Liver cancer's stubborn resistance to treatment has long been traced to its molecular diversity — no two tumors behave alike, and therapies that work in one patient often fail in another. A platform that could systematically map this heterogeneity and match it to effective drugs would represent a genuine shift in how oncologists approach one of the world's deadliest cancers. That prospect is now closer to reality.
Researchers built a biobank of patient-derived hepatocellular carcinoma organoids (HCCOs) — miniaturized, three-dimensional tumor models grown directly from diagnostic biopsies — capturing a wide spectrum of disease stages and clinical profiles. These living tumor proxies were then subjected to a high-throughput phenotypic drug screen covering more than 1,600 compounds, yielding multiple candidates with meaningful antitumor activity, including several already approved for other indications and thus eligible for repurposing. Beyond single-agent discovery, the team systematically tested doublet and triplet drug combinations anchored on regorafenib, the current standard second-line agent for advanced HCC. Several multi-drug regimens produced enhanced efficacy across diverse organoid lines while sparing non-tumoral cells — a critical selectivity threshold. A lead triplet combination was then validated in vivo, demonstrating improved tumor control alongside acceptable tolerability.
Organoid-based drug screening has been gaining traction across gastrointestinal cancers, but HCC has lagged partly because liver tumor tissue is notoriously difficult to propagate ex vivo. This biobank approach addresses that barrier at scale. The key limitation is translational: organoids lack immune components, stromal architecture, and vascular context — all factors that heavily influence HCC drug response in patients. The regorafenib-anchored combinations are clinically rational, since that drug is already in widespread use, but triplet regimens carry additive toxicity risks that organoid models cannot fully model. Still, as a platform for prioritizing candidates before costly clinical trials, this work is genuinely meaningful — incremental in concept but substantial in execution and scope.