IMAT-HCC model: Liver stiffness-enhanced integrated risk stratification for HCC recurrence after resection

Early enthusiasm for adjuvant immunotherapy with atezolizumab plus bevacizumab after curative treatment of hepatocellular carcinoma (HCC) has been tempered by longer follow-up, which questioned the durability of benefit. Beyond treatment efficacy, inadequate patient selection may have contributed. We aimed to externally validate the high-risk criteria adopted in the IMbrave050 trial and to investigate whether integrating liver stiffness measurement (LSM) could improve recurrence risk stratification and provide a pragmatic framework for future adjuvant trial enrolment.

We retrospectively analysed 147 patients who underwent curative liver resection for HCC between 2010 and 2020 at our institution. High-risk status was defined by IMbrave050 criteria. LSM was assessed preoperatively. Competing-risk analyses (Fine-Gray) evaluated predictors of recurrence and liver-related outcomes; we quantified the added value of LSM using category-based net reclassification improvement. We then defined an integrated selection rule (Integrated Model for Adjuvant Trial Selection in Hepatocellular Carcinoma [IMAT-HCC]) as IMbrave050-IN or IMbrave050-OUT with LSM ≥15 kPa.

Over a median 34 months, 88/147 patients (59.9%) recurred. Both IMbrave050 high-risk features and LSM ≥15 kPa were independently associated with recurrence (subhazard ratio [SHR] 1.66, 95% CI 1.09-2.54; and SHR 1.62, 95% CI 1.07-2.47, respectively). Patients classified as IMbrave050-OUT but with LSM ≥15 kPa showed recurrence rates comparable to IMbrave050-IN patients, whereas those with LSM <15 kPa had substantially lower risk. The IMAT-HCC rule provided clearer separation of cumulative incidence functions than IMbrave050 alone and improved patient reclassification (overall net reclassification improvement 8.0%), driven by correct up-classification of patients who eventually recurred.

Integrating LSM into tumour-based risk criteria refines identification of biologically high-risk patients after resection and improves exclusion of truly low-risk individuals. IMAT-HCC represents a simple, non-invasive and trial-ready selection rule that may enhance enrichment strategies for future adjuvant trials in HCC.

This study demonstrates that integrating liver stiffness measurement with tumour-based criteria refines the identification of patients at high biological risk of recurrence after curative resection for hepatocellular carcinoma. By addressing the limitations of purely tumour-centric selection strategies (i.e. IMbrave050 criteria), it highlights the prognostic relevance of the underlying liver disease in shaping post-resection outcomes. Incorporation of this approach into clinical research and trial design could improve risk stratification and patient selection for adjuvant therapies, while avoiding overgeneralisation given the retrospective, single-centre nature of the study.