Epidemiology of adult acute myeloid leukemia: Impact of exposures on clinical phenotypes and outcomes after therapy

Background: An increased risk of adult myeloid leukemia (AML) has recently been associated with lifestyle and environmental exposures, including obesity, smoking, some over the counter medications, and rural/farm habitats in case control studies. The association of these exposures with AML cytogenetic categories, outcomes after therapy, and overall survival is unknown. Methods: Relevant exposures were evaluated in a cohort of 295 consecutive AML patients diagnosed and treated at Mayo Clinic in Florida and Arizona. Standard cytogenetic risk categories were applied and reviewed in a central cytogenetic laboratory. The association of epidemiologic exposures with cytogenetic risk, complete remission after therapy, and overall survival was evaluated using logistic and Cox regression models. Results: A significant association between obesity and intermediate-abnormal cytogenetics was identified (OR: 1.94, P=0.025). Similarly, those with secondary AML were more likely to have poor risk (OR: 2.55, P< 0.001) and less likely to have intermediate normal (OR: 0.48, P=0.003) cytogenetics. In multivariate analysis, overall survival was improved for patients ≥60 years receiving intensive (RR: 0.21, P< 0.001) and non-intensive therapy (RR: 0.40, P< 0.001 compared to no treatment, and was lower for users of tobacco (RR 1.39, P=0.032), and those with poor risk cytogenetics (RR: 3.96, P=0.002) or poor performance status (RR: 1.69, P< 0.001). Furthermore, an association between statin use at the time of diagnosis (OR: 2.89, P=0.016) and increased complete remission after intensive chemotherapy was identified, while prior solid organ transplantation was associated with significantly lower complete remission rate after therapy (OR: 0.10, P=0.035). Conclusion: Our results provide evidence that specific epidemiologic exposures, including obesity, are significantly associated with unique AML cytogenetic risk categories and response to therapy. This supports a link between patient lifestyles, clinical exposures, and leukemogenesis.