AstraZeneca (LSE/STO/NYSE: AZN) reported that the combination of durvalumab (Imfinzi), tremelimumab (Imjudo), lenvatinib, and transarterial chemoembolisation met its primary endpoint in unresectable hepatocellular carcinoma eligible for embolisation, delivering a statistically significant improvement in progression-free survival versus TACE alone — a result that positions the regimen as a potential first systemic combination approved for this intermediate-stage population in the US and EU.
The EMERALD-3 trial is a randomised, open-label, sponsor-blinded, multicentre Phase III study enrolling 760 patients with unresectable HCC eligible for embolisation across 171 centres in 22 countries.
The trial met its primary endpoint of PFS for the four-drug combination — the STRIDE regimen (a single priming dose of tremelimumab 300mg added to durvalumab 1500mg, followed by durvalumab every four weeks) plus lenvatinib and TACE — versus TACE alone. At an interim analysis for overall survival, a key secondary endpoint, the combination also showed a trend toward OS improvement. A third arm evaluating STRIDE plus TACE without lenvatinib showed trends toward improved PFS and OS, though this comparison was not formally tested at this interim analysis. No quantitative data — hazard ratios, medians, or p-values — were disclosed.
The safety profile was consistent with the known profiles of each component, with no new findings identified.
Clinical context
The result carries weight because it is not focused on advanced or metastatic disease. Embolisation-eligible unresectable HCC maps broadly to intermediate-stage disease — a population for which TACE has remained the standard of care for more than two decades, with most patients experiencing disease progression or recurrence within six to ten months. No systemic combination has received FDA or EMA approval specifically for this setting. The closest attempt, the LEAP-012 trial evaluating Eisai's Lenvima (lenvatinib) plus Merck's Keytruda (pembrolizumab) plus TACE, missed its overall survival endpoint in the global analysis, receiving approval only from China's NMPA in 2025.
The STRIDE backbone — durvalumab blocking PD-L1 to counter tumour immune evasion, combined with tremelimumab's CTLA-4 blockade to prime T-cell activation — already holds FDA and EMA approval for advanced unresectable HCC based on the HIMALAYA Phase III trial, where it demonstrated durable OS benefit versus sorafenib. EMERALD-3 extends that immunotherapy backbone earlier in the disease course, layering in lenvatinib's multi-kinase inhibition of VEGFR and FGFR pathways alongside locoregional TACE. The mechanistic logic is that TACE-induced tumour antigen release may enhance the immunogenic response primed by dual checkpoint blockade, while lenvatinib's anti-angiogenic activity addresses vascular escape.
Competitive context
The competitive landscape in HCC has grown crowded at the advanced end. Atezolizumab plus bevacizumab (Roche) and nivolumab plus ipilimumab (Bristol Myers Squibb) are both approved for first-line advanced or metastatic HCC, and the dual PD-1/CTLA-4 class that BMS occupies is mechanistically adjacent to STRIDE. Cross-trial comparisons are limited by differences in patient populations, staging, and trial design, but EMERALD-3 targets a distinct disease stage where no approved systemic option currently exists in the US or EU — a differentiation that is structural rather than merely incremental. AstraZeneca estimates more than 200,000 patients will be eligible for embolisation in 2026.
The press release notes that AstraZeneca is in discussions with global regulatory authorities. Full data are to be presented at a forthcoming medical meeting, where the magnitude of the PFS benefit and the shape of the OS curves will be made public. The STRIDE-plus-TACE arm without lenvatinib will also be watched: if its unplanned trends toward improvement hold in formal testing, it could offer a chemotherapy-free locoregional-plus-immunotherapy option for patients who may not tolerate a TKI.
What EMERALD-3 does not yet answer is whether the PFS gain translates to the survival benefit that would cement a label change. Intermediate-stage HCC trials have historically struggled to convert locoregional control into OS improvements, partly because salvage options after TACE failure are varied and can confound survival analyses. The interim OS trend is directionally consistent but not yet definitive. AstraZeneca has indicated the trial will continue to follow OS and other key secondary endpoints in both investigational arms.
Spot something wrong? Report an issue with this article