A study on the application of metaproteomic and serum metabolomic integration analysis in TACE combined with targeted immunotherapy for unresectable hepatocellular carcinoma.

Abstract

e16207 Background: TACE combined with targeted therapy and immunotherapy represents a potentially effective therapy for uHCC. By evaluating the predictive value of gut microbiota-derived proteins and metabolites for treatment efficacy and prognostic outcomes, this study aims to provide evidence for prognostic stratification in patients in uHCC which received TACE combined targeted and immunotherapy in the real world. Methods: Prospective data of patients (pts) with uHCC who received TACE combined with targeted and immunotherapy were collected, and pts were grouped by therapeutic effect. The primary endpoint was integrated proteomics and metabolomics analysis based on plasma, urine, tumor tissue and faeces were conducted to evaluate the candidate biomarkers for prediction of prognostic outcomes in pts before treatment. Results: As of Jan 16, 2026, a total of 47 pts eligible for HCC were enrolled and 41 were evaluable, 13 (31.7%) had ECOG PS 1, 15 (36.6%) were BCLC C. The results show ORR was 43.9%, and DCR was 80.5%, mOS was 29.0m(95% CI, 22.7-35.3), No grade 5 AEs. The responders (CR+PR) had a longer OS (NR vs 20.6 m, p = 0.006) compared to non-responders (SD+PD). In proteomics analysis, our study indicated both plasma and urine proteomes can reflect the functional difference in tumor tissue between responder and non- responder groups. In serum, proteins associated with growth factor receptor, kinase activity, and programmed cell death exhibited differential expression. In urine, proteins related to metabolic processes demonstrate varying levels of expression. In the metabolomic analysis, all plasma, urine, and faeces can serve as indicators of functional difference in tumor tissue between the two groups. The differential metabolites were associated with amino acid metabolism in plasma, caffeine, ascorbate and aldarate metabolism in urine, and both amino acid and ascorbate/ aldarate metabolism in faeces. A panel including 4 plasma biomarkers and a panel containing 4 urine biomarkers could robustly predict responder group and non- responder group before treatment, with AUC = 0.89 and 0.95 respectively. Metabolites biomarker/biomarker panels in plasma, urine and faeces also demonstrated acceptable predictive performance (AUC = 0.768–0.821). Conclusions: The combination regimen is expected to be a effective treatment for uHCC. Biofluids and faeces multi-omics could reflect the functional differences in tumor tissues between the responder and non-responder pts, and could provide a non-invasive approach to predict treatment responsiveness. Clinical trial information: NCT06540508 . Baseline characteristics. n CR+PR (18) SD+PD (23) P Age (years) 62.2 (9.16) 62.0 (10.24) 0.950 Number of tumors 1 2 (11.1) 1 (4.35) 0.708 2-5 7 (38.9) 12 (52.2) ＞5 9 (50.0) 10 (43.5) Maximum tumor diameter (cm) 8.61 (3.68) 6.38 (4.47) 0.088 PVTT YES 8 (44.4) 6 (26.1) 0.369 NO 10 (55.6) 17 (73.9)