DECEMBER 9, 1997

I appreciate the opportunity to submit testimony to the Environment and Public Works Committee of the United States Senate. My curriculum vitae is attached. Briefly, I am a pharmacologist/toxicologist who has spent most of my career on the full time faculty at Georgetown University School of Medicine where I did research and taught courses in pharmacology and toxicology to medical students and undergraduates. In 1995, I retired from academics to move to Colorado. I am currently a Principal with the International Center for Toxicology and Medicine, where I work as a consultant on a variety of environmental and occupational health issues. As a consultant to the Oxygenated Fuels Association since 1993, I am very familiar with the health-related studies of oxygenated gasoline in general and MTBE specifically. I have served as a consultant and peer reviewer for the U.S. EPA, CDC and the National Academy of Science on this issue, and have written a paper on the acute health effects associated with exposure to oxygenated gasoline, which will be published in the December issue of the journal, Risk Analysis. A copy of this paper is also attached to this statement.

My testimony deals with the health implications of the continued use of MTBE in reformulated and oxygenated gasoline. In addressing this issue, the potential for toxicity of MTBE cannot be considered in isolation, but must be weighed against the benefits associated with its use in gasoline. Gasoline, itself, is known to contribute significantly to human exposures to numerous toxins, including carbon monoxide, ozone, and known human carcinogens such as benzene and 1,3-butadiene. The rationale behind the reformulation and addition of oxygenates to gasoline is to reduce these exposures. Thus, the focus in the consideration of health effects should be how the risks from MTBE exposure from oxygenated gasoline compare to the benefits associated with the decreased exposure to toxic gasoline-related emissions that occurs as a result of addition of MTBE to the gasoline.

The major route of human exposure to MTBE is through inhalation of air containing MTBE that has evaporated from gasoline or been released in the exhaust from vehicles. In addition, there can be human exposure associated with MTBE in water. The most significant source of MTBE in water is gasoline leaks and spills, including leakage of underground storage tanks. Gasoline contamination of water is a problem whether or not the gasoline contains MTBE. The question is, how does the movement of MTBE from gasoline to water affect the benefit risk equation for oxygenated gasoline vs. conventional gasoline?

We know a great deal about the toxicity of MTBE and the exposure concentrations necessary to cause toxicity. There has been extensive animal testing for acute and chronic toxicity, including carcinogenicity, as well as both experimental and epidemiological studies in humans. The animal studies involve exposures that are many orders of magnitude above the concentrations to which humans would be exposed. The results of these studies and their extrapolation in the prediction of human risk are considered separately for carcinogenic and non-carcinogenic endpoints since the approaches for extrapolating from animals to humans are different.

With respect to non-cancer endpoints, the thresholds for toxicity in animals are sufficiently high that toxicity in humans exposed to MTBE in air as a result of its use in oxygenated gasoline are not expected to occur. The epidemiological studies comparing health effects in areas using conventional vs. oxygenated gasoline, and experimental studies involving controlled exposure to MTBE at environmentally relevant concentrations support this conclusion. These data and conclusions are discussed much more fully in the attached paper.

Although the concentration of MTBE in water contaminated as a result of a gasoline leak or spill can be high, humans are not likely to be exposed at these levels because the presence of MTBE in water at very low concentrations impacts the taste and smell characteristics of the water such that exposure will be self-limiting. In situations where the MTBE concentration in water is high, there might be short-term exposures that result in irritant effects. However, longer exposures at these levels will not occur. Although there are no animal studies involving long-term drinking water exposure, the threshold for chronic, non-cancer toxicity can be extrapolated from a subchronic study involving oral gavage exposure (i.e., the chemical was delivered directly into the stomach by tube) or from the lifetime inhalation exposure studies. Using either approach for extrapolation, it is clear that the water safety level that would protect against chronic, non-cancer toxicity is well above the threshold for odor and taste changes. In other words, from a practical point of view, humans will not be chronically exposed to MTBE in water at concentrations associated with toxicity.

MTBE causes several types of tumors in animals exposed to high concentrations of the chemical. While it is generally assumed that a chemical that causes cancer in experimental animals poses some cancer risk to humans, the scientific and regulatory communities are recognizing that there are exceptions to this conservative assumption depending on the mechanism of action of the chemical. For example, when the mechanism of cancer induction is one that only occurs at high exposures where cell death and tissue damage occur, such an effect would not be expected to occur in humans since the exposure would be to far lower doses than in the experimental animals. Other mechanisms of cancer induction related to the effects of chemicals on hormonal balance or an animal-specific cellular component are similarly not necessarily relevant for predicting human risk. On the other hand, a chemical whose mechanism of action involves damage to DNA is likely to have a similar effect in humans. MTBE does not damage DNA, and there is some evidence that its carcinogenic effect in animals may involve mechanisms not relevant to predicting human risk; additional study is taking place to clarify this issue. For the purposes of this discussion, however, it will be assumed that the animal cancer response is a relevant predictor of human risk.

The cancer risk calculations contained in the September 2, 1996, California Environmental Protection Agency briefing paper on MTBE are as follows: the calculated increase in risk associated with breathing MTBE as a result of its use in gasoline is one to two lifetime cancer cases per million people exposed; balanced against this is a calculated decreased risk of about 60 per million that occurs because the use of reformulated gasoline reduces the opportunity for gasoline-associated exposure to known human carcinogens such as benzene and 1,3-butadiene. Adding the potential risk associated with exposure to MTBE through water at the upper limit of the threshold for taste and odor recognition, the net benefit of MTBE on human cancer risk remains above 50 per million.,

In summary, there is accumulating evidence that the projected health benefits of oxygenated and reformulated gasoline are, in fact, being realized. It is against this benefit that the risks of gasoline-related MTBE exposures need to be weighed. We know that there will be human exposure to MTBE as a result of its use in gasoline these exposures are primarily a result of breathing air containing evaporative and exhaust products of gasoline, but may also occur from gasoline-contaminated water supplies. However, the exposures from these sources are below the threshold for human toxicity. Whether or not MTBE exposure increases human cancer risk remains an area of scientific debate. But even if we make the assumption that MTBE is a potential human carcinogen, the predicted cancer risk associated with MTBE-containing reformulated gasoline is less than that associated with conventional gasoline. This is because compared to conventional gasoline, the use of reformulated gasoline results in decreased exposures to known human carcinogens such as benzene.

A recently published study has reported effects on the life cycle of white blood cells in a group of individuals exposed to water contaminated as a result of an underground storage tank leak. The water reportedly contained low levels of MTBE and benzene. There are some significant questions about the methods that were employed in the interpretation of this study, and the findings are seemingly implausible given the fact that the studies were done almost a year after the cessation of the exposure. In any case, however, the reported exposure was to both benzene and MTBE, making it impossible to conclude that MTBE was the causative agent. Given the fact that benzene is a known human carcinogen and its primary target in humans is the blood system, benzene is a much more likely candidate for causing the reported effects than is MTBE.

The scientific and regulatory communities will continue to study MTBE, and some questions do remain. These have been identified in several reviews that have been completed in the last year. While the toxicity of MTBE itself has been well studied, studies that directly compare the effects of gasoline, with and without MTBE, are planned but not yet completed. A question has also been raised as to whether there are some individuals who are uniquely sensitive to MTBE. Whenever a new chemical or drug is introduced, this possibility always exists. While nothing in MTBE's toxicological profile predicts that there will be such a sensitivity, at least one study is underway to investigate this possibility.

Another question that has been raised is whether it is necessary to do toxicological studies in animals exposed to MTBE in drinking water. With the use of a technique known as physiologically-based pharmacokinetic (PBPK) modeling, it is possible to identify the drinking water dose equivalents of the exposure regimens used in the inhalation studies. This extrapolation is based on the principle that it is the dose of a chemical delivered to the target tissue that determines the effect, independent of whether the dose was delivered by inhalation or by drinking water. The PBPK model is a computer simulation of the body, including the various organs (target tissues), each with its characteristic blood flow and pathways for handling the chemical; routes of elimination of the chemical are also included. Both inhalation and drinking water dosing can be simulated, and the target tissue concentrations of MTBE and its metabolites determined as a function of time. By doing this, the inhalation dose response data can be translated to target-tissue dose response data. The simulated drinking water exposure that results in similar target tissue doses can then be determined as a basis for the extrapolation. A PBPK model for MTBE and its major metabolite, TBA, has been published and is currently being validated for route-to-route extrapolation.

Use of PBPK modeling as the basis for route-to-route extrapolation has been used for a number of other chemicals and can be done with a high degree of confidence. In the case of MTBE, it may well be the only way to determine dose-response data for drinking water exposures since the odor and taste properties of MTBE are likely to prevent animal exposures at levels high enough to provide an adequate test of toxicological response. Some studies have been reported involving oral exposure using a gavage method, where a bolus of MTBE is introduced directly into the stomach. However, such studies are a poor simulation of a drinking water exposure because the dose is introduced all at one time rather than in increments over the course of the day. In this respect, inhalation exposure provides a better simulation of the exposure that occurs.

Continued examination and confirmation of the benefits and risks associated with the use of MTBE in reformulated gasoline is appropriate. However, there are adequate data at this point to support the safety and benefits of continued use of MTBE-containing reformulated gasoline as these studies are being done.