Pediatric Leukemia, Susceptibility Genes, and Drinking Water Contaminants

ISEE-334 Introduction: We conducted a case control study combining data from: (1) archived dried blood spots (DBS) collected to test newborns for inborn errors of metabolism; (2) a state geographic database on contaminant levels in public drinking water; (3) a population-based cancer registry; and (4) birth certificates. Glutathione-S-transferase isozymes GSTM1 and GSTT1 code for enzymes that promote detoxification and excretion from cells of some xenobiotics. The homozygous null genotypes are prevalent and associated with susceptibility to certain cancers. Methods: Population-based cases of acute lymphocytic leukemia (ALL) diagnosed before the age of six and born in [U.S. state] during 1979–1985 were identified through the [state] Cancer Registry. A control for each case was randomly selected from births on the same date. We determined residence address and demographic characteristics from birth certificates. Water contaminant concentrations for trihalomethanes (THMs) and three common solvents were assigned using birth date and residence. We retrieved DBS specimens, isolated DNA, amplified genes for GSTT1 and GSTM1 and assayed alleles on agarose gels. In accordance with the agency's Institutional Review Board, identifiers were irrevocably deleted, thereby removing the ability to reconstruct genetic information for any individual. Results: Only DNA from the years 1983–1985 could be amplified, yielding 152 analyzable subjects. Only one genotype, double null homozygosity for both GSTM1 and GSTT1, was associated with case status (OR 3.1 95% C.I. 0.92–10.0). High estimated concentrations of THMs or solvents were not statistically associated with pediatric leukemia in our sample, but addition of the water contaminant variables to the logistic regression model increased the OR for the double null genotype to 3.9 (95% C.I. 1.02–14.9). We did not detect an elevated interaction odds ratio of null susceptibility genes with water contaminants among the cases. Conclusion: The association of pediatric ALL with the absence of enzymes that detoxify and excrete some carcinogens is consistent with the hypothesis that exposures to such carcinogens may be risk factors for this disease. Our study also illustrates the potential use of newborn blood spots in investigating genetic susceptibility genes for pediatric diseases.