Treatment outcomes of the KEYNOTE-522 regimen in an ethnically diverse patient population: A real world experience.

e12663 Background: Pembrolizumab (pembro) in combination with neoadjuvant chemotherapy (NACT) became standard of care for high risk, early stage TNBC on July 26, 2021, as a result of the KEYNOTE-522 trial. Black women are disproportionately affected by TNBC with more advanced stage at diagnosis, however the KEYNOTE-522 did not collect data on race. This retrospective review evaluates response to pembro plus NACT in a racially diverse population to assess the generalizability of the trial findings to potentially under-represented patient populations. Methods: This is a single center, retrospective review of patients with stage II/III TNBC diagnosed between 08/2020 to 08/2023, treated with pembro plus NACT at Maimonides Medical Center, a safety-net hospital in Brooklyn, New York serving a diverse patient population. Exclusion criteria included age < 18 years, ER/PR/Her2 positivity, NACT without pembrolizumab and unreported race. Data were collected from electronic health records and analyzed using SPSS Statistics. Results: 50 patients met inclusion criteria, among whom 40 patients had complete information available for review. 52.5% (95% CI: 38% - 69%) of patients had a pCR. Nearly half of the patient population (47.5%) were African American. Subgroup analysis of ethnicity demonstrated a lower pCR amongst African American women, but this was not statistically significant (table). 5 patients had progression of disease with the development of distant metastasis (2 whilst on treatment and 3 following completion of therapy). 3 patients had grade 3/4 adverse events (pembro associated endocrine and cardiac dysfunction). Notably, lack of complete imaging response to treatment was significantly associated with lack of pCR (p=0.003). Conclusions: In this retrospective review of a racially diverse patient population, the pembrolizumab-NACT regimen was well tolerated with an acceptable side effect profile. While our rate of pCR is lower than the findings of the Keynote-522 trial; 52.5 % (95% CI: 38% - 69%) vs. 64.8% (95% CI: 59.9%-69.5%), our sample size is not large enough to detect a statistically significant difference. Sub-analysis of patient ethnicity demonstrated varying rates of pCR, however findings are limited by a small sample size. Additional data collection is in progress and future analysis will provide insight into a potential correlation between race and therapy response. Table: see text