Discovery of BMS-986238, a second-generation macrocyclic peptide inhibitor of programmed death-ligand 1 (PD-L1).

The discovery of BMS-986238, a second-generation macrocyclic peptide that is a potent inhibitor of PD-L1 and which progressed to the clinic, is described. This asset follows BMS-986189, a first-generation PD-L1 inhibitor, which upon QD administration to healthy volunteers demonstrated saturation of peripheral PD-L1 at low drug plasma levels, an observation consistent with the low picomolar binding affinity of BMS-986189 with PD-L1. The desired profile of a second-generation macrocyclic peptide inhibitor of PD-L1 included the property of tight binding kinetics to the PD-L1 receptor, but with a significantly modified pharmacokinetic (PK) profile and specifically a long plasma half-life. It was envisioned the combination of these properties, in a PD-L1-inhibiting macrocyclic peptide, would serve to reduce the frequency of QD-dosing as compared to the first-generation inhibitor BMS-986189. In addition, and most importantly, PK/PD models suggested a macrocyclic peptide with such properties had the potential for oral administration, given the relatively low Cmax that would be required to drive efficacy and the potential role of an extended half-life in reducing variability at steady state. The medicinal chemistry strategy employed to secure a second generation inhibitor included the evaluation of functionality at the solvent exposed regions of this ring system that would expectedly have some binding affinity with serum albumin. BMS-986238 emerged from this effort as a macrocyclic inhibitor of PD-L1, with low picomolar affinity for this receptor. This compound was highly protein bound in serum and demonstrated an extended half-life in both rat and cyno (>19h). Extensive preclinical formulation work on BMS-986238 resulted in a formulation that enabled oral bioavailability and provided a path to the clinic where this compound was dosed orally to healthy volunteers.