Clarification: chemical synergism.

several scientists held the hope that the "Maltoni" work would clarify questions about MTBE's carcinogenic potential. Despite the enthusiasm of some scientists for Belpoggi's results, the most cursory examination of the paper reveals critical issues that show the data have been grossly overinterpreted. Rather than predicting a health hazard, the data indicate that rats tolerate enormous daily oral doses of MTBE without exhibiting evidence of either tumor or nontumor pathology. Leydig cell tumors in high-dose male rats. According to Belpoggi et al., the administration of MTBE as an olive oil gavage to male rats (1.0 g/kg, 4 days per week) was associated with a significant increase in the incidence of Leydig cell tumors. But examination of the total information in the paper shows that the reported effect cannot be attributed to MTBE. The apparent association was due to a survival differential between control and dosed animals. Male rats administered the highest dose of MTBE survived longer than the control group. It is well known that Leydig cell tumor incidence is age related. The longer a rat survives, the more likely it is to have Leydig cell tumors. Claiming this survival-related effect to be indicative of a human health hazard strains the bounds of scientific logic. This is particularly true since Leydig cell neo-plasms are most likely unique to rats and appear to have no predictive utility for human carcinogenic responses (2). Lymphoma and leukemia (combined) in female rats. Belpoggi et al. reported that MTBE increased the incidence of lymphomas and leukemias (combined) in female rats. Since no mention was made of the incidences of these neo-plasms individually, one can only assume that neither was significantly elevated. The scientific validity of combining lymphomas and leukemias for statistical purposes is highly questionable. A National Toxicology Program working committee reviewed scientific guidelines and criteria for the combination of neo-plasms during the interpretation of rodent carcinogenesis studies (3). According to that group of experienced pathologists, combining certain tumors for statistical purposes is appropriate and might afford enhanced insights into the biological effects of the test chemical. In other cases, however, combinations are unjustifiable and can lead to overestimates of carcinogenic potential. According to the NTP panel, it is reasonable to combine different types of leukemias and to combine different types of lymphomas. But it is not appropriate to combine leukemias with lym-phomas. Treatment-associated increases in the incidence of one or the other of these …