PII-025 - ASCENDING DOSE IN VIVO QT STUDIES HAVE THE SENSITIVITY TO DETECT QT INTERVAL CHANGES THAT ARE SUFFICIENT TO SUPPORT A CLINICAL QT WAIVER: ASSESSMENT OF QT/QTC INTERVAL CHANGES USING AN ESCALATING DOSE DESIGN IN TELEMETERED BEAGLE DOGS.

M. Moye¹, A. Chaves², K. Yee², J. Ferraro², J. Yu², D. Steve², D. Lengel², C. Regan²; ¹Merck & Co., Inc., Rahway, NJ, United States, ²Merck & Co., Inc..

Background: The International Conference on Harmonization (ICH) S7B and recent E14/S7b Q&As outline the use of nonclinical assays to support risk assessment of delayed ventricular repolarization and can be used as part of a TQT substitution integrated risk assessment. In vivo canine studies are valuable to determine QT-prolonging drugs and may be used to evaluate translational QTc risk. The aim of this work is to establish the translatability of ascending dose (AD) design canine telemetry model with satellite animal PK using the positive control compounds Moxifloxacin, Ondansetron, and Dofetilide, and negative control Levocetirizine. While a less common nonclinical design approach, AD and separate cohort satellite PK can decrease study duration, limit tolerability concerns, and reduce colony sized of surgically implanted animals.

Methods: Compounds were administered at three dose levels in N=4 conscious, adult male beagle dogs chronically implanted with radiotelemetry devices to record ECG. Physiological signals were acquired continuously and analyzed with Notocord-HEM™ telemetry software. Toxicokinetic data was obtained with N=4 satellite dogs of similar age and body weight. Concentration QTci (cQTci) analysis on mean data was performed using linear regression in R and hysteresis analysis in NONMEM. For each compound the slope, intercept, and unbound concentration to achieve +10 ms ΔΔQTCi change was calculated.

Results: Exploratory data analysis revealed a counter-clockwise loop consistent with hysteresis for cQTci data from Moxifloxacin studies. Hysteresis modeling yielded a more sensitive threshold, 1034 (371-1536)* ng/mL, than through simple regression, 1719 ng/mL, of the same dataset; this value was consistent with the clinical critical concentration (1120 ng/mL) and a few fold more sensitive than canine studies detailed in E14/S7B training materials. Thresholds for Dofetilide, 0.65 (0.35-0.91) ng/mL, and Ondansetron, 95 (53-315) ng/mL, were computed without suggestion of hysteresis and compared against clinical critical concentrations (0.13 and 67 ng/mL). *90% CI.

Conclusion: An ascending dose paradigm with satellite animal PK can yield a concentration QTc relationship whose sensitivity is consistent with clinical expectations. This can be used as a positive benchmark for preclinical QTc risk assessment for candidate drugs.