Adding pembrolizumab to chemotherapy substantially increases progression-free survival (PFS) in treatment-naive advanced or metastatic triple-negative breast cancer (TNBC) regardless of chemotherapy type, suggests an analysis of the clinical trial KEYNOTE-355.
There was also a trend for improved outcomes with increasing programmed death–ligand 1 (PD-L1) expression in the tumor, as measured by combined positive score (CPS).
“These data further support a role for the addition of pembro to standard chemo for the first-line treatment of metastatic TNBC,” said study presenter Hope S. Rugo, MD, from the Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco.
The research was presented at the 2020 San Antonio Breast Cancer Symposium on Dec. 10.
Last month, pembrolizumab was granted accelerated approval by the Food and Drug Administration in combination with chemotherapy for the treatment of patients with locally recurrent unresectable or metastatic TNBC whose tumors express PD-L1 (CPS ≥10).
The approval was based on data from KEYNOTE-355, which involved almost 850 women with previously untreated locally recurrent inoperable or metastatic TNBC randomized 2:1 to pembrolizumab plus investigator’s choice of chemotherapy who were followed for 2 years.
For the current analysis, patients were stratified by PD-L1 CPS in the tumor, including over 320 patients with CPS ≥10, and by accompanying chemotherapy regimen.
In the overall intention-to-treat (ITT) population (n = 847), median PFS was longer with pembrolizumab plus chemotherapy versus placebo plus chemotherapy, at 9.7 months versus 5.6 months (hazard ratio, 0.82).
PFS improved step-wise with increased PD-L1 expression. In patients with PD-L1 CPS ≥1, the HR was 0.74, and in those with PD-L1 CPS ≥10, it was 0.65.
A similar incremental improvement by PD-L1 expression was seen in the overall response rate, with the rate topping out at 53.2% in the pembrolizumab plus chemotherapy arm, among the PD-L1 CPS ≥10 group.
Duration of response told a similar story, with the pembro-chemo combination providing superior results and the treatment effect increasing with PD-L1 enrichment.
Study discussant Sylvia Adams, MD, New York University Perlmutter Cancer Center, New York, said the “consistency of treatment effect” with different chemotherapy backbones seen in the study is “very important, as it is currently unknown what the optimal backbone is.”
She also noted that the chemotherapy analysis presented by Dr. Rugo was “exploratory” and “not powered to show the winner of the chemotherapy backbone.”
Nevertheless, in the postpresentation debate, Ian Krop, MD, PhD, Dana-Farber Cancer Institute, Boston, said that there are “several questions over the chemotherapy partner,” including whether there were differences in the populations who received each type of regimen.
Dr. Rugo replied that “because the trial wasn’t powered to look at the separate chemotherapy groups with any statistical significance ... it’s really impossible to draw any specific conclusions because it’s the overall population that’s evaluated.”
Asked about when overall survival results will be presented, Dr. Rugo said that “everybody is very interested” in that, “and we expect these results to be available next year.”
Study details
For KEYNOTE-355, researchers recruited women with previously untreated metastatic TNBC who had completed treatment with curative intent ≥6 months prior to their first disease recurrence.
They were randomized 2:1 to pembrolizumab or placebo plus investigator’s choice of chemotherapy from nab-paclitaxel, paclitaxel, or gemcitabine/carboplatin for up to 35 administrations of pembrolizumab or placebo or until progression, intolerable toxicity, or cessation of treatment.
Crossover was not allowed. Patients were stratified by type of chemotherapy, PD-L1 expression in the tumor, and prior treatment in the neoadjuvant or adjuvant setting with the same class of chemotherapy.
Response was assessed with imaging every 8 weeks until week 24, then every 9 weeks during the first year of follow-up, and then every 12 weeks.
Of 847 randomized patients, 566 received pembrolizumab plus chemotherapy and 281 were assigned to the placebo group. The median age was 53 years in both groups.
The majority (75.1%) of patients in both groups were PD-L1 positive with a centrally assessed CPS ≥1, while 38.9% of patients in the pembrolizumab arm and 36.7% of those given placebo had a CPS ≥10.
After a median follow-up of 25.9 months, 16 patients given pembrolizumab had completed the study and 33 were still ongoing.
This compares with five patients having completed the placebo arm, and 12 still ongoing, after a median follow-up of 26.3 months.
The overall response rate was higher with pembrolizumab plus chemotherapy in the ITT population, at 41.0% versus 35.9%, rising to 45.2% versus 37.9% in patients with PD-L1 CPS ≥1 and 53.2% versus 39.8% in the PD-L1 CPS ≥10 group.
Again, when the groups were stratified by on-study chemotherapy, the overall response rate was higher with pembrolizumab versus placebo regardless of the chemotherapy partner.
Finally, the duration of response with pembrolizumab plus chemotherapy was longer than that seen with placebo, at a median of 10.1 months versus 6.4 months in the ITT population.
In the PD-L1 CPS ≥1 group, the duration of response was 10.1 months versus 6.5 months, rising to 19.3 months versus 7.3 months in the PD-L1 CPS ≥10 group.
Dr. Adams nevertheless said that PD-L1 remains an “imperfect” biomarker in metastatic TNBC, although it is “the best to date.” Furthermore, the IMpassion130 trial, featuring atezolizumab, showed that there is “very poor” analytic and clinical concordance between assays, which “complicates clinical decision-making.”
This study was sponsored by Merck. Dr. Rugo, Dr. Adams, and Dr. Krop have disclosed financial ties to multiple pharmaceutical companies, including Merck.
A version of this article first appeared on Medscape.com.