Neoadjuvant PD-1–based immunotherapy plus chemotherapy produced pCRs across three cohorts in resectable locally advanced HNSCC.
Neoadjuvant ivonescimab, a PD-1/VEGF bispecific antibody, combined with cisplatin and nab-paclitaxel produced a pathologic complete response (pCR) rate of 60% in patients with resectable locally advanced head and neck squamous cell carcinoma (LAHNSCC), outperforming two comparator regimens built on cadonilimab and penpulimab, according to updated data from a randomized, open-label phase 2 trial (NCT06444009) presented in a poster session during the
Across the 3 cohorts (N = 59), a pCR was achieved in 12 of 20 patients (60%) treated with ivonescimab-based therapy (cohort 1), 8 of 19 patients (42%) treated with cadonilimab-based therapy (cohort 2), and 8 of 20 patients (40%) treated with penpulimab-based therapy (cohort 3). Major pathologic response (MPR) rates were similar across arms, at 15%, 16%, and 15%, respectively.
Patients with resectable LAHNSCC who were at least 18 years old and had an ECOG performance status of 0 to 1 were randomly assigned 1:1:1 to 3 cycles of neoadjuvant therapy with cisplatin and nab-paclitaxel plus 1 of 3 immunotherapy backbones: ivonescimab at 10 mg/kg on day 1 every 3 weeks; cadonilimab at 6 mg/kg on day 1 every 3 weeks; or penpulimab at 200 mg on day 1 every 3 weeks.
Patients achieving pCR after neoadjuvant therapy and surgery were exempted from radiotherapy; those without pCR received adjuvant chemoradiotherapy (high-risk features) or radiotherapy alone (medium-risk features). All patients then continued their assigned single-agent immunotherapy as maintenance for up to 16 cycles.
The primary end point was pCR; secondary end points included MPR, objective response rate (ORR), event-free survival, and overall survival.
Baseline characteristics were balanced across cohorts, with a median patient age of 57.4 years (range, 34-71) and a male predominance (75%-90% across arms). Most patients presented with AJCC stage IV disease (65%-80%) and clinical N2 nodal status (63%-85%). The oral cavity was the most common primary tumor site (55%-68%), and a majority of patients had PD-L1 combined positive score (CPS) of 10 or higher (53%-60%).
By RECIST 1.1 criteria, ORRs were high across all 3 arms, with complete response rates of 40% (cohort 1), 26% (cohort 2), and 20% (cohort 3), and partial response rates of 55%, 58%, and 60%, respectively; no patient in any cohort had progressive disease. Pathologic non-response was more frequent in the single-target penpulimab arm (40%) than in either bispecific-antibody arm (15% with ivonescimab, 31% with cadonilimab). One patient in cohort 3 died of influenza A infection while awaiting surgery and was not evaluable for pathologic response; investigators considered the death unrelated to study treatment.
Treatment-related adverse effects (AEs) occurring in at least 20% of patients in any cohort included cytopenias (neutrophil count decreased, anemia, white blood cell count decreased, platelet count decreased, lymphocyte count decreased), hepatic enzyme elevations (alanine and aspartate aminotransferase levels increased), and metabolic abnormalities (hypercholesterolemia, hypertriglyceridemia). Immune-related AEs including hypothyroidism (16%-35%) and pneumonitis (10%-21%) were reported across all 3 cohorts. Investigators characterized the safety profile as consistent with expected chemotherapy- and radiotherapy-associated toxicity and as manageable with continued follow-up.
The study authors concluded that neoadjuvant single- or dual-target immunotherapy combined with chemotherapy demonstrated promising antitumor activity in resectable LAHNSCC, with the PD-1/VEGF dual-target regimen achieving the highest pCR rate among the 3 cohorts, though longer-term efficacy and the optimal neoadjuvant regimen require validation with continued enrollment and extended follow-up.
