PurposeHere we assessed associations between null mutations in glutathione-S-transferase (GST)T1 and GSTM1 genes, and the rs1695 polymorphism in GSTP1 (Ile105Val), and risk of breast cancer-specific (n=45) and all-cause (n=99) mortality in a multiethnic, prospective cohort of 533 women diagnosed with stage I-IIIA breast cancer in 1995–1999, enrolled in the Health, Eating, Activity, and Lifestyle (HEAL) Study.MethodsWe measured the presence of the null mutation in GSTT1 and GSTM1, and the rs1695 polymorphism in GSTP1 by polymerase chain reaction. We assessed associations between breast-cancer specific and all-cause mortality using Cox proportional hazards models.ResultsParticipants with ER-negative tumors were more likely to be GSTT1 null (χ2=4.52; P=0.03), and African American women were more likely to be GSTM1 null (χ2=34.36; P<0.0001). Neither GSTM1 nor GSTT1 null mutations were associated with breast cancer-specific or all-cause mortality. In a model adjusted for body mass index, race/ethnicity, tumor stage and treatment received at diagnosis, the variant Val allele of rs1695 was associated with increased risk of all-cause (HR=1.81, 95% CI 1.16-2.82, P=0.008), but not breast cancer-specific mortality. The GSTT1 null mutation was associated with significantly higher levels of C-reactive protein.ConclusionsGSTM1 and GSTT1 null genotypes had no effect on outcome; however the variant allele of rs1695 appears to confer increased risk for all-cause mortality in breast-cancer survivors.Given the limited sample size of most studies examining associations between GST polymorphisms with breast cancer survival, and the lack of women undergoing more contemporary treatment protocols (treated prior to 1999), it may be helpful to re-examine this issue among larger samples of women diagnosed after the late 1990s, who all received some form of chemotherapy or radiotherapy.