Metabolic enzyme polymorphisms and susceptibility to acute leukemia in adults.

Genetic approaches to understanding the etiology of the acute leukemias are beginning to deliver meaningful insights. Polymorphic variants in xenobiotic metabolizer loci were a natural starting point to study the relevance of these changes. The finding that glutathione S-transferase (GST) T1 null variants increase leukemia risk has implicated oxidative stress in hematopoietic stem cells as an important etiological factor in acute myeloid leukemia (AML). The importance of these enzyme systems in handling specific substrates has also been confirmed by the finding of an increased risk of therapy-related leukemia in individuals with underactive variants of GSTP1 who have been exposed to a chemotherapeutic agent metabolized by this enzyme. Benzene is a well-recognized leukemogen, and genetic variants in its metabolic pathway can modulate the risk of leukemia following exposure. In particular, underactive variants of the NAD(P)H:quinone oxidoreductase 1 gene (NQO1) seem to increase the risk of AML. Other enzymes within the pathway are proving more difficult to study because of the absence of variants that significantly affect the biological activity of the enzyme under study. No effect of the myeloperoxidase (MPO) gene variants in altering the risk of AML has been seen in our studies. Another pathway recently shown to be important in determining leukemia risk is folic acid metabolism, particularly important in predisposition to acute lymphocytic leukemia (ALL). Polymorphic variants of the methylenetetrahydrofolate reductase gene (MTHFR) which impair its activity have been shown to be associated with a protective effect. This is thought to be due to an increased availability of nucleotide precursors for incorporation into DNA. This finding implicates misincorporation of uracil into DNA as an important mechanism of leukemic change in lymphoid precursors. Future studies will extend these observations but will require biological material collected from large well-controlled epidemiological studies. The technological challenges imposed by the high throughput of samples required by these studies are currently being addressed.