The prognosis of unresectable or [?? 0:03] gastric cancer remains poor and immune checkpoint
inhibitors, especially PD-1 blockade, shows some efficacy in later lines of treatment. So this
CheckMate 649 study was planned to investigate chemotherapy plus nivolumab as an anti-PD-1
therapy or nivolumab plus other immune checkpoint inhibitor, the CTLA-4 inhibitor ipilimumab, or
chemotherapy alone in the first line setting.

At the primary analysis after twelve months of minimum follow-up, first line nivolumab plus
chemotherapy demonstrated superior overall survival as well as a PFS benefit which led to approval
of this nivolumab plus chemo in several countries and changed the standard treatment. In this ASCO
GI meeting we reported our expanded analysis of efficacy and safety after the minimum follow-up of
24 months, which was one year longer than that of the interim analysis.

What was the methodology used in this study?

This CheckMate 649 study is an open-label globally conducted randomised phase III trial. As
mentioned before, in this ASCO GI presentation we mainly reported an updated analysis of
chemotherapy plus nivolumab versus chemotherapy alone. Here, more than 1,500 patients were
randomised to these two arms, regardless of PD-L1 expression and CPS score. Amongst them,
around 60% of patients, had a PD-L1 CPS 5 or higher.

What were your findings?

As the main result, clinically meaningful improvement in overall survival and PFS with nivolumab plus
chemotherapy was maintained in this longer follow-up. Actually, two year survival rates were 28%
with nivolumab plus chemotherapy and 19% with chemotherapy.

We also reported the PFS2, which was defined as time from randomisation to progression after the
first subsequent therapy, initiation of second subsequent therapy or death, whichever occurred earlier.
There was an improvement in PFS2 with a 25% reduction in risk of death or disease progression on
subsequent therapy.

Then, looking at the subgroup analysis, overall survival favoured nivolumab plus chemo across all key
subgroups with consistent results in patients with PD-L1 CPS 5 or higher. Then analysing efficacy in
various PD-L1 CPS subgroups, overall survival benefit with nivolumab plus chemo was enriched at
the higher CPS cut-off. Overall survival hazard ratio was 0.69 in patients with CPS 5 or higher versus
0.94 in patients with CPS less than 5.

Overall objective response rates were higher with nivolumab chemo versus chemotherapy, regardless
of PD-L1 CPS status and these responses were deeper in the waterfall plot and more durable with the
combination.

The safety profile of nivolumab chemo was consistent with the known safety profile of each agent and
in line with the previous reported twelve months of follow-up. We didn’t identify any new safety
signals.

How can these results impact the future treatment of gastrointestinal cancers?

I believe these results further support the nivolumab plus chemotherapy combination for its use as a
standard first line treatment in patients with advanced gastric, GJ or oesophageal adenocarcinoma.
 

What is next for this study?

Very importantly, additional biomarker analysis is already ongoing and I hope it may give us additional
insight about the optimal population for this combination. So, as mentioned before, already this
nivolumab chemo is approved in several countries, including the US, Japan and other countries, and
this indication is regardless of PD-L1 CPS status, so it is all comers.

On the other hand, this chemo nivolumab is mainly recommended for patients in the PD-L1 CPS 5 or
higher population, according to NCCN guidelines, because we observed a different magnitude of
survival benefit according to CPS level. Also, this chemo nivolumab is only approved for the CPS 5 or
higher population in EU countries. So this combination is definitive or the most recommended
treatment for patients with PD-L1 CPS 5 or higher but there are some controversies for its use in the
CPS <5 population. However, as shown in this presentation, some patients with CPS <5 achieved a
favourable outcome since we have a higher response rate even in this population. So clarification of
this population by additional biomarker might be very important work.

Is there anything else important that you would like to add?

I just want to mention that still several trials of immune checkpoint inhibitors are ongoing in the gastric
cancer field. For example, immune checkpoint inhibitor in combination with multi-kinase inhibitors
such as lenvatinib or regorafenib is under investigation in phase III trials. Furthermore, several
perioperative trials of checkpoint inhibitors are also still ongoing and will be reported within a couple of
years. So please stay tuned.