Although diabetes mellitus (DM) is one of the risk factors associated with increased breast cancer (BC) mortality, the effects of glycaemic control on the prognosis of BC have not been thoroughly evaluated. This retrospective study aimed to evaluate the relationship between glycaemic control and BC prognosis and to determine an optimal target of glycaemic control for BC patients with diabetes. We included 2812 stage 0-3 BC women, of whom 145 were diabetic and were 2667 non-diabetic. In those with diabetes, a mean haemoglobin A1C (HbA1C) < 7% (n = 77) was defined as well-controlled diabetes, while a mean HbA1C > 9% (n = 16) was defined as poorly controlled diabetes. All of the BC populations were followed from the date on which BC was diagnosed until 31 December 2015. Cox regression analysis was performed to estimate the adjusted hazards for all-cause mortality and BC-specific mortality. After controlling for the baseline and BC-related confounders, the adjusted hazard ratio (HR) for all-cause mortality and the HR for BC-specific mortality were 3.65 (95% confidence interval [95% CI] 1.13-11.82) and 8.37 (95% CI 1.90-36.91), respectively, for poorly controlled diabetic women and non-DM women. However, for the diabetic women with good glycaemic control, the HRs of all-cause mortality and BC-specific mortality were not significantly different (HR 0.91, 95% CI 0.42-1.01; HR 0.77, 95% CI 0.18-3.32, respectively) from those for both mortalities in non-DM patients. For moderate controlled diabetic women, the HRs for all-cause mortality and BC-specific mortality were 1.95 (95% CI 0.89-4.27) and 3.55 (95% CI 1.369-9.30), respectively. This pilot and retrospective cohort study reveals a relationship between glycaemic control and BC prognosis in diabetic women. In addition, well-controlled HbA1C, with maintained mean HbA1C values under 7%, may be associated with a better progression outcome of BC.
Keywords: Breast cancer; Diabetes mellitus; Haemoglobin A1C; Mortality.