What does this research mean for the field?

Achieving a lymph node yield of ≥20 nodes in unilateral neck dissection or an average of ≥18 nodes per side in bilateral dissection after neoadjuvant immunochemotherapy for oral squamous cell carcinoma is independently associated with significantly improved survival. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.ESTABLISHES_NEW_DIRECTION.

What question did this study set out to answer?

This research aims to establish benchmarks for lymph node dissection adequacy following neoadjuvant immunotherapy in oral squamous cell carcinoma.

June 6, 2026Open Access

Stratified lymph node yield thresholds after neoadjuvant immunochemotherapy: a surgical benchmark for survival in oral squamous cell carcinoma

Key Points

This research aims to establish benchmarks for lymph node dissection adequacy following neoadjuvant immunotherapy in oral squamous cell carcinoma.
A development and validation study with 256 training patients and 199 validation patients at two hospitals.
Patients were categorized based on the extent of neck dissection: unilateral or bilateral.
Optimal lymph node dissection thresholds were identified using statistical techniques for predicting overall survival.
Optimal thresholds were determined as 20 total lymph nodes for unilateral dissection and 18 nodes per side for bilateral dissection.
Inadequate lymph node dissection was linked to worse overall survival with aHR 2.42 (Group Un) and 2.29 (Group Bi).
The findings were validated with good discrimination in the external cohort, C-index for overall survival was 0.71.

Abstract

Background The adequacy of neck dissection following neoadjuvant immunotherapy combined with chemotherapy (NICT) for locally advanced oral squamous cell carcinoma (OSCC) lacks evidence-based benchmarks. This study aimed to define and validate extent-specific lymph node dissection (LND) adequacy thresholds associated with survival outcomes. Methods In a development and validation study, the training cohort comprised 256 consecutive patients with locally advanced OSCC treated with NICT and surgery at Henan Cancer Hospital (2019-2022). An independent cohort of 199 patients from the First Affiliated Hospital of Zhengzhou University (2020-2022) served for validation. Patients were stratified by surgical extent: unilateral (Group Un) or bilateral (Group Bi) neck dissection. Optimal LND thresholds predictive of 3-year overall survival (OS) were determined using restricted cubic splines and maximally selected rank statistics. For Group Un, the threshold (LND-Un) was based on total lymph node yield; for Group Bi, the threshold (LND-Bi) was based on the average yield per side. Survival and safety outcomes were compared between patients meeting (“Adequate”) or not meeting (“Inadequate”) these thresholds. Results The optimal thresholds were 20 total lymph nodes for unilateral dissections (LND-Un) and 18 nodes per side on average for bilateral dissections (LND-Bi). In the training cohort, inadequate LND was independently associated with worse OS (Group Un: adjusted hazard ratio aHR 2.42, 95% CI 1.47-3.99, p0.001; Group Bi: aHR 2.29, 95% CI 1.20-4.37, p=0.012) and disease-free survival (DFS). Inadequate LND was also linked to a higher risk of major complications in Group Un (adjusted odds ratio 2.15, 95% CI 1.06-4.38, p=0.034). These findings were robustly validated in the external cohort, where the LND-based model demonstrated good discrimination (C-index for OS: 0.71), excellent calibration, and positive net benefit on decision curve analysis. Conclusion This study establishes and validates stratified, surgical extent-specific thresholds for lymph node dissection adequacy after NICT in OSCC. Achieving a yield of ≥20 nodes in unilateral dissection or an average of ≥18 nodes per side in bilateral dissection is independently associated with significantly improved survival and an acceptable safety profile. These benchmarks provide a tangible, evidence-based guide for surgical quality assessment and decision-making in the post-NICT setting.

Ask AI

Helpful

Bookmark

View Full Paper