BMI at diagnosis and adverse outcomes among men with malignant testicular germ cell tumors.

Abstract

400 Background: Previous studies evaluating the relationship between measures of anthropometry, such as body mass index (BMI), and risk of testicular cancer have been conflicting, with most finding null or inverse associations. Methods: We conducted a retrospective review of 960 germ cell tumors among patients treated at the Dana-Farber Cancer Institute (DFCI) between 1997 and 2012 with information on BMI. BMI at diagnosis, tumor characteristics, IGCCC classification, treatment, relapse and cause of death was ascertained from electronic medical record review. We divided the men and conducted the analyses separately for two groups: 1) men who presented as clinical stage 1 (CS1), and 2) men who presented with metastases or were CS1 at diagnosis and developed metastases (lifemets). We conducted logistic regression to evaluate the association between BMI and characteristics at diagnosis, and Cox proportional hazards regression to investigate the risk of relapse and testicular cancer death. Results: 65% of men (n=628) were overweight or obese at diagnosis. We did not find an association between BMI and tumor characteristics at baseline, such as histology, lymphovascular invasion, or tumor size for either group. Among CS1 patients, men who were overweight or obese had a non-statistically significant reduced risk of relapse compared to men who were normal weight (HR: 0.83, 95% CI: 0.53-1.30). Among the lifemets group, 16% of men relapsed (n=83) and 8% died of testicular cancer (n=43). Although not statistically significant, men who were overweight or obese were less likely to present with intermediate (OR = 0.77, 95 CI: 0.39-1.52) and poor (OR = 0.66, 95% CI: 0.36-1.20) rather than good risk disease when compared to men with normal BMI. After adjusted for confounding variables, BMI was not associated with risk of relapse (HR: 0.94, 95% CI: 0.59-1.51) or risk of testicular cancer death (HR: 1.00, 95% CI: 0.54-1.84) among the lifemets group. Conclusions: We did not find an association between BMI at diagnosis and tumor characteristics at baseline, or risk of relapse or testicular cancer death among men with testicular cancer.