AI takes on liver cancer: could a computer help pick the right immunotherapy?

Copied word for word from the study's registry entry, so the wording is the study team's rather than ours.

Inclusion Criteria: * Age 18 years or older. * Hepatocellular carcinoma confirmed by histology or cytology, or diagnosed using accepted noninvasive radiologic criteria. * Unresectable hepatocellular carcinoma for which first-line systemic therapy is indicated, including Barcelona Clinic Liver Cancer stage B disease that is unsuitable for or no longer benefiting from locoregional therapy, or stage C disease. * No prior systemic anticancer therapy for unresectable hepatocellular carcinoma. * Child-Pugh class A liver function. * Eastern Cooperative Oncology Group performance status of 0 or 1. * Eligible and intended to receive at least one protocol-specified, guideline-concordant immune checkpoint inhibitor-based first-line regimen, as determined by the treating clinician before exposure to the study AI recommendation. * Eligibility confirmed and prospective participant registration completed before the final first-line treatment regimen is selected. * Baseline contrast-enhanced computed tomography or magnetic resonance imaging suitable for assessment according to Response Evaluation Criteria in Solid Tumors version 1.1 is available before initiation of first-line systemic therapy. * Required pretreatment clinical data are available within the protocol-specified assessment windows. * Able to provide written informed consent. Exclusion Criteria: * Known combined hepatocellular-cholangiocarcinoma or another primary liver malignancy other than hepatocellular carcinoma. * A contraindication or clinical condition that makes all protocol-specified immune checkpoint inhibitor-based first-line regimens inappropriate according to applicable prescribing information and routine clinical practice. * Initiation of systemic therapy or completion of the final first-line regimen decision before prospective participant registration. * Exposure of the treating clinical team to the participant-specific AI recommendation before confirmation of eligibility and registration. * Concurrent anticancer treatment for another malignancy that would materially interfere with treatment selection or study outcome assessment. * Planned participation in another interventional study that dictates first-line systemic treatment or would interfere with study outcome assessment. * Inability to undergo protocol-required tumor imaging or follow-up assessments.