MEMORANDUM

 

To:                        Douglas Throckmorton, M.D., Division Director

                              HFD 110, Division of Cardio‑Renal Drug Products

 

From:                    John Koerner, Ph.D., Pharmacologist

                              HFD 110, Division of Cardio‑Renal Drug Products

 

Through:               Charles Resnick, Ph.D., Supervisory Pharmacologist

                              HFD 110, Division of Cardio‑Renal Drug Products

 

Subject:                               Nonclinical Electrophysiological Effects of Ranolazine 

 

Date:                     23 October 2003

 

The sponsor (CV Therapeutics, Inc.) has submitted an amendment to this NDA[1] providing nonclinical study reports to address Pharmacology/Toxicology issues discussed in the discipline review letter.  Among other things, the sponsor attempts to prove that QT prolongation with ranolazine is not a concern.  This memo addresses study reports describing ranolazine’s electrophysiological properties in nonclinical studies.  Other study reports provided in the amendment were addressed by Dr. Elizabeth Hausner in her memo dated 29 September 2003. 

 

Although some of the evidence provided in this amendment is consistent with ranolazine induced QT prolongation and repolarization disturbances, other evidence suggests antiarrhythmic rather than proarrhythmic potential.  Nevertheless, these study results do not alter the previous conclusion (see original memorandum dated 4 September 2003) that the nonclinical electrophysiology findings do not preclude human risk, in large part due to lack of comprehensive information addressing the sensitivity and specificity of the in vitro assays cited.  Findings are summarized below.

 

The sponsor addressed effects of ranolazine, its enantiomers and metabolites on several ionic currents that modulate ventricular repolarization. 

 

·        Ranolazine did not inhibit the slowly activating delayed rectifier potassium current (I_Ks) through human channels (KvLQT1/minK) expressed in Xenopus oocytes. Hence, the sponsor concluded that ranolazine at concentrations up to 900 mM does not inhibit I_Ks. However, these data are not convincing since the Xenopus oocyte expression system greatly underestimates potency. The sponsor argued that results of previously submitted studies showing that ranolazine inhibited native I_Ks in isolated canine ventricular myocytes were due to current run-down and therefore an artifact of the test system. Given the contradictory study reports and assay limitations, the effects of ranolazine on I_Ks are presently unclear.

 

·        Ranolazine enantiomers, like racemic ranolazine, inhibited I_Kr and late I_Na with similar potencies in canine ventricular myocytes. Additionally, several ranolazine metabolites inhibited late I_Na when evaluated at a concentration of 10 mM.  The sponsor argues that late I_Na inhibition attenuates ranolazine’s effects on action potential duration in M-cells and other ventricular tissue, thereby preventing proarrhythmic activity.  However, the torsadogenic drugs terfenadine and cisapride also inhibit late I_Na, and terfenadine, like ranolazine, showed similar potencies on I_Kr and late I_Na.  Therefore, inhibition of late I_Na does not appear to preclude risk of drug-induced torsade de pointes.

 

·        Ranolazine increased the decay of I_Ca and therefore inhibited what the sponsor called late I_Ca.  Although this activity theoretically could shorten action potential duration, the actual significance of this finding is unknown.

 

The sponsor addressed proarrhythmic and antiarrhythmic potential of ranolazine in vitro.  The relevance of these findings is unknown since the sensitivity of these in vitro models has not been sufficiently well characterized. 

 

·        Ranolazine lengthened epicardial monophasic action potential duration in isolated female rabbit hearts. Ranolazine did not increase apex-base dispersion of action potential duration nor induce ventricular arrhythmias.  Additionally, ranolazine, at concentrations that inhibit I_Kr and late I_Na in canine ventricular myocytes, exerted actions consistent with antiarrhythmic activity, since it prevented pause-dependent ventricular arrhythmias induced by positive controls.  The specificity of these effects was not evaluated with other QT prolonging drugs that inhibit multiple ionic currents, e.g. terfenadine and cisapride.

 

·        Ranolazine prevented isoproterenol-induced delayed afterdepolarizations in isolated guinea pig ventricular myocytes.  This effect is consisted with ranolazine’s ability to block b1 adrenergic receptors at the concentration evaluated.

 

·        Ranolazine lengthened epicardial monophasic action potential duration in isolated guinea pig hearts, while showing effects consistent with antiarrhythmic activity. The sensitivity and specificity of this model has not been thoroughly characterized.

 

·        This sponsor argues that ranolazine lacks proarrhythmic activity since it does not induce early afterdepolarizations, or increase M-cell APD90 and transmural dispersion of repolarization in isolated canine cardiac wedge preparations.  Ranolazine was additionally negative for in vitro proarrhythmic effects in this preparation, since torsade like arrhythmias did not occur spontaneously and could not be elicited with a single extrastimulus in the drug’s presence.  Epicardial stimulation was utilized for these studies since stimulation at this site was necessary to capture cisapride’s proarrhythmic activity, presumably due to increased transmural dispersion of repolarization at baseline. 

 

In contrast to the present study results, a previous study showed ranolazine to lengthen M-cell action potential duration and increase transmural dispersion of repolarization, but only in the presence of       2 mM potassium.  The lowest potassium concentration evaluated in the present studies was 3 mM.  The difference in findings can be explained by the known enhancement of drug-induced I_Kr inhibition by hypokalemia.

 

The individual studies addressed in this memorandum are listed below.

Study Number: Title	Page
CVT303.069-P: Effect of Ranolazine on IKs in Isolated Canine Left Ventricle Myocytes	3
CVT303.063-P: Effects of Ranolazine Enantiomers on IKs, IKr, and Late INa, and Ranolazine Metabolites on Late INa	3
CVT303.059-P: Electrophysiologic Effects of Ranolazine on Late ICa in Isolated Canine Left Ventricular Myocytes	5
CVT303.065-P: Effects of Ranolazine on Ventricular Repolarization in Isolated Rabbit Hearts	6
CVT303.070-P: Effects of Ranolazine on lsoproterenol, Forskolin, and Ouabain Induced Delayed Afterdepolarizations and Triggered Activity of Guinea Pig Ventricular Myocytes	7
CVT303.061-P: Antiarrhythmic Effects of Ranolazine in a Human LQT Model: The In Vitro Guinea Pig Heart Perfused with the Proarrhythmic Sea Anemone Toxin ATX-II	7
CVT303.068-P: Electrophysiologic Effects of Ranolazine in Arterially Perfused Wedge Preparations from the Canine Left Ventricle: A comparison Between Epicardial and Endocardial Stimulation	10
Assessing Predictors of Drug-induced Torsade de Pointes	12

CVT303.069-P: Effect of Ranolazine on I_Ks in Isolated Canine Left Ventricle Myocytes

 

Ranolazine did not inhibit current (I_Ks) through human channels (KvLQT1/minK) expressed in Xenopus oocytes. Hence, the sponsor concluded that ranolazine at concentrations up to 900 mM does not inhibit I_Ks.

However, these data do not convincingly show lack of effect of ranolazine on I_Ks since the Xenopus oocyte expression system greatly underestimates potency.

 

 

 

CVT303.063-P: Effects of Ranolazine Enantiomers on I_Ks, I_Kr, and Late I_Na, and Ranolazine Metabolites on Late I_Na

 

Ranolazine enantiomers were evaluated for effects on ionic currents modulating ventricular repolarization in isolated canine ventricular myocytes.  Both enantiomers inhibited I_Kr and late I_Na, but only the

S-enantiomer inhibited I_Ks, and then only weakly.

 

                    ^ tested at concentrations up to 100 mM

The following table summarizes effects of racemic ranolazine and its enantiomers in the context of positive control drugs.  Racemic ranolazine and positive control drugs and were not evaluated concurrently. 

Note that terfenadine and cisapride, both torsadogenic drugs, inhibit both I_Kr and late I_Na, similar to ranolazine.  Furthermore, terfenadine, like ranolazine, is similarly potent on late I_Na and I_Kr.  Therefore, blockade of late I_Na does not appear to preclude torsadogenic risk. 

 

 

 

The following figure illustrates effects of several ranolazine metabolites on late I_Na.  At a concentration of

10 mM, several ranolazine metabolites inhibited late I_Na.  Inhibition was independent of basic cycle length (BCL).

 

 

 

 

CVT303.059-P: Electrophysiologic Effects of Ranolazine on Late I_Ca in Isolated Canine Left Ventricular Myocytes

 

Ranolazine increased the rate of decay of I_Ca in isolated left ventricular myocytes.  Late I_Ca, defined as current measured 300 ms after the start of the test pulse, was inhibited with an IC50 of 50 mM.  The figure below, which shows the I_Ca current trace vs time, illustrates that 5 mM ranolazine inhibited late I_Ca while only minimally affecting peak I_Ca.

 

 

 

 

 

CVT303.065-P: Effects of Ranolazine on Ventricular Repolarization in Isolated Rabbit Hearts

 

Ranolazine and its enantiomers were evaluated for effects on monophasic action potential duration (MAPD) in isolated female rabbit hearts paced at a constant rate of 1 Hz. Epicardial MAPDs were monitored in base and apex of the heart for determination of spatial dispersion.  E-4031, which inhibits I_Kr, and ATX-II, which enhances the late I_Na, were utilized as concurrent control test substances. 

 

Ranolazine at concentrations of 1-100 mM increased MAPD in both base and apex in a concentration-related manner.  Effects were similar in base and apex, such that the difference between these two sites, or what the sponsor refers to as dispersion, was not altered. Potencies (EC50s) were similar for both sites (4.3 and 4.8 mM for base and apex, respectively). In comparison, E-4031 and ATX-II increased MAPD and dispersion in a concentration dependent manner.   Ranolazine enantiomers also increased MAPD at a single site in a concentration dependent and manner (the site at which MAPD was monitored was not provided). The enantiomers’potencies on MAPD were similar (6.4 and 5.9 M for R and S enantiomers, respectively).

 

Ranolazine, in contrast to E-4031, did not induce ventricular ectopic beats at any concentration evaluated (data not shown). Ranolazine at concentrations of 5 and 10 mM reduced ventricular ectopic beats induced by E-4031 (data not shown). Additionally, ranolazine at 5 mM attenuated pause dependent ventricular arrhythmias induced by E-4031 and ATX-II (data not shown).  The specificity of this in vitro antiarrhythmic effect was not evaluated using other QT prolonging drugs that inhibit multiple ionic currents, e.g. terfenadine or cisapride.

CVT303.070-P: Effects of Ranolazine on lsoproterenol, Forskolin, and Ouabain Induced Delayed Afterdepolarizations and Triggered Activity of Guinea Pig Ventricular Myocytes

 

Ranolazine at a concentration of 10 mM reduced the amplitude of isoproterenol-induced delayed afterdepolarizations (DADs) and triggered activity in isolated guinea pig ventricular myocytes.  Ranolazine did not alter DADs induced by forskolin and ouabain. The effect on isoproterenol-induced DADs is likely due to ranolazine’s b1 adrenergic receptor blockade at the concentration evaluated.

 

 

 

 

CVT303.061-P: Antiarrhythmic Effects of Ranolazine in a Human LQT Model: The In Vitro Guinea Pig Heart Perfused with the Proarrhythmic Sea Anemone Toxin ATX-II

 

Ranolazine was evaluated for effects on epicardial monophasic action potential duration (MAPD₉₀) in isolated guinea pig hearts (gender not provided) paced at a constant rate of 1.5 Hz.  Acute AV block was induced by infusion of N⁶-cyclypentyladenosine, which blocks adenosine receptors.  E-4031, which selectively inhibits I_Kr, ATX-II, which selectively enhances the late I_Na, and chromanol 293, which inhibits I_Ks, were utilized as concurrent control test substances. In an additional experiment, rate dependence of ranolazine was compared to that of the positive control test substances.

 

Ranolazine at concentrations of 1-100 mM lengthened epicardial MAPD₉₀ in a concentration dependent manner, similar to positive control substances.   Ranolazine’s effects on MAPD₉₀ were independent of pacing cycle length, ranging from 400 to 1000 ms (data not shown).  In contrast, E-4031 and ATX-II but not chromanol 293B showed inverse rate dependence, with greater percentage increases at a basic cycle length of 1000 ms than at 400 ms.

 

 

 

 

Ranolazine attenuated MAPD lengthening and prevented early afterdepolarizations and polymorphic ventricular tachycardia induced by ATX-II in a concentration related manner.  Effects of positive control drugs that inhibit multiple ionic currents (including the late I_Na), e.g. terfenadine and cisapride, were not evaluated.

 

 

 

 

 

 

CVT303.068-P: Electrophysiologic Effects of Ranolazine in Arterially Perfused Wedge Preparations from the Canine Left Ventricle: A comparison Between Epicardial and Endocardial Stimulation

 

Endocardial stimulation appears to sensitize the isolated canine left ventricular wedge preparation to arrhythmias induced by QT prolonging drugs, since cisapride’s proarrhythmic effect in this model was captured with epicardial, but not with endocardial stimulation. Ranolazine was previously evaluated using endocardial stimulation; however, effects with epicardial stimulation have not yet been evaluated.

The purpose of the present study was to evaluate the in vitro electrophysiologic effects of ranolazine in canine wedges stimulated from the epicardial surface. Results were compared to those using endocardial stimulation in the same experiment. Both normokalemic (4 mM potassium) and hypokalemic (3 mM potassium) conditions were utilized, as was pacing the preparations at both short and long basic cycle lengths (500 and 2000 ms) to address rate dependence of effects.

 

With epicardial stimulation and normokalemia, ranolazine at concentrations of 1-100 mM did not lengthen M-cell or epicardial action potential duration (APD90). With epicardial stimulation and hypokalemia, ranolazine at 100 mM lengthened epicardial but not M-cell APD90. 

 

While ranolazine prolonged transmural QT interval and increased Tpeak –Tend in a concentration dependent manner, it did not lengthen M-cell action potential duration or transmural dispersion of repolarization (TDR) under any conditions. Indeed, ranolazine decreased TDR when evaluated with hypokalemia, likely due to a proportionately greater increase in epicardial vs M-cell APD90.

 

 

 

 

With endocardial stimulation, ranolazine at concentrations of 1-100 mM did not lengthen M-cell or  epicardial action potential duration (APD90). Effects were independent of potassium concentration. 

Ranolazine prolonged transmural QT interval in a biphasic manner, with peak effects at 5-10 mM. 

Ranolazine increased Tpeak –Tend over control values, but only at the highest concentration evaluated. Ranolazine did not significantly lengthen M-cell action potential duration or increase transmural dispersion of repolarization (TDR). Ranolazine’s electrophysiologic effects were similar at basic cycle lengths of 500 and 2000 ms (only effects at 2000 ms are shown).

 

 

 

 

 

Ranolazine did not induce torsade de pointes - type arrhythmias in vitro in canine left ventricular wedges using epicardial stimulation. 

 

           ^ The sponsor attempted to induce ventricular arrhythmias using a single extrastimulus applied to the

           epicardial surface at progressively shorter intervals until refractoriness was reached. This methodology was

           successful in eliciting ventricular arrhythmias in 2 of 6 wedge preparations exposed to cisapride (0.2 mM).

           Spontaneous arrhythmias were not observed with cisapride. Cisapride’s proarrhythmic effects were limited to a

           single concentration, with slightly higher and lower concentrations yielding no arrhythmias.

 

 

Assessing Predictors of Drug-Induced Torsade de Pointes

 

The sponsor provided a manuscript arguing that delayed ventricular repolarization and proarrhythmia are separable, i.e. drugs that do not induce early afterdepolarizations or increase dispersion of repolarization are unlikely to cause TdP, even in the setting of prolonged QT.

 

 

The sponsor supports the hypothesis that torsadogenic drugs are always associated with early afterdepolarizations and spatial dispersion of repolarization with data from several in vitro models.  The sponsor did not provide a comprehensive assessment of the sensitivity and specificity of any single model.