The relationship between diabetes and colorectal cancer prognosis: A meta-analysis based on the cohort studies

Abstract

Introduction: Though a meta-analysis reported the effect of diabetes on colorectal prognosis in 2013, a series of large-scale long-term cohort studies has comprehensively reported the outcome effect estimates on the relationship between diabetes and colorectal prognosis, and their results were still consistent.

Methods: We carried out an extensive search strategy in multiple databases and conducted a meta-analysis on the effect of diabetes on colorectal prognosis, based on the included 36 cohort studies, which contained 2,299,012 subjects. In order to collect more data, besides conventional methods, we used the professional software to extract survival data from the Kaplan-Meier curves, and analyzed both the 5-year survival rate and survival risk in overall survival, cancer-specific survival, cardiovascular disease-specific survival, disease-free survival, and recurrence-free survival, to comprehensively reflect the effect of diabetes on colorectal prognosis.

Results: The results found that compared to patients without diabetes, patients with diabetes will have a 5-year shorter survival in colorectal, colon and rectal cancer, with a 18%, 19% and 16% decreased in overall survival respectively. We also found similar results in cancer-specific survival, cardiovascular disease-specific survival, disease-free survival, and recurrence-free survival, but not all these results were significant. We performed the subgroup analysis and sensitivity analysis to find the source of heterogeneity. Their results were similar to the overall results.

Conclusions: Our meta-analysis suggested that diabetes had a negative effect on colorectal cancer in overall survival. More studies are still needed to confirm the relationship between diabetes and colorectal prognosis in cancer-specific survival, cardiovascular disease-specific survival, disease-free survival, and recurrence-free survival.

Fig 1. The study selection process for inclusion in our meta-analysis