Nomograms and an inflammation model predict immunotherapy prognosis in unresectable hepatocellular carcinoma: from survival and response.

Background: Screening out individuals who could benefit from immunotherapy is important in clinical practice. Methods: Clinical records of 199 individuals with unresectable HCC who underwent immunotherapy were retrospectively analyzed. Based on the effect of treatment, individuals were classified into two groups: the response group and the non-response group. NLR, LMR, and CAR were obtained before treatment. Kaplan-Meier method and Cox proportional risk regression model were used for survival analysis, and nomograms were established. A clinical prediction model was developed using logistic regression. Results: The cutoff values of NLR, LMR, and CAR were 5, 1.8, and 0.105, respectively. Individuals with NLR <5, LMR ≥1.8, and CAR <0.105 had higher ORR (43.09% vs. 13.16%, p < 0.001; 57.89% vs. 25.47%, p < 0.001; 41.49% vs. 22.86%, p < 0.01). In the multivariate analysis, it was determined that NLR and CAR independently served as prognostic factors for both PFS and OS. A combined parameter derived from these independent prognostic factors for ORR yielded an area under the curve (AUC) value of prediction probability of 0.825. Notably, the predictive performance of the model surpassed that of NLR (AUC = 0.794), LMR (AUC = 0.732), and CAR (AUC = 0.774). Conclusion: NLR, LMR, and CAR are potential predictive biomarkers of immunotherapy efficacy in individuals with advanced HCC.