It was a great delight to present at ASCO the EMERALD-3 trial, a phase III clinical trial looking into adding on a doublet of checkpoint inhibitors, tremelimumab plus durvalumab, with or without lenvatinib in combination with chemoembolisation for patients who are embolisation eligible hepatocellular carcinoma patients.

The question is why because, as we all know, whenever we do a chemoembolisation we are releasing tumour antigens that will upregulate immune checkpoints. What a great opportunity, then, to apply checkpoint inhibitors and, as such, we looked into adding on the anti-PD-L1 durvalumab plus the anti-CTLA-4 tremelimumab, and also we looked into what if we add on another mid-level manager like an anti-VEGF, being lenvatinib.

Why is this important? Because, as we know, hepatocellular carcinoma is sadly one of the most common diseases worldwide and, interestingly, chemoembolisation is one of the most commonly used therapies worldwide for that purpose but, sadly, with the median progression free survival only about ten months at most. On the other hand, the chemoembolisation would really probably be empowered by the double checkpoint inhibitors, as we know we are choosing specifically the tremelimumab anti-CTLA-4 plus durvalumab anti-PD-L1becuase of the already proven improvement in survival that already we report even at six years with the HIMALAYA study.

So, anyway, the study randomised patients with embolisation eligible hepatocellular carcinoma to either doing this trial regimen, which is, as we said, tremelimumab plus durvalumab plus lenvatinib plus TACE, compared to TACE. But also looked into STRIDE plus TACE compared to TACE without lenvatinib. The primary endpoint was the progression free survival for the arm A, which is STRIDE plus lenvatinib plus TACE, compared to arm C, TACE, but also looked at OS and PFS for arm B, STRIDE plus TACE compared to TACE as well.

The statistical model was pretty rugged and well designed where we would only be permitted to analyse formerly the next step, i.e. if we do the PFS and we meet it and it’s really significant then we will be able to analyse the OS and on. But nonetheless we were able to present data still on all those four questions – PFS for A versus C, and OS for A versus C and also looking to PFS of B versus C and also OS for B versus C.

The study was pretty rugged and pretty large – 760 patients all in all. We did see the demographics did not differ at all between which patients or arm. Of course we got to the primary endpoint, the progression free survival for the STRIDE, lenvatinib plus TACE versus TACE – it was positive, 13 months compared to 9.8 with a hazard ratio of 0.7 and p-value of 0.00007.

Now, we looked into, as I mentioned, all the other analyses, being the PFS and OS. PFS number one for arm B, STRIDE plus TACE versus TACE, was also in favour and it showed a trend – 12.9 months versus 8.1 with the hazard ratio of 0.0062. Interestingly here, before I carry on to the OS, we were intrigued by a question that we were asked multiple times by people – what about A versus B? In other words, what about STRIDE plus lenvatinib plus TACE compared to STRIDE plus TACE? In other words, does the lenvatinib contribute anything? Surprisingly, it did not. In fact, we looked at subgroup analyses and there was no difference at all between STRIDE, lenvatinib plus TACE compared to STRIDE/TACE. And, of course, who wants to add more toxicity of the lenvatinib to an already beneficial STRIDE plus TACE? A small caveat for that is we got to the publishing in Japan, they have an excellent control of their management of lenvatinib. It could be that because non-viral is a contributory factor there in Japan, this is yet to be analysed. Considering that it is a very small cohort and made, really, only 7.5% of the patient population.

However, to go back, as we said, no difference at all and this may bring me in to talk about OS, especially the OS between STRIDE plus TACE compared to TACE. Mind-blowing, we did not even meet a median of the OS versus 32.9 and currently the hazard ratio is 0.7 with a p-value of 0.02. Of course we will wait until maturity is complete – 45.4% at this stage but hopefully we will have it completely done soon and then will report official final data. Understandably, people ask will this change. Slightly, there’s already two separations of the Kaplan-Meier curves and at this point in time it is extremely unlikely that we’re going to have that changing or altering the outcome of the study.

Responses definitely were there, especially partial response and stable disease, about one third for each, for both, especially arm A and B, much less for arm C, understandably, with the chemoembolisation alone. Interestingly, that’s a very valuable point – the chemoembolisation number of interventions was way less in arm A and B compared to arm C. This does really speak one more time for the input of the value of adding the chemoembolisation to the STRIDE regimen, or the STRIDE regimen to the chemoembolisation.

Adverse events no concern at all. It was exactly what we know about those three regimens, being STRIDE, lenvatinib and TACE. I will note, though, here that there was a little bit more drop-off when we got to the lenvatinib toxicity and, if anything, at least 78.7% of patients really interrupted the lenvatinib and, at the same time, about 30% discontinued the lenvatinib as well.

So, what did we learn? We learned that STRIDE plus lenvatinib plus TACE improved PFS. We also learned that STRIDE plus lenvatinib plus TACE favoured OS. But, more importantly, there was no difference between STRIDE, lenvatinib and TACE versus STRIDE/TACE and as such we can say that the STRIDE plus TACE, with the clinically meaningful TACE plus STRIDE compared to TACE PFS and OS, is very valuable. We suggest that STRIDE is the driver for the efficacy of the ITT population that we studied.

A formal analysis will be there but for now safety is acceptable and, yes, people asked me and, yes, as of next day, tomorrow in clinic, the new regimen is STRIDE plus TACE rather than TACE by itself.

It’s a great opportunity to let you know that this is already accepted in Lancet Oncology, it’s in press and we’ll see it soon. With this, big thanks for all my co-authors who I am talking on their behalf. More importantly, special thanks for all the patients and their loved ones worldwide from all continents that included in that study and I will here specifically express, amplify, that, yes, Africa was also part of the study where we know it has a very high incidence of liver cancer as well. Thanks again for listening.