Concentration-effect modeling based on change from baseline to assess the prolonging effect of drugs on QTc together with an estimate of the circadian time course

• Published in: Journal of clinical pharmacology Vol. 54; no. 12; p. 1400
• Main Authors: Ferber, Georg, Wang, Duolao, Täubel, Jörg
• Format: Journal Article
• Language: English
• Published: England 01.12.2014
• Subjects: Analgesics - blood; Analgesics - pharmacokinetics; Analgesics - pharmacology; Anti-Bacterial Agents - adverse effects; Anti-Bacterial Agents - blood; Anti-Bacterial Agents - pharmacokinetics; Anti-Bacterial Agents - pharmacology; Cross-Over Studies; Dose-Response Relationship, Drug; Fluoroquinolones - adverse effects; Fluoroquinolones - blood; Fluoroquinolones - pharmacokinetics; Fluoroquinolones - pharmacology; Food-Drug Interactions; Humans; Long QT Syndrome - chemically induced; Long QT Syndrome - metabolism; Models, Biological; Morpholines - blood; Morpholines - pharmacokinetics; Morpholines - pharmacology; Pyrazoles - blood; Pyrazoles - pharmacokinetics; Pyrazoles - pharmacology; QTc; thorough QT study; assay sensitivity; concentration-response model; moxifloxacin
• ISSN: 1552-4604
• DOI: 10.1002/jcph.347

As ICH E14 was adopted by the US FDA and the EU CPMC in 2005, thorough QT studies have routinely been analyzed by looking at the time-matched difference between (baseline corrected) QTcF or QTcI under the supra-therapeutic dose and placebo. A study is considered negative, if the two-sided 90% confidence interval for this difference is below 10 ms for all investigated time points. ICH E14 suggests including a positive control, such as moxifloxacin, for assay sensitivity. Concentration-response analysis has been considered a more powerful alternative, but its application to parallel group studies was hampered as a double difference of QTcF per subject cannot be calculated. Recently, a new model based on change from baseline with fixed time and concentration effects has been proposed. It allows for a placebo-corrected prediction of the drug effect with an unbiased standard error, and the estimate of a time effect can be used for assay sensitivity. We demonstrate this approach, utilizing 2 studies reported elsewhere with a crossover design. We compare the results from a conventional concentration-response analysis based on the difference to placebo with results from the novel analysis based on the change from average baseline that includes a fixed time effect.