Real-world efficacy of cetuximab after immune checkpoint inhibitor and platin failure in recurrent/metastatic head and neck squamous cell carcinoma (R/M HNSCC): A 10-year retrospective study in 124 patients.

e18025 Background: In R/M HNSCC, cetuximab monotherapy historically yields an objective response rate (ORR) of 13%. Emerging data, including the INTERLINK-1 trial with 20% of ORR, suggest an IO-priming effect where prior immunotherapy (IO) enhances subsequent anti-EGFR sensitivity. We evaluated the real-world efficacy and prognostic factors of cetuximab following IO failure in a large HNSCC cohort. Methods: Since 2015, we retrospectively analyzed all medical records of patient treated with cetuximab in a single institution, the Centre Leon Berard in France, who were in failure after IO (pembrolizumab, nivolumab, ipilimumab, atezolizumab, tremelimumab, lirimumab, FS-118 or durvalumab) and platin for R/M HNSCC. Off-label panitumumab after hypersensitivity to cetuximab, or combination with monalizumab (based on the negative results of the INTERLINK-1 trial) were allowed. Patients were naive of anti-EGFR for R/M HNSCC ORR was analyzed by RECIST 1.1. PFS and OS were calculated from the first administration of cetuximab to the progression or death. Results: Between 2015 and 2025, 124 patients met the criteria. The median age was 65 years, with a majority of males (76.6%) and former smokers (74.2%). At the start of cetuximab, 47.6% of patients had an ECOG-PS ≥ 2. Cetuximab treatment was 2nd-line (45), 3rd-line (73), or 4th-line (6). Overall median PFS was 3.7 months (95% CI: 2.5-4.4) and median OS 7.1 months (95% CI: 6.0-9.3). Median OS was significantly better for ECOG-PS 0-1 (8.9 months; 95% CI: 7.0-12.1) compared to PS 2-3 (5.9 months; 95% CI: 3.6-8.9). At two months, ORR was 32.3% (95% CI: 24.0-40.5) and DCR 59.7% (51.0-68.3), including 5 CR (4.0%) and 35 PR (28.2%). Poor outcomes were linked to prior cetuximab-radiotherapy potentiation and pure metastatic disease. Prior taxane exposure or response to IO were favorable markers. Subgroup analyses do not appear to indicate any effect of the interval between cetuximab and IO. Grade ≥ 3 toxicities occurred in 16 patients; 5 severe allergies required switching to panitumumab. Conclusions: This study reports a high ORR of 32.3%after use of IO, significantly exceeding historical data. No impact of the time interval between IO and cetuximab was observed and ancillary studies are needed to undestand this higher ORR. That supports ongoing trials exploring IO and anti-EGFR combinations. Prior taxane use and IO response emerged as favorable prognostic markers. Median overal survival of patients with PS0-1 is 8.9 months and is confirmed as a standard treatment after platin and IO.