Abstract
BACKGROUND: Diabetes is associated with impaired breast cancer prognosis; however, the effectiveness of glycosylated hemoglobin (HbA(1c)) as a prognostic biomarker in breast cancer remains uncertain, especially for patients without diabetes. We aimed to determine whether elevated HbA(1c) is associated with a worse prognosis in breast cancer patients without known diabetes. METHODS: The study population comprised women with primary invasive stage I-III breast cancer between 2010 and 2020 surgically treated at Aarhus University Hospital, Denmark, without a diabetes diagnosis at baseline. We assessed HbA(1c) at breast cancer diagnosis as a categorical (quartiles; HbA(1c)-Q1 = 21-33 mmol/mol, HbA(1c)-Q2 = 34-36 mmol/mol, HbA(1c)-Q3 = 37-38 mmol/mol, HbA(1c)-Q4 = ≥ 39 mmol/mol) and log2-transformed continuous variable. Follow-up began at the date of primary breast cancer surgery and continued until the first occurrence of either a new breast cancer event (loco-regional or distant recurrence, or contralateral breast cancer), new primary cancer other than breast cancer, death, emigration, or end-of-follow-up (November 15th, 2021). Cox regression models estimated crude and adjusted hazard ratios and associated 95% confidence intervals (95% CIs) of a new breast cancer event and all-cause mortality, adjusting for patient characteristics based on a directed acyclic graph. The lowest HbA(1c) quartile (HbA(1c)-Q1) was used as reference. RESULTS: In total, 2514 women (median age 62 years) were included. During median 5.6 years follow-up for new breast cancer events, 230 (9.1%) events occurred. An escalating risk of new breast cancer events was observed with increasing HbA(1c) quartiles (adjusted hazard ratios, HbA(1c)-Q2: 1.09 [95% CI = 0.75-1.60]; HbA(1c)-Q3: 1.35 [95% CI = 0.88-2.07]; HbA(1c)-Q4: 1.69 [95% CI = 1.13-2.54]) compared to HbA(1c)-Q1. During median 6.0 years follow-up for all-cause mortality, 267 deaths (10.6%) occurred. No apparent association was evident between increasing HbA(1c) quartiles and all-cause mortality (adjusted hazard ratios, HbA(1c)-Q2: 0.75 [95% CI = 0.52-1.07]; HbA(1c)-Q3: 0.82 [95% CI = 0.55-1.21]; HbA(1c)-Q4: 1.06 [95% CI = 0.74-1.53]). Similarly, a log2(HbA(1c)) increase was associated with an increased risk of new breast cancer events, but not all-cause mortality. CONCLUSIONS: For women with primary breast cancer and no known diagnosis of diabetes, higher levels of HbA(1c) were associated with an increased risk of new breast cancer events, but not all-cause mortality. HbA(1c) may serve as a prognostic metabolic biomarker for breast cancer patients without diabetes.