WP1: The consortium successfully completed the largest-ever case-control dataset for ALD-HCC genetic discovery (4,418 cases/6,655 controls across 20 cohorts), enabling a genome-wide association study that replicated five known risk loci (PNPLA3, TM6SF2, TERT, WNT3A, HSD17B13) and identified a novel genome-wide significant signal at EPHA2. A polygenic risk score (PRS) incorporating these six loci demonstrated a two-fold increased HCC risk in high-risk UK Biobank cirrhotic patients with moderate-to-high alcohol consumption (sHR 2.08-2.26) with 10-year absolute risk differences up to 12.2% in heavy drinkers. Spatial transcriptomics analyses further revealed EPHA2 enrichment in biliary tract lineage cells, providing novel mechanistic insights.
WP2: Histopathological assessment of 45 ALD liver specimens was completed, with spatial omics and single-nuclei datasets now available for integration via UPCIT's MetaMAP and Histomap methods. Champalimaud uncovered novel neuro-immune axes in hepatic transformation, identifying potential first-in-class therapeutic targets for MASLD/HCC. The SCherlock method for robust cell-type marker identification in single-cell data has been finalized and is ready for submission alongside its validation paper.
WP3: FASTRAK trial recruitment reached 637 patients (102% of M36 target across 34 centers), confirming a 3% annual HCC incidence in this high-risk cohort. SERENA cohort (n=524 advanced MASLD/MetALD patients) reported a 7.6% 5-year HCC incidence, with FIB-4 emerging as a strong independent predictor (HR 1.34/unit AUC 0.81). Machine learning models benchmarked on CIRRAL/CirVir/NASH-AVC cohorts (n=2,544) achieved AUC 0.81 when combining clinical scores and pre-GENIAL PRS; ongoing integration of WP1 genetics, pathology, radiology, and FASTRAK/SERENA data targets AUC>0.90. Biobanks for circulating samples, digital pathology, and imaging are operationa
