Sara M. Tolaney, MD, MPH: In triple-negative breast cancer, we have seen that the benefit of checkpoint inhibitors as monotherapy is really quite modest. So, in looking at the KEYNOTE-086 study, we saw only about a 5% objective response rate in the second-line setting and beyond, regardless of PD-L1 status. More recently in KEYNOTE-119, chemotherapy was compared with pembrolizumab monotherapy where pembrolizumab was not found to be superior to chemotherapy in the overall population. We think what we’re seeing is that checkpoint inhibitors alone do have modest benefit.
There does seem to be greater benefit for checkpoint inhibitor monotherapy in the first-line setting and in a patient with PD-L1—positive disease where we see—in KEYNOTE-086—response rates of around 23% in that population. Potentially in those patients with high PD-L1 expression, checkpoint inhibition as monotherapy could potentially be performing better than chemotherapy. If you look at KEYNOTE-119 in those patients who had a CPS [combined positive score] greater than 20, you did start to see the curves separate favoring the pembrolizumab over the chemotherapy at that point.
But, currently, I don’t think there’s much of a role for checkpoint inhibitors alone until we’re able to find an optimal biomarker to select patients for them. And we do see data clearly showing that benefit with chemotherapy and checkpoint inhibition is much more robust as we’ve seen from IMpassion130 where we’re now seeing a survival benefit from the addition of checkpoint inhibition to chemotherapy.
At this time, the only biomarker we have to select patients for checkpoint inhibition with chemotherapy in metastatic triple-negative disease is PD-L1 … and that would be using the SP142 antibody for testing for PD-L1. I do think there are many other biomarkers that are being explored and that will need further work. We have seen data from both KEYNOTE-086 and KEYNOTE-119 that responses were higher in patients who had higher tumor-infiltrating lymphocytes. And that certainly seems to potentially be a biomarker to help predict benefit to checkpoint inhibition alone.
There are some very interesting newer data that have come out about tumor mutation burden. Preliminary data from a retrospective analysis of patients who had received either checkpoint inhibition alone or checkpoint with chemotherapy in which patients with high TMB [tumor mutation burden] defined as greater than 10 did seem to have better overall survival relative to patients who had low TMB. And this was regardless of PD-L1 status. So I think again, we still have a long way to go to develop optimal biomarkers for prediction of checkpoint inhibition. But, as of today, the only biomarker we have to predict benefit is PD-L1.
Transcript Edited for Clarity