Neoadjuvant immunotherapy in combination with chemotherapy in resectable locally advanced head and neck squamous cell carcinoma: Updated efficacy and safety data from a randomized phase II trial.

6091 Background: Neoadjuvant immunotherapy combined with chemotherapy is a promising strategy for resectable LAHNSCC. While most regimens employ single-target PD-1 inhibitors, dual-target agents (PD-1/CTLA-4 or PD-1/VEGF) have shown superior efficacy in recurrent/metastatic HNSCC. Building on our preliminary data (ASCO 2025) which suggested high pathological response rates, this ongoing randomized phase II trial aims to compare single- versus dual-target NAI regimens in an expanded cohort to identify the optimal treatment strategy. Methods: This is an ongoing randomized, open-label, phase II trial. Eligible pts were randomized (1:1:1) to receive 3 cycles of neoadjuvant therapy as follows: Cohort 1 will receive ivonescimab (PD-1/VEGF bispecific antibody, 10 mg/kg every 3 weeks), Cohort 2 will receive cadonilimab (PD-1/CTLA-4 bispecific antibody, 6 mg/kg every 3 weeks), and Cohort 3 will receive penpulimab (PD-1 antibody, 200 mg every 3 weeks), all in combination with cisplatin and nab-paclitaxel. After neoadjuvant treatment, Surgery was performed with the surgical margins based on pre-treatment (baseline) evaluation. Pts with pCR received adjuvant immunotherapy for up to 16 cycles. Pts without pCR received adjuvant radiotherapy or chemoradiotherapy, followed by 16 cycles of adjuvant immunotherapy. Results: Up to Dec. 2025, all 59 pts completed 3 cycles of neoadjuvant therapy and were evaluable, with a median follow-up of 12 months. The pCR rates were 60% (12/20) in Cohort 1, 42.1% (8/19) in Cohort 2, and 40% (8/20) in Cohort 3. The major pathologic response (MPR) rates were 75% (15/20), 57.9% (11/19), and 55% (11/20) in Cohort 1, 2, and 3, respectively. The ORR was 95% in Cohort 1 (CR: 8/20, PR: 11/20) and 84.2% in Cohort 2 (CR: 5/19, PR: 11/19), while Cohort 3 had an ORR of 80% (CR: 4/20, PR: 12/20). To date, pts have received a median of 6 cycles of adjuvant immunotherapy. The most common treatment-related adverse events (TRAEs) (>20%) were: leukopenia, anemia, neutropenia, thrombocytopenia, lymphocytopenia, hypothyroidism, hypertriglyceridemia, radiation dermatitis, stomatitis, vomiting, decreased appetite, and fatigue. Conclusions: Neoadjuvant dual-target (PD-1/VEGF) immunotherapy combined with chemotherapy demonstrated a higher pCR rate compared to dual-target (PD-1/CTLA-4) and single-target PD-1 regimens in resectable LAHNSCC. The treatment was well-tolerated, with all pts completing the intended neoadjuvant therapy and the observed most common TRAEs consistent with expected chemotherapy and radiotherapy profiles. Further analyses are ongoing with continued enrollment. Clinical trial information: NCT06444009 . RECIST and pathologic response. RECIST Pathologic Response CR PR SD PD pCR MPR pPR pNR (RVT=0) (0<RVT<10%) （RVT:10-49%） (RVT:≥50%) Cohort 1 (n=20) 8 11 1 12 3 2 3 Cohort 2 (n=19) 5 11 3 8 3 2 6 Cohort 3 (n=20) 4 12 4 8 3 0 9