Association of diabetes with colorectal cancer treatment and outcomes.

Abstract

e18112 Background: Diabetes mellitus (DM) is associated with an increased incidence of colorectal cancer, but the impact of diabetes on colorectal cancer (CRC) prognosis and treatment patterns remains unknown. As outcomes in both CRC and DM have improved, appropriate co-management of the two diseases has become a priority to help patients live longer and healthier. Methods: A systematic search of PubMed, Embase, and CINAHL databases was performed to evaluate/assess the association of diabetes with colon cancer treatment patterns and outcomes, including recurrence rates (RR), disease free survival (DFS), colon cancer-specific (CSM) and all-cause mortality (ACM). We included English language peer-reviewed studies through November 2016 following the preferred reporting items of systematic reviews and meta-analyses (PRISMA) method. We applied STROBE quality criteria and conducted a random-effects meta-analysis. An inconsistency index (I2) was used to estimate the heterogeneity between studies. Results: After reviewing 2,611 reports, 35 studies were included that described an association of DM with CRC treatment patterns or outcomes. A meta-analysis of 14 studies that assessed (ACM) among 146,129 patients demonstrated an increased risk for ACM for CRC patients with DM (OR 1.28, 95% CI 1.16-1.42, I2 = 80%, P < .00001). No statistical significance was found in comparing 4 studies that looked at CSM (OR 0.91, 95% CI .72-1.14, I2 = 89%, P = 0.42). Analysis of 4 studies that assessed DFS found lower DFS in patients with DM (OR 1.75, 95% CI 1.33-2.31, I = 3%, P < .0001). Aggregating 3 studies that evaluated RFS demonstrated no clear association of DM and RFS(OR 1.12, CI 95% 0.91-1.38, I = 64%, P = .27). No consistent treatment discrepancies were noted in 6 studies that assessed the impact of DM on various CRC treatment modalities (receipt of surgery, chemotherapy, radiation). Conclusions: CRC patients with diabetes are at a greater risk for ACM and have worse DFS compared to those without DM. Analysis was limited due to study heterogeneity. We did not find consistent evidence that CRC is treated differently in patients with DM. Further study is warranted to clarify the mechanism of poor outcomes in CRC patients with DM and to improve survival.