Abstract B218: Drug-drug interaction assessment of GDC-0032, a beta-sparing inhibitor of phosphoinositide 3-kinase (PI3K), using midazolam, a sensitive CYP3A4 substrate.

Background: GDC-0032 is an orally-bioavailable, potent, and selective inhibitor of Class IA PI3K alpha, delta, gamma, and p110 beta isoforms, with 30 fold less inhibition of the p110 beta isoform. GDC-0032 has shown promising preliminary clinical activity in PIK3CA-mutant cancers and is being explored in solid tumors and in combination with endocrine therapies in breast cancer. In vitro, GDC 0032 was not a potent inhibitor of CYP isoforms 1A2, 2B6, 2C8, 2C9, 2C19, 2D6, or 3A4 (IC50 > 30 µM) but displayed weak time-dependent inhibition of CYP3A4, with KI and kinact values of 77 µM and 0.030 min-1, respectively. In addition, GDC-0032 induced CYP3A4, but not CYP1A2 or CYP2B6 in human hepatocytes. Based on these data, the potential for GDC-0032 to impact the pharmacokinetics (PK) of midazolam (MDZ), a sensitive in vivo marker of CYP3A4 enzymatic activity, was assessed in a Phase I trial in patients with cancer. Methods: Twelve patients with advanced or metastatic solid tumors were enrolled. All patients received 5 mg MDZ on Day 1 (D1), followed by continuous daily dosing of GDC-0032 (9 mg) starting on D2. A second MDZ dose was given on D16. Predose, 0.5, 1, 1.5, 2, 3, 4, 8, and 24 hr post-dose plasma samples were collected on D1 and D16. Maximum concentration (Cmax) and area under the concentration-time curve (AUC0-24) of MDZ were estimated by non-compartmental methods using WinNonlin (version 5.2.1). MDZ plasma Cmax and AUC0-24 without GDC-0032 (D1) and after 15 daily doses of GDC-0032 (D16) were compared. The geometric mean ratios (GMRs) of Cmax and AUC0-24 and the corresponding 90% CIs were estimated (WinNonlin version 5.2.1). Physiologically-Based Pharmacokinetic (PBPK) modeling was used to prospectively predict the DDI potential of GDC-0032 (Simcyp version 11.0 and Gastroplus version 8.0). Results: No trend in individual Cmax or AUC0-24 of MDZ was observed post-GDC-0032 treatment relative to pre-treatment. The change in MDZ exposure (Cmax and AUC0-24) before and after GDC-0032 treatment was less than 2-fold for all patients (median: 1.1 fold; range: 0.4-1.9). The GMRs (90% CIs) for MDZ+GDC-0032 relative to MDZ alone for Cmax and AUC0-24 were 0.98 (0.70-1.37) and 1.04 (0.68-1.60), respectively. Prospective PBPK modeling accurately predicted a lack of marked DDI between GDC-0032 and MDZ (1-1.4 fold change). Conclusions: Despite in vitro results suggesting that GDC-0032 may impact the PK of concomitant medications which are metabolized by CYP3A4, co-administration with MDZ did not change the PK of MDZ in patients with cancer. PBPK modeling predicted little to no impact of GDC-0032 on the PK of MDZ, consistent with the in vivo observation. These results indicate that GDC-0032 may be given concomitantly with CYP substrates with minimal risk of a PK drug-drug interaction. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):B218. Citation Format: Richard A. Graham, Sravanthi Cheeti, Laurent Salphati, Tong Lu, Carla Kurkjian, Daniel Von Hoff, Jasgit Sachdev, Dejan Juric, Jose Baselga, Hema Parmer, Jerry Hsu, Jeffrey Infante. Drug-drug interaction assessment of GDC-0032, a beta-sparing inhibitor of phosphoinositide 3-kinase (PI3K), using midazolam, a sensitive CYP3A4 substrate. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr B218.