Paclitaxel, Carboplatin, and Cetuximab Induction Chemotherapy for Locally Advanced Head and Neck Squamous Cell Carcinoma: A Retrospective Evaluation of Treatment Outcomes and Safety

Background/Purpose Definitive chemoradiotherapy (CRT) for locally advanced head and neck squamous cell carcinoma (HNSCC) has limitations in advanced disease. Paclitaxel, carboplatin, and cetuximab (PCE) offers a potentially safer induction chemotherapy alternative. This study evaluated the feasibility and outcomes of PCE induction followed by CRT compared with CRT alone among patients with Stage IV HNSCC who received definitive CRT. Materials and Methods This retrospective study included 128 patients with Stage IV HNSCC without distant metastases treated at a single institution (PCE group, n=52; CRT group, n=76). Survival analyses were restricted to the 127 patients who initiated definitive CRT, with time measured from CRT initiation to reduce immortal time bias; CRT completion was analyzed as an outcome, not an eligibility criterion. An adjusted Cox model accounted for baseline imbalances in T classification and Stage. Results Median follow-up was 17.0 months (PCE) and 65.0 months (CRT). The PCE group had more advanced disease (T3-4: 90.4% vs. 65.8%; Stage IVB: 34.6% vs. 15.8%). PCE achieved an 82.7% response rate, CRT completion in 51 of 52 patients (98.1%), and significantly higher cisplatin ≥200 mg/m² achievement (86.5% vs. 59.2%, p=0.002). PFS was longer with PCE in both unadjusted (HR: 0.515, 95% CI:0.271–0.980, p=0.043) and adjusted analysis (adjusted HR: 0.477, 95% CI:0.247–0.924, p=0.028). OS did not differ significantly (adjusted HR: 0.553, p=0.147), and its interpretation is limited by the shorter follow-up in the PCE group. Exploratory inflammatory and nutritional markers showed no reliable prognostic association. Conclusions PCE induction was feasible, with a high rate of CRT completion and acceptable tolerability. PFS was longer in the PCE group after limited adjustment, whereas OS did not differ significantly. Given the retrospective, nonrandomized design, residual confounding, and shorter follow-up in the PCE group, these findings are hypothesis-generating and require confirmation in larger, prospective studies.