OIII-A-1FDA evaluation of cardiac repolarization data for 19 drugs and drug candidates

BACKGROUND/AIMS Drug-induced QT interval prolongation, which can lead to torsade de pointes, is a concern in drug development. The present study evaluated the predictability of nonclinical assays (in vitro hERG and in vivo safety pharmacology studies) for drug-induced QT interval prolongation. METHODS Nonclinical and clinical data on 19 drugs/candidates were evaluated. Data included that submitted to the FDA in support of INDs and NDAs, and published data. Clinical studies included positive controlled QT studies and those showing QT interval prolongation without a positive control arm. Drugs were considered positive if the mean QTc increase was greater than 5 msec. Nonclinical criteria for a positive assay were: IC50 ≤1 μM for hERG inhibition; 10% increase in QTc in vivo or significant at P<0.05. RESULTS Of 10 drugs that were positive clinically (QTc increases of 5.5 to 62 ms), false negative rates were 50% and 60%, for hERG and in vivo studies, respectively. Of 9 drugs that were negative clinically, false positive rates were 22% and 14% for hERG and in vivo studies, respectively. For drugs positive in either hERG or in vivo studies, false negative and false positive rates were 20% and 22%, respectively. Changing the IC50 cutoff or normalizing IC50's to therapeutic free plasma levels did not change predictability in a usable manner. CONCLUSION In this database, data from in vitro hERG and in vivo safety pharmacology studies was insufficient to fully predict whether drugs will prolong QT interval clinically. Clinical Pharmacology & Therapeutics (2005) 79, P29–P29; doi: 10.1016/j.clpt.2005.12.106