Event-free survival by residual cancer burden after neoadjuvant pembrolizumab plus chemotherapy versus placebo plus chemotherapy for early TNBC: Exploratory analysis from KEYNOTE-522.

503 Background: KEYNOTE-522 (NCT03036488) tested the benefit from adding pembrolizumab (pembro) to chemotherapy (chemo) in patients (pts) with early TNBC. The primary results showed statistically significant and clinically meaningful improvements in pCR and EFS with pembro.Prior studies have shown the prognostic value of the residual cancer burden (RCB) method to quantify the extent of residual disease after neoadjuvant chemo. In this exploratory analysis, we assessed EFS by RCB in KEYNOTE-522. Methods: 1174 pts with previously untreated, nonmetastatic, stage T1c/N1-2 or T2-4/N0-2 TNBC were randomized 2:1 to pembro 200 mg Q3W or placebo (pbo) given with 4 cycles of paclitaxel + carboplatin, then 4 cycles of doxorubicin or epirubicin + cyclophosphamide. After definitive surgery, pts received pembro or pbo for 9 cycles or until recurrence or unacceptable toxicity. Dual primary endpoints are pCR and EFS. RCB was assessed by the local pathologist at the time of surgery. The association between RCB categories (RCB-0, -1, -2, -3, corresponding to increasingly larger residual cancer) and EFS was assessed based on a Cox regression model with treatment as a covariate. Results: Median follow-up was 39.1 months at data cutoff (23 MAR 2021). Pembro shifted RCB to lower categories across the entire spectrum (Table). The HRs (95% CI) for EFS were 0.70 (0.38 - 1.31) for RCB-0 (equivalent to pCR), 0.92 (0.39 - 2.20) for RCB-1, 0.52 (0.32 - 0.82) for RCB-2, and 1.24 (0.69 - 2.23) for RCB-3. The most common EFS event in both arms was distant recurrence, which occurred in fewer pts in the pembro arm in all RCB categories. Conclusions: Increased RCB score was associated with worse EFS. Pts with residual disease had lower RCB values in the pembro arm, including fewer pts with RCB-3. Pembro + chemo prolonged EFS vs chemo alone in the RCB-0, -1, and -2 categories; the small sample size limits interpretation in the RCB-3 category. The small subset of pts with extensive residual disease (RCB-3) in both arms, 5.1% and 6.7%, respectively, had a poor prognosis. These results highlight the importance of neoadjuvant treatment with pembro for improving survival in pts with early TNBC, and identified a subset of pts for whom additional therapies will be needed. Clinical trial information: NCT03036488. [Table: see text]