Safety margins from ex vivo APD to human plasma exposure?
December 10, 2009 7:31 am Research, ScienceDear members
I would like to hear your thoughts on what safety margins you believe we need when evaluating results from ex vivo action potential recordings (e.g. GP papillary muscle APD90).
Any references would also be greatly appreciated.
Also, how big an effect do you consider to be relevant?
Best regards, Morten - H. Lundbeck A/S
December 11th, 2009 at 9:16 am
Dear Member
There is some published information available on this topic - an ABPI QT Task Force (chaired by Tim Hammond) published a large , innovative review (Redfern et al ) on the topic of Safety Margins relative to cardiac risk for a wide panel of agents with a wide range of QT liability - the bigger the margin , the better !!
December 19th, 2009 at 9:25 am
Morten,
we know from sources like Redfern et al (2002) and Rob Wallis’ previous presentations to the society that the positive predictive power of the APD assays. That is if you have an effect on APD90 it will very likely translate at the same unbound plasma concentrations in man. There is not a great publication to further elaborate on your question but using Jonker et al (2005) which examines the translation from hERG in vitro to QTc prolongation in man for dofetilide we know that even 5% block of hERG is associated with QTc prolongation in man for that compound. My personal experience is that this amount of Purkinje fibre APD90 is that 5% block of hERG is already associated with some prolongation. Because a 30th of the IC50 is an IC3 (a 3% inhibition) you can see pharmacologically why 30-fold below the hERG IC50 was a key value in the Webster et al (2001) and Redfern et al (2002) papers. These papers used actual clinical Cmax concentrations in the calculation, and my preceeding comments mean that you don’t want to have unbound concentrations in the clinic higher than a 30th of the hERG IC50. How far below that you need your mean therapeutic concentrations depends on your PK variability, your peak-to-trough ratio and the likelihood of drug-drug interactions. As a rule of thumb I’d use another 30-fold ( makes it 30-fold below a 30th of the hERG IC50 or 900-fold below the hERG IC50). So overall I advise the teams to try and create 1000-fold separation between target potency and hERG block. The short answer to your question focussed on APD90; I’d certainly be worried about 10% prolongation and would try and remain 30-fold below this. I’d further hope that my highest clinical exposures did not get within 10-fold of the concentration. That way I think you would have a low risk of arrhythmia but also a low risk of a positive thorough QT study.