BackgroundFor decades, stratification criteria for first-line clinical studies have been highly uniform. However, there is no principle or consensus for restratification after systemic treatment progression based on immune checkpoint inhibitors (ICIs). The aim of this study was to assess the patterns of disease progression in patients with advanced hepatocellular carcinoma (HCC) who are not eligible for surgical intervention, following the use of immune checkpoint inhibitors.MethodsThis is a retrospective study that involved patients with inoperable China liver stage (CNLC) IIIa and/or IIIb. The patients were treated at eight centers across China between January 2017 and October 2022. All patients received at least two cycles of first-line treatment containing immune checkpoint inhibitors. The patterns of disease progression were assessed using RECIST criteria 1.1. Different progression modes have been identified based on the characteristics of imaging progress. The study’s main outcome measures were post-progression survival (PPS) and overall survival (OS). Survival curves were plotted using the Kaplan-Meier method to compare the difference among the four groups. Subgroup analysis was conducted to compare the efficacy of different immunotherapy combinations. Variations in the efficacy of immunotherapy have also been noted across patient groups exhibiting alpha-fetoprotein (AFP) levels equal to or exceeding 400ng/mL, in contrast to those with AFP levels below 400ng/mL.ResultsThe study has identified four distinct patterns of progress, namely p-IIb, p-IIIa, p-IIIb, and p-IIIc. Diverse patterns of progress demonstrate notable variations in both PPS and OS. The group p-IIb had the longest PPS of 12.7m (95% 9.3-16.1) and OS 19.6m (95% 15.6-23.5), the remaining groups exhibited p-IIIb at PPS 10.5 months (95%CI: 7.9-13.1) and OS 19.2 months (95%CI 15.1-23.3). Similarly, p-IIIc at PPS 5.7 months (95%CI: 4.2-7.2) and OS 11.0 months (95%CI 9.0-12.9), while p-IIIa at PPS 3.4 months (95%CI: 2.7-4.1) and OS 8.2 months (95%CI 6.8-9.5) were also seen. Additional stratified analysis was conducted and showed there were no differences of immunotherapy alone or in combination in OS (HR= 0.92, 95%CI: 0.59-1.43, P=0.68) and PPS (HR= 0.88, 95%CI: 0.57-1.36, P=0.54); there was no significant difference in PPS (HR=0.79, 95% CI: 0.55-1.12, P=0.15) and OS (HR=0.86, 95% CI: 0.61-1.24, P=0.39) for patients with AFP levels at or over 400ng/mL. However, it was observed that patients with AFP levels above 400ng/mL experienced a shorter median progression of PPS (8.0 months vs. 5.0 months) after undergoing immunotherapy.ConclusionIn this investigation of advanced hepatocellular carcinoma among Chinese patients treated with immune checkpoint inhibitors, we identified four distinct progression patterns (p-IIb, p-IIIa, p-IIIb and p-IIIc) that showed significant differences in PPS and OS. These findings demonstrate the heterogeneity of disease progression and prognosis after immunotherapy failure. Further validation in large cohorts is necessary to develop prognostic models that integrate distinct progression patterns to guide subsequent treatment decisions. Additionally, post-immunotherapy progression in patients with AFP levels ≥400ng/mL indicates a shortened median PPS. These findings provide valuable insights for future personalized treatment decisions.