497

Acute myeloid leukemia (AML) is the more common form of leukemia and the most common cause of leukemia death. Although 50 to 75 percent of adults with AML achieve compete remission (CR) with chemotherapy, the majority of patients will die either of relapse or of complications associated with the therapy. Therefore, more specific and less toxic therapy for AML patients is needed. Glutathione S-transferases (GSTs) are polymorphic enzymes that catalyze detoxification of some chemotherapeutic metabolites and neutralization of reactive oxygen species generated by therapeutic regimens. Several studies have reported differences in survival for cancer patients who have null or variant genotypes for GSTM1, GSTT1 or GSTA1 enzymes. We evaluated the role of genetic polymorphisms in GSTM1, GSTT1 and GSTA1 on occurrence of toxicity secondary to chemotherapy and on survival among 200 patients treated with standard dose cytosine arabinoside and daunorubicin in the Southwest Oncology Group (SWOG) 9031 and 9333 clinical trials. GSTM1 and GSTT1 genotypes were determined by multiplex polymerase chain reaction (PCR) and GSTA1 was determined by MALDI-TOF mass spectrometry. Logistic regression analysis was used to explore if the GST polymorphisms were associated with toxicity and Kaplan-Meier methods and Cox proportional hazards models were used to evaluate survival in relation to genotype. Logistic regression analysis revealed no significant associations between GST genotypes and toxicity in any organ group. GSTM1 and GSTT1 gene deletions were not associated with risk of death (GSTM1 present: hazard rate [HR]=1.08, 95% confidence interval [CI]=0.80-1.47; GSTT1 present: HR=0.92, 95% CI=0.64-1.32). There was a non-significant trend for decreased risk of death for patients with the GSTA1 low activity *B/*B genotype compared to the common *A/*A genotype (GSTA1*B/*B: HR=0.84, 95% CI= 0.61-1.15), although the confidence interval included unity. Adjustment for co-variates such as cytogenetic group, race, gender, previous cancer, FAB, WBC count and age at diagnosis did not significantly alter the risk estimates. Our results indicate that the GSTs do not play a significant role in treatment outcomes in adult patients with AML. It is possible that other polymorphic genes involved in drug metabolism and DNA repair may play a role in treatment outcome, and further research into this area is warranted.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]