Abstract 111P
Background
This study aimed to identify a new prognostic index by applying recursive partitioning analysis (RPA) in hepatocellular carcinoma (HCC) patients treated with atezolizumab plus bevacizumab (AB).
Methods
RPA was applied on 784 consecutive HCC patients treated with AB.
Results
RPA allowed the identification of the atezolizumab bevacizumab prognostic (ABE) index, comprising three groups of patients: low risk, [(i) Child-Pugh A (CPA) patients without macrovascular invasion (MVI) but with albumin-bilirubin (ALBI) 1, aspartate aminotransferase (AST) normal value (NV), and alpha-fetoprotein (AFP) < 400 ng/mL, (ii) CPA patients without MVI but with ALBI 1, AST increased value (IV), and neutrophil-lymphocyte ratio (NLR) < 3, and (iii) CPA patients with MVI, ALBI 1, and AFP < 400 ng/mL]; intermediate risk, [(i) CPA patients without MVI but with ALBI 1, AST NV, and AFP ≥ 400 ng/mL, and (ii) CPA patients without MVI but with ALBI 1, AST IV, and NLR ≥ 3]; high risk [(i) CPA patients with ALBI 2, (ii) CPA patients with ALBI 1, MVI, and AFP ≥ 400 ng/mL, and (iii) CPB patients]. Overall survival was 22.5 months (95% CI 17.0-22.5 months) in patients with low risk (60.1%), 14.2 months (95% CI 12.4-15.7 months) in intermediate risk (19.1%), and 7.0 months (95% CI 6.0-8.7 months) in high risk (20.8%); low risk HR 1, intermediate risk HR 1.76 (95% CI 1.26-2.46), high risk HR 3.99 (95% CI 2.76-5.77); P < 0.01. Progression-free survival was 9.4 months (95% CI 8.4-10.8 months) in patients with low risk, 6.1 months (95% CI 5.5-8.1 months) in intermediate risk, and 5.3 months (95% CI 3.7-5.8 months) in high risk; low risk HR 1 (reference group), intermediate risk HR 1.47 (95% CI 1.14-1.89), high risk HR 1.79 (95% CI 1.37-2.35); P < 0.01. In the three groups, differing profiles of toxicity have been highlighted, notably in terms of hypertension (low risk 27.4%; intermediate risk 22.7%; high risk 17.2%, P = 0.03), proteinuria (low risk 28.7%; intermediate risk 35.3%; high risk 22.7%, P < 0.05), and hypothyroidism (low risk 6.1%; intermediate risk 2.7%; high risk 1.8%; P = 0.03).
Conclusions
The ABE index is an easy-to-use tool able to stratify HCC patients undergoing first-line therapy with AB.
Legal entity responsible for the study
M. Persano.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.