What I did at the virtual ESMO congress, I presented the first results of CheckMate 649 which is a worldwide randomised phase III study on nivolumab plus chemotherapy versus chemotherapy as first line treatment for advanced gastric cancer, gastroesophageal junction and oesophageal adenocarcinoma.
As you know, standard first line chemotherapy in this disease results mostly in less than one year. Nivolumab provided superior overall survival versus placebo in the ATTRACTION-2 study in heavily pre-treated advanced gastric cancer. Additionally, nivolumab and chemo showed promising anti-tumour activity in first line chemotherapy in the ATTRACTION-4 study. Now, PD-L1 expression by a CPS cut-off of 5 or higher has shown enrichment for efficacy with checkpoint inhibitors in studies before.
Our trial I presented is the CheckMate 649 which had three arms: there was one arm which is nivolumab plus ipilimumab that is immunotherapy alone. This will be reported maybe next year since this part is still closed and blinded. What I reported at ESMO is the two arms on chemotherapy, XELOX or FOLFOX, plus or minus nivolumab in 1,581 patients. 60% of these patients were PD-L1 CPS 5 or higher. CPS is the amount of PD-L1 positive cells, not only tumour, but also in the tumour microenvironment – lymphocytes and macrophages. I reported on the dual primary endpoints of overall survival and PFS, progression free survival.
The baseline characteristics in both arms were consistent and well balanced. Mostly non-Asians and patients with metastatic disease were included. With regards to the duration of treatment, the patients who were treated with nivolumab plus chemo had a longer duration of treatment with around 7 months versus chemotherapy with only 5 months. Disease progression in around two-thirds of all patients was the most common reason for treatment discontinuation.
Now I come to the overall survival which was the first primary endpoint. The overall survival was highly statistically better for nivolumab plus chemo versus chemotherapy alone with 14.4 months for nivolumab plus chemo versus 11.1 months for chemotherapy alone with a hazard ratio of 0.71, a p-value of 0.0001, so that was highly statistically significant.
Since you cannot see the Kaplan-Meier curves, I’ll just describe them. 11% of patients had higher one year survival which was 57% with nivolumab plus chemo versus 46% with chemotherapy alone. The separation of the Kaplan-Meier curve was continuous from the beginning to the end of the Kaplan-Meier curve. Again, always around 10% difference from the beginning to nearly three years. However, at the end of the Kaplan-Meier curves it is still not yet fully clear since we still have to wait for later and longer median follow-up.
Now, with regards to the overall survival of the CPS1 or higher, and all randomised patients, again the median OS was in favour of the nivolumab plus chemo and again statistically significant with a hazard ratio of 0.77 versus 0.8 in all randomised patients. With regards to progression free survival, again in the primary population with CPS 5 or higher there was a highly statistical benefit with median PFS from 7.7 for nivolumab chemo versus chemo with 6.0 months. So, again, a hazard ratio of 0.68, statistically significant with a p-value of 0.0001.
So, taken together, the PFS benefit was seen in all subgroups. With regards to the subgroups, for overall survival there was also a clear favourable prognosis for patients with nivolumab plus chemo. Across multiple pre-specified subgroups like region, ECOG performance status or primary tumour location but also MSI and MSS, overall survival consistently favoured nivolumab plus chemo.
With regards to response and duration, there was again a clear advantage with nivolumab plus chemo with a response rate of 60% with the combination versus 45% with chemotherapy alone. So that means 15% benefit of response with chemo plus nivolumab. Again, the duration was clearly longer in the patient sub-population with CPS 5 or higher with the combination of nivolumab plus chemo.
With regards to safety there were no new safety signals. Additionally, the treatment related adverse events were numerically higher for nivolumab plus chemo but the risk-benefit ratio was also favourable for the combination.
With regards to treatment related adverse events with potential immunologic aetiology, according to organ groups most of them were low and grade 3/4 events of immunologic aetiology occurred in less than 5%. There were no grade 5 events.
So, taken together, nivolumab is the first PD-1 inhibitor to demonstrate superior overall survival and progression free survival in the combination with chemotherapy versus chemotherapy alone in previously untreated patients with advanced gastric, gastroesophageal junction and oesophageal adenocarcinoma. There was a clear, highly clinically meaningful overall survival benefit in these patients, particularly for CPS 5 or higher but also for CPS 1 and higher and all randomised patients. There was a statistically significant benefit for the PFS.
Therefore, I concluded at ESMO and I’m very confident that nivolumab plus chemotherapy presents and will present a new potential standard first line treatment for patients in this disease.
Are there next steps planned?
Yes, of course. Since we don’t have all biomarkers tested and also many additional analyses we need for approval at FDA and EMA, the analysis will go on, particularly for MSI, MSS, for other subgroups like TMB and so on. Then we hope soon to have this combination approved by the authorities for our patients.