Nonclinical pharmacological characterization of a focal adhesion kinase inhibitor, PF-562,271

ND-4 Cancer cells are characterized by the ability to grow in an anchorage-independent manner. This characteristic phenotype is due, in large part, to the activity of the non-receptor tyrosine kinase, focal adhesion kinase. FAK localizes in focal adhesion plaques and has a role as a scaffolding and signaling protein for other adhesion molecules. As such, FAK acts as a signaling molecule for many integrins which do not have intrinsic kinase activity, as well as, complimenting signal transduction through other RTKs (e.g. EGFR, VEGFR).Recent studies demonstrate a strong correlation between increased FAK expression and the invasive phenotype of aggressive human tumors. There has been some controversy regarding the importance of the kinase activity relative to the scaffolding functions of FAK.Potent inhibitors of FAK have been identified based upon a discovery approach that combined compound screening, structure-based drug design and traditional medicinal chemistry. Two different series of compounds, 2,4-di-anilino pyrimidines and 3,5-di-substituted indoles, were found to be modest inhibitors of focal adhesion kinase. Modeling and co-crystal structures of these inhibitors with FAK led to specifically substituted 2,4-diamino pyrimidines, which were found to inhibit FAK in both kinase and cell assays (1 nM-900 nM). Within this particular class of molecules, the anchoring hydrogen bond donor-acceptor motif for kinase activity was identified. Small, seemingly minor changes to inhibitor structure caused major conformational changes in the way the inhibitor bound and co-crystallized with FAK. Structure-based drug design (SBDD) paved the way for design of novel inhibitors with optimal ADME, selectivity, and potency properties by specific substitution at the C2, C4, and C5 positions of the 2,4-diamino pyrimidine core. Replacement of the 3,5-di-substituted indole moiety with a 5-amino oxindole at the pyrimidine C2 position allowed us to retain a key hydrogen bond interaction to Arg 426, while reducing MW and removing a potential metabolic liability in the original dehydropiperidine. Co-crystal structures of C4 derivatives demonstrated the plasticity of the kinase active site in this region, and furthermore suggested a possible route to achieve selectivity over other kinases by regioselective substitution off of the amino-methyl aryl ring with a RSO2R substituent. Finally, SBDD led to the replacement of the original (pyrimidine C5) Br atom with a CF3 group. PF-562,271 is a potent ATP competitive, reversible inhibitor of FAK and Pyk2 kinase with IC50 of 1.5 and 14 nM, respectively. PF-562,271 is potent in an inducible cell based assay* measuring phospho-FAK with an IC50 of 5 nM. PF-562,271 was evaluated in a number of kinase screens and panels and displays [[Unsupported Character - Codename s]]>100x selectivity against a long list of non-target kinases. PF-562,271 inhibits FAK phosphorylation in vivo in a dose dependent fashion (calculated EC50 35 ng/mL, free, 0.5-4 hrs post dose) following oral administration to female athymic (nu/nu) mice bearing human glioblastoma, U87MG subcutaneous tumors. In vivo inhibition of FAK phosphorylation was sustained (56%) for over 4 hours with a single oral dose of 33 mg/kg. The antitumor efficacy of PF-562,271 was evaluated in the following human s.c. xenograft models: PC-3M (prostate), BT474 (breast), BxPc3 (pancreatic), LoVo (colon), U87MG (glioblastoma), and H460 (lung). Regressions were observed in PC-3M, BT474, BxPc3, and LoVo models at doses of 25-50 mg/kg, BID corresponding to Cmax (free) ranges of 77-885 ng/ml, Cave (free) of 14-40 ng/ml, and inhibition of phospho-FAK of 31-76% for >4 hours. Maximum tumor growth inhibition in the U87MG tumor was 51% using an osmotic mini-pump corresponding to a Css of 0.7 ng/mL (free) with a concomitant decrease in phospho-FAK of 45%. No weight loss, morbidity, or mortality were observed in any TGI experiment (up to 50 mg/kg BID x 28 days or 100 mg/kg QD x 25 days). Tumor growth inhibition was dose and drug exposure dependent. BID dosing and mini-pump experiments (Css) resulted in greater tumor growth inhibition in multiple models compared to QD dosing (equivalent total daily dose) suggesting Cave and the time above the Cave are more relevant to efficacy than Cmax. Taken together, these data demonstrate that kinase inhibition with an ATP competitive small molecule inhibitor of FAK results in robust anti-tumor activity. This compound is presently in Phase I clinical trials (clinical data to be presented at 2007 ASCO).