Phase II study of rucaparib and nivolumab in patients with leiomyosarcoma

Abstract

Background: Objective responses to immune checkpoint inhibitors (ICI) in leiomyosarcoma (LMS) are rare. Response rates may be increased by combination with other drugs known to promote immune infiltration, such as poly(ADP-ribose) polymerase (PARP) inhibitors, which have led to benefit in BRCA-altered uterine LMS. We therefore evaluated the combination of a PARP inhibitor, rucaparib, and the anti-programmed death receptor-1 monoclonal antibody, nivolumab, in patients with advanced LMS and investigated its effects on the tumor immune microenvironment.

Methods: This was an open-label, single-center, single-arm, phase II study in patients with advanced refractory LMS. Full protocol available Patients were treated with rucaparib 600 mg orally, two times daily, continuously and nivolumab 480 mg intravenously on day 1 of a 28-day cycle. Re-staging scans were performed every 8 weeks. Blood and tissue samples were collected at baseline and at week 8 on treatment. The primary objective was the best objective response rate by 24 weeks using Response Evaluation Criteria in Solid Tumour (RECIST V.1.1). Secondary objectives included treatment-related toxicity, progression-free survival, overall survival, and changes in immune pathways in blood and tumor.

Results: 20 patients with LMS were enrolled. There was one partial response (PR) (5%) in a patient with uterine LMS and a somatic BRCA deep deletion. 19 (95%) patients had a treatment-related adverse event (TRAE) and 7 (35%) had a grade 3 or higher TRAE. Interferon (IFN) α and γ hallmark pathways were more highly expressed in patients who derived benefit from treatment (at least stable disease by 16 weeks) vs those who did not in both baseline (adjusted p=0.005 for IFN-α, 0.03 for IFN-γ) and on-treatment biopsies (adjusted p=0.0002 for IFN-α, 0.0001 for IFN-γ), but the abundance of tumor immune cell populations did not differ between these groups at either time point.

Conclusion: The addition of a PARP inhibitor did not improve the efficacy of ICI in LMS. Adverse events, especially due to overlapping toxicities, were frequent and often led to dose delays and modifications.

Keywords: Immune Checkpoint Inhibitor; Solid tumor.

Figure 1. Treatment outcomes. (A) Best tumor response by RECIST; (B) duration of treatment; (C) progression-free survival by primary site; (D) overall survival by primary site. RECIST, Response Evaluation Criteria in Solid Tumors; PD, progressive disease; PR, partial response; STLMS, soft tissue leiomyosarcoma; ULMS, uterine LMS.

Figure 2. Oncoprint of targeted tumor and germline sequencing. Oncoprint summarizing targeted somatic next-generation sequencing (MSK-IMPACT) and germline sequencing results in the study participants. Top, progression-free survival (PFS) in months and best overall response. Bottom, germline mutations.

Figure 3. Circulating tumor DNA (ctDNA) results. (A) Concordance between ctDNA and targeted tumor sequencing by count (left) and percent (right) in 19 patients with MSK-IMPACT performed on archival tumor specimens; ctDNA variant calls at any time point are shown. (B–D) Spider plots of responses of patients who discontinued study due to (B) RECIST progression, (C) clinical progression, or (D) toxicity or withdrawal of consent, each annotated with ctDNA changes. RECIST, Response Evaluation Criteria in Solid Tumors.

Figure 4. Gene expression. (A) Heatmap of relative expression at baseline of genes that differed significantly by nominal p value between patients with versus without progression at 16 weeks. (B) Heat plot of expression of enriched hallmark pathways that were upregulated or downregulated at baseline in patients with PFS≥16 weeks compared with levels in patients with PFS<16 weeks. (C) Heatmap of enrichment of immune and stromal cell populations at baseline between patients with versus without progression at 16 weeks. IFN, interferon; NK, natural killer.