An Excel file containing all small molecule drugs presented in Tables 1, 2 and 4 of the printed manuscript, including additional fields containing mechanism of action/target, MW, dose schedule, Tmax, Clearance, Volume of distribution, year approved, full references to source data and drug product label, approved indications, additional notes.

An Excel file containing all biological drugs presented in Table 3 of the printed manuscript, including additional fields containing mechanism of action/target, MW, dose schedule, Clearance, Volume of distribution, year approved, full references to source data and drug product label, approved indications, additional notes.

This file contains Supplementary Tables S1 and S2 (pharmacokinetic data) and Supplementary Figures S1 (chemical structures of the four MET kinase inhibitors examined in this study), S2 (pharmacokinetic and pharmacodynamic data for ARQ197), S3 (body weight data for pharmacodynamics-guided dosage regimens), S4 (total full-length MET levels in the single-dose, dose-response time course experiment), S5 (total full-length MET levels in single- and multi-dose experiments using the pharmacodynamics-guided dosage regimens), and S6 (mouse tolerability information for XL880, XL184, and EMD1214063).

Abstract The development of molecularly targeted agents has benefited from use of pharmacodynamic markers to identify “biologically effective doses” (BED) below MTDs, yet this knowledge remains underutilized in selecting dosage regimens and in comparing the effectiveness of targeted agents within a class. We sought to establish preclinical proof-of-concept for such pharmacodynamics-based BED regimens and effectiveness comparisons using MET kinase small-molecule inhibitors. Utilizing pharmacodynamic biomarker measurements of MET signaling (tumor pY1234/1235MET/total MET ratio) in a phase 0–like preclinical setting, we developed optimal dosage regimens for several MET kinase inhibitors and compared their antitumor efficacy in a MET-amplified gastric cancer xenograft model (SNU-5). Reductions in tumor pY1234/1235MET/total MET of 95%–99% were achievable with tolerable doses of EMD1214063/MSC2156119J (tepotinib), XL184 (cabozantinib), and XL880/GSK1363089 (foretinib), but not ARQ197 (tivantinib), which did not alter the pharmacodynamic biomarker. Duration of kinase suppression and rate of kinase recovery were specific to each agent, emphasizing the importance of developing customized dosage regimens to achieve continuous suppression of the pharmacodynamic biomarker at the required level (here, ≥90% MET kinase suppression). The customized dosage regimen of each inhibitor yielded substantial and sustained tumor regression; the equivalent effectiveness of customized dosage regimens that achieve the same level of continuous molecular target control represents preclinical proof-of-concept and illustrates the importance of proper scheduling of targeted agent BEDs. Pharmacodynamics-guided biologically effective dosage regimens (PD-BEDR) potentially offer a superior alternative to pharmacokinetic guidance (e.g., drug concentrations in surrogate tissues) for developing and making head-to-head comparisons of targeted agents. Mol Cancer Ther; 17(3); 698–709. ©2018 AACR.

Abstract Approved and marketed drugs are frequently studied in nonclinical models to evaluate the potential application to additional disease indications or to gain insight about molecular mechanisms of action. A survey of the literature reveals that nonclinical experimental designs (in vitro or in vivo) often include evaluation of drug concentrations or doses that are much higher than what can be achieved in patients (i.e., above the maximally tolerated dose or much higher than the clinically relevant exposures). The results obtained with these high concentrations may be particularly helpful in elucidating off-target effects and toxicities, but it is critical to have a dose–response curve that includes the minimally effective or clinically effective concentration for comparison. We have reviewed the clinical literature and drug product labels for all small molecules and biological agents approved by the FDA for use in oncology to identify and compile the available pharmacokinetic parameters. The data summarized here can serve as a guide for selection of in vitro concentrations and in vivo plasma exposures for evaluation of drug effects in nonclinical studies. Inclusion of drug concentrations or exposures that are relevant to those observed in clinical practice can improve translation of nonclinical mechanism of action findings into potentially relevant clinical effects. Clin Cancer Res; 23(14); 3489–98. ©2017 AACR.

Abstract Background: A direct comparison of drug efficacy for the multiple agents currently in clinical development targeting MET-driven cancers would be useful for the selection of optimal treatment options. Previously, we utilized validated MET pharmacodynamic (PD) assays to compare the time course of phosphorylated-MET (pMET) suppression for five MET inhibitors (ASCO 2013). In the current study, we selected three candidates that demonstrated potent MET inhibition to compare anti-tumor efficacy. Methods: PD time course and tumor PK data were utilized to simulate a dosing schedule anticipated to produce >90% pMET suppression in a SNU5 gastric cancer xenograft model. From these data, dosing schedules of 44 mg/kg (Q12H) cabozantinib, 12.5 mg/kg (Q12H) EMD1214063, and 16.5 mg/kg (QD) foretinib were chosen to achieve necessary tumor exposure to suppress pMET. Once tumors reached a 150±50 mm3 size, drugs were administered continuously for 21 days and tumor volumes were measured intermittently for 62 days. To measure pMET suppression, tumor quadrants were collected at 4, 12, and 24 hrs from all treatment groups after dose 1 (day 1) and on day 8 (after dose 8 of foretinib or dose 14 of cabozantinib and EMD1214063). Results: Intact MET levels (pM/μg protein) were approximately 70% lower than vehicle controls at all day 8 collection points for all three drugs. The pY1234/35MET/MET ratios were 89%-99% (p<0.001) lower than vehicle controls on day 8 at 4 and 12 hrs post-dose for all three drugs. Compared to pretreatment levels, tumor volumes were reduced by 80-90% for all three drugs within 10-15 days post therapy and remained regressed until 13-20 days after treatment was stopped. Tumors then slowly regrew, but remained approximately 80% smaller than the vehicle group on day 62 (end of study). Conclusions: These studies demonstrate that PD response-guided regimens delivered drug doses that were lower than efficacious doses described previously (except cabozantinib), but effectively reduced tumor volume in SNU5 xenografts. The three MET inhibitors showed comparable anti-tumor efficacy when accompanied by equivalent pMET inhibition. Our data also affirms the utility of the MET PD assays to guide dose ranging studies. Funded by NCI Contract No HHSN261200800001E. Citation Format: Apurva K. Srivastava, Melinda G. Hollingshead, Jeevan P. Govindharajulu, Joseph M. Covey, Dane Liston, James Peggins, Donald P. Bottaro, John J. Wright, Robert J. Kinders, Joseph E. Tomaszewski, James H. Doroshow, Ralph E. Parchment. Met target inhibition-guided efficacy in preclinical models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3691. doi:10.1158/1538-7445.AM2014-3691

11103 Background: Several MET inhibitors are currently being developed that block aberrant HGF/MET signaling in different cancers. We utilized validated MET pharmacodynamic (PD) assays to compare time course, magnitude, and reversal of MET suppression by 5 MET inhibitors in preclinical models. Methods: Immunoassays (total MET, pY1234/35MET, and pY1356MET) were developed and validated to measure modulation of MET by 5 MET inhibitors (crizotinib, tivantinib, cabozantinib, foretinib, and EMD1214063). The comparison was implemented in 3 sequential stages: 1) establish time course and magnitude of MET inhibition after single drug administration of 4 different doses; 2) determine dose(s) and schedule for sustained MET inhibition and downstream signaling at optimal levels; and 3) compare efficacy of MET inhibitors at MTD and equal MET inhibition. The preclinical models include an autophosphorylation gastric tumor (SNU5) model and a paracrine MET activation model in hHGF knock-in mice. Plasma and tumor exposures were measured using LC-MS/MS to correlate with PD effects. Results: We completed phase one in the SNU5 model and determined inhibition of pY1234/35MET and total MET in tumor tissues after single administration of MET inhibitors. Time course and magnitude of pY1234/35MET inhibition varied considerably among MET inhibitors, with the most rapid (>80% suppression in 30 min) and sustained inhibition (up to 48 h) observed with EMD1214063 at a dose of 30 mg/kg. The maximal inhibition of pY1234/35MET and time taken for biomarker recovery were wide-ranging among MET inhibitors. Tumor drug exposures were concomitantly higher than plasma for all drugs and correlated inversely with pY1234/35MET, except for tivantinib which, unlike other drugs, is not ATP competitive inhibitor. Conclusions: We applied validated PD assays to directly compare similarities and differences in extent and duration of MET inhibition by 5 MET inhibitors. Our results provide important foundation for head-to-head comparison of efficacies of MET inhibitors at MTD and equal MET inhibition. Funded by NCI Contract No HHSN261200800001E.