Comprehensive analysis on reactive cutaneous capillary endothelial proliferation following camrelizumab-based therapy in patients with solid tumors: A large-scale pooled analysis of nine phase 2 or phase 3 registration trials.

Abstract

11037 Background: Previous studies demonstrated the positive association between cutaneous immune-related adverse events and long-term survival in advanced cancer patients treated with immunotherapy. As a unique adverse event related to camrelizumab, the association of reactive cutaneous capillary endothelial proliferation (RCCEP) with patient prognosis may also exist. Here we comprehensively analyzed the characteristics of RCCEP and this association. Methods: This was a pooled analysis based on individual patient-level data derived from seven phase 3 and two phase 2 registration trials for new drug application in China. Patients with advanced non-small cell lung cancer, hepatocellular carcinoma, esophageal squamous cell carcinoma, nasopharyngeal carcinoma, and gastric cancer treated with camrelizumab monotherapy (Camre), camrelizumab plus apatinib (Camre-Apa), or camrelizumab plus chemotherapy (Camre-Chemo) were included. Landmark analyses taking the median time to the first RCCEP onset as cutoff reference were performed for survival. Results: RCCEP occurred in 74.5% (251/337) of patients with Camre, 30.6% (120/392) with Camre-Apa, and 73.2% (737/1007) with Camre-Chemo. The severity was grade 1 or 2 in almost all patients with RCCEP (97.9%; 1085/1108), and the median frequency of RCCEP events was 1 (IQR, 1-2). Median time to the first RCCEP onset was 1.0 months [IQR, 0.7-1.2] with Camre, 4.7 months (IQR, 2.8-7.8) with Camre-Apa, and 1.5 months (IQR, 1.0-2.6) with Camre-Chemo; 1-month, 5-month, and 2-month landmark analyses of survival were performed for the three treatment groups, respectively. Patients with RCCEP showed better clinical outcomes than those without (objective response rate: 22.7% vs 2.3% with Camre, 42.5% vs 18.0% with Camre-Apa, and 73.7% vs 45.9% with Camre-Chemo; median progression-free survival [landmark analysis]: 3.0 vs 1.9 months [HR = 0.51, 95% CI, 0.38-0.69], 13.8 vs 12.6 months [HR = 0.88, 95% CI, 0.61-1.26], and 9.7 vs 6.9 months [HR = 0.58, 95% CI, 0.47-0.71]; median overall survival [landmark analysis]: 11.8 vs 3.9 months [HR = 0.44, 95% CI, 0.33-0.59], 35.2 vs 23.7 months [HR = 0.66, 95% CI, 0.49-0.90], and 23.4 vs 12.0 months [HR = 0.45, 95% CI, 0.37-0.54]). Discontinuation of camrelizumab treatment due to RCCEP barely occurred (0.3%; 5/1736). Conclusions: Although RCCEP occurred commonly, most events were mild without impact on camrelizumab treatment. RCCEP occurred early with 1-2 events mainly in each patient. The occurrence of RCCEP was positively associated with both short-term response and long-term survival, regardless of camrelizumab monotherapy or combination therapy. These findings can enhance patient confidence in continuing camrelizumab treatment.