KEYNOTE-756 trial shows increased response in ER+/HER2- breast cancer

Dr Fatima Cardoso - Champalimaud Cancer Center, Lisbon, Portugal

In KEYNOTE-756 patients were randomised between a backbone of chemotherapy with the regimen paclitaxel 12 weeks followed by four cycles of anthracyclines and were randomised between having pembrolizumab or placebo. This was a double-blinded phase III randomised trial. Very importantly, this trial has two primary endpoints, one is pCR rate and the other one is event free survival. It’s also important to characterise the type of patients that enter the trial. So this trial was for high-risk early breast cancer defined as T1c/T2 with positive nodes, so N1 or N2, or T3/T4 with either N0 until N2. All tumours had to be grade 3 and obviously no treatment naïve.

So, 1,278 patients were randomised and there were several stratification factors. It’s also important to note that this trial is powered to detect a difference both in pCR and event free survival. Actually, the alpha was split between the two, 0.5 for pCR, 2% for EFS. If pCR is positive, as was the case, then the alpha is transferred for event free survival and if event free survival is positive it will be transferred to overall survival.

In the characteristics of the patients in general, highlighting that about 35% had T3/T4 tumours, 90% had node positive disease, 75% of tumours were PD-L1 positive and the vast majority, 95%, had ER positivity above 10%. So the ER low population represents only 5%.

What did you find?

This analysis that was presented at ESMO is the final analysis for pCR. It was run with a cut-off date of May 2023, at the time ten months after the last patient was randomised. All patients had gone through the neoadjuvant phase and only 58 patients were still on treatment in the adjuvant phase. It was also the first interim analysis for event free survival but these results were not released by the IDMC.

Regarding the pCR rate, we have seen an increase with the addition of pembrolizumab with a delta of 8.5%. So from 15.6 to 24.3% according to the definition of pCR. The benefit was seen with or without DCIS included in that definition with a singular magnitude. In terms of subgroups, in all subgroups there was a benefit with the addition of pembrolizumab. There was no subgroup where we didn’t see this benefit, including in PD-L1 positive and negative, albeit the magnitude of benefit is bigger when PD-L1 is positive. Also regarding this population of ER low we see what seems to be a bigger benefit but with a very large confidence interval due to the fact that this represents only 5% of the population.

For safety we have not seen any new safety signals. We see in the neoadjuvant phase the majority of side effects are chemotherapy related with some small differences between the two arms with a slightly higher discontinuation rate in the pembrolizumab arm. Only one death, related to myocardial infarction, considered by the investigator related to chemotherapy.

In what concerns the immune-mediated events, nothing also new here. The majority were hypo- or hyperthyroidism as it would be natural since what we know about pembrolizumab. Very small percentages of the other immune-mediated events, no deaths and a slightly higher discontinuation rate in the pembrolizumab arm.

So, in conclusion, we can say that KEYNOTE-756 is the first fully approved phase III randomised trial of immunotherapy in high-risk early breast cancer with ER positive HER2 negative tumours. It met its dual primary endpoint pCR with a statistically significant increase of 8.5%, regardless of PD-L1 status. No new safety signals were seen. We continue to have follow-up and the study continues to be powered to test event free survival as a primary endpoint. Follow-up will continue and more results will be presented in the future.

What impact could these results have?

We should not change our standard of care based on pCR alone. From my perspective, we should not change in any subtype of breast cancer but in particular in this subtype we need to have the event free survival data before we change the standard of care. However, in the same session another study was presented also with immunotherapy in a very similar population with a different agent and the same results in terms of pCR were seen. So we have a consistency of increased pCR rates in this high-risk early breast cancer ER positive HER2 negative. So we can have some hope that we will see an impact in event free survival.

When will you be reporting on EFS?

It will depend on the analysis but in principle in about one or two years.