D3S-001, a highly potent, selective, and differentiated covalent inhibitor of KRAS G12C: Human dose prediction and first‐in‐human (FIH) trial design.

Abstract

e15087 Background: Effective inhibition of G12C mutated KRAS is believed to be critical for anti-tumor efficacy in KRAS-driven cancers. D3S-001 is a potent and highly differentiated KRAS G12C inhibitor, which demonstrates rapid and near complete target inhibition and strong anti-tumor activity in preclinical cell line and animal models [Preclinical characterization of D3S-001, a highly potent, selective, and differentiated covalent inhibitor of KRAS G12C, ASCO 2022 Abstract #378099]. The current study utilizes pharmacokinetic (PK) modeling to predict the extent of KRAS G12C target inhibition at the proposed clinical dose range. Methods: Allometric scaling and non-linear fixed effect modeling were used to predict human PK profiles from DMPK data in animals. Human in vivo target inhibition was calculated based on human PK predictions, free drug fraction and results from in vitro target engagement assay. A First-in-Human (FIH) study was designed to investigate D3S-001 monotherapy in subjects with advanced solid tumors having a KRAS G12C mutation. Six dose escalation cohorts will be investigated, starting from 50mg QD. Results: Based on a GLP toxicology study in dogs, the highest non-severely toxic dose (HNSTD) was determined to be 20 mg/kg/day. Per ICH S9, a 20 mg/kg/day dose in dog supports a FIH starting dose of 110 mg. To ensure an additional safety margin, a lower starting dose of 50 mg was proposed. At 50 mg QD, the predicted human exposure (Cmax 60 nM, Caverage 49 nM, Ctrough 31 nM) is expected to achieve ̃70% target inhibition at 2 hours after dosing. In NCI-H358 mouse xenograft model, this exposure is associated with 94% tumor growth inhibition (TGI). Safety margins based on Cmax and AUC comparing to dog HNSTD are 17-fold and 9-fold, respectively. At 100 mg QD, the predicted human exposure (Cmax 120 nM, Caverage 98 nM, Ctrough 63 nM) is expected to achieve ̃95% target inhibition. This exposure is associated with tumor regression of 72% in the NCI-H358 mouse xenograft model. At dose levels higher than 100 mg QD, predicted human exposures are expected to consistently achieve >95% target inhibition starting 1 hour after dosing. These exposures are associated with deeper levels of tumor regression of 97%. Moreover, even at 400 mg QD, predicted human plasma concentrations are still below dog plasma concentration of 443 nM, Caverage at HNSTD. Conclusions: D3S-001 is predicted to achieve rapid and consistently efficacious KRAS G12C target inhibition in humans at proposed clinical dose range. This high degree of KRAS G12C inhibition over the dosing interval may translate into greater clinical benefit to patients with KRAS G12C mutations comparing with other clinical-stage KRAS G12C inhibitors. Clinical data from the FIH study will be evaluated and compared to preclinical predictions.