Abstract P2-09-05: LCCC 1525: A phase II study of a priming dose of cyclophosphamide prior to pembrolizumab to treat metastatic triple negative breast cancer (mTNBC)

Background: While immunotherapy holds promise in the treatment of mTNBC, response rates (RR) in unselected patients (pts) are approximately 20%. Strategies to augment response to immunotherapy include depletion of regulatory T cells (Tregs). A single dose of cyclophosphamide (Cy) given prior to checkpoint inhibition achieved this goal in preclinical models of TNBC (Taylor et al., JCI, 2017). Thus, we designed a phase II study to evaluate this strategy in the clinical setting of mTNBC. Patients/Methods: In cycle 1 (C1), eligible pts with mTNBC received a single dose of Cy 300mg/m2 IV on day 1 (C1D1) followed by pembrolizumab 200mg IV on day 2 (C1D2), then every 3 weeks thereafter. The co-primary objectives were (1) progression free survival (PFS, null 1.9 mos vs. 2.9 mos, 80% power, alpha 0.05) and (2) reduction in Tregs in peripheral blood measured by flow cytometry. Secondary endpoints were response rate (RR), survival (OS), and RNA-based correlative endpoints. Results: 40 patients were evaluable for efficacy: mean age 54.5 yrs (33 – 82 yrs), 75% white, 22% black, 3% American Indian. All patients had received 1 prior line of chemotherapy in the metastatic setting; 29% received 5 or more prior lines. The most common grade 3 adverse events (AE9s), all 5%, were neutropenia, anemia, elevated AST, and fatigue. Immune-related grade 3 AE9s, all 3%, included colitis, dry mouth, pneumonitis. Overall RR was 21% (0 CR, 8 PR), 3 pts had stable disease. Median PFS was 1.8 months (mos) (95% CI 1.4–2.5) and OS was 6.3 mos (95% CI 2.8–8.4). There was a non-significant decrease in Tregs from C1D1 to C1D2 (-3.3%, p=0.19); but from C1D2 to C2D1, Tregs increased 21.7% (p=0.005). There was no association between changes in Tregs or number of prior lines of therapy with RR (p>0.09), while immune-related AE9s were associated with response (p=0.02). Correlatives studies illustrate B cell immune gene signature expression and B cell receptor repertoire diversity were enriched in responders, while genes/pathways associated with neutrophils, anti-apoptosis, PI3K/AKT and down-regulation of MHC class 1 were associated with non-response. Conclusions: While pembroliuzumab plus Cy was well-tolerated among pts with mTNBC, efficacy was similar to historical control, likely due to minimal effect of Cy on Tregs. Correlative analyses illustrate that study of adaptive immune features, including B cell biology, is a promising strategy for understanding response to PD-1 inhibition in breast cancer. Further strategies to deplete Tregs in a more sustained manner are worthy of future exploration (NCT02768701). Citation Format: Anders CK, Moore D, Sambade M, Cuaboy L, Garrett A, Woodcock M, McKinnon K, Cowens K, Bortone D, Calhoun B, Carey L, Dees C, Jolly T, Muss H, Reeder-Hayes K, Kaltman R, Jankowitz R, Gudena V, Olajide O, Perou C, Vincent B, Serody J. LCCC 1525: A phase II study of a priming dose of cyclophosphamide prior to pembrolizumab to treat metastatic triple negative breast cancer (mTNBC) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-09-05.