Abstract P2-14-18: A randomized phase II trial of carboplatin with or without nivolumab in metastatic triple-negative breast cancer

Abstract

Background: Platinum agents induce DNA crosslinking and cause accumulation of genotoxic stress, which leads to immune activation via IFN-γ signaling, making the combination with nivolumab (PD-1 antibody) an attractive strategy to enhance the benefit of either agent alone in metastatic triple-negative breast cancer (mTNBC). Methods: In this phase II open-label, investigator-initiated, multicenter trial, patients with unresectable locally advanced or mTNBC treated with 0-1 prior lines of chemotherapy in the metastatic setting were randomized 1:1 to carboplatin (AUC 6) with or without nivolumab (360 mg) IV every 3 weeks. Stratification factors included: germline BRCA (gBRCA) status, prior neo/adjuvant platinum, and number of prior lines of metastatic therapy. After approval of PD-L1 inhibition for mTNBC, the study was amended to include first-line mTNBC only and PD-L1 status was added as a stratification factor. Patients randomized to carboplatin alone were allowed to crossover at progression to receive nivolumab (+ nab-paclitaxel post-amendment). The primary objective was to compare progression-free survival (PFS) per RECIST 1.1 criteria of carboplatin with or without nivolumab in first-line mTNBC in the intent-to-treat (ITT) population. Key secondary objectives were objective response rate (ORR), overall survival (OS), clinical benefit rate, and duration and time to objective response. PD-L1 status was confirmed centrally using the SP142 Ventana assay (positive, ≥1% IC). Paired research biopsies at baseline, on-treatment and at progression were performed, if safely accessible. The trial closed to accrual prior to reaching target accrual due to approval of PD-1 inhibition in combination with platinum-based chemotherapy for PD-L1+ mTNBC. Results: Between 1/30/2018 and 12/9/2020, 78 patients enrolled. Three patients did not receive protocol treatment, and the safety analysis was conducted among the 75 that received any treatment; 37 received carboplatin + nivolumab (Arm A), 38 received carboplatin alone (Arm B). Median age was 59.1 yrs (range: 25.4-75.8). Four patients (5.3%) had a known gBRCA1/2 mutation. Sixty-two (82.7%) patients received 0 prior lines (ITT population) and 13 (17.3%) 1 prior line of metastatic therapy. Sixty-seven patients (89.3%) experienced any grade ≥2 treatment-related adverse event (AE). The most frequent AE were platelet count decrease (n=40; 53.3%), anemia (n=36; 48.0%), neutrophil count decrease (n=33; 44.0%) and fatigue (n=24; 32.0%). Grade 3/4 AE were observed in 46 (61.3%) patients, and there was one grade 5 AE (COVID19 pneumonia). Any grade ≥2 immune-related AE (irAE) were observed in 25 of the 37 (67.6%) patients treated with carboplatin + nivolumab. Grade 3/4 irAE were observed in 11 (29.7%) patients. In the ITT population (32 on Arm A; 30 on Arm B), median PFS was 4.2 months with carboplatin + nivolumab, and 5.5 months with carboplatin (stratified HR 0.98, 95% CI [0.51 - 1.88]; p=0.95). ORR was 25% vs. 23.3%, respectively. At a median follow-up of 23.5 months, median OS was 17.5 months vs. 10.7 months (stratified HR 0.63, 95% CI [0.32 - 1.24]; p=0.18). In patients with PD-L1+ mTNBC (13 on Arm A; 11 on Arm B), median PFS was 8.3 months and 4.7 months, respectively (stratified HR 0.63, 95% CI [0.21 - 1.89]; p=0.41). ORR was 23.1% vs. 27.3%, respectively. Median OS was 17.5 months vs. 9.6 months (stratified HR 0.59, 95% CI [0.20 - 1.75]; p=0.34). Conclusions: Addition of nivolumab to carboplatin in patients with previously untreated mTNBC, unselected by PD-L1 status, did not significantly improve PFS. A trend toward improved PFS and OS was observed in patients with PD-L1+ mTNBC. Tissue, blood and intestinal microbiome biomarker analyses are planned; bulk tumor and single-cell sequencing, and TCR sequencing in peripheral blood are ongoing. Clinical trial information: NCT03414684. Citation Format: Ana C Garrido-Castro, Noah Graham, Kevin Bi, Jihye Park, Jingxin Fu, Tanya Keenan, Edward Thomas Richardson, Ricardo Pastorello, Paulina Lange, Victoria Attaya, Robert Wesolowski, Natalie Sinclair, Zarah Lucas, Steve Lo, Nadine Tung, Meredith Faggen, Peter A Kaufman, Caroline C Block, Fred Briccetti, Madhavi Toke, Wendy Chen, Kai Wucherpfennig, Sascha Marx, Ye Tian, Judith Agudo, Jennifer L Guerriero, Stuart Schnitt, Nancy U Lin, Eric P Winer, Elizabeth A Mittendorf, Nabihah Tayob, Eliezer Van Allen, Sara M Tolaney. A randomized phase II trial of carboplatin with or without nivolumab in metastatic triple-negative breast cancer [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P2-14-18.

Comparative efficacy and safety of PD-1/PD-L1 inhibitors in triple negative breast cancer: a systematic review and network meta-analysis of randomized controlled trials

“…There were seven phase III trials and five phase II trials. Six studies looked at the effect of PD-1/PD-L1 inhibitors on unresectable, invasive, or metastatic (advanced) TNBC [ 33 , 35 – 37 , 40 , 43 ], four looked at non-metastatic/early-stage TNBC [ 39 , 41 , 42 , 44 ], and two looked at treated metastatic TNBC for maintenance therapy [ 34 , 38 ]. Atezolizumab (n = 5 trials) was the most commonly studied ICI [ 33 , 36 , 39 , 40 ], followed by Pembrolizumab (n = 4 trials) [ 34 , 35 , 41 , 42 ], Durvalumab (n = 2 trials) [ 37 , 38 ], and Nivolumab (n = 1 trial) [ 43 ].…”

“…Six studies looked at the effect of PD-1/PD-L1 inhibitors on unresectable, invasive, or metastatic (advanced) TNBC [ 33 , 35 – 37 , 40 , 43 ], four looked at non-metastatic/early-stage TNBC [ 39 , 41 , 42 , 44 ], and two looked at treated metastatic TNBC for maintenance therapy [ 34 , 38 ]. Atezolizumab (n = 5 trials) was the most commonly studied ICI [ 33 , 36 , 39 , 40 ], followed by Pembrolizumab (n = 4 trials) [ 34 , 35 , 41 , 42 ], Durvalumab (n = 2 trials) [ 37 , 38 ], and Nivolumab (n = 1 trial) [ 43 ]. Six trials used multiple-agent chemotherapy regimens in combination with PD-1/PD-L1 inhibitors [ 36 , 37 , 39 , 41 , 42 , 44 ], and four used mono-chemotherapy regimens with PD-1/PD-L1 inhibitors, including two Taxane-based [ 33 , 40 ], one Platinum-based [ 43 ], and one Investigator's choice chemotherapy [ 35 ].…”

“…Atezolizumab (n = 5 trials) was the most commonly studied ICI [ 33 , 36 , 39 , 40 ], followed by Pembrolizumab (n = 4 trials) [ 34 , 35 , 41 , 42 ], Durvalumab (n = 2 trials) [ 37 , 38 ], and Nivolumab (n = 1 trial) [ 43 ]. Six trials used multiple-agent chemotherapy regimens in combination with PD-1/PD-L1 inhibitors [ 36 , 37 , 39 , 41 , 42 , 44 ], and four used mono-chemotherapy regimens with PD-1/PD-L1 inhibitors, including two Taxane-based [ 33 , 40 ], one Platinum-based [ 43 ], and one Investigator's choice chemotherapy [ 35 ]. The other two trials compared PD-1/PD-L1 inhibitors alone to chemotherapy for maintenance therapy in patients with previously treated metastatic TNBC [ 34 , 38 ].…”

“…The OS was reported in nine RCTs [ 33 – 35 , 37 – 41 , 43 ], three of which used Atezolizumab [ 33 , 39 , 40 ], two used Pembrolizumab [ 35 , 41 ], and one used either Durvalumab or Nivolumab as a neoadjuvant to chemotherapy (Fig. 3 A) [ 37 , 43 ].…”

“…The OS was reported in nine RCTs [ 33 – 35 , 37 – 41 , 43 ], three of which used Atezolizumab [ 33 , 39 , 40 ], two used Pembrolizumab [ 35 , 41 ], and one used either Durvalumab or Nivolumab as a neoadjuvant to chemotherapy (Fig. 3 A) [ 37 , 43 ]. Pembrolizumab in a neoadjuvant setting had a pooled HR of 0.82 (95% CI 0.65 to 1.03, SUCRA = 46%, n = 2 trials, 1449 patients), which was comparable to Atezolizumab’s HR of 0.92 (95% CI 0.74 to 1.15, SUCRA = 28%, n = 3 studies, 1886 patients), demonstrating a prolonged but insignificant OS in PD-1/PD-L1 inhibitors arms (see SUCRA Additional file 1 : Table S3).…”