Efficacy of pembrolizumab in pre-treated breast cancer harboring high mutational load 140-290-results from the Drug Rediscovery Protocol (DRUP)

Pembrolizumab combined with chemotherapy is approved for treatment-naive, advanced triple-negative breast cancer (BC), but has also demonstrated efficacy in subsets of pre-treated BC patients (pts). However, patient and biomarker selection remain suboptimal. In the Drug Rediscovery Protocol (DRUP), we studied efficacy of pembrolizumab in pts with metastatic BC who exhausted standard of care treatment and had a tumor mutational load (ML) between 140-290. DRUP is an ongoing, non-randomized, platform trial evaluating efficacy of commercially available targeted- and immunotherapies outside their labelled indications, in multiple parallel cohorts of patients with cancer (NCT02925234). Eligible pts had treatment-refractory, metastatic mismatch repair proficient BC harboring 140-290 somatic missense variants across the tumor genome, as found with whole genome sequencing (WGS). Pembrolizumab 200 mg monotherapy every three weeks was administered until disease progression or unacceptable toxicity. Primary endpoints were clinical benefit (CB: objective response (OR) or stable disease (SD) ≥ 16 weeks according to RECIST1.1) and safety. Pts were enrolled using a Simon like 2-stage model, with 8 pts in stage 1 and up to 24 pts in stage 2 if at least 1/8 pts had CB in stage 1. Fresh frozen biopsies for biomarker analysis, including WGS, were obtained at baseline. From November 2017 to September 2021, twenty-four evaluable pts were enrolled. CB was observed in three pts (12.5% (95% CI 2.7-32.4), all had partial response. At data cut-off, the (near complete) response of one patient was ongoing for 21 months, other responses lasted 10 and 6 months. Median progression-free survival was 1.8 months (95% CI 1.7-1.9) and overall survival 6.5 months (95% CI 3.1-10.8). Height of ML was unrelated to CB; median ML in pts with CB was 160 (range 143-176) vs. 183 (148-274) in pts without CB. No unexpected toxicity was observed. Even in relatively low mutational loads, pembrolizumab induced objective and clinically relevant responses in a small subset of heavily pre-treated BC pts. To identify those pts who will benefit from this treatment, additional predictive biomarkers are urgently needed.