Assessment of Drug‐induced QT Prolongation: To Bin or Not to Bin?

“ i o p The possibility of torsades-type proarrhythmia of oth antiarrhythmic and nonantiarrhythmic drugs is a ell-recognized problem. Several preclinical models ave been proposed to identify drugs inducing torsades requently enough to require regulatory assessment of he relevant risk-benefit ratio. However, none of these reclinical models has been accepted by regulators as ufficiently sensitive and specific to replace clinical nvestigations. Moreover, the clinical frequency of torades induction that is believed to require regulatory wareness is very low (above approximately 1 in 10 to 0 exposures), and thus surrogates applicable to clincal studies are needed. Although it is generally undertood that QT interval prolongation is a rather impreise surrogate of torsadegenic toxicity, no other test pplicable in clinical studies has been studied suffiiently to replace the test of drug-induced QT interval rolongation. Because QT interval prolongation is indeed a rather eak surrogate (with especially poor specificity), the egulators presently require the tests of drug-induced T interval prolongation to be carried out with a very igh precision. It is believed that drugs that can be hown not to cause even small QT interval prolongation ill cause torsades either not at all or so infrequently hat their regulatory assessment can be freed from deailed torsades-related risk-benefit evaluation. The required precision of the detection of drugnduced QT prolongation is so high that all factors ecreasing the precision need to be carefully elimiated. An imprecise heart rate correction of the QT nterval is one of the most important of these factors. It s well recognized that all ad hoc correction (Bazett,