PEER REVIEW DRAFT, DO NOT CITE OR QUOTE
SEPA
United States
Environmental Protection Agency
Office of Chemical Safety and
Pollution Prevention
Risk Evaluation for
Methylene Chloride
Systematic Review Supplemental File:
Data Quality Evaluation of Human Health Hazard Studies -
Human Controlled Experiments
CASRN: 75-09-2
H
October 2019, DRAFT
1

-------
PEER REVIEW DRAFT, DO NOT CITE OR QUOTE
Human Controlled Experiments
Acute health reference values such as the SMAC, interim AEGL 1 and 2 and California REL
have been based on evidence of neurological effects in controlled human experimental studies
(Putz et al.. 1979: Winneke. 1974: Stewart et al.. 1972s) and evidence for increased
formation of carboxyhemoglobin (COHb) in blood in these and other studies of humans
exposed to methylene chloride (Andersen et al.. 1991: Divincenzo and Kaplan. 1981:
Peterson. 1978: Astrand et al.. 1975: Ratnev et al.. 19741 EPA also identified additional
human experimental studies that evaluated CNS and related effects through backwards
searches of AEGL, SMAC and the California REL documents. Gamberale (1975) evaluated
methylene chloride's association with reaction time, short-term memory and subjective
symptoms. DiVincenzo et al. (191T) evaluated cerebral and motor functions using a wooden
pegboard task. Kozena et al. (1990) evaluated reactions to weak auditory stimuli and
subjective feelings before, during and after methylene chloride exposure. Winneke and Fodor
(1976^) described two separate experiments. In the first experiment, females conducted tasks
(such as adding numbers), which were then interrupted to determine performance on another
task (critical flicker frequency, or CFF). In the second experiment, females alternately
performed tests of auditory vigilance and CFF. Winneke (1974) appears to describe the
second experiment that is identified by Winneke and Fodor (1976), and therefore EPA did
not re-evaluate this second experiment because EPA already evaluated it (as presented by
Winneke (1974)).
As noted above, several studies elevated COHb levels in blood, which can serve as a
biomarker for exposure to carbon monoxide (CO), in this case formed by metabolism of
methylene chloride. These studies developed quantitative relationships between blood levels
of COHb and various health effects. Acute health reference values for CO are based on blood
levels of COHb.
EPA has not developed formal data quality criteria to evaluate these studies. Instead, EPA
qualitatively evaluated individual aspects of data quality. Of the human controlled
experimental studies evaluating CNS effects associated with methylene chloride (Kozena et
al.. 1990: Putz et al.. 1979: Winneke and Fodor. 1976: Gamberale et al.. 1975: Winneke.
1974: Divincenzo et al.. 1972: Stewart et al.. 1972Y all included objective tests to measure
neurological endpoints, such as visual critical flicker frequency, visual evoked response to
strobe light, auditory vigilance tasks and others. Although EPA evaluated studies that
examined methylene chloride directly as well as the toxicokinetic studies that determined the
amount of COHb formed, EPA narrowed its focus for the risk evaluation to studies that
directly tested methylene chloride's effect on CNS outcomes because there is evidence that
methylene chloride exposure has a greater effect on CNS effects than just through COHb
alone.
The studies differed regarding the use of blinding. Putz et al. (1979) and Kozena (1990^)
stated that both the volunteers and investigators were blinded to the subjects' exposures.
Winneke (1974) and Winneke and Fodor (1976) used a single blind method that was not
fully described; it is assumed that the volunteers (not the investigators) were blinded to their
exposure status. Stewart et al. (1972) and DiVincenzo et al. (1972) did not mention whether
blinding was employed, and both measured subjective symptoms. Gamberale et al. (1975)
used menthol to hide the odor of methylene chloride and thus, subjects were blinded to their
2

-------
PEER REVIEW DRAFT, DO NOT CITE OR QUOTE
methylene chloride status. These authors did not mention whether the investigators were
blinded, and it is not known how the introduction of menthol affected any of the results.
Blinding was also not mentioned for the strictly toxicokinetic studies that evaluated changes
in COHb (Andersen et al.. 1991: Divincenzo and Kaplan. 1981: Peterson. 1978: Astrand
et al.. 1975: Ratnev et al.. 1974Y although that is not expected to have had any measurable
influence on the results in these studies, where outcomes consisted of measured levels of
methylene chloride, CO and COHb in expired air and/or blood at various time points.
Most studies described the methods of methylene chloride atmosphere generation, and
measurements (made by gas chromatography, infrared spectrophotometry or hydrocarbon
analyzer) showed exposure concentrations varied little from target values for several of these
studies. Comparing the inhalation method of generation against the TSCA data evaluation
criteria described in Application of Systematic Review in TSCA Risk Evaluations HJ.S. EPA.
20181 most authors used appropriate methods of test substance identification and generation
of inhalation exposures (e.g., chambers with continuous metering into the room, purity
information). However, Gamberale et al. (1975) used a breathing valve and gave very few
details on the generation of the test concentrations; no measurements were made except for
methylene chloride in alveolar air. Kozena (1990) also provided few details and used a half
mask, making it more difficult to determine methylene chloride exposure concentrations.
Given differences in exposure methods and limited information, these studies are difficult to
compare with the other studies that measured CNS effects.
The authors across several studies chose exposure durations and concentrations to provide
information useful for evaluating the relationships being investigated, including dose-
dependence of findings. For the studies with toxicokinetic components, sample collection
methods, timing and analytical methods were described and were appropriate.
Exposure groups were generally small, as would be expected for human experimental studies,
although Winneke (1974) included group sizes as large as 18 or 20 for some exposure levels
in some neurological performance tests and Kozena et al. (1990) exposed 16 males. Stewart
et al. (1972) inadvertently exposed subjects to methylene chloride immediately before the
experiment was supposed to begin, but this exposure did not result in a change in COHb
levels. In the controlled exposure neurological performance studies, neurological findings are
presented in relation to negative controls, but the source of the control data was not always
clear. Subjects served as their own controls in several of these studies, a typical design for
human experimental studies.
As a group, many of the studies provide a consistent picture of the acute neurological effects
and kinetics of COHb formation associated with methylene chloride exposure in humans,
which reinforces the reliability of the data from the individual studies. Most used an
acceptable measure of test substance generation including analyzing the test concentrations.
Most studies describe results in sufficient detail. However, DiVincenzo et al. (1972) did not
provide any details regarding their results.
Putz et al. (1979) received a medium confidence level due to use of blinding but only a single
concentration. Although Winneke (1974) used multiple concentrations, the authors used a
single blind method that was not well described and EPA gave this study a medium data
quality rating. Because Stewart (1972) did not mention whether the investigators and
3

-------
PEER REVIEW DRAFT, DO NOT CITE OR QUOTE
volunteers were blinded to their exposure status, EPA has low confidence in the subjective
symptoms reported in this study whereas the objective measures are given a medium
confidence rating.
EPA gave four studies low data quality ratings. The method of exposure, lack of measured
exposure concentrations and use of menthol by Gamberale (1975) resulted in a low
confidence rating. Similarly, Kozena et al. (1990) provided no details on their exposure
generation and furthermore, describe use of a half mask and thus, EPA gave this study a low
quality rating. Primarily because DiVincenzo et al. (1972) provided no details regarding
results and limited information on whether negative controls were used, EPA gave this study
a low confidence rating. Finally, the first experiment by Winneke and Fodor (1976) provided
details that were too limited regarding the outcome assessment methodology and did not
describe the outcomes regarding adding of numbers; therefore, EPA also gave this a low
confidence rating.
The studies evaluating generation of COHb were not given confidence levels because these
studies haven't been further considered for this risk evaluation.
4

-------
PEER REVIEW DRAFT, DO NOT CITE OR QUOTE
References
Andersen. ME: Clewell. HI. Ill: Gargas. ML: Macnaughton. MG: Reitz. RH: Nolan. RI:
Mckenna. MI. (1991). Physiologically based pharmacokinetic modeling with
dichloromethane, its metabolite, carbon monoxide, and blood
carboxyhemoglobin in rats and humans. Toxicol Appl Pharmacol 108: 14-27.
Astrand. I: Ovrum. P: Carlsson. A. (1975). Exposure to methylene chloride: I. Its
concentration in alveolar air and blood during rest and exercise and its
metabolism. Scand J Work Environ Health 1: 78-94.
Divincenzo. GD: Kaplan. CI. (1981). Uptake, metabolism, and elimination of methylene
chloride vapor by humans. Toxicol Appl Pharmacol 59: 130-140.
http: //dx.doi.org/10.1016/0041 -008Xf81190460-9
Divincenzo. GD: Yanno. FI: Astill. BP. (1972). Human and canine exposures to methylene
chloride vapor. Am Ind Hyg Assoc J 33: 125-135.
http: //dx.doi.org/10.1080/0002889728506622
Gamberale. F: Annwall. G: Hultengren. M. (1975). Exposure to methylene chloride: II.
Psychological functions. Scand J Work Environ Health 1: 95-103.
Kozena. L: Frantik. E: Vodickova. A. (1990). Methylene chloride dose not impair vigilance
performance at blood levels simulating limit exposure. Activitas Nervosa
Superior 32: 35-37.
https://heronet.epa.gov/heronet/index.cfm/reference/download/reference id/
29233
Peterson. IE. (1978). Modeling the uptake, metabolism and excretion of
dichloromethane by man. Am Ind Hyg Assoc J 39: 41-47.
http: //dx.doi.org/10.1080/0002889778507711
Putz. VR: lohnson. BL: Setzer. IV. (1979). A comparative study of the effects of carbon
monoxide and methylene chloride on human performance. J Environ Pathol
Toxicol 2: 97-112.
Ratnev. RS: Wegman. DH: Elkins. HB. (1974). In vivo conversion of methylene chloride to
carbon monoxide. Arch Environ Occup Health 28: 223-226.
Stewart. RD: Fisher. TN: Hosko. MI: Peterson. IE: Baretta. ED: Dodd. HC. (1972).
Experimental human exposure to methylene chloride. Arch Environ Occup Health
25: 342-348. http://dx.doi.org/10.1080/00039896.1972.10666184
U.S. EPA. (2018). Application of systematic review in TSCA risk evaluations. (740-P1-
8001). Washington, DC: U.S. Environmental Protection Agency, Office of Chemical
Safety and Pollution Prevention.
https://www.epa.gov/sites/production/files/2Q18-
06/documents/final application of sr in tsca 05-31-18.pdf
Winneke. G. (1974). Behavioral effects of methylene chloride and carbon monoxide as
assessed by sensory and psychomotor performance. In C Xintaras; BL Johnson; I
De Groot (Eds.), Behavioral toxicology: Early detection of occupational hazards
(pp. 130-144). Cincinnati, OH: U.S. Department of Health, Education, and Welfare,
National Institute for Occupational Safety and Health.
Winneke. G: Fodor. GG. (1976). Dichloromethane produces narcotic effect. Occup Health
Saf 45: 34-35.
https://heronet.epa.gov/heronet/index.cfm/reference/download/reference id/
29075
5

-------