Neoadjuvant BMS-813160, Nivolumab, Gemcitabine, and Nab-Paclitaxel for Patients with Pancreatic Cancer

Abstract

Purpose: Targeting tumor-associated macrophages through C-C chemokine receptor type 2 (CCR) in pancreatic ductal adenocarcinoma (PDAC) improves the efficacy of chemotherapy and restores T-cell immunity in preclinical models.

Patients and methods: We conducted a phase I/II single-institution study (NCT03496662) combining chemotherapy gemcitabine and nab-paclitaxel (GnP), CCR2/5 inhibitor BMS-813160, and nivolumab for four 28-day cycles for patients with borderline resectable (BR) or locally advanced (LA) PDAC. The recommended phase II dose of BMS-813160 was established in the 3 + 3 design. Primary endpoints were safety and objective response rate (ORR). Secondary endpoints included resection rate, median progression-free survival, and median overall survival.

Results: Eight patients were treated with GnP alone (control arm), and 32 patients (29 response evaluable) were treated at the recommended phase II dose. One instance each of grade 3 diarrhea and QTc prolongation occurred which were atrributed solely to BMS-813160. After four cycles of study treatment (N = 22), the ORR was 42% and 20% among patients with BR and LA PDAC, respectively, compared with 0% of control patients. A total of 83.3% of patients with BR and 20% of patients with LA PDAC who completed study treatment underwent surgical resection. For intent-to-treat analyses, patients with BR PDAC had a median progression-free survival and median overall survival of 11.9 and 18.2 months, respectively, whereas patients with LA PDAC had 14.7 and 17 months, respectively. Biomarker analyses showed decreased intratumoral monocytes and macrophages and enhanced T-cell proliferation and effector gene expression.

Conclusions: Neoadjuvant BMS-813160/nivolumab/GnP was well tolerated and seems to achieve comparable ORR and resectability with historical data; however, with prolonged PFS and OS in LA-PDAC patients, warranting a larger phase II study with a more efficacious CCR2-targeted therapeutic.

Figure 1.

Enrollment and clinical results of the study. A, CONSORT diagram showing enrollment, treatment, and analyses of study patients. B, Swimmer plot depicting treatment course and time on study for all patients with PDAC enrolled in this study. C, Waterfall plot showing changes of target lesion sum (best response) for all 37 response-evaluable study patients enrolled in each cohort based on RECIST 1.1 criteria. Kaplan–Meier curves depicting the PFS and OS for ITT patients (red-dotted line) and patients who completed four cycles of study treatment (red solid line) on investigational arm and all patients on control arm (black line) with (D) BR or (E) LA PDAC. ARDS, acute respiratory distress syndrome; SBRT, stereotactic body radiotherapy; tx, treatment.

Figure 2.

Correlative studies of pre- and posttreatment tumor samples. A, Analysis of patient tumor tissue biopsy samples at pretreatment (pre) and the end of cycle 2 (post). Tumor biopsy tissues were analyzed by flow cytometry for CD45⁺CD11b⁺CD14⁺CSF1R⁺CCR2⁺CD33⁺ IM/monocytic myeloid-derived suppressor cells (Mo-MDSC) and CD45⁺CD11b⁺MHCII⁺CD14^midCD33⁽⁻⁾CD15⁽⁻⁾ TAM and CD45⁺CD11b⁺CD14⁽⁻⁾CD15⁽⁺⁾ neutrophils/granulocytic myeloid-derived suppressor cells (G-MDSC). B, Analysis of neutrophil-to-TAM ratios using data from (A). The percent of CD45⁺ cells is depicted as mean ± SEM. n.s.-denoted groups are not statistically different. C and D, Uniform Manifold Approximation Projection (UMAP) plot showing different subsets of CD45⁺ immune cells categorized based on marker gene expression from all pre- and posttreatment tumor samples using single-cell RNA sequencing. E, Gene set enrichment analysis (GSEA) showing regulated signature changes in CD8+ and CD4+ T-cell clusters (from D) in postttreatment tumor samples. NES, normalized enrichment score.