Effect of glycated haemoglobin A1c on the survival of patients with oral squamous cell carcinoma and diabetes: a multicentre, retrospective, database cohort study

• Published in: The lancet oncology 2022-07, Vol.23, p.S23-S23
• Main Authors: Chao, Chun-Yuan, MD, Luo, Sheng-Dean, MD, Chen, Wei-Chih, MD, Wu, Shao-Chun, MD, Chiu, Tai-Jan, MD, Wang, Yu-Ming, MD, Yang, Yao-Hsu, PhD, Fang, Fu-Min, Prof, Li, Chung-Yi, Prof, Wu, Ching-Nung, MD
• Format: Article
• Language: English
• Subjects: Cancer; Cohort analysis; Diabetes; Diabetes mellitus; Hematology, Oncology, and Palliative Medicine; Hemoglobin; Medical prognosis; Mortality; Oral cancer; Oral carcinoma; Oral squamous cell carcinoma; Patients; Regression analysis; Squamous cell carcinoma
• ISSN: 1470-2045; 1474-5488
• DOI: 10.1016/S1470-2045(22)00422-3

AbstractBackgroundFew studies have evaluated the effect of glucose concentrations on cancer prognosis. We investigated the association between haemoglobin A1c (HbA1c) concentrations and survival in patients with oral squamous cell carcinoma. MethodsA 19-year retrospective cohort study of patients with oral squamous cell carcinoma was done using the Chang Gung Research Database to identify 7279 patients diagnosed with oral cancer between Jan 1, 2001, and June 30, 2020. A 1:2 frequency-matching ratio between patients with diabetes and patients without diabetes was used. Adjusted and unadjusted Cox's regression models were used to evaluate the relative hazards of all-cause mortality and disease-specific mortality in relation to HbA1c concentrations. Findings3600 patients were included. An unadjusted Cox's regression model indicated that diabetes and high concentrations of HbA1c were significantly prognostic of poor survival. An adjusted hazard ratio (aHR) of an HbA1c interval of at least 8% at the initial diagnosis of oral squamous cell carcinoma (with the non-diabetes group as the reference) was significantly higher for disease-specific mortality (1·25 [95% CI 1·00–1·56] to 2·24 [1·65–3·03]) than for the non-diabetes group in different regression models. Considering the effect of sustained HbA1c control, the mean aHR of HbA1c of at least 9% (reference: 7 ≤ HbA1c < 8) interval was significantly higher for all-cause mortality (1·78 [1·12–2·84] to 2·13 [1·38–3·31]) than for other intervals in 699 patients. In addition, increased hazards of all-cause mortality (2·09 [1·47–2·98] to 2·18 [1·54–3·08]) and disease-specific mortality (2·20 [1·40–3·46] to 2·41 [1·51–3·83]) were consistently observed in the highest quartiles of average real variability of HbA1c. InterpretationPoor and unstable control of HbA1c can predict risk of mortality in patients with oral squamous cell carcinoma and diabetes. FundingKaohsiung Chang Gung Memorial Hospital, Taiwan (grant numbers CFRPG8K0131 and CORPG8L0481).