Abstract
Aim: Conduct a systematic review with meta-analysis to determine the association between incident chemotherapy-induced neutropenia (CIN) and either diabetes mellitus (DM) or hyperglycemia in patients with cancer.
Methods: Observational studies in cancer patients of any age receiving chemotherapy and having diabetes or hyperglycemia either during or before chemotherapy induction were included. Studies were retrieved by searching four databases (PubMed, EBSCO, ProQuest, and Cochrane) and cross-referencing. The metric for combining studies was the odds ratio (OR). Results were pooled using a random-effects model, while heterogeneity and inconsistency were assessed using the Q and I-2 statistic, respectively. Potential small-study effects were assessed using the funnel plot.
Results: Ten studies met the criteria for inclusion. Overall, the odds of having CIN were 32% higher among cancer patients with either DM or hyperglycemia compared with those without DM or hyperglycemia (OR = 1.32, 95% CI, 1.06-1.64). Statistically significant heterogeneity and inconsistency were found (Q = 33.15, p < 0.05, I-2 = 72.9%). Funnel plot asymmetry reflecting potential small-study effects was observed.
Conclusions: Diabetes mellitus and hyperglycemia may be associated with an increased rislc for CIN among cancer patients. However, additional well-designed studies are needed before any final and definitive recommendations can be made. (C) 2017 Elsevier Inc. All rights reserved.