Background: The global Phase 3 EMERALD-1 study (NCT03778957) primary endpoint was met: durvalumab (D) + bevacizumab (B) + transarterial chemoembolization (TACE) significantly improved progression-free survival vs placebos (pbos) + TACE (median 15.0 vs 8.2 mo; HR, 0.77; 95% CI, 0.61–0.98; p=0.032 [threshold 0.0435]) in participants (pts) with embolization-eligible unresectable hepatocellular carcinoma (uHCC) with a manageable safety profile. This post hoc, exploratory analysis assessed safety in the study’s two treatment (tx) periods (pds): D-TACE (D-T) and D-B. Methods: Pts were randomized 1:1:1 to D+TACE, D+B+TACE, or pbos+TACE. In the D-T pd, pts received 1–4 TACE (cTACE or DEB-TACE [investigator choice]) plus D (1500 mg Q4W) or pbo for D. In the D-B pd, post-last TACE, pts received D (1120 mg Q3W) plus pbo for B, D (1120 mg Q3W) plus B (15 mg/kg Q3W), or pbos for D and B. Adverse events (AEs), start/end date, maximum/change in CTCAE grade, and causality were assessed in the D-T and D-B pds in pts who received any study tx in the arm to which they were randomized, until end of follow-up. Results: The Table shows duration of exposure (DoE). In the D+TACE, D+B+TACE, and pbos+TACE arms, AEs were reported by 144 (74.6%), 139 (72.0%), and 148 (74.0%) pts in the D-T pd and 133 (68.9%), 147 (76.2%), and 132 (66.0%) pts in the D-B pd; were possibly related (pos rel) to study tx in 59 (30.6%), 56 (29.0%), and 41 (20.5%) pts in the D-T pd and 76 (39.4%), 114 (59.1%), and 69 (34.5%) pts in the D-B pd; and were provoked by TACE in 72 (37.3%), 90 (46.6%), and 85 (42.5%) pts in the D-T pd and 16 (8.3%), 18 (9.3%), and 21 (10.5%) pts in the D-B pd, respectively. Conclusions: D+B+TACE had a manageable safety profile across the D-T and D-B pds, consistent with the individual agents and underlying disease. These data further support D+B+TACE as a new potential standard of care in embolization-eligible uHCC. Clinical trial information: NCT03778957.

*Excluding TACE. †Pos rel to D or pbo for D. ‡Pos rel to B or pbo for B.