Prognosis of patients with hepatocellular carcinoma treated with immunotherapy – development and validation of the CRAFITY score

•Baseline serum AFP ≥100 ng/ml and CRP ≥1 mg/dl were independently associated with worse OS in ICB-treated patients with HCC.•A score based on these two variables predicts disease control rate and overall survival in ICB-treated patients with HCC.•The score was validated in an independent cohort of ICB-treated patients with HCC.•In sorafenib-treated patients with HCC, the score was prognostic for overall survival but not predictive for disease control rate. Background & AimsImmunotherapy with atezolizumab plus bevacizumab represents the new standard of care in systemic front-line treatment of hepatocellular carcinoma (HCC). However, biomarkers that predict treatment success and survival remain an unmet need.MethodsPatients with HCC put on PD-(L)1-based immunotherapy were included in a training set (n = 190; 6 European centers) and a validation set (n = 102; 8 European centers). We investigated the prognostic value of baseline variables on overall survival using a Cox model in the training set and developed the easily applicable CRAFITY (CRP and AFP in ImmunoTherapY) score. The score was validated in the independent, external cohort, and evaluated in a cohort of patients treated with sorafenib (n = 204).ResultsBaseline serum alpha-fetoprotein ≥100 ng/ml (hazard ratio [HR] 1.7; p = 0.007) and C-reactive protein ≥1 mg/dl (HR, 1.7; p = 0.007) were identified as independent prognostic factors in multivariable analysis and were used to develop the CRAFITY score. Patients who fulfilled no criterion (0 points; CRAFITY-low) had the longest median overall survival (27.6 (95% CI 19.5-35.8) months), followed by those fulfilling 1 criterion (1 point; CRAFITY-intermediate; 11.3 (95% CI 8.0-14.6) months), and patients meeting both criteria (2 points; CRAFITY-high; 6.4 (95% CI 4.8-8.1) months; p <0.001). Additionally, best radiological response (complete response/partial response/stable disease/progressive disease) was significantly better in patients with lower CRAFITY score (CRAFITY-low: 9%/20%/52%/20% vs. CRAFITY-intermediate: 3%/25%/36%/36% vs. CRAFITY-high: 2%/15%/22%/61%; p = 0.003). These results were confirmed in the independent validation set and in different subgroups, including Child-Pugh A and B, performance status 0 and ≥1, and first-line and later lines. In the sorafenib cohort, CRAFITY was associated with survival, but not radiological response.ConclusionsThe CRAFITY score is associated with survival and radiological response in patients receiving PD-(L)1 immunotherapy. The score may help with patient counseling but requires prospective validation.Lay summaryThe immunotherapy-based regimen of atezolizumab plus bevacizumab represents the new standard of care in systemic first-line therapy of hepatocellular carcinoma (HCC). Biomarkers to predict treatment outcome are an unmet need in patients undergoing immunotherapy for HCC. We developed and externally validated a score that predicts outcome in patients with HCC undergoing immunotherapy with immune checkpoint blockers. Immunotherapy with atezolizumab plus bevacizumab represents the new standard of care in systemic front-line treatment of hepatocellular carcinoma (HCC). However, biomarkers that predict treatment success and survival remain an unmet need. Patients with HCC put on PD-(L)1-based immunotherapy were included in a training set (n = 190; 6 European centers) and a validation set (n = 102; 8 European centers). We investigated the prognostic value of baseline variables on overall survival using a Cox model in the training set and developed the easily applicable CRAFITY (CRP and AFP in ImmunoTherapY) score. The score was validated in the independent, external cohort, and evaluated in a cohort of patients treated with sorafenib (n = 204). Baseline serum alpha-fetoprotein ≥100 ng/ml (hazard ratio [HR] 1.7; p = 0.007) and C-reactive protein ≥1 mg/dl (HR, 1.7; p = 0.007) were identified as independent prognostic factors in multivariable analysis and were used to develop the CRAFITY score. Patients who fulfilled no criterion (0 points; CRAFITY-low) had the longest median overall survival (27.6 (95% CI 19.5-35.8) months), followed by those fulfilling 1 criterion (1 point; CRAFITY-intermediate; 11.3 (95% CI 8.0-14.6) months), and patients meeting both criteria (2 points; CRAFITY-high; 6.4 (95% CI 4.8-8.1) months; p <0.001). Additionally, best radiological response (complete response/partial response/stable disease/progressive disease) was significantly better in patients with lower CRAFITY score (CRAFITY-low: 9%/20%/52%/20% vs. CRAFITY-intermediate: 3%/25%/36%/36% vs. CRAFITY-high: 2%/15%/22%/61%; p = 0.003). These results were confirmed in the independent validation set and in different subgroups, including Child-Pugh A and B, performance status 0 and ≥1, and first-line and later lines. In the sorafenib cohort, CRAFITY was associated with survival, but not radiological response. The CRAFITY score is associated with survival and radiological response in patients receiving PD-(L)1 immunotherapy. The score may help with patient counseling but requires prospective validation.