Diabetes, obesity, and survival in a large cohort of early-stage breast cancer patients.

Abstract

1503 Background: Diabetes and obesity have risen to epidemic proportions globally. Both conditions, which share overlapping etiologies, are associated with worse breast cancer (BC) outcomes in some but not all studies. To address these inconsistencies, we examined the relationship between diabetes, body mass index (BMI), and survival in a large cohort of early-stage BC patients. Methods: We retrospectively studied patients treated at M.D. Anderson from 1996-2005. For comparison, the patients were divided into diabetic (DM) and non-diabetic (ND) groups and into 3 BMI classes: normal or underweight (N, BMI < 25), overweight (Ov, BMI 25-30), and obese (Ob, BMI ≥ 30). Recurrence-free survival (RFS) and overall survival (OS) were estimated by the Kaplan-Meier product and compared among groups via the log-rank test. Multivariate analysis was conducted via Cox proportional hazards models. Results: There were 6,106 patients. Median age was 52 (range 19-91). 93% had stage I-II disease, 7% stage III. 77% were hormone-receptor (HR) positive, 18% HER2-positive. 476 patients had DM (7.8%). Patients distributed into BMI classes as follows: N 40%, Ov 31%, Ob 29%. By chi-squared tests, diabetes was associated with increasing age, stage, and BMI; and higher BMI was associated with increasing age, stage, grade, and diabetes. At a median follow-up of 5.4 years, there were 488 recurrences and 921 deaths. RFS and OS were significantly worse in DM vs. ND patients (p < 0.001 and p < 0.001, respectively), and in the high vs. low BMI groups (p < 0.001 and p > 0.002, respectively). After adjustment for DM status, BMI, age, stage, grade, and HR status, DM patients had decreased RFS (hazard ratio [hr], 1.22; 95% CI > 0.99,1.50; p > 0.065) and OS (hr, 1.40; 1.11,1.78; p > 0.005). In the same model, Ob vs N patients had decreased RFS (hr, 1.15; 0.99,1.33; p > 0.065) and OS (hr, 1.26; 1.06,1.50; p > 0.010). Conclusions: In this large cohort of early-stage BC patients, diabetes and obesity each correlated with a significant decrease in OS, as well as a borderline significant decrease in RFS, after adjustment for DM status, BMI, age, stage, grade, and HR status. Further studies to discover the reasons for these poorer outcomes in diabetic and obese breast cancer patients are underway. No significant financial relationships to disclose.