387468
6
3
3
2
0.007334
0.4163
0.4163
0.8239
The p-value for Test 1 is less than .05. There appears to be a
difference between response and/or variances among the dose levels
It seems appropriate to model the data
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 7.36887
BMDL = 5.12669
B-14
-------
Decreased Testicular Spermatid Concentration in Rats Exposed to NaCN in Drinking Water for
13 Weeks (NTP, 1993)
Table B-7. Decreased testicular spermatid concentration in F344 rats
following administration of NaCN in drinking water for 13 weeks
Male rats
Dose (mg/kg-d CN")
Mean/10"4 mL suspension ± SD
Spermatid count
0 ppm
0
89.28 ±9.64
30 ppm
1.4
84.68 ±12.74
100 ppm
4.5
82.90 ±9.99
300 ppm
12.5
77.10 ±6.96
Table B-8. BMD modeling results for decreased testicular spermatid
concentration in rats
Study
NTP
(1993);
male rats
Endpoint
Spermatid
concentration (testis)
Model
Linear (1° polynomial)
Polynomial (2°)
AIC
227.04
228.73
Goodness-
of-fit/7-
value
0.70
0.53
BMD
11.2
8.5
BMDL
6.9
2.9
Linear Model with 0.95 Confidence Level
95
90
CD
CO
§85
or
CO
03 on
^ oO
75
70
Li
<
-
near
>
<
-
-
^^^
:
-_
^__^^
^^—^_ :
:
_ . . , , , BMDL , , BMD, :
0 2 4 6 8 10 12
dose
16:40 02/21 2008
Figure B-4. Observed and predicted decrease in testicular spermatid
concentration in F344 rats following administration of NaCN in drinking
water for 13 weeks.
B-15
-------
The computer output from the polynomial model of testicular spermatid concentration follows:
Polynomial Model. (Version: 2.12; Date: 02/20/2007)
Input Data File: C:\BMDSl-4-lC\UNSAVEDl.(d)
Gnuplot Plotting File: C:\BMDSl-4-lC\UNSAVEDl.plt
Thu Feb 21 16:40:27 2008
HMDS MODEL RUN
The form of the response function is:
Y[dose] = beta 0 + beta l*dose + beta 2*doseA2 + ...
Dependent variable = MEAN
Independent variable = COLUMN1
rho is set to 0
Signs of the polynomial coefficients are not restricted
A constant variance model is fit
Total number of dose groups = 4
Total number of records with missing values = 0
Maximum number of iterations = 250
Relative Function Convergence has been set to: le-008
Parameter Convergence has been set to: le-008
Default Initial Parameter Values
alpha = 100.87
rho = 0 Specified
beta_0 = 87.4548
beta 1 = -0.861906
Asymptotic Correlation Matrix of Parameter Estimates
( *** The model parameter(s) -rho
have been estimated at a boundary point, or have been
specified by the user,
and do not appear in the correlation matrix )
alpha
alpha 1
beta_0 l.le-010
beta 1 -2.5e-013
beta_0 beta_l
l.le-010 -2.5e-013
1 -0.69
-0.69 1
Confidence Interval
Parameter Estimates
95.0% Wald
B-16
-------
Variable
Upper Conf. Limit
alpha
132.879
beta_0
91.563
beta_l
-0.246857
Estimate
92.3886
87.4548
-0.861906
Std. Err.
20.6587
2.09605
0.313806
Lower Conf. Limit
51.8982
83.3466
-1.47695
Table of Data and Estimated Values of Interest
Dose
Res .
0
1.4
4.5
12.5
N
10
10
10
10
Obs
89
84
82
77
Mean
.3
.7
.9
.1
Est
87
86
83
76
Mean
.5
.2
.6
.7
Obs S
9.
12
9.
6.
td Dev
64
.7
99
96
Est
9
9
9
9
Std Dev
.61
.61
.61
.61
Seal
-0
-0
0
ed
0.6
.516
.222
.138
Model Descriptions for likelihoods calculated
Model Al: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A2: Yij = Mu(i) + e(ij)
Var{e(ij)} = Sigma(i)A2
Model A3: Yij = Mu(i) + e(ij)
Var{e(ij)} = SigmaA2
Model A3 uses any fixed variance parameters that
were specified by the user
Model R: Yi = Mu + e(i)
Var{e(i)} = SigmaA2
Likelihoods of Interest
Model
Al
A2
A3
fitted
R
Log (likelihood)
-110.169386
-108.417108
-110.169386
-110.520064
-113.975552
# Param's
5
8
5
3
2
AIC
230.338773
232.834216
230.338773
227.040129
231.951103
Explanation of Tests
Test 1: Do responses and/or variances differ among Dose levels?
(A2 vs. R)
Test 2: Are Variances Homogeneous? (Al vs A2)
Test 3: Are variances adequately modeled? (A2 vs. A3)
Test 4: Does the Model for the Mean Fit? (A3 vs. fitted)
B-17
-------
(Note: When rho=0 the results of Test 3 and Test 2 will be the same.
Tests of Interest
Test -2*log(Likelihood Ratio) Test df p-value
Test 1
Test 2
Test 3
Test 4
11.1169
3.50456
3.50456
0.701356
6
3
3
2
0.08483
0.3202
0.3202
0.7042
The p-value for Test 1 is greater than .05. There may not be a
difference between responses and/or variances among the dose levels
Modeling the data with a dose/response curve may not be appropriate
The p-value for Test 2 is greater than .1. A homogeneous variance
model appears to be appropriate here
The p-value for Test 3 is greater than .1. The modeled variance appears
to be appropriate here
The p-value for Test 4 is greater than .1. The model chosen seems
to adequately describe the data
Benchmark Dose Computation
Specified effect = 1
Risk Type = Estimated standard deviations from the control mean
Confidence level = 0.95
BMD = 11.1519
BMDL = 6.85
B-18
------- |