Background Intermediate hepatocellular carcinoma is often treated with locoregional therapy (LRT), such as transarterial chemoembolization (TACE) or transarterial radioembolization (TARE). Despite advances in TACE and TARE delivery, median progression-free survival following treatment is <1 year, highlighting a need for new treatment options.

Immune checkpoint inhibitor combinations (tremelimumab + durvalumab and atezolizumab + bevacizumab) have shown promise as first-line treatments for unresectable hepatocellular carcinoma (uHCC) not eligible for LRT. The Phase 3 EMERALD-1 study is evaluating the efficacy and safety of TACE + durvalumab ± bevacizumab versus TACE alone in participants with uHCC eligible for embolization. TARE and TACE are recommended in participants with uHCC in the US; a need exists for evidence to support additional treatment options in settings where TARE is the preferred treatment. The EMERALD-Y90 study will evaluate the efficacy and safety of TARE with durvalumab monotherapy, followed by durvalumab + bevacizumab in participants with uHCC eligible for embolization.

Methods EMERALD-Y90 (NCT06040099) is a Phase 2, single-arm study that will enroll ~100 participants aged ≥18 years with uHCC (Child-Pugh class A, Eastern Cooperative Oncology Group performance status 0–1) amenable to embolization who are ineligible for or have declined treatment with resection and/or ablation or liver transplant. Exclusion criteria include prior LRT (TACE, TARE, or stereotactic body radiotherapy associated with curative setting ≥6 months prior to study, and radiofrequency ablation if target lesion not treated or subsequently progressed, are permitted), prior systemic therapy, evidence of extrahepatic spread, or major portal vein invasion. Participants will receive partition-based dosing of TARE using yttrium-90 glass microspheres. Following TARE, participants will receive durvalumab 1500 mg (one dose) followed by durvalumab 1120 mg + bevacizumab 15 mg/kg every 3 weeks until study completion or discontinuation criteria are met.

The primary endpoint is progression-free survival (time from start of TARE until date of disease progression [investigator-assessed per modified Response Evaluation Criteria in Solid Tumors (mRECIST)] or death due to any cause [DAC]). Secondary endpoints include safety and tolerability, and 6-, 12-, and 24-month progression-free survival, objective response rate (% of participants with confirmed complete or partial response [investigator-assessed per mRECIST]), overall survival (time from start of TARE until DAC), and duration of response (time from date of first documented response until date of progression or DAC). An early safety review is planned when ≥30 participants complete ≥2 cycles of durvalumab + bevacizumab. Study enrollment will occur at ~20 US sites.

Results Median duration of exposure to durvalumab or placebo for durvalumab was highest in the durvalumab+bevacizumab+TACE arm during the durvalumab-bevacizumab period. In the durvalumab+bevacizumab+TACE and placebo+TACE arms of the IP-treated set, adverse events were reported by 139 (72.0%) and 148 (74.0%) participants in the durvalumab-TACE period and 147 (76.2%) and 132 (66.0%) participants in the durvalumab-bevacizumab period (table 1); were possibly related to treatment in 56 (29.0%) and 41 (20.5%) participants in the durvalumab-TACE period and 114 (59.1%) and 69 (34.5%) participants in the durvalumab-bevacizumab period; and were provoked by TACE in 90 (46.6%) and 85 (42.5%) participants in the durvalumab-TACE period and 18 (9.3%) and 21 (10.5%) participants in the durvalumab-bevacizumab period, respectively. Incidence of adverse events with the outcome of death was low across treatment arms in both periods, with none reported with durvalumab+bevacizumab+TACE during the durvalumab-bevacizumab period. Regardless of ALBI Grade, baseline characteristics were generally consistent across treatment arms. In the intent-to-treat population, PFS and TTP were numerically improved with durvalumab+bevacizumab+TACE versus placebo+TACE regardless of baseline ALBI Grade (table 2).

Acknowledgements Medical writing support, under the direction of the authors, was provided by Andrea Hough, PhD, CMC Connect, a division of IPG Health Medical Communications, and was funded by AstraZeneca, in accordance with Good Publication Practice (GPP 2022) guidelines. This abstract was originally presented at the 2024 Society of Interventional Oncology Annual Scientific Meeting.

Trial Registration Clinical trial identification: NCT06040099.

Ethics Approval The trial protocol was approved by local institutional review boards.