Changes in High-Risk Pathologic Features across Sequential Resections for Recurrent Oral Squamous Cell Carcinoma: A Longitudinal Cohort Study
Background: This study evaluated whether high-risk pathologic features escalate across sequential resections for recurrent oral squamous cell carcinoma (OSCC), and whether change is associated with transition number. Methods: In this retrospective, single-center study, 14 of 160 patients undergoing head and neck cancer surgery between January 2022 and August 2026 had repeat resection for recurrent OSCC; all their resections (n = 33; index resections from 2017 onward) were reviewed, yielding 19 surgery-to-surgery transitions. The index and first subsequent resections (n = 14) were compared for seven pathologic parameters—histologic differentiation, pathologic T category, lymphovascular invasion, perineural invasion, worst pattern of invasion type 5, depth of invasion (DOI), and cortical bone invasion—using the Wilcoxon signed-rank or McNemar's exact test with Holm-Bonferroni adjustment. All 19 transitions were analyzed via linear regression with cluster-robust standard errors (patient as cluster) to assess associations between transition number and direction of change, with sensitivity analyses (single-transition subset, leave-one-cluster-out). Results: No parameter changed significantly between index and first subsequent resection (all adjusted p = 1.000). Across all 19 transitions, DOI showed a significant inverse association with transition number (β = −21.6 mm/transition, adjusted p = 0.028), remaining significant in all leave-one-cluster-out iterations. No other parameter was significantly associated with transition number. Conclusions: This preliminary cohort showed no consistent escalation of high-risk pathologic features across sequential resections for recurrent OSCC, except for an exploratory DOI decrease possibly reflecting earlier detection rather than biological indolence. Larger multi-institutional cohorts with molecular clonality assessment are needed to clarify how pathologic risk evolves across sequential resections.