Pathologic complete response rates improved with the neoadjuvant and adjuvant treatment of pembrolizumab in TNBC patients

Prof Peter Schmid - St Bartholomew's Hospital London, London, UK

As we’re all aware, patients with triple negative breast cancer who get neoadjuvant chemotherapy and they achieve what we call a pathological complete response, which means the complete disappearance of the cancer prior to surgery, they have a sustained clinical benefit. The meta-analysis data show a strong link between pathCR and long-term outcome for long-term event free survival with a hazard ratio of 0.24 and overall survival with a hazard ratio of 0.16. We know if you give neoadjuvant chemotherapy with anthracyclines and taxanes the pathCR rates are around 40%. If you add in platinum they’re around 50-55%. We had, prior to this study, smaller trials that demonstrated that if we add pembrolizumab or other checkpoint inhibitors to neoadjuvant chemotherapy that the treatment is safe and showed promising early activity in triple negative breast cancer.

So we designed a phase III trial, KEYNOTE-522, in patients with operable triple negative breast cancer with either T1c tumours and node positive disease or T2-T4 tumours irrespective of the nodal status. Patients were randomised between pembrolizumab or placebo given in combination with twelve weeks of carboplatin paclitaxel chemotherapy followed by twelve weeks of AC or EC chemotherapy. Patients then underwent surgery and after surgery they carried on with pembrolizumab or with placebo for a total of one year.

The trial had two main primary endpoints and at this point in time we have mature data and definitive data for the first primary endpoint, pathCR, the complete disappearance of all cancer. We also have early data for event free survival but the follow-up is very short at this point in time. If you look at the definitive pathCR data based on 602 patients it is a statistically significant and, in my opinion, meaningful difference in pathCR rates of 13.6% between the two groups with 51.2% pathCR in patients receiving the control chemotherapy and 64.8% of patients receiving chemotherapy and pembrolizumab.

With a short-term, this is short from an adjuvant point of view, follow-up of only 15.5 months we see a separation of the curves but not statistically significant at this point in time, considering the pre-calculated p-value boundary of significance of 0.000051. We have an estimated difference in the estimated eighteen months event free survival rates of 19.3% to 85.3% in the placebo group.

If you look at the benefit in terms of pathCR by different disease stages, and the trial included patients with stage 2 disease and stage 3 disease, as you can see consistently across all stages there is a benefit from adding immune therapy to chemotherapy. But what is impressive, in my opinion, is looking at the right side of the slide where you see patients with stage 3a and stage 3b, in other words higher risk cancers, where the delta, the difference in pathCR rates, is around 25% for stage 3a and stage 3b patients. Obviously in the control arm, because these are higher risk patients, we see that the pathCR rate is going down but we maintain a high pathCR rate in patients with the addition of pembrolizumab.

If you look at patients with and without lymph node involvement, again lymph node involvement being associated with a higher risk, also associated with lower pathCR rates, we can see that there’s a consistent benefit in patients who are, on the left, lymph node negative or, on the right, lymph node positive. But I would again like to point out that patients with higher risk disease have a difference of benefit of 20.6% from the addition of pembrolizumab to chemotherapy.

One of the key questions is do all patients benefit equally from immune therapy and one of the markers in metastatic disease that’s linked with response is something called the PD-L1 status. Now, we checked this out in this trial using an assay called CPS and patients who were defined as CPS less than one are PD-L1 negative; patients who had more CPS, so CPS higher than one, are PD-L1 positive. As you can see on the left side of the slide, patients with PD-L1 negative tumours, as per this definition, have a similar benefit from the addition of immune therapy to chemotherapy compared to patients with PD-L1 positive patients. We also looked, because other studies showed if patients have a very high expression of PD-L1 and have very immunogenic tumours they may have an even higher benefit of immune therapy, and we wondered whether most of the benefit was driven by this group. But, as you can see, the absolute difference in the delta between chemotherapy and the combination is consistent whether we use a cut-off of one, of ten or twenty. Of note, if you look at the group of patients with a CPS of ten or more, inflamed tumours, or CPS of twenty we saw a pathCR rate of 78% and 81% in these patients, a pathCR rate we have never seen in such an aggressive disease group.

One of the questions that came up when we first discussed the trial results was was the chemotherapy exposure possibly different and affected by the addition of immune therapy. We saw that if you have very stringent definitions of what is full dose of chemotherapy approximately 80% of patients in each arm had full exposure of chemotherapy. If you look at, and this is an exploratory analysis, if you look at the pathCR rates in patients who had a full exposure of chemotherapy or patients who had less than full exposure of chemotherapy we again see the difference or the benefit from adding pembrolizumab is around 14% for full exposure, 15% for less than full exposure, meaning that the benefit from adding immune therapy is maintained even in those patients who, for whatever reason it was, received less than full exposures of chemotherapy.

So our summary at this point in time is that the addition of pembrolizumab to neoadjuvant chemotherapy produces a significant and substantial benefit in pathCR rates and we see a larger magnitude of the pCR benefit versus chemotherapy alone in patients with higher stage disease, in this instance stage 3 disease, and node positive disease.

We also saw that the benefit of neoadjuvant chemotherapy and pembrolizumab on pathCR was observed in patients who received less than the planned chemotherapy although, of course, the absolute pathCR rates were lower in those patients who received less than chemotherapy, also irrespective of the PD-L1 threshold, the CPS threshold for the assay used in this trial. Neoadjuvant pembrolizumab and chemotherapy were also associated with a higher rate of RCB01 data we’ll show later this morning and we’ll see that immune mediated adverse events were consistent with the known profiles of each regimen and we didn’t see any new safety concerns.

We feel at this point in time that further follow-up is needed to confirm the EFS benefit and we also need longer term safety profile data. We also present in the future additional biomarker data which will be helpful, including TILs and BRCA1 status. Thank you very much.