{
  "abstract": "Background Early relapse remains a major challenge in early-stage hepatocellular carcinoma (HCC), with only 30% achieving complete pathological responses to immunotherapy. While transcriptomic profiling has identified HCC subgroups linked to relapse-free survival (RFS) and immunotherapy responses, its clinical applicability is limited by complexity and turnaround time. Here we report a comprehensive analysis of HCC’s immunopathologic landscape and show the value of fresh tumor tissue analysis for rapid biological classification.Methods We prospectively collected fresh/fixed tumor, margin and surrounding liver tissue from 32 resected untreated HCC patients. Transcriptomic profiling of fixed tissue identified subgroups based on established prognostic and immunotherapy-related signatures. Correlations between subgroups and pathology, immune cells (flow cytometry), secreted factors (multiplex assays) and RFS were assessed. Random forest analysis identified fresh tissue features strongly associated with transcriptome subgroups. A classification model was developed and tested on an extended cohort with 13 additional resected patients.Results Transcriptomic analysis validated previously described HCC subgroups: ‘excluded, ‘intermediate’, ‘non- cytotoxic’, and ‘cytotoxic’ ( figure 1A). ‘Excluded’ and ‘Intermediate’ subgroups displayed a non-inflamed gene signature, higher chromosomal instability, and minimal immune infiltration. ‘Excluded’ HCC showed specific elevated WNT-beta-catenin signalling. ‘Non-cytotoxic’ and ‘cytotoxic’ subgroups displayed an inflamed gene signature correlated with steatohepatitic histology and moderate/high immune infiltration. ‘Non-cytotoxic’ HCC showed an elevated gene signature of poor survival post-immunotherapy, while ‘cytotoxic’ HCC had elevated cytolytic signature and improved RFS (mRFS not reached vs. 18.5 months, p=0.004; figure 1B). Flow cytometry and secreted factor analysis revealed distinct immune profiles: non-inflamed HCCs had increased myeloid cell infiltration and elevated SCF. ‘Cytotoxic’ HCC showed increased T and B cells, decreased myeloid cells, and elevated IL-16 and sCD25, while ‘non-cytotoxic’ was uniquely elevated in IL- 6, IL-1beta and CCL2.To challenge the predictive value of the ‘cytotoxic’ transcriptomic signature (figure 2A), a classification tree was developed using significant features identified by random forest, distinguishing ‘cytotoxic’ HCC from all other subgroups using rapid secretome data (AUC=0.96; figure 2B). Using an extended cohort with 13 new patients, we confirmed the combined score of secreted factors IL16 and sCD25 effectively predicted RFS (p=0.006) (figure 2C). Multivariate Cox analysis showed that this combined secretome score outperformed conventional clinical prognostic markers, including microvascular invasion, AFP, and tumor burden.Conclusions Immune secretome from fresh tumor tissue enables rapid and sensitive prediction of RFS in early-stage HCC, supporting its potential clinical utility for risk stratification and treatment guidance compatible with the short screening period timeframes of clinical trials.Ethics Approval This study was conducted in accordance with the Declaration of Helsinki. Patient information and sample collection were authorized by the Ministry of Higher Education and Research (approval number DC-2021-4572). The study protocol complied with the reference methodology MR-004 established by the French Data Protection Authority (Commission Nationale de l’Informatique et des Libertés, CNIL). All participants provided informed consent prior to inclusion in the study.Abstract 183 Figure 1Transcriptomic validation of immune-based hepatocellular carcinoma (HCC) subgroups and their prognostic relevanceAbstract 183 Figure 2Comparative prognostic performance of transcriptome- versus secretome-derived immune classification in HCC",
  "authors": [
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Heloise Halse"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Mélodie Bonvalet"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France"
      ],
      "name": "Léonie Bui"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France"
      ],
      "name": "Rim Kalala"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Chifaou Mohamed-Djalim"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Delphine Bredel"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Aminata Drame"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Séverine Mouraud"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Andrey Yurchenko"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France",
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Sylvie Job"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France",
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Jamila Faivre"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France",
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Olivier Rosmorduc"
    },
    {
      "affiliations": [
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Astrid Laurent-Bellue"
    },
    {
      "affiliations": [
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Catherine Guettier"
    },
    {
      "affiliations": [
        "Gustave Roussy, Villejuif, Ile-de-France, France"
      ],
      "name": "Amélie Bigorgne"
    },
    {
      "affiliations": [
        "Centre Hépato-Biliaire Paul Brousse, Villejuif, Ile-de-France, France",
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France"
      ],
      "name": "Eric Vibert"
    },
    {
      "affiliations": [
        "Université Paris-Saclay, Le Kremlin-Bicêtre, Ile-de-France, France",
        "Gustave Roussy and Paris Saclay University, Villejuif, France"
      ],
      "name": "Aurélien Marabelle"
    }
  ],
  "title": "183 Fresh tumor analysis enables rapid characterization of prognostic subgroups and identifies novel therapeutic targets for early-stage hepatocellular carcinoma",
  "uid": "44c40556-fc9d-5ebb-aa65-c189937f0042"
}