Ching-Ping Wang

Background/Objectives: The prognostic significance of pre-existing diabetes mellitus (DM) in surgically treated oral cavity cancer remains uncertain. We evaluated the association between recorded pre-existing DM and 5-year overall survival. Methods: Adults with a first recorded oral cavity cancer diagnosis who underwent cancer-directed surgery within 6 months were identified in the TriNetX Research Network. The recorded DM was defined by ICD-10-CM codes E08–E13 before cancer diagnosis. Survival was evaluated from cancer diagnosis and, in a complementary analysis addressing guarantee-time bias, from the first qualifying surgery. Multivariable Cox regression and 1:1 propensity score matching were applied. Results: The diagnosis-indexed cohort included 2304 patients with recorded DM and 25,195 without recorded DM. DM was not associated with mortality in the unmatched analysis (HR 1.05, 95% CI 0.953–1.146), after multivariable adjustment (adjusted HR 1.081, 95% CI 0.977–1.196), or after matching (2242 patients per group; HR 0.997, 95% CI 0.878–1.131). In the surgery-indexed matched analysis (2298 patients per group), recorded DM was associated with a lower observed mortality hazard (HR 0.83, 95% CI 0.734–0.934). An E11-only sensitivity analysis yielded consistent findings. Conclusions: Recorded pre-existing DM was not associated with inferior 5-year overall survival. DM alone should not justify withholding or delaying surgery in otherwise appropriate candidates, although metabolic and systemic comorbidities should be optimized. The lower surgery-indexed mortality hazard likely reflects selection and residual confounding and should not be interpreted as a protective effect of DM.