Transarterial chemoembolisation is standard for hepatocellular carcinoma that cannot be resected, ablated or transplanted, and it induces tumour immune responses — which is the rationale for adding immunotherapy. EMERALD-3 randomised 760 participants with Child-Pugh A liver function and ECOG status 0 to 1 across 177 sites in 21 countries to three arms: durvalumab with tremelimumab plus lenvatinib plus TACE, the same immunotherapy plus TACE without lenvatinib, or TACE alone. Seventy-two per cent were Asian and 83% male.
The primary comparison met its endpoint. Median progression-free survival was 13.0 months (95% CI 12.2 to 16.7) with the triple combination against 9.8 months (8.0 to 11.4) with TACE alone, hazard ratio 0.70 (0.57 to 0.86, p=0.0007). The immunotherapy-plus-TACE arm without lenvatinib performed similarly against its matched comparator, 12.9 against 8.1 months, hazard ratio 0.71 (0.56 to 0.91) — which suggests lenvatinib is contributing little beyond what the immunotherapy delivers.
Overall survival has not separated. At a median follow-up of about 24 months it was 39.5 months against 34.7 months, hazard ratio 0.84 (95% CI 0.65 to 1.09, p=0.18). The authors are explicit that further follow-up is under way.
The toxicity gradient across the three arms is the number to carry into a consultation. Serious adverse events occurred in 23% on TACE alone, 51% on immunotherapy plus TACE, and 64% on the triple combination. Seven treatment-related deaths occurred in the triple-combination arm — two from myocarditis, and one each from hepatic failure, haemophagocytic lymphohistiocytosis, septic shock, cardiac failure and an unknown cause — against none in the immunotherapy-plus-TACE arm and two on TACE alone.
Read together, those two findings point somewhere specific. Immunotherapy plus TACE achieved essentially the same progression-free survival benefit as the triple combination, with serious adverse events at 51% rather than 64% and no treatment-related deaths. If this regimen is adopted, that is the version to look at first. Hepatocellular carcinoma is common in India on a background of hepatitis B and increasingly of metabolic liver disease, and these patients are often less fit than a trial cohort restricted to Child-Pugh A and good performance status.
- Progression-free survival improved from 9.8 to 13.0 months; overall survival has not separated.
- Adding lenvatinib gave little extra benefit over immunotherapy plus TACE, at higher toxicity.
- Serious adverse events rose steeply across arms: 23%, 51% and 64%.
- Seven treatment-related deaths occurred only in the triple-combination arm, including two from myocarditis.
- Eligibility was Child-Pugh A with ECOG 0 to 1; most Indian patients presenting with HCC will not meet it.
The statistics, in plain English
Three arms with an enrolment design that changed partway through makes the comparisons less clean than they look. The immunotherapy-plus-TACE arm stopped recruiting at 175 patients while the other two continued, so its comparison uses only the first 175 patients randomised to TACE — a smaller, earlier subset rather than the full control arm. That is a legitimate prespecified approach but it means the two progression-free survival results are not directly comparable with each other. The overall survival hazard ratio of 0.84 with an interval of 0.65 to 1.09 is immature at this follow-up; medians of 39.5 and 34.7 months with follow-up around 24 months mean many patients are still alive and the curves can still move.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for oncology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free