Durvalumab with or without bevacizumab with transarterial chemoembolisation in hepatocellular carcinoma (EMERALD-1): a multiregional, randomised, double-blind, placebo-controlled, phase 3 study

Abstract

Background: Transarterial chemoembolisation (TACE) is standard of care for patients with unresectable hepatocellular carcinoma that is amenable to embolisation; however, median progression-free survival is still approximately 7 months. We aimed to assess whether adding durvalumab, with or without bevacizumab, might improve progression-free survival.

Methods: In this multiregional, randomised, double-blind, placebo-controlled, phase 3 study (EMERALD-1), adults aged 18 years or older with unresectable hepatocellular carcinoma amenable to embolisation, an Eastern Cooperative Oncology Group performance status of 0 or 1 at enrolment, and at least one measurable intrahepatic lesion per modified Response Evaluation Criteria in Solid Tumours (RECIST) were enrolled at 157 medical sites including research centres and general and specialist hospitals in 18 countries. Eligible patients were randomly assigned (1:1:1), stratified by TACE method, region, and portal vein invasion, using an interactive voice response or web response system, to TACE plus either durvalumab plus bevacizumab (1500 mg intravenous durvalumab once every 4 weeks, then 1120 mg durvalumab plus 15 mg/kg intravenous bevacizumab once every 3 weeks), durvalumab plus placebo (same regimen using placebo instead of bevacizumab), or placebo alone (same regimen using placebo instead of durvalumab and instead of bevacizumab). Participants, investigators, and those assessing outcomes were masked to treatment assignment until data analysis. The primary endpoint was progression-free survival, by blinded independent central review (BICR), and per RECIST version 1.1, with durvalumab plus bevacizumab versus placebo alone in the intention-to-treat population (ITT; ie, all participants assigned to treatment). Key secondary endpoints were progression-free survival by BICR per RECIST version 1.1 with durvalumab plus placebo versus placebo alone, overall survival, and time to deterioration in select patient-reported outcomes. Participants continue to be followed up for overall survival, and overall survival and patient-reported outcomes will be reported in a later publication. Safety was assessed in the safety analysis set, which included all participants assigned to treatment who received any study treatment (ie, any durvalumab, bevacizumab, or placebo) by treatment received. This study is registered with ClinicalTrials.gov, NCT03778957, and is closed to accrual.

Findings: Between Nov 30, 2018, and July 19, 2021, 887 patients were screened, of whom 616 were randomly assigned to durvalumab plus bevacizumab (n=204), durvalumab plus placebo (n=207), or placebo alone (n=205; ITT population). Median age was 65·0 years (IQR 59·0-72·0), 135 (22%) of 616 participants were female, 481 (78%) were male, 375 (61%) were Asian, 176 (29%) were White, 22 (4%) were American Indian or Alaska Native, nine (1%) were Black or African American, one (<1%) was native Hawaiian or other Pacific Islander, and 33 (5%) were other races. As of data cutoff (Sept 11, 2023) median follow-up for progression-free survival was 27·9 months (95% CI 27·4-30·4), median progression-free survival was 15·0 months (95% CI 11·1-18·9) with durvalumab plus bevacizumab, 10·0 months (9·0-12·7) with durvalumab, and 8·2 months (6·9-11·1) with placebo. Progression-free survival hazard ratio was 0·77 (95% CI 0·61-0·98; two-sided p=0·032) for durvalumab plus bevacizumab versus placebo, and 0·94 (0·75-1·19; two-sided p=0·64) for durvalumab plus placebo versus placebo. The most common maximum grade 3-4 adverse events were hypertension in participants who received durvalumab and bevacizumab (nine [6%] of 154 participants), anaemia in participants who received durvalumab and placebo (ten [4%] of 232 participants), and post-embolisation syndrome in participants who received placebo alone (eight [4%] of 200 participants). Study treatment-related adverse events that led to death occurred in none of 154 participants who received durvalumab and bevacizumab, three (1%) of 232 who received durvalumab and placebo (n=1 for arterial haemorrhage, liver injury, and multiple organ dysfunction syndrome), and three (2%) of 200 who received placebo alone (n=1 for oesophageal varices haemorrhage, upper gastrointestinal haemorrhage, and dermatomyositis).

Interpretation: Durvalumab plus bevacizumab plus TACE has the potential to set a new standard of care. With additional follow-up of the EMERALD-1 study, future analyses, including the final overall survival data and patient-reported outcomes, will help to further characterise the potential clinical benefits of durvalumab plus bevacizumab plus TACE in hepatocellular carcinoma amenable to embolisation.

Funding: AstraZeneca.

Figure 1:. Study schematic

TACE=transarterial chemoembolisation. *1500 mg durvalumab, or placebo, once every 4 weeks. †1120 mg durvalumab, or placebo, once every 3 weeks. ‡15 mg/kg bevacizumab, or placebo, once every 3 weeks.

Figure 2:. Trial profile

Categories for participants who discontinued study treatment, are no longer ongoing in study, or who received subsequent anticancer therapy are not exclusive, and participants might be included in more than one of these categories. ITT=intention-to-treat. *One participant discontinued study treatment because their second lesion did not respond to TACE and one participant discontinued study treatment due to deteriorating performance status. †One participant discontinued study treatment due to COVID-19 pandemic-related logistics and one participant discontinued study treatment before receiving placebo for bevacizumab due to clinical judgement after COVID-19 infection resolution. ‡Reasons for discontinuation of study treatment only include reasons for durvalumab discontinuation. §As assessed by the investigator.

Figure 3:. Progression-free survival, assessed by BICR per RECIST version 1.1, in the durvalumab plus bevacizumab group versus the placebo alone group (primary endpoint; A) and in the durvalumab plus placebo group versus the placebo alone group (key secondary endpoint; B), in the intention-to-treat population

BICR=blinded independent central review. RECIST=Response Evaluation Criteria in Solid Tumours. *Threshold of significance for this analysis was 0·044, based on the α spent at the progression-free survival interim analysis (2·27%) and the actual number of events at the final progression-free survival analysis.

Figure 4:. Subgroup analysis of progression-free survival, assessed by BICR per RECIST version 1.1, in the durvalumab plus bevacizumab group versus the placebo alone group in the intention-to-treat population

Size of circles are proportional to the number of events. Participant subgroups for region, transarterial chemoembolisation technique, and portal vein invasion are per values entered into the interactive voice response system at randomisation. For subgroups with fewer than 20 events, hazard ratios were not calculated. BICR=blinded independent central review. ECOG=Eastern Cooperative Oncology Group. NC=not calculated. RECIST=Response Evaluation Criteria in Solid Tumours. ULN=upper limit of normal. *One participant in each group had both hepatitis B virus and hepatitis C virus; neither of these participants had a progression-free survival event. †PD-L1 tumour area positivity data were missing for 50 participants in the durvalumab plus bevacizumab group and 53 participants in the placebo alone group; the hazard ratio in the missing subgroup was 0·66 (95% CI 0·39–1·11). ‡α-fetoprotein data were missing for one participant in the durvalumab plus bevacizumab group; this participant had an event; the normal range for α-fetoprotein per local laboratory ranges was 0–25 IU/mL. §Number of target lesions at baseline data were missing for one participant in the durvalumab plus bevacizumab group; this participant did not have an event. ¶Hepatoma arterial-embolisation prognostic score was missing for three participants in the durvalumab plus bevacizumab group. ‖Baseline tumour burden data were missing for one participant in the durvalumab plus bevacizumab group. **Post-hoc subgroup analysis.