Primary Tumor Resection Plus Systemic Therapy Versus Systemic Therapy Alone in Hepatocellular Carcinoma With Extrahepatic Metastases: A Multicenter Prospective Randomized Controlled Trial

This is a multicenter, prospective, open-label, randomized, parallel-group controlled trial in patients with hepatocellular carcinoma (HCC) with extrahepatic metastases whose primary intrahepatic tumor is considered resectable with curative intent (R0 intent). A total of 280 participants will be randomized in a 1:1 ratio to either primary tumor resection plus protocol-allowed first-line systemic therapy or systemic therapy alone.

Randomization will be performed using block randomization with a block size of 4 through a centralized system managed by an independent data management team. Stratification factors include maximum tumor diameter (<5 cm vs. >=5 cm), category of systemic therapy (immunotherapy plus anti-angiogenic therapy vs. immunotherapy plus targeted therapy), and liver function status (Child-Pugh A vs. B7).

All participants will receive guideline-concordant first-line immune-based systemic therapy. Preferred regimens include PD-1/PD-L1 inhibitor plus anti-angiogenic therapy. PD-1 inhibitor plus multikinase targeted therapy is allowed when anti-angiogenic therapy is contraindicated or not feasible. The use of more than one PD-1/PD-L1 agent in combination is not permitted. Optional hepatic arterial infusion chemotherapy (HAIC) may be used according to clinical need in either arm and will be recorded for exploratory and adjusted analyses.

In the experimental arm, participants will undergo resection of the primary liver tumor before systemic therapy. The surgical goal is R0 resection, and the procedure may be anatomic or non-anatomic hepatectomy based on tumor location, liver reserve, and operative risk. Systemic therapy will be initiated after adequate postoperative recovery. In the control arm, participants will receive systemic therapy alone. If a participant in the control arm later becomes a candidate for surgery, such surgery will be recorded as an unplanned intervention, but the participant will remain in the originally assigned group for the primary intention-to-treat analysis.

Tumor response will be assessed by contrast-enhanced CT or MRI according to RECIST version 1.1. Imaging will be performed every 6 weeks through week 54 and every 9 weeks thereafter until disease progression or treatment discontinuation. To preserve comparability of response assessment between the two arms, lesions planned for resection in the experimental arm will not be selected as target lesions at baseline.