Clinical implications of hyperprogression with immune checkpoint inhibitors in patients with head and neck squamous cell carcinoma (HNSCC).
6034 Background: Hyperprogressive disease (HPD) refers to paradoxical acceleration of tumor growth kinetics (TGK) after initiation of treatment with anti- PD-1/PD- L1 agents and has been reported across tumor types in 4-29% of patients using different definitions. Preliminary data suggest that HPD might affect response to subsequent therapies. Methods: We compared TGK prior and TGK upon immunotherapy (IO) in 62 patients (pts) with recurrent/metastatic (R/M) HNSCC treated with PD-1/PD-L1 inhibitors. The TGK ratio (TGKR, ratio of tumor growth velocity before and upon treatment) was calculated. The first imaging assessment was performed 3 months (mo) after IO initiation. HPD was defined as 1. Radiological HPD (TGKR≥2) or 2. Clinical HPD (Disease-related rapid clinical deterioration post IO). Results: After median follow-up of 12.3 mo (range, 0.4-28.1), 43 pts progressed and 38 died. Median PFS was 2.8 mo (95%CI, 2.2-3.4) and median OS 8.6 mo (95%CI, 4.2-12.9). HPD was observed in 16 pts (25.8%), while 15 pts had early PD (Time to Treatment failure, TTF < 3 mo) and 31 late PD (TTF > 3mo). Among 16 pts with HPD, 11 had radiological HPD and 10 had clinical HPD. 4 pts had both clinical and radiological HPD. Pts with late PD had median OS 11.3 mo (95%CI, 9.3-13.3), those with early PD 5.2 mo (95%CI, 3.1-7.3 months) and those with HPD 5.1 mo (95%CI, 4.4-5.9) (p < 0.005). Regarding post-progression OS, pts with late PD had median 11.3 mo (95%CI 0-22.8), those with early PD 2.5 mo (95%CI 0.6-4.4) and those with HPD 4.2 mo (95%CI 1.7-6.7) (p = 0.001). Pts with HPD had a trend for longer median post-progression OS compared to pts with early PD (p = 0.121). Median PFS with chemotherapy after immunotherapy failure was 3.0 mo (95%CI 2.4-3.6) for pts with late PD, 2.1 mo (95%CI 0.9-3.4) for pts with early PD and 6.1 mo (95%CI 3.0-9.3) for those with HPD (p = 0.040). HPD was associated with longer median PFS with chemotherapy compared to pts with early PD (p = 0.016), while the difference in median PFS with chemotherapy between pts with HPD and late progressors was non-statistically significant (p = 0.260). Conclusions: Radiological or clinical HPD was observed in 25.8% of patients with R/M HNSCC treated with IO. Early progression to immunotherapy is an important predictor of short survival, while HPD was associated with improved PFS to subsequent chemotherapy.