It began in late 1992--the quiet introduction of methyl tertiary butyl ether (MTBE) into unleaded gasoline as a means of reducing carbon monoxide emissions during the winter months in targeted U.S. cities. Shortly afterwards, scattered health complaints, including headaches, dizziness, and nausea, from residents of some areas, along with reports of more specific health effects from residents of Fairbanks and Anchorage, Alaska, began to surface. These reported health effects led EPA technical experts and policy-makers to take a closer look at the MTBE program. The reports also launched an era of assiduous research by scientists and public health officials across the country to learn more about MTBE's short-term and long-term, and possibly carcinogenic, health effects.
A broad spectrum of opinion characterizes the MTBE debate. The petroleum industry touts the chemical's benefit as an air quality enhancer, while some environmentalists and a few scientists deplore it as an unnecessary additive whose health risks outweigh any purported benefits to public health. Most scientists' views lie somewhere in between. Rather than enter a fruitless debate, they continue to ask questions, collect data, and await the results of statistically viable, peer-reviewed research that they hope will clarify the MTBE dilemma.
Environmental Benefits
Most scientists, public health authorities, and even environmental groups agree that, as a fuel oxygenate, MTBE does decrease emissions of carbon monoxide, a toxic gas that poses a real health risk, especially to individuals who suffer from coronary artery disease. Because carbon monoxide emissions are highest in cold weather, refiners began adding 15% MTBE and other oxygenates including ethyl tertiary butyl ether (ETBE), tertiary amylmethyl ether (TAME), and ethanol to gasolines used during the winter months in 39 cities across the United States to meet the air quality standards set in the 1990 Clean Air Act Amendments.
MTBE is the most widely used oxygenate in the national reformulated gasoline (RFG) program, an extension of the oxyfuels program. The RFG program debuted in January of this year and has been hailed by Mary Nichols, the EPA's assistant administrator for air and radiation, as "the most important environmental fuels program since lead was banned from gasoline." The RFG program, which was developed with cooperation from the oil and chemical industries, automakers, environmental and citizen groups, and the EPA, promises to significantly reduce emissions of carbon monoxide, volatile organic compounds, nitrous oxides, and other air toxics. The EPA estimates that the use of reformulated gasoline in the nine target metropolitan areas plus other areas that opt into the program will reduce smog-producing emissions by 300,000 tons in 1995 alone--the equivalent of removing 8 million vehicles from the road.
Gasoline with MTBE has been sold in the United States since 1979, when the compound was added to fuels as an octane-enhancer after lead was phased out of motor fuels. According to the MTBE Task Force, made up of companies that cooperated with the EPA to initiate MTBE testing, the oxygenate is one of the most extensively tested and best-understood components of gasoline. The fuel industry continues to proclaim the relative safety of MTBE to humans and lauds the chemical's performance in reducing carbon monoxide levels and ozone-producing air toxics in most RFG areas. It also points out that gasoline formulated with or without MTBE contains carcinogens such as benzene and that 15% MTBE in gasoline reduces benzene emissions by 11%.
Despite what MTBE's detractors may say about the fuel industry's bias, safety remains a top priority for the industry, according to John Kneiss, director of health and product stewardship for the Oxygenated Fuels Association (OFA). "There is no scientific evidence that exposure to MTBE in its context is harmful to humans," Kneiss says. "Our intent is to make gas safer and less polluting. We're supporting additional research programs to gain greater understanding about MTBE," he adds, referring to the $2 million in additional research the OFA is underwriting over the next two years. The OFA has funded $4 million on MTBE research since 1987, along with millions the petroleum industry has spent.
Linking Symptoms to Exposure
MTBE supporters say the scientific research confirms that, at concentrations the average person is exposed to when refueling a vehicle, an amount less than 1 part per million (ppm), MTBE has a wide safety margin. Nevertheless, there is still the matter of health complaints. Motorists have complained of nausea, dizziness, lightheadedness, eye, nose, and throat irritation, headaches, and mostly, the distinct odor. However, aside from the odor of MTBE which most agree is unpleasant, none of these health symptom complaints has been replicated successfully by scientists in a laboratory setting.
After complaints from Alaskans during the winter of 1992-93, the Centers for Disease Control collaborated with the State of Alaska, the National Institute for Occupational Safety and Health (NIOSH), and the EPA to conduct a two-phase field epidemiology study in Fairbanks. The study investigated the possible relationship between MTBE oxyfuel exposure and symptomatic responses and between MTBE oxyfuel exposure and blood levels of MTBE and its major metabolite, tertiary butyl alcohol (TBA).
Survey responses showed an increase in symptom-reporting rates during phase I, when MTBE-formulated gasoline was in use, over those reported in phase II, when MTBE was not in use. According to the EPA's Office of Research and Development, causal relationships between MTBE and symptoms could not be determined, partly because the sample size may have been too small to demonstrate statistical significance and partly because there may have been bias in symptom reporting due to extensive negative publicity about MTBE at the onset of the oxyfuel season. The study did confirm a relationship between exposure and blood levels of MTBE and TBA, even at very low concentrations of MTBE.
Following the Alaska studies, the CDC sponsored two similar studies. One was in Stamford, Connecticut, an oxyfuels location where there were no widespread consumer complaints. The other was in Albany, New York, chosen for comparison purposes because the area did not participate in the oxyfuels program.
The Stamford study yielded no apparent relationship between symptom prevalence by occupational category and the median blood MTBE or TBA concentrations associated with each category. In fact, the similarity of responses across job categories in both studies suggested that the symptoms may not have been due to gasoline exposure.
Another epidemiological study soon followed, this one conducted by the Environmental and Occupational Health Sciences Institute (EOHSI) in New Jersey. Designed by epidemiologist Sandra N. Mohr, the study attempted to explore the relationship between the symptomatic responses of 237 garage workers from the New Jersey Departments of Transportation and Treasury and exposure to high and low concentrations of MTBE. Participants were divided into two groups: one cohort worked in northern New Jersey and was sampled during the winter oxyfuel season; the other group worked in the southern part of the state and was sampled 10 weeks after oxyfuels were phased out in the area. The study, published in 1994 in Inhalation Toxicology, found no differences in the reporting of symptom frequency over a 30-day period and no differences between the groups across the work shift. The results led researchers to conclude that "no untoward health effects clearly attributable to MTBE exposure could be demonstrated."
Though these studies have failed to directly link adverse health symptoms with MTBE exposure, at least one scientist researching MTBE believes the health complaints are legitimate. Myron Mehlman, a staff scientist at EOHSI, stoutly contends the additive is dangerous to human health. He also remains doubtful about some of the interpretations and conclusions of scientists whose research receives financial support from the oil industry. "Many reports [have been] submitted by people who are suffering as a result of exposure to MTBE," he says. "It will be interesting to see how many more reports are submitted as more people become familiar with the symptoms associated with MTBE," adds Mehlman, who says he has conducted two studies that confirm an association between symptoms and exposure.
In 1994, Mehlman studied a group of people in New Jersey who come into daily contact with presumably higher concentrations of MTBE: members of the Oil, Chemical & Atomic Workers Union, who work in refineries blending MTBE with gasoline. Over a period of several months, Mehlman recorded symptoms reported by the workers including neurotoxic, respiratory, and ear-nose-throat symptoms. In one survey, 91% of the workers reported headaches, and 50% complained of breathing problems.
Mehlman also studied more than 200 New Jersey drivers in the winter of 1995 who reported symptoms after exposure to MTBE. "Eighty percent complained of headaches, sixty-three percent reported lightheadedness, and forty percent reported an inability to concentrate," Mehlman says. "A few complained of rashes--between five percent and ten percent."
George Lucier, director of the Environmental Toxicology Program at the NIEHS, concedes these health symptom reports can't be ignored altogether, even though most currently published research doesn't directly link MTBE to those symptoms. "Whenever you look at anecdotal reports, you have to take them for what they are and do a careful evaluation," says Lucier. He also says researchers need to consider whether there may be a subgroup of the population that is sensitive to MTBE, or if the reported symptoms are tied to something else altogether.
Two controlled inhalation studies have attempted to identify a causal relationship between MTBE exposure and symptoms in humans. The first was performed in the EPA's Health Effects Research Laboratory (HERL) in Research Triangle Park, North Carolina. Researchers exposed 40 healthy male and female subjects (all approximately 25 years old) for 1 hour to 1.39 ppm (5 mg/m3) of pure MTBE--a concentration that falls within the range measured during refueling, but is 10 times higher than measurements found in vehicles during a 30-minute urban commute.
The study, published in Inhalation Toxicology in 1994, found neither increases in symptom reporting due to MTBE exposure nor increases in objective biomarkers of ocular and nasal inflammation, even though the subjects' blood-MTBE levels were similar to those of occupationally exposed workers who reported symptoms. James Prah, an EPA research psychologist who collaborated on the study with scientists from the EPA, the CDC, and the University of North Carolina, calls the findings "relentlessly nonsignificant."
"Bill Cain took our protocol, replicated it, and found the same results," Prah says, referring to an inhalation study conducted by Cain and colleagues at Yale University, which measured human reactions and pharmacokinetic responses to low-level MTBE exposure.
Pharmacokinetic analysis on two subjects in the HERL study, one male and one female, showed a rapid rise of blood-MTBE to 8.2 and 14.7 parts per billion, with a rapid decline and clearance half-time of about 36 minutes. Levels of TBA increased gradually to 7-10 parts per billion (ppb) and persisted at half the levels of MTBE for seven hours after the exposure to MTBE. "If there are toxic effects," Prah says, "it's possible that TBA is responsible, rather than MTBE."
As to why adverse symptoms have been reported by individuals during oxyfuel seasons but haven't been confirmed in the controlled inhalation studies, no one can give a clear answer. "Maybe it's a combination of exposures that contributes to the effects, or there may be people who are more sensitive to MTBE," suggests Maria Costantini, senior staff scientist and toxicologist at the Health Effects Institute in Cambridge, Massachusetts. "The clinical studies have been done using the pure substance," she explains, referring to the concentrations of MTBE used in the HERL and Yale chamber studies. "Maybe it's a problem of the mix," Costantini says.
Her colleague, HEI President Daniel Greenbaum, agrees: "There has been relatively little research done on the effects of MTBE in mixture," Greenbaum says. To that end, researchers at the EPA's human studies facility in Chapel Hill, North Carolina, hope to concoct a "surrogate gasoline" from 20-25 of the components found in gasoline, says Prah. Human subjects in a controlled environment would be exposed to a 1% vapor of that mixture, both with and without MTBE. "Then we'll see if we can provoke symptoms," Prah explains. "If we can, we'll group the subjects by symptom observed."
The hypothesis suggested by both Costantini and Lucier--that a subpopulation especially sensitive to MTBE may exist--has been raised by others in MTBE research. Nancy Fiedler, along with Mohr and other researchers at the Robert Wood Johnson Medical School in Piscataway, New Jersey, attempted to identify a sensitive subpopulation in a study they conducted in 1993. They administered a symptom questionnaire to 13 subjects who reported having multiple chemical sensitivity. Multiple chemical sensitivity sufferers report symptoms and illnesses in response to low-level exposure to a variety of chemicals and substances commonly encountered in the environment. Questionnaires were also administered to five subjects with chronic fatigue syndrome, as well as to six healthy control subjects. Symptom scores from multiple chemical sensitivity subjects suggested that they indeed experienced greater discomfort in conjunction with refueling.
The Carcinogenicity Debate
While there is speculation about the factors that may influence an individual's exposure to MTBE, there is also much discussion about the chemical's carcinogenic effects. According to C.B. Hiremath, a toxicologist in the EPA's Office of Health and Environmental Assessment in Washington, DC, the EPA classifies MTBE as a "possible carcinogen," based on current studies. "We're waiting for more information so we can make a complete assessment," says Hiremath, who has been reviewing data on MTBE for two years.
Two chronic animal cancer bioassays of MTBE provide the basis for the few conclusions that have been drawn regarding MTBE's carcinogenicity. Conducted by J.S. Chun and H.D. Burleigh-Flayer and colleagues at Bushy Run Research Center in Pennsylvania, the bioassays were completed in 1992. One study exposed groups of rats, both male and female, to MTBE concentrations of 1,400, 10,800, or 28,800 mg/m3 in filtered air for six hours a day, five days a week, for 24 months. Control animals inhaled filtered air only.
Researchers observed an increase in rare kidney tumors in the mid- and high-dose groups of male rats. Exposure-related increases in liver and kidney weights were reported in females in the mid- and high-exposure groups. The kidney tumors in male rats raised the question of whether these tumors were attributable to the accumulation of a species- and sex-specific protein, alpha-2u-globulin, which has been associated with a pattern of damage in the kidney tubule cells of male rats. According to EPA reports, little evidence exists that MTBE causes this protein to accumulate, though the agency is awaiting results from ongoing studies.
The mid- and high-dose groups of male rats also exhibited a statistically significant, dose-related increase in testicular tumors compared to controls. But some scientists question the significance of the tumor response due to historically high background incidences of interstitial cell tumors in this strain of rat. "Ninety-five percent of the control rats got this tumor," says John Mennear, a professor of toxicology at the Campbell University School of Pharmacy who follows MTBE research as a consultant for ARCO Chemical Company, the largest producer of MTBE.
In the second Bushy Run study, groups of male and female mice were exposed to the same concentrations of MTBE used in the rat study, for the same exposure durations, for 18 months. MTBE exposure caused an increased incidence of liver tumors in mice at the high dose.
Mennear points out that "it takes a toxic dose" of MTBE to elicit a tumor response in laboratory animals. "Nobody is saying MTBE is innocuous," he admits. "However, the concentrations people are exposed to routinely are not toxic." Mennear and Larry Andrews, ARCO Chemical's manager of toxicology and regulatory compliance, believe that the chronic animal study results do not imply a risk for humans: "From what we know about the mechanisms that cause tumors [in laboratory animals], there is a large margin of safety for humans," said Andrews.
Scientists hope new research conducted by Cesare Maltoni and colleagues at the Ramazzini Foundation of Oncology and Environmental Sciences will shed more light on MTBE's potential as a human carcinogen. This study differs from the other chronic animal studies in that MTBE is being administered to rats by gavage in an olive oil medium, as opposed to inhalation. According to Mehlman, the study shows MTBE exposure causes increased levels of lymphomas and leukemias, which confirms his view that MTBE is a "poison." North Carolina's state toxicologist, Ken Rudo, recently reviewed the Maltoni study. He too, says the study found leukemias and lymphomas, adding that there was a "very, very clear dose response."
"The study also verified testicular tumors, but in a different kind of rat with low background tumor incidence," Rudo reports. The Maltoni study is weak is some areas, Rudo claims, because it includes "no statistics for tumor incidence and no information on pathology." In short, he says, "it doesn't meet the basic criteria for a two-year bioassay."
Future Research Needs
While opinions on MTBE's carcinogenicity may vary among scientists associated with the issue, most agree that further study is necessary. According to NIEHS's Lucier, the National Toxicology Program has been studying the effects of TBA. This research so far has shown "the same kind of tumors produced by MTBE in the EPA-mandated studies," says Lucier.
Studies currently underway should prove useful to scientists who must extrapolate the animal study findings to predict human risk. Susan Borghoff, a staff scientist at the Chemical Industry Institute of Toxicology (CIIT) is developing a physiologically based pharmacokinetic model for MTBE and TBA in rats. The model will be used to predict the blood levels of MTBE in rodents and enhance the understanding of the kinetics of MTBE and TBA. According to Borghoff, knowledge of the kinetic behavior of MTBE and TBA will aid in assessing human risk. "This is just a first-stage model," says Borghoff. "Studies have been proposed to collect this information in humans."
Other CIIT researchers are trying to find clearer answers to why a number of compounds in unleaded gasoline--MTBE included--cause liver tumors in mice exposed to high levels. "We're testing the hypothesis that MTBE in unleaded gasoline is inducing a liver tumor response through hormone modulations, specifically estrogen," explains Thomas Goldsworthy, a staff scientist at CIIT. The study should yield a clearer understanding of the mechanism that is critical for the tumor response, as well as the dosage at which the response occurs, he says.
Results of the study, which should be complete by the end of the year, may reveal if tumor response to MTBE is similar to the tumor response elicited by unleaded gasoline formulated without MTBE. A previous CIIT study suggested that high-level exposures to unleaded gasoline produce an increase in liver tumors in female mice by interfering with estrogen hormone function.
This finding, according to the CIIT, seems to suggest two things: that an association between unleaded gasoline and increased incidence of liver tumors is a high-dose phenomenon that may be unique to female mice, and that such an association wouldn't likely be observed in humans, due to low exposure levels typical for humans. Previous CIIT research had concluded that the sensitivity of male rats to kidney tumors stems from the interaction of unleaded gasoline components with a kidney protein found in male rats, but not in humans.
Unleaded gasoline and MTBE have similar chronic and subchronic effects, says Goldsworthy. So far, he and his colleagues have tested unleaded gasoline and MTBE separately. The next step, Goldsworthy says, is "to see what happens when you put MTBE and gasoline together."
Weighing Risks and Benefits
New research should help weigh the risk of MTBE as a possible carcinogen and the effectiveness of MTBE-blended fuels in reducing carbon monoxide levels. The question is whether, in minimizing one risk, is another risk--however small--being introduced? "Scientists have to think in these terms of benefit versus risk," Mennear concedes.
This kind of analysis may lead at least one oxyfuels state, North Carolina, to petition the EPA to relieve the state from participating in the program. The state's Department of Environment, Health, and Natural Resources has collected data for two years showing that the use of fuels with MTBE has not reduced ambient carbon monoxide during the oxyfuel season below carbon monoxide levels prevalent during the remaining months of the year (typically 7-8 ppm based on an 8-hour exposure). "All we were left with was risk, but no benefit," says Rudo.
Although reductions in carbon monoxide levels have been reported in many states, most notably in the northeast and California, Rudo would like to see the EPA give states more latitude in determining the effectiveness of using oxygenated fuels from the standpoint of health risks and benefits.
More testing of the other oxygenates should be done, according to HEI's Costantini. "Ethanol has been studied for ingestion, not inhalation," she says. She would like to see a comparative analysis of the pharmacokinetics of ethanol inhalation versus ingestion. She also believes a comparative study of all the oxygenates on their relative toxicity would be helpful.
According to Lucier, any risk- benefit analysis should include an evaluation of alternatives to MTBE. "I think there is a search for alternatives," he says, adding a word of caution: "We just don't want to rush into using another chemical before it is thoroughly tested."
Lucier says that, in addition to conducting toxicological evaluations of different chemicals in the interest of public health, the National Toxicology Program also provides regulatory agencies with important data to strengthen risk assessments. In his role as chair of the North Carolina Scientific Advisory Board, Lucier helped formulate an analysis of MTBE's effectiveness as a carbon monoxide reducer versus its human carcinogenic potential. Extrapolating from the animal data, Lucier says, "we estimated that the risk was between zero and one cancer per 100,000 people exposed."
Statistically speaking, MTBE poses a very low cancer risk to humans, Lucier says, though that means for 250 million people exposed over their lifetimes, as many as 2,500 could develop cancer. Which Lucier points out, shows that "a risk assessment decision can have tremendous impact," says Lucier. "The idea is to make gasoline as safe as possible."
Jennifer Medlin
National Reformulated Gasoline Hotline Emission Reduction Predictions
Carbon Monoxide (CO)--The average car emits 557 pounds of carbon monoxide per year. Use of reformulated gasoline will reduce CO emissions by 13%.
Volatile Organic Compounds (VOCs)--The average car emits 75 pounds of volatile organic compounds per year. Use of reformulated gasoline will reduce VOC emissions by 15%.
Nitrogen Oxides (NOx) --The average car emits 39 pounds of nitrogen oxides per year. Use of reformulated gasoline will reduce NOx emissions by 3%.
Air Toxics--The average car emits 1.74 pounds of air toxics per year. Use of reformulated gasoline will reduce air toxics emissions by 24%.
Source: RFG Issue Brief, January/February 1995 |
[Table of Contents]
Last Update: May 16, 1997