Diabetes mellitus (DM), serum glucose, and outcome in gastroesophageal (GE) cancer

Abstract

e15661 Background: Diabetes mellitus (DM) and impaired glucose tolerance (IGT) are frequent comorbidities in cancer patients. Their diagnosis is predictive of early recurrence following resection of colorectal cancer and of increased treatment related toxicity in localized breast cancer. Impaired glucose tolerance is a risk factor for the development of GE cancer. As nutrition and performance status are key determinants of outcome, we hypothesized that DM would influence treatment outcome in GE cancer. Our primary objective was to analyze the role of serum glucose, DM, serum albumin and BMI on treatment outcome and toxicity amongst a cohort of uniformly treated patients with GE cancer. Methods: 211 patients with GE cancer enrolled on 3 localized (2 esophagus, 1 gastric) and 4 metastatic consecutive prospective clinical trials (3 gastric, 1 esophagus) were examined. The studies involved treatment with chemotherapy combinations of taxanes, platinum, irinotecan, 5-fluorouracil, bevacizumab and cetuximab. Diagnosis of DM was based on clinical data, random glucose or fasting-glucose levels. Study parameters were tested for their association to time to treatment failure (TTF), overall survival (OS) and toxicity using the X2 test. Results: 132 metastatic pts were included: median age 57(range 32–76), M/F 96/36, median BMI 25.3 (16.2–40.1), 19 diabetic / 113 non-diabetic; overall survival was 10.8 months. 79 patients had localized disease: median age 61(32–80), M/F=53/26, median BMI 25.8 (range 16.1–38.5), 10 diabetic/69 non-diabetic; overall survival was 23.2 months. There was no statistically significant association of random glucose (p=0.06), DM (p=0.9), or BMI(p=0.2) with survival or with treatment related toxicity. Low serum albumin was associated with reduced survival, hazard ratio 2.12 (95% CI1.5–3.2), p<0.001. Conclusions: In GE cancer patients on protocol therapy, we confirm the negative prognostic significance of serum albumin on patient survival. This is remarkable considering all patients were protocol eligible with adequate organ function and performance score. We did not observe a clinically relevant impact of IGT or DM on treatment outcome or toxicity. No significant financial relationships to disclose.