Abstract
IMPORTANCE: Numerous phase 2 trials have evaluated the efficacy of neoadjuvant immune checkpoint inhibition (ICI) for mucosal head and neck squamous cell carcinoma (HNSCC), using some degree of pathologic treatment response as a primary or secondary end point. However, whether pathologic treatment response is a meaningful surrogate end point for survival has yet to be determined. OBJECTIVE: To systematically assess the association between pathologic treatment response and overall survival (OS) and disease-free survival (DFS) after neoadjuvant ICI. DATA SOURCES: A systematic search of the PubMed, OVID Medline, Embase, CINAHL, and Cochrane databases was performed from January 1, 2000, through May 31, 2025. STUDY SELECTION: Peer-reviewed studies investigating neoadjuvant ICI for the treatment of mucosal HNSCC in patients 18 years and older were identified. Full-length English-language articles that presented pathologic treatment response and survival data (OS and/or DFS) and any association between the 2 were included. DATA EXTRACTION AND SYNTHESIS: Three blinded reviewers independently extracted study characteristics, pathologic treatment response data, and survival data, including hazard ratios (HRs) and CIs when available, according to PRISMA guideline. Data were compiled for statistical analysis to calculate DFS, OS, and HRs using a random-effects model. The I2 index was used to report data heterogeneity. MAIN OUTCOMES AND MEASURES: HRs for the association of pathologic treatment response with DFS and OS. RESULTS: Eleven trials involving 451 patients met inclusion criteria, with 368 patients included in this meta-analysis. Nine nonrandomized and 2 randomized studies were included, including 7 cohort studies, 2 randomized clinical trials, and 2 retrospective cohort studies, each with a different neoadjuvant ICI regimen. Pooled analysis demonstrated that overall (primary tumor plus lymph node) partial pathologic response (PPR; ≤50% residual viable tumor; HR, 0.53; 95% CI, 0.28-0.97; I2 = 2.1%) and major pathologic response (MPR; ≤10% residual viable tumor; HR, 0.34; 95% CI, 0.12-0.93; I2 = 0.0%) were both associated with improved DFS up to 2 years. PPR and MPR were not associated with improved OS. Nine of 11 studies were at low risk of bias. CONCLUSIONS AND RELEVANCE: Study findings suggest that overall PPR and MPR are associated with improved DFS. These data provide additional support for the potential use of pathologic treatment response as a surrogate for DFS after neoadjuvant ICI in resectable mucosal HNSCC.