It is a great honour and delight on behalf of all dear colleagues to present the phase III final results of
the HIMALAYA study. As we all know, the HIMALAYA study was based on the perspective that a
combination of therapy is needed for improving on outcome with regard to patients with advanced or
unresectable hepatocellular carcinoma. Many other combinations have been tried – anti-VEGF, anti-
FGF etc. – but over here we are going all the way up to the top of the chain of command and the
lymph node itself where we’re looking at anti-CTLA-4 plus anti-PD-L1.
The HIMALAYA study was a standard study for patients with advanced unresectable hepatocellular
carcinoma – good performance status, Child-Pugh score A – and the stratification factors were
macrovascular invasion, aetiology, hepatitis C, hepatitis B and other non-viral, and performance
The study originally had four arms; one of them was ultimately taken out or removed because the
75mg of tremelimumab times four plus durvalumab, as we know from the AZ22 study that was
already published in JCO, did not fare any benefit compared to single agent durvalumab. As such, the
HIMALAYA study ended up with three arms – 300mg of tremelimumab only for one dose plus
durvalumab that’s given every four weeks; durvalumab as single agent every four weeks; and
sorafenib as a standard of care, 400mg twice a day.
The primary endpoint was to compare the tremelimumab 300mg, T300 +D, durvalumab, to sorafenib,
looking for superiority for overall survival. A secondary endpoint was durvalumab single agent versus
sorafenib looking for non-inferiority.
The study, if anything, had the standard background with regard to the demographics of the patients.
Of note here it’s important to restress that about 40% of patients were from Asia, 60% from the rest of
the world, including Japan, and the aetiologies were very much global – 30% hepatitis B, 30%
hepatitis C and 40% for the others non-viral.
The outcome showed an improvement in survival in favour of the tremelimumab 300 plus durvalumab,
T300+D – 16.4 months median overall survival compared to 13.8 for the sorafenib.
Now, it’s very important to note that the delineation of the survival did not start until about 4 months or
so and, as you can see, at one example at nine months when we did the cut-off and we have seen
the hazard ratio below nine months was 0.87, after nine months it was 0.7. That kind of suppression
continued to evolve over time with the landmark analyses at 18 months, at 24 months and at 36
months. By 36 months, three years, 31% of the patients on the T300+D were still alive compared to
20% on the sorafenib arm.
The secondary endpoint looking at durvalumab versus sorafenib showed no difference with regard to
the overall survival – 16.6 months for the durvalumab, 13.8 for the sorafenib, hazard ratio is 0.86.
Despite that, another separation of the two curves was noted as per nine months and it evolved over
time enough that by three years 24% of the patients on durvalumab were still alive, compared to 20%
It’s very important to look into the Forest plots. The benefit for tremelimumab 300+D was noted
among all different subgroups, independent of sex, age, region, aetiology and extent of disease. One
might delineate that no difference was noted in the female group but, remember, this was a very small
group, considering the biology of the disease was less than 100 patients. And with regard to the
hepatitis C also there was no difference but this was because of a certain imbalance that in
retrospective we reanalysed and found that if we correct for intrahepatic spread as well as ALBI score
the hazard ratio for the hepatitis C also improves to 0.89 in favour of the T300+D.
The progression free survival did not differ between the two arms of the durvalumab plus
tremelimumab, or T300+D, versus sorafenib. If anything, some separation happened afterwards over
time. Most importantly, though, it is expected that the progression free survival will not necessarily
matter because of the priming of the T-cells it takes time. Despite that, the response was noted as
early as two months for the T300+D but took about four months for the sorafenib.
More importantly, by the time of receiving only one cycle or more, and even though patients
progressed, they continued on therapy: 47% of them on the tremelimumab 300+D versus 34% for the
sorafenib. This again tells us that the checkpoint inhibitors used, especially dual checkpoint inhibitors,
will require some priming time. At the same time we still can see the effect of the T-cells being present
there with the immediate response at two months but, of course, it will evolve over time.
Was the combination safe? By all means. If anything, we have seen that the grade 3 and 4 treatment
related adverse events were the highest in sorafenib – 37% compared to 26% for the T300+D. The
treatment related hepatic and haemorrhage [?? 5:44] events were the one that we mostly could be
concerned about, [?? 5:49] hepatic and haemorrhage. There were multiple in the T300+D but not to
any concern of raising alarming concerns per se – 7% only for the hepatic ones and barely apparent
for the haemorrhage, 0.5%, and none of them were oesophageal varices haemorrhages. Immune
adverse events were the classic and expected hepatic events – diarrhoea and dermatitis. Yes, they
were more present in the T300+D but not to any concerning status.
If anything we concluded that the HIMALAYA study, a phase III trial which was, number one, global,
including patients with different aetiology of disease from all over the world and showed an
improvement in survival in favour of the T300+D compared to sorafenib. That improvement in survival
continued over time and we have seen that by 36 months, or three years, there is still 31%, one third,
of patients on T300+D were still alive. If anything, the T300+D and durvalumab single agent shows
both an acceptable safety profile and, of note, the durvalumab compared to sorafenib showed a
comparator non-inferiority to the sorafenib and still, however, with the favourable benefit/risk profile.
By all means this was a great effort from all over the world – 1,325 patients from throughout the world
– and thanks for all the co-investigators and colleagues from all over the world for being part of the
How can these result impact the future treatment of hepatocellular carcinoma?
By all means the HIMALAYA study is a positive study, showed a direct and impressive improvement
in outcome, and no doubt that we anticipate and hope that the agencies throughout will accept
tremelimumab plus durvalumab as a new standard of care.