Pembrolizumab in advanced osteosarcoma: results of a single-arm, open-label, phase 2 trial

Abstract

Aim: To evaluate the activity and safety of the PD-1 antibody pembrolizumab in adult patients with advanced osteosarcoma.

Material and methods: The study was a single-arm, open-label, phase 2 trial in patients with unresectable, relapsed osteosarcoma. The primary endpoint was clinical benefit rate (CBR) at 18 weeks of treatment, defined as complete response, partial response, or stable disease using RECIST v1.1. The trial had a Simon´s two-stage design, and ≥ 3 of 12 patients with clinical benefit in stage 1 were required to proceed to stage 2. The trial is registered with ClinicalTrials.gov, number NCT03013127. NanoString analysis was performed to explore tumor gene expression signatures and pathways.

Results: Twelve patients were enrolled and received study treatment. No patients had clinical benefit at 18 weeks of treatment, and patient enrollment was stopped after completion of stage 1. Estimated median progression-free survival was 1.7 months (95% CI 1.2-2.2). At time of data cut-off, 11 patients were deceased due to osteosarcoma. Median overall survival was 6.6 months (95% CI 3.8-9.3). No treatment-related deaths or drug-related grade 3 or 4 adverse events were observed. PD-L1 expression was positive in one of 11 evaluable tumor samples, and the positive sample was from a patient with a mixed treatment response.

Conclusion: In this phase 2 study in advanced osteosarcoma, pembrolizumab was well-tolerated but did not show clinically significant antitumor activity. Future trials with immunomodulatory agents in osteosarcoma should explore combination strategies in patients selected based on molecular profiles associated with response.

Keywords: NanoString; Osteosarcoma; PD-1 inhibitor; PD-L1 expression; Pembrolizumab.

Conflict of interest statement

K. Boye has served on advisory board for Bayer and has received research support from Eli Lilly and Merck. E. Palmerini has served on an advisory board for Takeda, Amgen, Daiichi Sankyo, Eli Lilly, Eusa Pharma, and Deciphera, and has received other research support from Bristol Myers Squibb, Pfizer, PharmaMar, and Daiichi Sankyo.

© 2021. The Author(s).

Figures

Fig. 1

Waterfall plot showing best RECIST response. Individual patients are represented by vertical bars and the change in tumor size according to RECIST v1.1 is depicted on the Y-axis. Two patients did not undergo radiological evaluation and are not included

Fig. 2

Photomicrographs of immunohistochemical staining with monoclonal anti-PD-L1 antibody (clone 22C3; Dako). a > 50% positive membranous staining of tumor cells. Scale bar 50 μm. b Negative staining. Scale bar 100 μm

Fig. 3

Baseline images and response evaluation after treatment with pembrolizumab in study patient 12 (PROMO-12). a CT images and b 18F-FDG PET/CT images at baseline (left panel) and after two cycles of pembrolizumab (right panel). Arrows indicate reduced uptake of 18F-FDG and tumor size in lung and kidney metastases, and arrowheads indicate progressive abdominal and pelvic metastases

Fig. 4

Gene expression analysis of tumor samples using the NanoString platform. a Abundance of various cell type populations across different samples. The cell type score is calculated based on cell-type specific gene expression markers. b Pathways showing differential behavior in PROMO-12 compared to the remaining samples. Pathway scores are calculated based on gene expression data using nSolver V4 software and oriented such that increasing score corresponds to increased expression. c Normalized gene expression values of top 28 differential expressed genes involved in these pathways. For both plots, rows are centered and unit variance scaling is applied to rows. Both rows and columns are clustered using correlation distance and average linkage. In b and c, samples were analyzed in duplicate and both duplicates are shown