V. Health Effects of Exposure to MDA

A. Summary of the Committee's Recommendations

1. Introduction

The Committee reviewed the record evidence concerning the acute and chronic effects of exposure to MDA in both animals and humans and concluded that MDA should be treated as a hepatotoxic agent and as a suspect human carcinogen. The Committee also concluded that an occupational standard regulating worker exposure to MDA should be developed. The following discussion provides the Committee's findings with respect to the hepatotoxic and carcinogenic hazards posed by occupational exposure to MDA.

2. Acute Effects of Exposure to MDA

A. Hepatotoxicity. The record evidence on the acute effects of occupational exposure to MDA indicates that occupational exposure to MDA may result in hepatotoxicity (poisoning of the liver). The Committee relied on an abundance of human and animal data to support this finding. (See Hepatotoxicity Section of the Committee's Document, Ex. 9.) The Committee found that one or a few exposures to high doses of MDA may result in toxic hepatitis. However, in all cases the clinical signs and symptoms of hepatitis produced by this exposure were reversible. The Committee's discussion concerning the acute effects resulting from acute exposures can be found at 52 FR 26779 and 26780. In summary the Committee stated clearly that "The predominance of data reflect the induction of disease as a result of dermal absorption of MDA" and further provides a data analysis from Kopelman, McGill and Motto, and Brooks et al. An analysis of the data did not rule out the possibility that liver toxicity might result from low doses. Furthermore, the analysis did not determine the effects long term low doses might have on liver function. However, the Committee tentatively did conclude that at the present occupational levels the clinically observed non-neoplastic effects of exposure to MDA appear to be totally reversible. This conclusion was based solely on review of the data found in the acute human studies (human chronic exposure studies are not available). Animal data however, did indicate that long term MDA dosing at low levels produced various levels of liver damage. Thus while making a finding that occupational exposure to MDA may result in liver toxicity, the Committee was unable to develop dose- response data which could predict with some certainty the exposure necessary to produce liver toxicity. More precisely, the Committee was unable to conclude that at 5 ppb, liver toxicity would not occur.

In an effort to make these findings, the Committee extensively reviewed the record evidence to determine the levels of exposure at which a No Observed Effect Level (NOEL) for the clinical observation of hepatitis could be expected to occur in a worker population. The available literature on workers occupationally exposed to MDA provided limited data on the occupational doses to which the workers were exposed. This is due in part to a lack of ambient sampling data but more often because the primary mode of exposure was through the skin and not through inhalation. The Committee further acknowledged that in the case of MDA, unlike many acutely toxic chemicals which are associated with acute inhalation effects such as irritation and pulmonary edema, the primary effect has been liver damage following ingestion or skin absorption. The only available data the Committee could use to estimate a NOEL for liver toxicity due to occupational exposure to MDA are the data reported by Kopelman et al. from the Epping Jaundice incident. This data suggested that levels in excess of 100 ppb would be necessary to produce acute hepatitis in worker populations. The Committee relied on these findings in making its recommendations for the TWA and the STEL.

B. Dermal Irritation. The Committee believed that the ability of MDA to induce contact sensitization has not been studied sufficiently to conclude that MDA causes sensitization.

C. Retinal Effects. The Committee reviewed the record evidence concerning the effects which might result from eye contact with MDA. The Committee concluded that direct contact between MDA and the eye should be avoided. In addition, the Committee noted that ingestion of MDA might also result in damage to the eye and as such should also be avoided.

3. Chronic Effects of Exposure to MDA

A. Hepatotoxicity. The Committee found that at the present occupational levels, the observed or clinical non-neoplastic effects resulting from exposure appear to be totally reversible (Ex. 9). This conclusion is based on review of the data found in the acute human studies. Animal data indicate that long term MDA dosing at low levels produces various levels of liver damage, but since most of the studies have involved the dosing of the animals until sacrifice, it is difficult to determine if the observed effects would or could have been reversed if sufficient time had been allotted for healing.

B. Carcinogenicity. The Committee concluded that MDA is a carcinogen in F344/N rats and B6C3F1 mice of each sex. Furthermore, it appears that carcinogenicity is induced either through ingestion, inhalation, or dermal absorption of the substance.

The Committee considered extensively the type of data needed to determine carcinogenicity in animals and to relate the observed effects in animals with that expected in humans. The Committee generally accepted the policy set forth by public health agencies, that test results in mammalian species (including the mouse), are acceptable data for predicting potential hazards to exposed humans.

The Committee also recognized that confounding factors associated with long term bioassays could cause carcinogenicity findings to be questioned. First, the Committee recognized the need to use control groups, as was done in the NTP and ORNL bioassays, and to validate the carcinogenic findings in rodent species that normally exhibit a high spontaneous incidence of tumors.

Second the Committee also discussed the effect that high dosing and subsequent acute toxicity may have on the production of liver and thyroid tumors found in the female mice of the NTP study. The Committee determined that the observed incidence of tumors found in the female mice of the NTP study did not occur as a result of high dosing but occurred from exposure to MDA.

Third the Committee noted that the presence of tumor viruses in mice does not necessarily invalidate the identification of MDA as a carcinogen.

In making this determination, the Committee made use of the policies advocated by numerous health agencies, including OSHA, which requires that, to make a viral etiology finding, the virus must be established to be the sole direct mechanism producing the carcinogenic effect. Not finding this necessary evidence, the Committee agreed that the carcinogenic response was not the result of viral etiology.

The Committee found that the NTP study was conducted properly, and therefore used this study as the principal basis for its carcinogenicity findings.

In addition, a majority of the Committee members concluded that MDA induces cancer by a genotoxic rather than a non-genotoxic mechanism and, as such, a threshold level for the carcinogenic response did not exist. The Committee concluded that the evidence offered for the existence of thresholds for this carcinogen was insufficient to overcome the extant evidence for a genotoxic mechanism. The Committee relied on two basic concepts to make this decision. First, the members required that if a threshold was to be considered, data indicating at what level a threshold would occur must be provided. Secondly, once a threshold is established in experimental animals, the threshold must be shown to be applicable to any exposed group of workers. No evidence was offered which meets these minimum criteria and thus the Committee made a recommendation that a no-threshold-effect be used to predict the risk associated with occupational exposure to MDA. In addition, the Committee believed that, even if a threshold for specific carcinogens could be demonstrated in experimental test animals or even in a specific human population, it might not be applicable to any given human population at risk. No data were furnished which equated a threshold observed in animals with that expected in humans.

Other concerns raised by some of the Committee members involve the use of MDA dihydrochloride rather than MDA itself as the administered dose in the NTP bioassays. The Committee noted that test animals in the Oak Ridge National Laboratories study were exposed dermally to MDA and not the hydrochloride as in the NTP study. Furthermore, the Committee noted that in the Oak Ridge test animals the carcinogenic response seen in the female mouse livers was approximately double that noted in the NTP study. Thus, the Committee concluded that exposure to MDA produced the carcinogenic effect, and not exposure to the salt.

The Committee also examined supportive evidence of carcinogenicity derived from short term mutagenicity tests. The Committee recognized that the various short-term tests do not measure the same mutagenic endpoint; thus positive and negative findings are not uncommon, since no single short-term test can measure all the events which might lead to mutagenesis. The Committee agreed that there is a wide variety of opinions on the reliability of using short-term studies as indicators of potential carcinogenicity. Many of the Committee members believed, however, that such tests provide meaningful indicative results and that substances which give positive results in well validated systems are likely to be carcinogenic. Further, it appears that the probability of a false-positive result for a chemical which is positive in one well conducted bioassay and one well validated short term-test is extremely small. Thus, based on record evidence consisting of both bioassays and short-term tests the Committee concluded that MDA causes cancer in experimental animals.

The Committee also analyzed the data to relate the findings of "pooled tumors" incidence in mice to some common site in man. The Committee acknowledged that scientific investigations have shown that target sites for the carcinogenic action of a substance in humans are not necessarily the same as those found in animal experiments. There were basically three pieces of suggestive evidence examined by the Committee to link the carcinogenic response in animals to the expected response in humans (bladder cancer):

(1)  The NIOSH-Vertol Health Hazard Evaluation (HHE) study (Ex. 1-255);
(2)  The presence of bladder transitional cell papillomas in three MDA
       treated rats in the NTP-Bioassay (Ex. 1-36); and
(3)  Some structure-activity links with the proven human and animal
       bladder carcinogen, benzidine, and the dog bladder carcinogen,
       methylenebis-(2-chloroaniline)(MBOCA).

The only available human data implicating MDA as a human carcinogen were from the HHE. The Committee reviewed these data thoroughly before concluding that the data were insufficient to positively identify MDA as a human bladder carcinogen or to use the data contained in this report to establish permissible exposure limits. However, the Committee did not exclude the fact that the report did develop a hypothesis regarding MDA exposure and bladder cancer which warrants further investigation using the more rigorous epidemiologic methods.

The Committee also found that the development of bladder transitional cell papillomas in the female rats in the NTP bioassays to be significant.

These relatively rare tumors were benign although progression to malignancy in this class of tumors may occur. Furthermore, the Committee recognized that the appearance of transitional cell papillomas in MDA treated rats was unique and demonstrated the chemical specificity of the results observed.

The Committee analyzed the structure activity relationships between MDA and several other substances identified by EPA as structural analogs. A majority of the Committee members maintained that there are significant structural differences between benzidine and MDA and that a strong analogy does not exist. The Committee generally believed, however, that while the structural analogy data are not conclusive, nonetheless these data should be relied upon to suggest that MDA may cause bladder cancer in humans. Although the Committee could not positively link occupational exposure to MDA with bladder cancer in workers, the Committee recommended stringent standard provisions to protect workers against the carcinogenic potential posed by MDA regardless of the target site.

4. Reproductive Effects

The majority of the Committee members concluded that, while the data suggest that there may be hormonal changes at relatively high doses, the occupational significance of these changes could not be assessed.

5. Teratogenic Effects

The Committee has reviewed the data on the teratogenic effects of exposure to MDA and could not relate the significance of these observed effects in animals with those anticipated in the occupational setting.

6. Absorption, Distribution, and Deposition

The majority of the Committee members agreed that where sufficient data exist which are MDA specific (e.g., dermal absorption data), these data should be used to determine the biological activity of the chemical. However, the Committee found that data obtained through the El-Hawari study (Ex. 1-251) were not sufficient to make determinations concerning the gastrointestinal and respiratory absorption of MDA. The Committee anticipated that future research on the gastro intestinal and respiratory absorption of MDA will also substantiate the findings made from the structural analogue comparisons and demonstrate that these assumptions are also conservative.

The Committee agreed that a 100 percent absorption through the gastrointestinal tract of the mouse be used in generating the risk assessment model rather than 50 percent absorption proposed by EPA. The Committee realizes that this is a conservative approach because it assumes that the observed effect is a result of absorption of the entire dose administered and not a result of the absorption of a lesser portion of the administered dose. This assumption has the effect of reducing the expected risks predictable from occupational exposure to MDA by 50 percent.

The Committee agreed, however, with EPA's assumption that absorption through lung tissue is roughly equivalent to gastro- intestinal absorption (50 percent), especially if MDA is in the vapor phase or has a particle size of less than 2 microns.

The Committee also concluded that MDA is actually dermally absorbed at approximately 2 percent per hour and not 1 percent as previously assumed (Ex. 1-251). Therefore, an absorption rate of 2 percent can also be applied to MDA exposure which occurs through dermal deposition and absorption.

The Committee also stressed the significance which the hazard of dermal exposure posed. Data from the Oak Ridge National Laboratory study (Ex. 8) heightened the Committee's concern over these hazards. The Committee concluded that when a chemical is ingested, it is transported through the hepatocellular detoxification system and is not generally diluted as a result of passing through the general circulatory system. In the case of chemicals applied to the skin, however, a significant dilution takes place as a result of the absorbed chemical passing through the general circulatory system before passing through the hepatocellular detoxification system. Compared with the findings of the NTP study in which animals were exposed through ingestion, the Oak Ridge data reported almost a two-fold increase in the liver tumor incidence observed in the female test animals dermally exposed to MDA. These findings are additional evidence that occupational dermal exposure to MDA should be prevented.

In addition, the Committee was concerned with the findings of El-Harawi (Ex. 1-251) indicating that once deposited on the skin, MDA cannot be completely removed by cleansing. The data suggest that the use of solvents to remove MDA from the skin actually increases the absorption of MDA. It also appears that soap and water provide the best medium for removing the substance from the skin, but only remove approximately 60 percent of the material deposited on the skin. These findings support the provisions of the final standard which require the use of personal protective clothing and equipment to prevent MDA exposure and medical surveillance to assure that the integrity of the protective equipment and clothing is being maintained.

B. OSHA's Findings

The following discussion of the health effects associated with occupational exposure to MDA is merely a summary account of the extensive analysis and findings made by [the Committee and] OSHA. Complete discussions of the health effects conclusions reached by the Committee [and accepted by OSHA] are found at 52 FR 26779 et seq. (July 16, 1987) [and 54 FR 20677 et seq. (May 12, 1989), respectively]. All of these health effects findings were essentially unchallenged by commenters and parties at the hearing.

The record evidence on the acute effects of occupational exposure to MDA indicates that exposure may result in hepatotoxicity (poisoning of the liver). These findings are based on an abundance of human and animal data. (52 FR 26779).

Evidence also indicates that direct contact between MDA and the eye as well as ingestion might result in damage to the retina of the eye.(52 FR 26780).

OSHA also finds that at the present occupational levels, the observed non-neoplastic effects on the liver resulting from exposure appear to be reversible (Ex 9). This conclusion is based on review of the data found in studies of acute liver disease in humans.

OSHA concludes that MDA is a carcinogen based on studies of F344/N rats and B6C3F1 mice of each sex. Furthermore, it appears that carcinogenicity is induced either through ingestion, inhalation, or dermal absorption of the substance. There were basically three pieces of evidence examined by OSHA which related the carcinogenic response of MDA in animals to the expected response in humans (bladder cancer):

(1) The NIOSH-Vertol Health Hazard Evaluation (HHE) study which demonstrated a significantly elevated PCMR for bladder cancer among workers exposed to MDA (Ex.1-255);

(2) The presence of bladder transitional cell papillomas in three MDA treated rats in the NTP-Bioassay (Ex. 1-36); and

(3) Some structure-activity links with benzidine, a proven human and animal bladder carcinogen, and with methylenebis-(2-chloroaniline) (MBOCA) a substance known to cause bladder cancer in the dog and suspected of causing bladder cancer in humans (52 FR 26787).

Although the evidence was not conclusive in demonstrating a causal link between occupational exposure to MDA and bladder cancer, OSHA nonetheless developed standard provisions to protect workers against the carcinogenic potential posed by MDA regardless of the target site.

OSHA also finds that, while the data suggest that there may be hormonal changes at relatively high doses, the occupational significance of these changes can not be assessed. (52 FR 26783).

Furthermore, OSHA has reviewed the data on the teratogenic effects of exposure to MDA and can not relate the significance of these observed effects in animals with those anticipated in the occupational setting. (52 FR 26784).

OSHA has also determined that the available data on the ability of MDA to induce contact sensitization has not been studied sufficiently to conclude that MDA causes sensitization. (52 FR 26786).

In reviewing the record evidence concerning the acute and chronic effects of exposure to MDA in both animals and humans, OSHA concludes that MDA must be regulated as both a hepatotoxic agent and a human carcinogen. OSHA tentatively made these findings in its NPRM and the Agency's conclusions remain unchanged. In fact, there was no evidence submitted in response to the NPRM which would cause OSHA to amend its earlier conclusions that MDA should be treated as a hepatotoxic agent and a suspect human carcinogen.

[57 FR 35630, Aug. 10, 1992]