Phase I safety, efficacy, and biomarker response evaluations of three oral PD-L1 inhibitors: INCB086550, INCB099280, and INCB099318.

BACKGROUND: Orally administered small-molecule programmed death ligand 1 (PD-L1) inhibitors may have the potential to improve patient outcomes in the treatment of a range of cancers compared with their antibody-based counterparts. A small molecule might achieve better tumor tissue penetration, and oral administration could significantly improve convenience and access for patients. METHODS: Three phase 1 open-label, non-randomized, dose escalation, and expansion studies evaluated the safety, preliminary efficacy, pharmacokinetics (PK), and pharmacodynamics (PD) of three agents in patients with advanced solid tumors: INCB086550 (NCT03762447), INCB099280 (NCT04242199), and INCB099318 (NCT04272034). RESULTS: Overall, 138, 182, and 104 patients received INCB086550, INCB099280, and INCB099318, respectively. Most had previously received ≥2 lines of cancer therapy for advanced or metastatic disease; 9.6%-16.5% had received prior immunotherapy. All three agents were rapidly absorbed and showed stable dose-dependent PK. With INCB086550, 88 patients (63.8%) had ≥1 treatment-related treatment-emergent adverse event (TEAE), and 19 (13.8%) had ≥1 treatment-related grade ≥3 TEAE. In total, 14 patients (10.1%) had a nervous system-associated TEAE for which an immune-mediated etiology could not be ruled out; events were predominantly peripheral sensory and motor neuropathies. With INCB099280 and INCB099318, 144 (79.1%) and 69 (66.3%) of patients had ≥1 treatment-related TEAE, and 25 (13.7%) and 12 (11.5%) had ≥1 treatment-related grade ≥3 TEAE, respectively. The most frequent immune-related adverse events were skin reactions (INCB099280 and INCB099318) and hepatitis (INCB099280). No dose-limiting toxicities (DLTs) occurred during dose escalation with INCB086550 or INCB099318; two DLTs occurred in two patients with INCB099280 (grade 2 vomiting with 600 mg once daily and grade 2 maculopapular rash with 800 mg two times per day). Overall objective response rates for INCB086550, INCB099280, and INCB099318 were 10.9% (95% CI 6.2% to 17.3%; n=15), 8.8% (95% CI 5.1% to 13.9%; n=16), and 8.7% (95% CI 4.0% to 15.8%; n=9), respectively. Target engagement and PD activity were demonstrated, including PD-L1 binding, and increases in cytokine and chemokine production, as well as T-cell activation and proliferation. CONCLUSIONS: Both INCB099280 and INCB099318 had an acceptable safety profile, with preliminary evidence of antitumor activity. The risk of immune-mediated neuropathy led to discontinuation of the clinical program for INCB086550.