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


CC:                       Elizabeth Hausner, D.V.M., Pharmacologist

                              HFD 110, Division of Cardio‑Renal Drug Products


                              Albert DeFelice, Ph.D., Supervisory Pharmacologist

                              HFD 110, Division of Cardio‑Renal Drug Products


NDA: 21526

Sponsor: CV Therapeutics, Inc.

Drug: RANEXA, Ranolazine

Proposed Clinical Indication:  Angina


Date Completed: 09/03/03             


Re: Nonclinical Electrophysiological and Proarrhythmic Effects of Ranolazine 


Executive Summary


Ranolazine’s nonclinical electrophysiological effects are generally consistent with QT interval prolongation observed in clinical trials. However, the sponsor provided information arguing against proarrhythmic properties of ranolazine.


According to theory, torsade de pointes with QT prolonging drugs arises from increased transmural dispersion of repolarization and triggered activity (early afterdepolarizations) rather than from QT interval prolongation, per se.  The elegant models of Antzelevitch argue that findings in isolated canine ventricular strips and wedge preparations, in particular action potential lengthening and early afterdepolarizations in M-cells, transmural dispersion of repolarization, and induction of torsade-like arrhythmias in vitro can predict proarrhythmic potential of a drug.  Theory is supported in part utilizing data from positive control drugs such as amiodarone, cisapride, d-sotalol, erythromycin, quinidine and terfenadine.  


While findings with ranolazine are primarily negative in the isolated canine ventricular wedge preparation, ranolazine increases M-cell action potential duration and transmural dispersion of repolarization under conditions of hypokalemia, consistent with its ability to inhibit repolarizing currents, IKr and IKs, at concentrations similar to those required for its proposed mechanism of action.  Moreover, proarrhythmic potential was not adequately evaluated since ranolazine was not tested under conditions necessary to capture cisapride-induced arrhythmias in this model. Additionally, in vivo evaluation for proarrhythmia was not comprehensive. 


Additional regulatory concerns reflect the complex nature of any proarrhythmia model, lack of evaluation of metabolites in this in vitro model, and adequacy of testing for proarrhythmia, utilizing known risk factors for clinical arrhythmias, e.g. female gender, different pacing modalities (pause, acceleration),  adrenergic influences, and heart failure. 


For these reasons, study results provided by the sponsor cannot preclude the risk of proarrhythmia with ranolazine.

Review and Evaluation


Ionic Currents


Ranolazine inhibited both repolarizing currents (IKr and IKs), depolarizing currents (INa-Late, and ICa-L), and

INaCa in isolated canine ventricular myocytes.  


Rank order for potency on native currents was IKr = IKs > INa-Late > INaCa > ICa-L.


Ionic Current

(Canine Ventricular Myocytes)

IC50 (mM)






> 100









Ranolazine was also shown to weakly inhibit cloned human potassium channels (hERG and KvLQT1), with IC50s of 86 mM and 1.46 mM, respectively.  While this data demonstrates the ability of ranolazine to block human repolarizing currents, it is not useful for potency estimation since studies were performed using a Xenopus oocyte expression system, which greatly underestimates potency. For example, IC50s for concurrent positive controls dofetilide and verapamil were approximately 20 -30 fold greater in oocytes than in mammalian expression systems (1,2).  Additionally, KvLQT1 channels were not coexpressed with minK, which likely further underestimates ranolazine potency on this repolarizing channel (3,4).



Ionic Current

(cloned human channel  expressed in Xenopus oocytes)

IC50 (mM)












* IC50s for inhibition of hERG expressed in mammalian expression system:

dofetilide, 0.015 mM; verapamil 0.143 mM (literature reference values (1,2). 


The sponsor evaluated effects of several metabolites at a single concentration of 50 mM on IKr in canine ventricular myocytes. Several metabolites inhibited IKr by about 50% at this concentration.  The sponsor did not evaluate effects of metabolites on other ionic currents, including IKs. 
Action Potential Duration


Ranolazine was evaluated for effects on action potential duration in several systems, including epicardium and M-cells from canine left ventricular strips.  Concentrations evaluated ranged from 1 to 100 mM.  In canine epicardial strips, ranolazine (³ 5 mM) lengthened APD90 in a concentration dependent way. Effects were potassium and rate dependent, with lengthening slightly more pronounced with hypokalemia

(2 vs 4 mM potassium) and a more rapid stimulation rate (500 ms vs 2000 ms basic cycle lengths; i.e.,

120 vs 30 beats per minute).  Ranolazine did not lengthen APD90 in M-cell strips.


Ranolazine was also evaluated for effects on action potential duration in isolated, buffer perfused canine ventricular wedge preparations.  Ranolazine was evaluated in epicardium and M-cells at concentrations of 1- 100 mM, under conditions of normokalemia and hypokalemia (4 and 2 mM potassium) using basic cycle lengths of 500 and 2000 ms (stimulation frequencies of 120 and 30 beats per min, respectively).  


In the presence of normokalemia, ranolazine did not lengthen APD90 in epicardium and M-cells. However,  in the presence of hypokalemia, ranolazine lengthened APD90 in a concentration dependent way in both of these tissues.



Ranolazine also prolonged transmural QT interval, Tpeak to Tend, and transmural dispersion of repolarization, and altered T-wave morphology (wide, low, notched; see below).  Ranolazine (10 and

100 mM)  also lowered maximum upstroke velocity, indicating sodium channel blockade at concentrations evaluated. Maximum upstroke velocity was lowered by approximately 50% at 100 mM.  Exposure duration and time course of effects were not provided.






In comparison, most but not all positive control drugs showed pronounced M-cell APD90 lengthening and increased transmural dispersion of repolarization in canine ventricular strips or wedges (5-12).



Pronounced Effects


Transmural Dispersion of Repolarization

Amiodarone (chronic)





















* Biphasic concentration – response relationship, with prolongation at low concentrations, and shortening or attenuation of effect at high concentrations.

^ Requires extended perfusion period (120 min) and hypokalemia (2 mM potassium) for effects.


In canine Purkinje fibers, ranolazine (1 to 100 mM) shortened APD90 and depressed maximal upstroke velocity. APD90 effects appeared independent of extracellular potassium (2 vs 4 mM) and basic cycle length (500 vs 2000 ms).  Similar to findings in M-cells, ranolazine (10 and 100 mM) decreased maximum upstroke velocity, with approximately 50% inhibition at 100 mM.  Ranolazine (5 and 10 mM) shortened

d-sotalol induced APD lengthening and suppressed EADs in a concentration related way.  Lack of action potential duration lengthening in canine Purkinje fibers with ranolazine is not impressive considering  the poor sensitivity of this model to torsadogenic drugs such as terfenadine and terodiline (13, 14).





Ranolazine (1- 100 mM) did not induce early afterdepolarizations (EADs) in canine ventricular M-cell strips and Purkinje fibers. Ranolazine (5 and 10 mM) reversed d,l-sotalol induced APD90 lengthening and prevented EADs in both canine M-cell strips and Purkinje fibers. Lower ranolazine concentrations were not evaluated for this property.


In comparison, the historical positive control erythromycin induced EADs in 2 of 10 M-cell strips at the highest concentration evaluated (8). Quinidine (3.3 mM)  induced EADs in 3 of 5 M-cell strips and EAD-induced triggered activity in 2 of 5 M-cell strips with hypokalemia (2 mM potassium) but not with normokalemia (4 mM potassium) (9).  M-cell strips from chronically amiodarone treated dogs did not show EADs or triggered activity (5).  Additionally, chronic amiodarone attenuated d-sotalol induced APD90 lengthening and EADs.


The sponsor evaluated ranolazine for torsade-like polymorphic arrhythmias (TdP, spontaneous and programmed electrical stimulation (PES) inducible) in isolated canine ventricular wedge preparations.  Ranolazine was evaluated in the presence and absence of hypokalemia (2 mM potassium).  PES consisted of a single extrastimulus (S2) applied to the epicardial surface of the wedge preparation with the basic stimuli (S1) applied to the endocardium at basic cycle lengths of 500 and 2000 ms.  


Ranolazine did not induce torsade like arrhythmias in these studies.  However, the sample size was quite small (n =  4 for normokalemia, and n = 3 for hypokalemia), and positive historical control data from the Antzelevitch laboratory shows TdP incidences for d-sotalol, erythromycin and cisapride to be fairly low in this in vitro model (7, 8, 10, 11). Additionally, cisapride’s proarrhythmic effects were limited to a single concentration, with negative proarrhythmia findings at both lower and higher concentrations, consistent with cisapride’s biphasic effect on APD90 and transmural dispersion of repolarization.  Most importantly, induction of TdP with cisapride required epicardial pacing (S1), which doubled baseline transmural dispersion of repolarization.  As only endocardial pacing (S1) was utilized in the ranolazine experiments, cisapride’s positive torsadogenic effect in this model is not applicable to the ranolazine experiments.  Indeed, the inability of cisapride to induce TdP in this model using endocardial pacing (S1), argues against the model’s sensitivity to torsadogenic drugs when evaluated in this manner. 



TdP Incidence



Ranolazine (1-100 mM)

0 of 4

0 of 3

d-Sotalol (100 mM)

2 of 6

2 of 8

3 of 6

4 of 6

Erythromycin (100 mM)

no data

3 of 4

Cisapride (0.1-5 mM)


2 of 6*

                              * TdP inducible at 0.2 mM cisapride (only); required epicardial pacing (S1)



In Vivo Effects: QT Interval Prolongation and Proarrhythmia


Acute intravenous infusion of ranolazine was evaluated for effects on QT interval and proarrhythmia in anesthetized dogs with acute AV block.  Ranolazine was administered at a dose of 0.5 mg/kg plus 1, 3 and 15 mg/kg/hr iv for 30 minutes/dose level (n=6).  In one additional dog, ranolazine was administered at a dose of 1.5 mg/kg plus 15 and 30 mg/kg/min iv for 30 minutes per dose level. Sotalol (8 mg/kg followed by 4 mg/kg/hr iv for 20 minutes served as a positive control (n=5). Left ventricular effective refractory periods, QT interval, and QRS duration were determined at basic cycle lengths of 300, 400, 600 and

1000 ms following each dose of ranolazine or sotalol.  All dogs then received bolus intravenous doses of phenylephrine (10, 20, 30, 40 and 50 mg/kg iv) in order to elicit arrhythmias.


In this study, ranolazine increased QT interval by approximately 10% (not statistically significant).  The QT effect was biphasic, with the maximum increase observed at 3 mg/kg/hr.  Ranolazine increased left ventricular refractoriness but only at the fastest stimulation rate (300 ms basic cycle length).  Ranolazine increased QRS interval by 10% (statistically significant) at the next higher dose of 15 mg/kg/hr.  Proarrhythmia was not observed with ranolazine in the presence of phenylephrine.  The highest ranolazine dose of 30 mg/kg/hr induced acute heart failure and death.


In contrast to findings with ranolazine, sotalol induced a large increase in QT interval (33% increase at 1000 ms basic cycle length), and increased left ventricular refractoriness without altering QRS duration.  Phenylephrine induced tachyarrhythmias, including TdP and ventricular fibrillation, in sotalol treated dogs.


Limitations of this study design include its acute nature, since chronic but not acute AV block has been shown to promote proarrhythmia by electrical remodeling and downregulation of ion channels, including IKr (15, 16). Additionally, ranolazine displaces radiolabeled ligand from a-1 adrenergic receptors,  suggestive of potential functional antagonism (17).  If ranolazine is indeed a functional a-1 adrenergic antagonist at doses administered above, then the in vivo canine model used for this study is inappropriate.  Finally, positive control data is quite limited for this model, lessening the regulatory importance of negative findings.



1.      Snyders DJ and Chaudhary A, High-affinity open-channel block by dofetilide of HERG expressed in a human cell line, Mol Pharmacol 49: 949-955 (1996) [PMID: 8649354].


2.      Zhang S, Zhou ZF, Gong QM, Makielski C, and January CT, Mechanism of block and identification of the verapamil binding domain to HERG potassium channels, Circ Res 84: 989-998 (1999) [PMID: 10325236].


3.      Yang T, Hideaki K, Roden D,  Phosphorylation of the IKs channel complex inhibits drug block: Novel mechanisms underlying variable antiarrhythmic drug actions, Circulation 108: 132-134 (2003) [PMID 12835205].


4.      Busch A, Busch GL, Ford E, Suessbrich H, Lange HJ,  etal, The role of IsK protein in the specific pharmacological properties of the IKs channel complex, Br J Pharmacol 122: 187-189 (1997) [PMID 9313924].


5.      Sicouri S, Moro S, Litovsky S, Elizari MV, Antzelevitch C,  Chronic amiodarone reduces transmural dispersion of repolarization in the canine heart,  J Cardiovasc Electrophysiol 8:1269-79 (1997) [PMID: 9395170].


6.      Burashnikov A, Antzelevitch C, A combination of IKr, IKs, and ICa or INa block produced a relatively homogeneous prolongation of repolarization of cells spanning the canine left ventricular wall, PACE 20: II-1216, (1997).


7.      DiDiego J, Belardinelli L, Antzelevitch C, Cisapride-induced transmural dispersion of repolarization and torsade de pointes in the canine left ventricular wedge preparation during epicardial stimulation, Circulation, 108:1027-33 (2003) [PMID: 12912819].


8.      Antzelevitch C, Sun Z, Zhang Z, Yan G, Cellular and ionic mechanisms underlying erythromycin-induced long QT intervals and torsade de pointes, J Amer Coll Cardiol 28: 1836-1848 (1996) [PMID 8962574].


9.      Sicouri S, Antzelevitch C, Drug-induced afterdepolarizations and triggered activity occur in a discrete subpopulation of ventricular muscle cells (M cells) in the canine heart: quinidine and digitalis,  J Cardiovasc Electrophysiol 4: 48-58,  (1993) [PMID: 8287236].


10.   Shimizu W, Antzelevitch C, Effects of a K(+) channel opener to reduce transmural dispersion of repolarization and prevent torsade de pointes in LQT1, LQT2, and LQT3 models of the long-QT syndrome,  Circulation, 102:706-12 (2000) [PMID: 10931813].


11.   Shimizu W, Antzelevitch C, Differential effects of beta-adrenergic agonists and antagonists in LQT1, LQT2 and LQT3 models of the long QT syndrome, J Am Coll Cardiol. 35:778-86, (2000) PMID: 10716483


12.   Antzelevitch C, (personal communication).


13.   Gintant GA, Limberis JT, McDermott JS, Wegner CD, Cox BF, The canine Purkinje fiber: an in vitro model system for acquired long QT syndrome and drug-induced arrhythmogenesis,  J Cardiovasc Pharmacol. 37:607-18 (2001) [PMID: 11336111].


14.   Pressler ML, Warner MR, Rubart M, Rardon DP, Zipes DP, In vivo and in vitro electrophysiologic effects of terodiline on dog myocardium,  J Cardiovasc Electrophysiol 6: 443-54 (1995) [PMID: 7551314].


15.   Vos MA, de Groot SH, Verduyn SC, van der Zande J, Leunissen HD, Cleutjens JP, van Bilsen M, Daemen MJ, Schreuder JJ, Allessie MA, Wellens HJ, Enhanced susceptibility for acquired torsade de pointes arrhythmias in the dog with chronic, complete AV block is related to cardiac hypertrophy and electrical remodeling, Circulation: 98:1125-35 (1998) [PMID: 9736600].


16.   Volders PG, Sipido KR, Vos MA, Spatjens RL, Leunissen JD, Carmeliet E, Wellens HJ,  Downregulation of delayed rectifier K(+) currents in dogs with chronic complete atrioventricular block and acquired torsades de pointes, Circulation 100:2455-61 (1999) [PMID: 10595960]


17.   Hausner EA, Pharmacology Toxicology Review, NDA 21526.