PD-1 Inhibitor Falls Short in Triple-Negative Breast Cancer

Single-agent pembrolizumab (Keytruda) failed to improve survival versus chemotherapy in previously treated metastatic triple-negative breast cancer (TNBC), a large randomized trial showed.

The primary analysis in patients with a PD-L1 combined positive score (CPS) ≥10 showed a median overall survival (OS) of 12.7 months with pembrolizumab and 11.6 months with investigator’s choice of chemotherapy. Analyses of patients with CPS ≥1 and the overall population also showed no significant advantage for pembrolizumab.

Pembrolizumab led to fewer grade 3/4 treatment-related adverse events (TRAEs), reported Eric Winer, MD, of Dana-Farber Cancer Institute in Boston, and colleagues in Lancet Oncology.

“These findings might inform future research of pembrolizumab monotherapy for selected subpopulations of patients, especially those with PD-L1-enriched tumors, and inform a combinatorial approach for the treatment of patients with metastatic triple-negative breast cancer,” the authors concluded.

The study reflected the ongoing efforts to define checkpoint inhibitors’ role in TNBC. The IMpassion130 trial showed improvement in progression-free survival (PFS) and OS when atezolizumab (Tecentriq) was added to chemotherapy as initial treatment for locally advanced or metastatic TNBC. However, the IMpassion131 trial failed to duplicate the findings in previously treated metastatic TNBC.

In two previous trials, single-agent pembrolizumab demonstrated activity in previously treated metastatic TNBC. In the phase I KEYNOTE-012 trial, the immune checkpoint inhibitor resulted in an objective response rate (ORR) of 18.5%, and the phase II KEYNOTE-086 trial yielded an ORR of 5.3%. In both trials, responses proved durable in most cases.

The favorable results with pembrolizumab continued in the phase III KEYNOTE-355 trial, which showed significant improvement in PFS with the addition of the PD-1 inhibitor to chemotherapy for patients with previously untreated locally advanced/metastatic TNBC and a CPS ≥10. Follow-up for OS continues.

Winer and colleagues reported findings from the randomized, phase III KEYNOTE-119 trial, comparing pembrolizumab monotherapy versus single-agent chemotherapy as second or third-line therapy for metastatic TNBC. Investigators in 31 countries screened 1,098 patients and randomized 622 to the two treatment arms. Randomized treatment continued until disease progression or development of unacceptable toxicity. The primary endpoint was OS in patients with a CPS ≥10, in patients with CPS ≥1, and in the total study population, evaluated in a hierarchical manner.

The study population had a median age of 52, and 47% had an ECOG performance status of 1. About two-thirds of the patients had a PD-L1 CPS ≥1, and 31% had a CPS ≥10.

After a median follow-up of about 31 months, analysis of the CPS ≥10 subgroup showed the pembrolizumab arm had a 22% reduction in the survival hazard, which did not achieve statistical significance (95% CI 0.57-1.06, P=0.057). The CPS ≥1 analysis yielded median OS values of 10.7 months for the pembrolizumab arm and 10.2 months for the chemotherapy arm (HR 0.86, 95% CI 0.69-1.06, P=0.073). Analysis of the overall population showed a median OS of 9.9 months with pembrolizumab and 10.8 months with chemotherapy (HR 0.97, 95% CI 0.82-1.15).

A post hoc exploratory analysis did suggest that pembrolizumab activity might increase with higher CPS values. The analysis, which involved fewer than 20% of the study population, showed that patients with CPS ≥20 had a median OS of 14.9 months with pembrolizumab versus 12.5 months with chemotherapy (HR 0.58, 95% CI 0.38-0.88).

The most common grade 3/4 TRAEs all occurred more often with chemotherapy: anemia (3% vs 1%), decreased white blood cells (5% vs <1%), decreased neutrophil count (10% vs <1%), and neutropenia (13% vs 0%). Serious AEs occurred in 20% of each group.

The findings are consistent with the history of single-arm anti-PD-1/L1 therapy for breast cancer, said Eitan Amir, MD, and David W. Cescon, MD, PhD, both of Princess Margaret Cancer Center in Toronto, in an accompanying commentary. Regardless of the type of breast cancer, checkpoint inhibitors have produced low response rates, but responses have been durable.

“Given the low response rates observed in the overall population with pretreated triple-negative breast cancer in previous studies of anti-PD-1 or anti-PD-L1 monotherapy, the primary results of KEYNOTE-119 are unsurprising,” they wrote. “Since KEYNOTE-119 was launched, clinical development has focused principally on combinations of chemotherapy and immunotherapy in the first-line setting.”

The post hoc analysis is intriguing but should be interpreted cautiously.

“The finding that this higher PD-L1 expression threshold might be a predictor of pembrolizumab monotherapy benefit adds to previously observed associations with single-drug immunotherapy benefit, including de novo metastatic disease, absence of previous chemotherapy, normal lactate dehydrogenase, lung or nodal involvement, and absence of liver metastases … . It would be intriguing to see if similar results can be validated in triple-negative breast cancer,” Amir and Cescon added.