A phase 2 study of an anti-PD-L1 antibody (atezolizumab) in dedifferentiated chondrosarcoma

11533 Background: Chondrosarcoma (CS) is one of the most common bone malignancies in adults. Contrary to the indolent nature of low-grade CS, the dedifferentiated subtype (dCS), representing 5-10% of all CS, is known for aggressive behavior, high risk of relapse following resection, and poor prognosis. Expression of PD-L1 has been demonstrated in dCS samples and correlated with high numbers of tumor-infiltrating lymphocytes (Kostine, et al., Mod Pathol, 2016). Response to anti-PD1 therapy has not been evaluated prospectively. We report here the outcomes of a dCS cohort treated with the anti-PD-L1 agent, atezolizumab (atezo). Methods: Patients (pts) 2 years of age or older received intravenous atezolizumab 1200 mg (15 mg/kg with a 1200 mg cap in pediatric pts) once every 21 days. Prior immune checkpoint inhibitor therapy was not allowed. Primary objective was response rate (ORR). Imaging was carried out at the end of cycle 3 and then every two cycles; responses were evaluated per RECIST 1.1. The study employed a Simon two-stage design. If no responses were observed within 9 months of the ninth patient being enrolled, the cohort was to be terminated early. Research biopsies for immuno-pharmacodynamic (IO-PD) studies were collected at baseline, prior to C3D1, and optionally at progression. Results: Nine pts were enrolled to the dCS cohort. Three pts were female, 8 pts were White (1 unknown), all had an ECOG performance score ≤1, and their median age was 63 years (range, 53-85). Primary disease sites were pelvis (2); sternum (2); femur, hip, chest, scapula, and lung (1 each). Median duration of treatment for all pts was 9 weeks. Seven pts were evaluated for response, of whom 3 (42.9%) were documented to have stable disease (SD) as best response, lasting a median of 25.9 weeks (range, 15-38.3 weeks). Four pts had a best response of disease progression. Two pts died prior to first response assessment. No RECIST objective responses were observed; the cohort was closed due to futility. Reasons for treatment discontinuation included progression (n = 6), death (1, respiratory failure unrelated to treatment), withdrawal of consent (1), and SARS-CoV-2 infection (1). Treatment-related adverse events (TrAE), grades 1-3, occurred in 7 pts (78%). Grade 3 TrAEs occurred in 2 pts (22%), included infusion reaction, myonecrosis, and anemia. IO-PD studies are ongoing to elucidate changes within the tumor microenvironment. Conclusions: Though objective response was not seen, atezo showed stabilization of disease in 1/3 of the patients with this aggressive tumor. IO-PD results will be critical to identify determinants of atezolizumab resistance within this dCS cohort and to identify possible partners for combination therapy. Funded by NCI Contract No. HHSN261201500003I. This project was also supported in with funding and drug supply from Genentech Inc (a member of the Roche group). Clinical trial information: NCT04458922 .