Indocyanine Green Fluorescence Guidance in Surgery for Locally Advanced Oral Squamous Cell Carcinoma After Neoadjuvant Immunotherapy

OBJECTIVE: To explore the feasibility of indocyanine green (ICG) near-infrared fluorescence guidance and margin-related outcomes in patients undergoing surgery for locally advanced oral squamous cell carcinoma (OSCC) after neoadjuvant immunotherapy. METHODS: This retrospective single-center cohort included 30 consecutive patients who underwent curative-intent surgery between May 2024 and March 2025. Fifteen underwent ICG-guided surgery and 15 underwent conventional surgery, according to routine clinical implementation and imaging availability. The primary outcome was patient-level positive margin status on final pathological assessment. Recorded margin-to-tumor distances, intraoperative margin findings and perioperative outcomes were assessed separately. Comparisons were exploratory and unadjusted. RESULTS: Tumor-associated fluorescence was visualized in all 15 ICG-treated patients (100%; exact 95% confidence interval [CI], 78.2%-100%). Mean recorded margin-to-tumor distance was 9.30 ± 1.21 mm with ICG guidance and 16.40 ± 2.41 mm with conventional surgery (mean difference, -7.10 mm; 95% CI, -8.55 to -5.65; p < 0.001). Corresponding medians were 9.0 and 16.0 mm, and minima were 8.0 and 13.0 mm. The operative summary describes additional intraoperative resection in two conventional-group patients following frozen-section assessment. No statistically significant between-group differences in perioperative outcomes were detected, but confidence intervals were wide. Final pathological margin assessment identified no positive margins in either group (ICG-guided 0/15, 0.0%; conventional 0/15, 0.0%; Fisher's exact p = 1.000). The two intraoperative frozen-section-positive events in the conventional group were resolved to negative final margins after same-operation resection extension. CONCLUSION: ICG fluorescence imaging was feasible in this small retrospective cohort and was associated with shorter recorded margin-to-tumor distances. These findings are hypothesis-generating and do not establish improved oncologic margin control, perioperative equivalence or functional benefit. Larger prospective studies with standardized pathological assessment and adequate follow-up are needed.