KEYNOTE-756 results
For the ITT analysis, 635 patients assigned to pembrolizumab and 643 assigned to placebo were evaluable.
At the first interim assessment, conducted at a median follow-up of 33.2 months, with the longest follow-up out to 51.8 months, the co-primary endpoint of an improvement in pCR with immunotherapy was met. The pCR rate with pembrolizumab was 24.3%, compared with 15.6% with placebo, an absolute difference of 8.5% (P = .00005).
Data for the other co-primary endpoint, EFS, were not mature at the time of data cutoff, and will be reported at a future date, Dr. Cardoso said.
An analysis of pCR rates in subgroups showed that pembrolizumab benefited most patients, with the exception of those 65 years and older, patients with Eastern Cooperative Oncology Group performance status of 1 (vs. 0), patients who received their anthracycline regimen every 2 weeks rather than every 3, and node-negative patients.
The benefit was particular pronounced among patients with less than 10% ER positivity, she pointed out.
Adverse events in the neoadjuvant phase were primarily related to chemotherapy, with no major differences between the arms, although grade 3 or greater events were slightly more frequent with pembrolizumab (52.5% vs. 46.4%), and two patients in the pembrolizumab arm died (one death was from acute myocardial infarction considered related to the long QT syndrome; cause of the other patient’s death was not specified).
Adverse events leading to discontinuation were also more common with pembrolizumab (19.1% vs. 10.1%, respectively).
Immune-mediated adverse events of any grade were also higher in the immunotherapy arm, occurring in 32.8% of patients vs. 7% of patients in the placebo arm.
There were no deaths from immune-related adverse events.
Eye on safety
In his discussion, Dr. Johnston emphasized that “it’s important in a curative population that we don’t harm patients in a setting where we have a variety of other therapies available.”
Recalling the deaths of patients in the immunotherapy arm of each trial, he commented that “deaths in early breast cancer in a treatment setting are always a disaster, and we have to make sure that we manage these adverse events as we can best, and we know how to do that now.”
Checkmate 7FL was supported by Bristol Myers Squibb. Dr. Loi reported financial and nonfinancial interests with BMS and with other companies. KEYNOTE-756 was supported by Merck Sharp & Dohme. Dr. Cardoso disclosed consulting and institutional research support from Merck and others. Dr. Johnston reported consulting or advisory roles, honoraria, and research funding from several companies, not including either BMS or Merck.