Proof Negative

The present regulatory practice requires practically every new pharmaceutical entity to be assessed for potential propensity to modify cardiac ventricular repolarisation and to cause torsade de pointes tachycardia. Guidance document E14 of the International Conference of Harmonisation (1) specifies regulatory requirements for such an assessment.

In brief, the clinical programme of every new compound needs to include a so-called ‘thorough’ QT/QTc study that can verifiably answer the question of whether the drug does or does not prolong the QT interval duration. It is well recognised that a drug-induced QT/QTc interval prolongation is a rather imprecise surrogate of torsadegenic toxicity and that more accurate indicators are highly desirable. Unfortunately, although proposals have been made for deriving more accurate surrogates in both preclinical and clinical studies (2,3), consensus is still lacking on their value and on the possibility of replacing investigations of drug-induced QT/QTc prolongation with more precise studies. From a regulatory point of view, a drug has no meaningful effects on cardiac repolarisation if the upper singlesided 95 per cent confidence interval of QTc change on treatment (corrected for baseline and placebo) is below 10ms.

REGULATORY AWARENESS THRESHOLD

This threshold of 10ms has been derived from simulation studies of data distribution in previous clinical investigations that contained electrocardiographic measurements. While it might be questionable whether this threshold is still in agreement with the present technologies for electrocardiographic measurements, it is more important to recognise that a threshold defining the presence or absence of drug-induced changes on QT interval duration has to be derived from regulatory decisions rather than from scientific observations and their statistical analyses. This is because, scientifically speaking, a distinction cannot be made between drugs that do and do not cause torsade de pointes tachycardia.