MEDICAL REVIEW OF SAFETY

 

NDA#21,526

Drug Name:  ranolazine (Ranexa^TM)

Sponsor: CV Therapeutics, Inc.

Review start date: February, 2003

Medical Reviewer: Maryann Gordon, M.D.

 

 

Table of Contents

Summary of safety…………………………………………………………..	2
1.0   Overall clinical program………………………………………………... Extent of exposure Adequacy of clinical program	5
2.0  Demographics…………………………………………………………… Exclusion criteria	12
3.0 All adverse events……………………………………………………….. Dose response	14
4.0 Serious safety…………………………………………………... Deaths Withdrawals for adverse events Serious adverse events Syncope	18
5.0 ECG changes…………………………………………………………... ECG intervals T-wave morphology	26
6.0 Vital signs………………………………………………………………	38
7.0 Laboratory evaluations…………………………………………………	40
8.0 Special populations………………………………………………… Age, Race, Gender Hepatic impairment Congestive heart failure Other concomitant diseases	42
9.0 Drug-drug interactions ………………………………………………….	45
10.0 Abrupt withdrawal…………………………………………………..	46
11.0 Safety Update………………………………………………………….	46
12.0 Placebo study 3031…………………………………………………..	56
13.0 Placebo study 3033…………………………………………………	66

Summary of safety

This is an agent that was developed for chronic stable angina. The number of patients who were included in the database and received ranolazine is around 2700; approximately 280 subjects received the drug for at least 1 year. Several formulations were studied: immediate release (IR) with nearly 1300 subjects, sustained release (SR) with nearly 1360 subjects, and intravenous (IV) with less than 80 subjects. There were 3 placebo controlled SR angina clinical trials (designated as Phase 2/3 controlled angina) with a total of 749 ranolazine and 455 placebo subjects.  One of these trials was a cross over with doses up to 1500 mg bid. The other trial was a parallel group, 12 weeks duration with the highest dose being 1000 mg bid.  The third trial enrolled only 11 patients. Targeted SR dose range was 500 mg- 1500 mg bid.

 

QT interval prolongation and T wave morphology changes

The sponsor found out early in development that ranolazine increases the QT interval on ECG and changes the morphology of the T wave.  The drug effect at peak concentration is greater than at trough. The mean changes by dose are shown below.

 

Mean change from baseline in QT[1]/QTc interval (msec) at peak

	Placebo N=432	Ranol 500 N=177	Ranol 750 N=269	Ranol 1000 N=428	Ranol 1500 N=170
Mean change from baseline	-3.7/-2.0	-1.0/3.3	7.3/3.5	6.7/5.0	8.5/11.0
Max mean change from baseline	0.9/1.1	-1.0/3.3	16.3/8.9	11.5/8.1	8.5/11.0

Table N-1.3.2.1vol 1.0376

 

The table below shows the number and percent of patients, by dose, who had selected QTc interval changes from baseline at endpoint at peak drug concentration.

 

No. and (percent) of patients

Change from baseline	Placebo N=433	Ranol 500 N=177	Ranol 750 N=271	Ranol 1000 N=433	Ranol 1500 N=170
0-30 msec	167 (38.6)	67 (37.9)	160 (59.0)	242 (55.9)	71 (41.8)
31-60 msec	21 (4.8)	20 (11.3)	6 (2.2)	29 (6.7)	28 (16.5)
>61 msec	4 (0.9)	6 (3.4)	1 (0.4)	1 (0.2)	10 (5.9)

Table N-15.3.1 vol 1.0377

 

There also were changes in the morphology of the T-wave during ranolazine use. The frequencies of notched T waves are shown below by treatment group at peak and trough concentrations (study CVT 3031).

 

% of subjects with notched T waves

	Placebo	Ranol 500	Ranol 1000	Ranol 1500
peak	2	1	3	6
trough	<1	<1	5	5

 

There were more notched T waves were reported in the Ranolazine 1000 mg and 1500 mg doses than in the placebo and ranolazine 500 mg dose groups.

 

The number and percent of patients in CVT 3033 with notched T waves at weeks 2 and 12 by drug group are shown below.

 

% of subjects with notched T waves (at peak)

Placebo		Ranolazine SR  750 mg		Ranolazine SR  1000 mg
Week 2	Week 12	Week 2	Week 12	Week 2	Week 12
0.4	0	4.1	1.2	2.0	3.4

 

Genetic studies have shown that long-QT syndrome (LQTS) is a primary electrical disease caused by mutations in specific ion channels.[2] LQTS patients exhibit QT prolongation on the ECG and are at risk of arrhythmogenic syncope and sudden death. In addition to duration, T-wave morphology is often abnormal, and notched T waves have been included in diagnostic criteria.[3] This pattern has been associated with a poor prognosis.[4]

 

Drug interactions

CYP3A4 is a major determinant for ranolazine clearance. There was an average increase of plasma concentration of 3- to 4-fold in the presence of the potent CYP3A4 inhibitor ketoconazole (200 mg bid)[5].  The effect on QTc is shown below.

 

 

 

 

Concomitant use with diltiazem resulted in increases in ranolazine plasma concentrations of 1.5- to 2.4-fold over the diltiazem total daily dose range (180-360 mg)[6]. Ranolazine 1,000 mg bid at steady-state caused a less than two-fold increase simvastatin exposure dosed at 80 mg qd[7].

 

Hepatic impairment

Subjects with moderate hepatic impairment had increases in AUC and Cmax. This resulted in increases in QTc.

 

Renal impairment

Subjects with creatinine clearance decreasing from 100 mL/min to 30 mL/min had increases in AUC and Cmax.

 

Adverse events

Commonly reported events in the SR controlled angina studies were dizziness (6.8% placebo subtracted), constipation (6.1%), and nausea (5.0%). Events reported mostly by subjects receiving 1500 mg bid included syncope, sweating, and vomiting. Syncope and assorted events that could be related to syncope were reported by 19.2% of the overall ranolazine population compared to the 4.4% of the placebo population. There is orthostatic hypotension reported with the higher doses[8].

 

The survival curves of chronic angina patients on ranolazine versus those on placebo over a 3 month period were similar. The Cox proportional hazards regression model rules out that ranolazine is more than 3.27 times worse than placebo or more than 8.2 times better than placebo.

Changes in laboratory values were unremarkable and included small decreases in hematocrit/

hemoglobin and small increases BUN and serum creatinine.

 

1.0 Overall clinical program

Eighty-one clinical studies were conducted in support of the safety and efficacy of ranolazine. The studies sponsored by Syntex are identified in the NDA with the prefix “RAN” and those sponsored by CVT with the prefix “CVT.” Clinical reports for studies sponsored by Syntex were also assigned a report number by Syntex that begins with the prefix “CL.”

 

As agreed with the Division at the pre-NDA teleconference of 20 December 2001, 64 of the 81 studies are included in the Integrated Safety Summary (ISS) database. The 17 studies that were not integrated include 16 early, low-dose studies conducted by Syntex and 1 bioequivalence study (CVT 301-15). These studies were discussed in the narrative of the ISS, but the data were not integrated.

 

The safety data were generated with several formulations of ranolazine that were used throughout the course of the development program. All of these formulations resulted in systemic exposure to the same ranolazine moiety (i.e. ranolazine base). Ranolazine SR is the proposed commercial formulation.

The 64 studies that comprise the ISS database were categorized as follows:

 

•Thirteen Phase 2/3 controlled studies: 11 angina (3 SR, 8 IR), and 2 intermittent claudication studies (SR);   

• Five Phase 2/3 uncontrolled open-label extension studies in angina (2 SR and

3 IR);

 

•Forty-six Phase 1 and clinical pharmacology studies, including two studies in CHF (one SR and one IV), one study in patients with renal failure (SR) and one study in patients with hepatic impairment (SR).

 

The overview of the program including sample sizes is shown below.

There were 1299 subjects who received the immediate release, 1359 received sustained release, and 77 received the IV formulations. Subjects could receive more than 1 formulation.

 

Of the 2985 patients in the ISS data base, 2682 received ranolazine (any formulation) and 1529 received placebo (some patients received both ranolazine and placebo, about 800 received only placebo or placebo first[9]). Of the 1529 placebo subjects, 947 received placebo IR, 13 received placebo IV, and 569 received placebo SR (vol 1.0343 page 3).

 

Duration of exposure ranged from single dose to more than 2 years of treatment. Doses ranged from 10 mg once daily to 2000 mg bid. More than 45% (1359/2985) of the patients in the ISS data base  received the SR formulation. The mean duration of exposure for patients receiving open label ranolazine SR as of the cut off date is 448 days.

 

“All” subjects (ISS data base)

A total of 2985 patients from 64 studies are included in this database. The formulations used were IR and SR. Two studies are still ongoing: CVT 3032 and CVT 3034, both are uncontrolled follow up studies.

 

The mean duration of exposure, the number and percent of the ISS patients who discontinued early and reasons for the discontinuations are shown below.

 

 

 

 

Mean duration of exposure was 160 days for ranolazine and 25 days for placebo. Overall, more than 4 times as many ranolazine patients discontinued treatment compared to placebo patients (18.3% vs. 4.2%).  Of the 492 ranolazine patients who discontinued, 212 (7.9%) did so because of an adverse event, 31 (1.2%) for  non-compliance, 31 chose to withdraw, 27 (1.0%) died (plus 4 not included in list), 75 (2.8%) were stopped because of the sponsor, and 100 (3.7%) discontinued for other reasons.

 

Controlled trials

There were 11 studies with 2103 angina patients receiving either the IR or the SR formulations. A list of the Phase 2/3 controlled angina studies by individual study number and number of subjects by dose is shown below.

 

 

 

 

IR doses ranged from 10 mg qd to 400 mg tid. SR doses ranged from 500 mg bid to 1500 mg bid. The 2 major studies that used the SR formulation were CVT 3031 and CVT 3033. These studies together enrolled 1102 ranolazine subjects (63.4% of the ranolazine population).

 

Controlled SR angina trials

A total of 1025 patients were treated in Phase 2/3 SR controlled angina studies. Of these patients, 749 received ranolazine (570 patients received only ranolazine SR and 179 received both ranolazine SR and placebo) and 276 received only placebo[10].

 

The table below shows the mean duration of exposure and the patient disposition for this selected patient population, by dose.

 

The mean durations of exposure were 66 days for the total SR population (n=749) and 53 days for placebo (n=455).  The dose of ranolazine with the largest number of patients is 1000 mg bid (459 patients). Discontinuation rates for any reason was 12.1% for any dose of ranolazine compared to 7.7% for placebo. The reason with the largest percent of discontinuations in the total SR group was for an adverse event (7.9%). There is no obvious dose response for noncompleters but sample sizes and length of exposure are unequal.

 

Long term, open label trials

CVT 3032 and CVT 3034

Of the 550 patients enrolled in these studies[11], 440 are still ongoing and 110 (20%) were discontinued. Of the patients who discontinued, 58 did so because of an adverse event. In addition, there were 262 subjects who received the IR formulation during one of 5 uncontrolled IR studies.

 

The disposition of these subjects (and IR patients from earlier trials) is shown in the table below.

 

 

Twenty percent of long term subjects withdrew early from ranolazine treatment. More than half of these withdrew for unacceptable adverse events.

 

Other populations

There were 6 studies with patient populations with diseases other than angina. These include  CHF (2 studies, 96 patients), intermittent claudication (2 studies, 48 patients), renal impairment (1 study, 29 subjects), hepatic impairment (1 study, 32 subjects). Data from these populations were integrated into the ISS database.

 

Extent of exposure

 

“All” subjects (ISS data base)

The numbers of subjects who received treatment (ranolazine, placebo, or both) for a specified amount of time are shown below.

 

The mean duration of exposure was 160 days for ranolazine compared to 25 days for placebo.

 

Phase 2/3 SR controlled angina studies

Extent of exposure for this subpopulation is shown below.

 

The average duration of exposure for all ranolazine was 66 days compared to 53 days for  placebo. The 750 mg dose had the longest average duration of exposure (82 days).

Long term, open label studies

CVT 3032 and CVT 3034

The mean duration of exposure to ranolazine SR in uncontrolled studies was 448 days. A total of 276 patients received the drug for at least one year and 101 received it for at least 2 years.

 

Adequacy of clinical experience

The development program was excessively large in number but had only a limited number of clinical trials that were helpful in assessing safety. That said, there are enough patients studied under controlled conditions to comfortably determine the major safety effects of ranolazine.

 

Case report forms were submitted and spot checked for subjects who died or were withdrawn for adverse event.

 

2.0 Demographics

The table below shows demographic and baseline characteristics for the ISS database population and the Phase 2/3 SR controlled studies.

 

 

 

 

“All” subjects (ISS data base)

The 2 treatment groups were well balanced.  The majority of subjects were male, less than 65 years of age, and  white. Less than 7% of subjects were 75 years of age or older.

 

Overall, around 14% of subjects had diabetes mellitus and about 14% had congestive heart failure. The majority had coronary artery disease, around 8% had prior unstable angina,

and around 40% had had a previous myocardial infarction. About 10% had ventricular arrhythmias, about 4% had valvular heart disease, less than 1% had had a prior cardiac arrest, about 40% had hypertension, about 3% had had a prior stroke, about 14% had had angioplasty and about 16% had had cardiac revascularization.

 

Commonly used concomitant medication includes ACE inhibitors, alpha/beta blockers, calcium channel blockers, and HMG CoA reductase inhibitors.

 

Phase 2/3 SR controlled angina studies

The majority of subjects was male, less than 75 years of age, and mostly white.

About 11% of subjects were at least 75 years of age. The 2 treatment groups were well balanced.

Around 23% of subjects had diabetes mellitus and about 25% had congestive heart failure. All had coronary artery disease, around 20% had prior unstable angina, and around 55% had had a previous myocardial infarction. About 10% had ventricular arrhythmias, about 5% had valvular heart disease, around 2% had had a prior cardiac arrest, 64% had hypertension, about 5% had had a prior stroke, less than 25% had had angioplasty and about 20% had had cardiac revascularization. The 2 treatment groups were well balanced.

 

Commonly used concomitant medication includes ACE inhibitors, alpha/beta blockers, calcium channel blockers, and HMG CoA reductase inhibitors. The 2 treatment groups were well balanced.

 

Exclusion criteria

The list below outlines the patients who were excluded from the 2 of the largest placebo controlled efficacy trials (CVT 3033 and 3031).

 

-presence of electrocardiographic or other factors that might interfere with ECG interpretation or may cause a false positive stress test

- New York Heart Association Class III-IV CHF;

-Clinically significant valvular heart disease or congenital cardiac defects;

-Unstable angina pectoris within the 2 months prior to study entry;

-Second or third degree atrio-ventricular block or uncontrolled clinically significant cardiac arrhythmias or a history of life-threatening ventricular arrhythmias unassociated with acute MI;

-Corrected QT interval (QTc) > 0.50 sec at Visit 1;

-Required medications known to prolong the QT interval;

-Required medications that inhibit or induce cytochrome P450 3A4,

-Unwillingness to refrain from grapefruit/grapefruit juice consumption for the duration of the study

-Requirement for digoxin;

-MI, CABG, PTCA, or other catheter-based rcvascularization procedures within 2 months before study entry;

-Active acute myocarditis or pericarditis;

-Hypertrophic cardiomyopathy;

-Uncontrolled hypertension;

-Systolic BP< 100 mmHg;

 

3.0 All adverse events

Methodology

According to the sponsor, adverse event data were collected by routine monitoring and reporting while the patient was on-study. An adverse event was defined as any unfavorable or unintended sign (including laboratory values), symptom, or disease that appeared or worsened during the clinical trial, whether or not deemed causally associated with the study drug. Investigators identified and graded adverse events by direct observation, questioning, and spontaneous reports from patients. Investigators identified the action taken regarding the adverse event, and they assigned a causality descriptor. An adverse event could result in the patient’s premature discontinuation from the study. Adverse events were reported as verbatim terms in the case report forms (CRFs); these terms were subsequently mapped (using a COSTART thesaurus) to a preferred term and body system.

 

The table below shows the reporting of all adverse events in the ISS database and the Phase 2/3 SR controlled angina studies.

 

“All” subjects (ISS data base)

The mean duration of exposure is more than 4 times for the ranolazine treatment group than the placebo group (25 days vs. 160 days, respectively). Therefore, it is not unusual that the group with the longer exposure (ranolazine, in this case) would have a higher reporting rate. The usefulness of conclusions drawn from these data is questionable.

 

Phase 2/3 SR controlled angina studies

The mean duration of exposure was similar for the 2 treatment groups (66 days and 53 days for ranolazine and placebo, respectively). There were more reports of any adverse event for the ranolazine group (36.7%) compared to placebo (22.2%), and the ranolazine group reported more serious events (6.8% vs. 3.5%), and more events resulting in treatment interruption (3.1% vs. 0.7%)/discontinuation (8.0% vs. 3.7%).

 

Individual adverse events

The table below shows the reporting of adverse events that were reported by at least 2% of the subjects in the ISS database on any treatment.

 

“All” subjects (ISS data base)

Since the duration of use was more than 6 times longer in the ranolazine group compared to the placebo group, the usefulness of examining the individual events in the ISS data base is questionable.

 

 Phase 2/3 SR controlled angina studies

The numbers and percents of patients in the Phase 2/3 controlled angina studies who reported an adverse event (limited to those events reported by more than 1% of the total ranolazine group and reported more in the ranolazine group than the placebo group) are shown in the table below, by treatment group. The placebo subtracted rate is also shown.

 

No. and (percent) of patients

event	Total ranolazine N=749	Total placebo N=455	Placebo subtracted %
Any event	275 (36.7)	101 (22.2)	14.5
Dizziness	61 (8.1)	6 (1.3)	6.8
Constipation	49 (6.5)	2 (0.4)	6.1
Nausea	43 (5.7)	3 (0.7)	5.0
Asthenia	31 (4.1)	10 (2.2)	1.9
Dyspepsia	16 (2.1)	4 (0.9)	1.2
Abdominal pain	13 (1.7)	3 (0.7)	1.0
Cough increased	8 (1.1)	1 (0.2)	0.9
Headache	22 (2.9)	9 (2.0)	0.9
Pain	11 (1.5)	3 (0.7)	0.8
Dyspnea	14 (1.9)	6 (1.3)	0.6
Infection	9 (1.2)	4 (0.9)	0.3
Rhinitis	1 (0.1)	0	0.1

Table 12 vol 1.0340 pg 62

 

The placebo subtracted rate for reporting any adverse event was 14.5%. Those events with placebo subtracted rates greater than 2% includes dizziness (6.8%), constipation (6.1%), and nausea (5.0%).

 

Dose response

Events possibly associated with dose are shown in the table below for the SR formulation, by dose.

 

No. and (percent) of patients reporting events

	Ranol 500 mg N=181	Ranol 750 mg N=279	Ranol 1000 mg N=459	Ranol 1500 mg N=187
Any event	28 (15.5)	87 (31.2)	134 (29.2)	63 (33.7)
Dizziness	2 (1.1)	10 (3.6)	29 (6.3)	22 (11.8)
Asthenia	0	5 (1.8)	16 (3.5)	11 (5.9)
Nausea	1 (0.6)	9 (3.2)	17 (3.7)	16 (8.6)
Syncope	0	0	5 (1.1)	3 (1.6)
Sweating	0	3 (1.1)	5 (1.1)	5 (2.7)
Vomiting	0	2 (0.7)	5 (1.1)	4 (2.1)

Table G2.2 vol 1.0364

 

Unfortunately, the sample sizes for the SR formulation are small. However, the most convincing dose related adverse events are shown above. For example, syncope was only reported with doses 1000 mg and above.

 

4.0 Serious safety

Methodology

According to the sponsor, a serious adverse event (SAE) was characterized by one of the following criteria: resulted in death, was life threatening, required hospitalization or prolongation of an existing hospitalization, resulted in persistent or significant disability or incapacity, caused congenital anomaly or birth defect, and/or was considered medically significant by the investigator. Prior to 1997, the criteria for an SAE also included cancer and drug overdose. For some older (Syntex) studies, SAEs were not identified on the case report forms. Consequently, for those studies, SAEs were identified by medical monitors after review of the safety database. Summary tables regarding SAEs included: the type of SAE, concomitant medications received, action taken, outcome, and treatment assignment. All treatment-emergent SAEs were included in the summary tables, regardless of their severity or relationship to the study drug. SAEs occurring in the run-in period were excluded.

 

Deaths

There were 37 reported deaths (33 ranolazine and 4 placebo) in all 81 ranolazine studies[12].  For the total ranolazine group (including IR, SR, and IV formulations), the mortality rate was 1.2% (33/2682). The controlled trials randomized 749 subjects to ranolazine with a mortality rate of 0.7% (5/749) and 455 subjects to placebo also with a mortality rate of 0.7% (3/455).

 

Subject ID	Dose SR bid and IR tid/duration (days)	Cause of death
Controlled studies
3031/133-1017	Dosing schedule: Ranol SR 1000 mg for 9 days, placebo for 7 days, ranol 500 mg for 3 days	V fib/collapsed at home
3033/177-9027	Ranol SR 750 mg/33	Acute MI
3033/704-7600	Ranol SR 750 mg/18	Sudden death
3033/706-9575	Ranol SR 1000 mg/13	Sudden death
3033/710-7631	Placebo/18	Sudden death (elevated ethanol level)
3033/717-8668	Placebo/6	Sudden death while driving
3033/751-9386	Placebo/95	Dissection of coronary arteries followed by cardiac arrest during elective PTCA
054/6858-414	Ranol IR 120 mg/41	MI
Uncontrolled studies
054/6858/415+	Dosing schedule: ranol IR 240 mg for 28 days, placebo for 28 days, ranol IR 120 mg for 28 days.	Sudden death (died 2 days after last dose)
1513/3073-4609	Ranol IR 30 mg/died 45 days after last dose	Pulmonary embolism
3032/133-1018	Ranol SR 1000 mg/26	Malignant melanoma
3032/133-1019	Ranol SR 1000 mg/24	Lung carcinoma
3032/149-1193	Dosing schedule: Ranol SR 750 mg for 70 days, 1000 mg for 261 days	Sudden death. Developed a fib/flutter earlier, QTc was increased to 525 msec on day 173. QTc was 386 msec about 6 weeks prior to death
3032/153-1249	Ranol SR 750 mg/86	Esophageal carcinoma
3032/162-1281	Ranol SR 1000 mg/19	Sudden death. QTc interval up to 533 msec. Had complaints of dyspnea and chest pain immediately prior to death
3032/180-1462	Ranol SR 750 mg/366	CVA
3032/182-1458	Ranol SR 750 mg/210	Sudden death
3032/501-1441	Ranol SR 750 mg/167	CHF
3032/512-1366	Ranol SR 750 mg/256	AMI
3032/515-1392	Ranol SR 750 mg/342	Cardiovascular insufficiency S/P revascularization
3034-181-8445	Ranol SR 1000 mg/84	AMI with arrhythmia
3034/182-9269	Ranol SR 500 mg/428	Sudden death
3034/185-8374	Ranol SR 1000 mg/176	Sudden death
3034/190-8007	Ranol SR 1000 mg/298	Sudden death proceeded by complaints of angina and dyspnea
3034/195-8051	Ranol SR 750 mg/315	Sudden death (heart arrest)
3034/204-9024	Ranol SR 1000 mg/224	Sudden death (VF)
3034/236-8480	Ranol SR 750 mg/8	MI
3034/510-8353	Ranol SR 1000 mg/53	Pulmonary embolism proceeded by nausea and vomiting
3034/562-9186	Ranol SR 750 mg/332	unknown
081/6810/181	Ranol IR 400 mg/325	MI
1515/3838/2210	Ranol IR 120 mg/1319	MI
1515/3435/3702	Ranol IR 60 mg/187	Smoke inhalation
2074/3953/7001	Ranol IR 400 mg/526	Sudden death
2074/3971/15008	Ranol IR 400 mg/168	Died while undergoing CABG with balloon pump support
2074/1807/28002	Ranol IR 400 mg/364	Ruptured aortic aneurysm
1789/3645/2302^	Placebo IV/13	Complication of PTCA

+not included in clinical data base

^not included in ISS

 

Many causes of death were reported as sudden or other cardiovascular related; such deaths are not unexpected in this patient population.

 

Survival curve: all patients from Phase 2/3 controlled angina studies

 

There was no difference in the survival curves  between the placebo and ranolazine groups. Also, there was no difference between the groups when patients who received <120 mg ranolazine were removed from the analysis.

 

The survival data were analyzed using a Cox proportional hazards regression model[13] with a single effect for treatment. The hazard ratio and the 95% two sided confidence interval are shown in the table below.

 

The estimated hazard ratio of 0.633 corresponds to a 36.7% reduction in the risk of death in the ranolazine treatment group. However, this estimate is highly variable because of the small number of events in the dataset. Therefore, the confidence interval for the hazard ratio is very wide. The analysis rules out that ranolazine is more than 3.27 times worse than placebo, or more than 8.2 times better than placebo, with regard to patient survival.

 

Long term, open label trials

CVT 3032 and CVT 3034

Of the 550 patients enrolled in these studies[14], 440 are still ongoing and 110 (20%) were discontinued. Of the patients who discontinued, 58 did so because of an adverse event. In addition, there were 262 subjects who received the IR formulation during one of 5 uncontrolled IR studies.

 

There were 25 deaths (table F-1.3 vol. 1.0378).  Causes of death are shown below.

Cause of death	Number of subjects
Cancer	2
AMI	5
Sudden death	7
CVA	1
Congestive heart failure	2
V fib/tach	1
Cardiac arrest	2
Pul embolism	1
Unknown	2
House fire	1
AAA rupture	1
total	25

Table F-1.3 vol 1.0378

 

Causes of death included 7 sudden deaths, 1 ventricular fibrillation, and 5 acute myocardial infarctions.

 

Withdrawals for adverse events

 

The table below shows the numbers and percents of subjects who withdrew from a study because of an adverse event.

The percents of dropouts for adverse events in all but the smallest categories were at least twice as high in the ranolazine group compared to placebo. One explanation is that the mean duration of exposure is much higher for the treated compared to the placebo groups (in the ISS database the ranolazine and placebo groups mean duration of exposure were 160 days and 25 days, respectively). However, in the Phase 2/3 controlled angina trials where the mean duration were similar (ranolazine 66 days and placebo 53 days), the dropout rate was more than 2 times higher in the ranolazine group (8.4%) compared to placebo (4.0%).

 

ISS database

The table below shows the adverse events leading to discontinuation in at least 0.5% of subjects in the ranolazine group and reported more often in the ranolazine group than the placebo group.

 

No. and (percent) of patients

Event	Placebo N=1529	Ranolazine N=2682	Placebo subtracted
Any event	31 (2.0)	226 (8.4)	6.4
Dizziness	1 (0.1)	30 (1.1)	1.0
Nausea	1 (0.1)	26 (1.0)	0.9
Angina	11 (0.7)	36 (1.3)	0.6
Asthenia	0	13 (0.5)	0.5
Headache	1 (0.1)	17 (0.6)	0.5
Myocardial infarct	0	14 (0.5)	0.5
Constipation	0	14 (0.5)	0.5

Table G-6.1

 

The adverse events leading most often to discontinuation in the ranolazine group compared to placebo included dizziness, nausea, and angina.

 

Phase 2/3 controlled trials

The table below shows the adverse events leading to discontinuation in at least 0.5% of subjects in the ranolazine group and reported more often in the ranolazine group than the placebo group.

 

No. and (percent) of patients

Event	Placebo N=455	Ranolazine N=749	Placebo subtracted
Any event	18 (4.0)	62 (8.3)	4.3
Dizziness	1 (0.2)	13 (1.7)	1.5
Nausea	0	10 (1.3)	1.3
Headache	0	6 (0.8)	0.8
Constipation	0	6 (0.8)	0.8
Asthenia	0	5 (0.7)	0.7
Myocardial infarct	0	4 (0.5)	0.5
Syncope	0	4 (0.5)	0.5
Vomiting	0	4 (0.5)	0.5

Table G-6.3

 

A total of 62 (8.3%) subjects discontinued ranolazine compared to 18 (4.0%) placebo subjects. The adverse events leading most often to discontinuation in the ranolazine group compared to placebo included dizziness, nausea, headache, and constipation.

 

By dose

Specific adverse events that led to discontinuation and are suggestive of a dose response are shown in the table below.

 

No. and (percent) of patients

	Ranol 500 mg N=181	Ranol 750 mg N=279	Ranol 1000 mg N=459	Ranol 1500 mg N=187
Any event	2 (1.1)	22 (7.9)	28 (6.1)	10 (5.3)
Dizziness	0	2 (0.7)	8 (1.7)	3 (1.6)
Nausea	0	1 (0.4)	5 (1.1)	4 (2.1)
Headache	0	1 (0.4)	2 (0.4)	3 (1.6)
Constipation	0	2 (0.7)	2 (0.4)	2 (1.1)
Vomiting	0	1 (0.4)	1 (0.2)	2 (1.1)
Syncope	0	0	3 (0.7)	1 (0.5)
Asthenia	0	1 (0.4)	3 (0.7)	1 (0.5)

Table J-1.3 vol 1.0366

 

Although the sample sizes are relatively small, the highest ranolazine group (1500 mg) had the highest reporting rates for these events.

 

Serious adverse events

The numbers and percents of patients reporting serious events in the 81 ranolazine studies are shown below.

In all cases except the 16 early studies, the percents of ranolazine patients reporting serious events were at least twice as high compared to placebo patients.

 

Phase 2/3 SR controlled angina studies

Only angina was identified as a serious adverse event that was reported by at least 1% of subjects in any treatment group. Out of 749 ranolazine subjects in the controlled angina studies, 1.7% reported angina compared to 1.8% in the placebo group.

 

Syncope

The number and percent of subjects reporting adverse events of syncope or suggestive of syncope or pre syncope are shown below from the ISS data base.

The percent of total ranolazine subjects reporting one or more of these events was 19.2%, this is more than 4 times the percent of placebo patients reporting the same events (4.4%). The percents of subjects reporting syncope and/or near syncope were 0.1% (1) for placebo and 0.8% (21) for ranolazine. Reports of dizziness and light headed were also more common with ranolazine. 

 

Doses associated with syncope (orthostatic changes)

There were 10 reports of syncope (called syncopal episode, vasovagal episode, loss of consciousness, collapse) in the Phase 2/3 controlled trials. The doses being used at the time the event was reported were 1500 mg SR (3), 5 at 1000 mg SR (5), 120 tid IR (1), and placebo (1). In study RANSO201, doses of 1500 and 2000 mg bid produced significant mean orthostatic changes (-9.8 mmHg and -8.4 mmHg, respectively, compared to placebo) at 4-6 hours after dosing. Some of the subjects could not have their erect blood pressures recorded because of these changes. The figure below shows the mean erect systolic blood pressures profiles by dose.

 

Erect systolic blood pressure decreased from baseline at around 4 hours after dose intake, the time of peak drug concentration. The drop was greater for the 2000 mg dose. Three subjects could not undergo erect blood pressure recordings because of symptoms of lightheadedness.

 

5.0 ECG

Methodology

All ECGs from CVT-conducted multiple dose studies were read by a central Core ECG Laboratory (St. Louis University Core ECG Laboratory). The parameters summarized are the ECG intervals as measured in msec: PR, QRS, QT, and the corrected QT (QTc).

 

Each QT value from CVT-conducted multiple dose studies is the maximum QT value among the 12 leads in that ECG, corrected as indicated for heart rate.

 

ECG intervals

 

ISS database

Mean changes from baseline at endpoint for ECG intervals are shown below. The ECG recordings were not necessarily obtained at peak effect.

 

Compared to placebo, there were larger changes from baseline for all of the ECG intervals listed above. The changes from baseline for the QT and QTc intervals for the ranolazine SR group were 3.6 msec and 2.0 msec, respectively, and 2.7 msec and 1.5 msec, respectively, for total ranolazine. Comparative changes for the placebo groups were negative.  These measurements are independent of time of last dose so results are underestimates of the true effect.

 

Shift tables

The following table shows the number and percent of subjects who received either placebo or ranolazine SR only (in the ISS database) and had a normal QT/QTc interval at baseline that became abnormal at endpoint.

 

 No. and (percent) of subjects

	Placebo SR N=569	Ranolazine SR N=1359
Normal QT at baseline abnormal QT at endpoint	49 (8.6)	210 (15.5)+
Normal QTc at baseline abnormal QTc at endpoint	28 (4.9)	70 (5.2)^

+includes 2 subjects whose abnormal QT interval was judged not to be clinically significant

^includes 65 subjects whose abnormal QTc interval was judged not to be clinically significant

Table N 8.1 vol1.0363 pg 379-380

 

Phase 2/3 SR controlled angina studies

The table below shows the mean changes from baseline for ECG intervals by dose.

There were dose related increases in all of the ECG intervals and these changes were greater than those for placebo. Mean increases from baseline for QT/QTc were 1.3/2.1 msec for the 500 mg dose and 10.5/8.7 msec for the 1500 mg dose. 

 

Maximum mean QT changes were 11.4,  9.9, 25.7, 20.5, and 18.4 msec for placebo, ranolazine 500 mg, 750 mg, 1000 mg, and 1500 mg, respectively (table N-1.3.1 vol 1.0376).

 

The measurements discussed above are independent of time of last dose (and peak effect is greater than trough effect) so they are underestimates of the true effect.

 

Peak effect

Mean change from baseline in QTc interval (msec) at peak

	Placebo 455	Ranol 500 N=177	Ranol 750 N=269	Ranol 1000 N=428	Ranol 1500 N=170
Mean change from baseline	-2.0	3.3	3.5	5.0	11.0
Max mean change from baseline	1.1	3.3	8.9	8.1	11.0

Table N-1.3.2.1vol 1.0376

 

Effects of ranolazine on QTc interval are greater when measured at peak drug concentration than effects measured at trough (or at random).

 

The shift table below shows the number and percent of patients, by dose, who had a normal QTc interval at baseline and an abnormal one at endpoint. ECG measurements were made at peak drug concentration.

 

No. and (percent) of patients

	Placebo N=455	Ranol 500 N=181	Ranol 750 N=279	Ranol 1000 N=459	Ranol 1500 N=187
Normal at baseline and abnormal+ at endpoint	24 (5.3)	18 (9.9)	10 (3.6)	28 (6.1)	40 (21.4)

+includes those that the sponsor identified as not clinically significant.

Table N-8.3.2.1 vol 1.0377

 

A total of 40 subjects (21.4%) of subjects who received the 1500 mg dose of ranolazine developed an abnormal QTc interval that was “normal” at baseline.

 

The table below shows the number and percent of patients, by dose, who had selected QTc interval changes from baseline at endpoint at peak drug concentration.

 

No. and (percent) of patients

Change from baseline	Placebo N=433	Ranol 500 N=177	Ranol 750 N=271	Ranol 1000 N=433	Ranol 1500 N=170
0-30 msec	167 (38.6)	67 (37.9)	160 (59.0)	242 (55.9)	71 (41.8)
31-60 msec	21 (4.8)	20 (11.3)	6 (2.2)	29 (6.7)	28 (16.5)
>61 msec	4 (0.9)	6 (3.4)	1 (0.4)	1 (0.2)	10 (5.9)

Table N-15.3.1 vol 1.0377

 

Major placebo controlled trials and their follow up studies

A total of 980 patients were randomized into Studies CVT 3031 or CVT 3033 and had baseline and at least one post-randomization ECG. Of these patients, 550 also elected to continue ranolazine treatment in the respective open-label studies, Studies CVT 3032 or CVT 3034 as of the NDA cut-off date (October 15, 2001). Of this selected group, those with either an increase from baseline > 60 msec in the QTc interval or a QTc value > 500 msec using Bazett’s correction are summarized by treatment in the table below.

 

 

Overall, the effect of ranolazine on the QTc interval was worse compared to placebo. For patients receiving the highest dose, 2.3% had both an increase in QTc of at least 60 msec over baseline and had QTc >500 msec compared to 0.7% of placebo patients. Again, this is probably an underestimate because the effect of ranolazine is worse at peak concentration.

 

The sponsor states that the Fridericia correction is “better [because it] eliminates the relationship between QT and heart rate as well as reduces variability.”  The table below shows the number and percent of QT outliers using Fridericia’s correction.

 

 

Regardless of the correction used, ranolazine has been shown to prolong the QT interval.

Plasma concentration and QT effect

Study CVT 3111 was designed to characterize the relationship between the plasma ranolazine  concentration and the effect of ranolazine (and its major metabolites) on the QTc interval by achieving plasma ranolazine concentrations higher than those typically generated.

 

This was a randomized, placebo-controlled, single IV infusion, dose escalation, 4 period study involving 30 female and male healthy subjects. Period I consisted of a 2 hour IV infusion of ranolazine aiming for a target peak plasma ranolazine concentration of 2,000 ng/mL. Period 2 consisted of a 72 hour IV infusion of ranolazine/placebo aiming for a target peak plasma ranolazine concentration of 4,000 ng/mL. Period 3 consisted of a 72 hour IV infusion of ranolazine/placebo aiming for a target peak plasma ranolazine concentration of 10,000 ng/mL. Period 4 consisted of a 72 hour IV infusion of ranolazine/placebo aiming for a target peak plasma ranolazine concentration of 15,000 ng/mL. In each period the infusion of ranolazine/placebo was preceded by a 24 hour placebo infusion.

 

The figures below show the mean difference from baseline in QTc versus time for each target plasma concentration.

 

 

 

Only 1 subject completed the ranolazine infusion in period 4.

 

A linear relationship between ranolazine plasma concentration and change in QTc interval from baseline was estimated to have an average slope of 2.29 msec per 1000 ng/mL.  Three subjects were discontinued prematurely because of the attainment of protocol specified stopping rule of a greater than 30% increase in QTc from baseline or a value >500 msec.

 

There was no delay between QTc interval change and the achievement of steady-state ranolazine concentration. Therefore, it is likely that QTc changes are related to ranolazine and not its metabolites.

 

Higher oral dose

Subjects in Study RAN0201 received the highest single dose of oral ranolazine administered (ranolazine SR 2,000 mg b.i.d.). These doses resulted in significant orthostatic changes in systolic blood pressure compared to placebo 4 to 6 hours post-dosing (about 10 mmHg drop in standing systolic blood pressure).  The mean plasma ranolazine concentration at the same time points ranged from 7,223 to 6,328 ng/mL in those receiving 2,000 mg bid. Three of eight volunteers administered ranolazine SR developed severe symptoms of lightheadedness upon standing.

 

PR Interval: the mean value was higher with ranolazine with a statistically significant increase over placebo of 8.8 msec at 2 hr post-dose on Day 5 with the 2000 mg dose (p = 0.024). This difference disappeared by 24 h after the final dose.

 

QT Interval: by Day 5 QT interval tended to be prolonged with both doses of ranolazine

compared to placebo at all time points. The statistically significant increases seen with the 2000 mg dose were at 2 h post-dose (+ 14.2 msec), 4 h post-dose (+ 19.2 msec), and 6 h post-dose (+ 18.2 msec) on Day 5.

 

The mean QTc/QT intervals and mean differences from baseline profiles at steady state are shown below.

 

 

Changes in QT were large (up to 19.2 msec) and statistically different from placebo at hours 2, 4, and 6 (for the 2000 mg dose), the time of peak drug concentrations.

 

Substantial differences between active and placebo treatment n the mean QTc intervals were evident on Day 5, with statistical significant differences from placebo at every time point with both doses of ranolazine, except at 24 h after the final dose of 2000 mg. The maximum mean differences seen were at 4 h post-dose, when the mean value for the 1500 mg treatment was 20.6 msec greater than for placebo and the mean value for the 2000 mg treatment was 28.2 msec greater than for placebo.

 

T wave Morphology:  repolarization abnormalities, represented by blunting or notching of the T wave, were detected in 1/8 subjects during placebo treatment, 5/8 subjects during treatment with 1500 mg ranolazine and 7/8 subjects during treatment with 2000 mg ranolazine SR. The only subject without any observed changes in T wave morphology also had the lowest plasma ranolazine concentrations.

 

T wave morphology

Along with QT interval changes, there were changes in T-wave morphology. These changes include both a reduction in amplitude as well as change in its configuration. The T-waves were classified by the Core EGG Laboratory according to the following system: amplitude (positive, negative, flat) and configuration (biphasic +/-, biphasic -/+, and notched).

 

The figure below shows the logistic regression function fitted to all data from active treatment to determine the likelihood of a notched T-wave at various ranolazine concentrations.

A decrease in amplitude was seen up to a concentration of approximately 3,000 ng/mL.  Of the 109 ECGs with reported T wave notching, the average ranolazine concentration was 3,020 ng/mL (SD 2,310). The remaining ECGs (about 5700) with no T wave notching reported, the average ranolazine concentration was 1,360 ng/mL (SD 1583).

 

There were 3 episodes of syncope, one was considered to be serious and 2 that occurred during standing for blood pressure measurements. These episodes tended to be preceded by nausea, dizziness, sinus bradycardia (despite hypotension), blurring of vision, diplopia.

 

There were 3 withdrawals for QTc greater than 500 msec.

 

In total, 10 out of the 15 subjects who reached plasma ranolazine concentrations higher than 8,000 ng/mL in any of the dosing periods had their dosing interrupted because of CNS-related AEs, including visual disturbances and altered sensation. One additional subject also had dosing interrupted because of adverse event at a plasma ranolazine concentration of 5,510 ng/mL. None of the subjects’ ECGs showed signs of cardiac arrhythmia.

 

 

In study CVT 3031, there were changes in the morphology of the T-wave during ranolazine use. These changes include both a reduction in amplitude (positive, negative, flat) as well as changes in its configuration (biphasic +/-, biphasic -/+, and notched).

 

The frequency of notched T waves is shown below by treatment group at peak and trough concentrations.

 

% of subjects with notched T waves

	Placebo	Ranol 500	Ranol 1000	Ranol 1500
peak	2%	1%	3%	6%
trough	<1%	<1%	5%	5%

 

There were more notched T waves were reported in the Ranolazine 1000 mg and 1500 mg doses than in the placebo and ranolazine 500 mg dose groups.

 

Genetic studies have shown that long-QT syndrome (LQTS) is a primary electrical disease caused by mutations in specific ion channels.[15] LQTS patients exhibit QT prolongation on the ECG and are at risk of arrhythmogenic syncope and sudden death. In addition to duration, T-wave morphology is often abnormal, and notched T waves have been included in diagnostic criteria.[16] This pattern has been associated with a poor prognosis.[17]

 

6.0 Vital signs

The tables below show the mean change from baseline for vital signs at trough and peak drug concentration. The data are grouped into ISS database, Phase 2/3 SR controlled trials, and open labeled studies.

 

 

 

Phase 2/3 SR controlled angina studies

 

The tables below shows mean changes in pre-exercise standing vital signs at peak and trough drug concentrations, by dose.

 

 

 

 

7.0 Laboratory values

Methodology

 

Laboratory assessments collected at baseline and the end visit of each study phase were used. Laboratory parameters were in three panels:

 

•           Hematology: white cell count, red cell count, hemoglobin, hematocrit, platelets, granulocytes, lymphocytes, and monocytes;

 

•           Chemistry: BUN, creatinine, sodium chloride, total carbon dioxide, potassium, glucose, SGOT, SGPT, alkaline phosphatase, total bilirubin, creatinine kinase (OK), CK-MB isoenzymes;

 

•           Urinalysis: color, pH, protein, glucose, ketones, bilirubin, urobilinogen, and nitrates.

 

•           Adrenocorticotropic hormone (ACTH) stimulation testing was performed in some studies.

 

Laboratory parameters were evaluated for abnormalities in the 64 studies in the ISS data base as

well as in the Phase 2/3 SR controlled angina studies plus their long term follow up studies. Only data from the controlled angina are presented in this review, although the results in the ISS data base were examined.

 

Hematology

 

The table below shows the mean changes from baseline for selected hematology parameters for the Phase 2/3 SR formulation, by dose (numbers of patients are approximates). The parameters were selected for the table if they appeared to be consistently different from placebo.

 

Mean change from baseline

	Placebo N=260	Ranol 500 mg N=41	Ranol 750 mg N=237	Ranol 1000 mg n=264	Ranol 1500 mg n=52
Eosinophils %	0	1.1	0	0.2	1.1
Hematocrit %	-0.1	-1.5	-0.9	-1.2	-1.0
Hemoglobin g/dL	-0.1	-0.5	-0.5	-0.6	-0.4
Lymphocytes %	0.3	-1.7	-1.5	-1.4	-1.1
RBC 106/ul	-0.1	-0.2	-0.2	-0.3	-0.1

Appendix V B table L-1.3 vol 1.0367

 

Compared to placebo, there are small mean increases from baseline in eosinophils and small mean decreases in hematocrit/hemoglobin and lymphocytes. These changes were not dose related. The sponsor states that there was no evidence of occult blood loss. (vol 340 page 11).

 

Shift changes for these parameters are shown below.

 

These changes are small and seemingly unrelated to dose.

 

The summary of mean changes for selected hematology values for the ISS data base, the Phase 2/3 SR controlled studies (all doses combined), and the open label studies.

In the ranolazine group[18], there was 1 withdrawal for anemia in the ranolazine group (IR formulation), 2 withdrawals for leukopenia (1 each IR and SR formulations). There were no  placebo patients dropping out for these reasons.

 

Regarding serious adverse events, there were 3 reports of anemia (1 IR formulation and 2 SR formulation. There was 1 serious report  of leukopenia (SR formulation). There were no placebo patients with these reports.

 

Blood chemistries

 

The table below shows the mean changes from baseline for selected blood chemistries parameters from the Phase 2/3 SR formulation, by dose (numbers of patients are approximates).  The parameters were selected for the table if they appeared to be consistently different from placebo.

 

Mean change from baseline

	Placebo N=300	Ranol 500 mg N=41	Ranol 750 mg N=262	Ranol 1000 mg n=301	Ranol 1500 mg n=55
BUN mg/dl	-0.1	0.5	1.0	1.2	2.5
Creatinine mg/dl	0	0.1	0.1	0.1	0.2
Chloride mEq/l	0	0	-0.7	-0.6	-1.7
Glucose mg/dl	0.1	4.3	2.6	1.1	8.3
Sodium	-0.1	-0.7	-0.6	-0.8	-1.8

Appendix V B table L-1.3

 

Compared to placebo there were small increases in mean changes from baseline for both BUN and creatinine.

 

In the ranolazine groups from the ISS database, there was 1 withdrawal for increased BUN (IR formulation).[19]

 

8.0 Special populations

Age, race, gender

There were no clinical trials specifically designed to determine if there are age, gender, and/or race differences in safety.

 

The frequencies of reported adverse events, serious adverse events, and adverse events leading to discontinuation are shown below for the overall ISS database and Phase 2/3 SR controlled angina studies population.

 

 

 

The placebo subtracted incidence rates are shown below.

 

Percent of patients

	age			gender		race
	<65	65-<74	>75	Female	Male	White	Non white
Total aes	25.4	29.7	34.0	30.4	26.5	28.0	34.2
Serious aes	4.8	11.3	18.2	10.3	6.8	8.1	7.3
Discontinued	3.9	8.8	20.6	12.1	5.0	6.4	12.2

 

Although there are differences in incidence rates, it’s difficult to know if they are just the result of sample size discrepancies.

 

Placebo subtracted incidence rates-Phase 2/3 SR controlled angina data base are shown below.

 

Percent of patients-placebo subtracted

	age			gender		race
	<65	65-<74	>75	Female	Male	White	Non white
Total aes	8.5	16.6	31.5	14.5	14.6	13.3	40.9
Serious aes	2.6	3.4	5.5	2.7	3.4	2.9	11.8
Discontinued	1.5	5.4	12.0	7.2	3.5	3.9	14.7

 

As with the ISS database, it’s difficult to know if the differences in incidence rates are just the result of sample size discrepancies.

 

Hepatic impairment

Subjects with moderate hepatic impairment had an AUC and Cmax that were 76% and 51% higher, respectively, compared to healthy volunteers[20], when receiving ranolazine 875 mg followed by 500 mg bid. Subjects with mild impairment were similar to their healthy counterparts. The sponsor advises that this drug should not be used in patients with Child-Pugh category B or worse hepatic impairment.  The table below shows the increase in QT with increasing hepatic impairment.

 

The subjects had larger than expected prolongation of QTc. This could be the result of small sample sizes.

 

Renal impairment

With creatinine clearance decreasing from 100 mL/min to 30 mL/min, the average increase in ranolazine AUC and Cmax was approximately 80%[21].

 

Congestive heart failure

Study CVT 3031 enrolled 85 subjects with stable NYHA class III or IV heart failure and an ejection fraction <35%. Study design was double blind, placebo controlled, and randomized with patients receiving placebo or ranolazine SR 750 mg bid with or without digoxin.  There were no deaths. Seven subjects reported a total of 8 serious events:  cerebral ischemia, neuropathy, heart failure (2), myocardial ischemia, syncope, atrial flutter, and ventricular tachycardia. Half of these events were reported by subjects not receiving ranolazine. There were 2 discontinuations for adverse events: myocardial ischemia and heart transplant surgery.

 

Hypotension and/or postural hypotension were reported by 10 patients (all randomized to ranolazine). There was no evidence in this small study that ranolazine 750 mg bid worsens heart failure in patients with advanced CHF.

 

Regarding the entire ISS database, only 0.2% (5/2682) reported congestive heart failure or heart failure as an event that resulted in study drug discontinuation. There were 0/1529 placebo patients dropping out for this reason (table G-6.1).

 

Other concomitant diseases[22]

Adverse events, laboratory abnormalities, ECG changes, and vital signs were inspected in patients with concomitant reactive airway disease (N=153), or diabetes (383), or low BP, low HR and/or prolonged AV conduction (N=381). No studies were conducted to determine if there were effects of these concomitant diseases in patients taking ranolazine. There is no indication that patients with one or more of these concomitant diseases taking ranolazine are at increased risk compared to patients without additional disease.

 

 

9.0 Drug-drug interactions

CYP3A4 is a major determinant for ranolazine clearance. There was an average increase of plasma concentration of 3- to 4-fold in the presence of the potent CYP3A4 inhibitor ketoconazole (200 mg bid)[23]. Concomitant use with diltiazem resulted in increases in ranolazine plasma concentrations of 1.5- to 2.4-fold over the diltiazem total daily dose range (180-360 mg)[24]. Ranolazine 1,000 mg bid at steady-state caused a less than two-fold increase simvastatin exposure dosed at 80 mg qd[25]. Concomitant use of ranolazine and drugs that inhibit as well as those that are metabolized by CYP3A4 should be contraindicated.

 

In study CVT 3021, ranolazine SR 1000 mg bid was taken by healthy volunteers in conjunction with either placebo, or diltiazem 180, 240, or 360 mg qd for 8 days. There were statistically significant increases in ranolazine Cmax and AUC_0-12.

 

Verapamil (120 mg t.i.d.) increased ranolazine average plasma concentrations 2.25-fold at steady-state[26]. The primary cause of this effect could be the inhibition of P-glycoprotein (P-gp) in the gut, increasing the bioavailability of ranolazine. Concomitant use of ranolazine and drugs that inhibit P glycoprotein in the gut should be contraindicated.

 

Digoxin concentrations increased by 1.2-1.6 fold when used with ranolazine.

 

10.0 Abrupt withdrawal

In study CVT 3033, a 2-day rebound assessment for possible increase in anginal events, as measured by exercise treadmill test duration, was included in the study design. Ranolazine patients were discontinued from doses of 750 mg twice a day or 1000 mg twice a day compared to patients who were maintained on placebo during a 12 week treatment period. Trough exercise testing was obtained in all patients.

 

No patients were withdrawn from the study during the 2-day assessment phase and there were no deaths. There were 2 serious adverse events (myocardial infarction and myocardial ischemia) in patients randomized to ranolazine SR 750 mg with diltiazem as the concomitant medication.

 

11.0 Safety update

The 4-month safety update summarizes data collected from CVT’s two ongoing open-label studies (CVT 3032 and CVT 3034) during the period of 15 October 2001 to 31 October 2002. It includes data from 194 new patients enrolled in CVT 3034, and additional data from 440 patients whose participation was ongoing as of the 15 October 2001 NDA data cut-off date. This safety update does not include any new information from controlled clinical studies.

 

As agreed with the Agency, the data presented in this submission includes updated information on the following:

•           deaths;

•           serious adverse events (SAEs);

•           withdrawals due to adverse events (AEs); and

•           electrocardiogram (ECG) data.

 

With this update, the total exposure has increased from 1171 to 1714 subject/patient years. The ongoing studies used the ranolazine SR formulation in doses ranging from 500 mg bid to 1000 mg bid. Currently, 219 patients have been exposed to ranolazine for 6-12 months, 402 for 12-24 months, and 293 for more than 24 months. Mean duration of exposure for the ISS data base is now 321 days; for the phase 2/3 controlled angina studies, it is now 612 days.

 

Patient disposition

 

Serious adverse events

There were 74 additional patients reporting a serious adverse event with submission of the safety update. The reports were mostly angina and myocardial infarction.

 

 

Patient 3034/180/ 180 8213
 was discontinued for elevated BUN (64 mg/dl), serum creatinine (2.0 mg/dl) , and serum uric acid (9.5 mg/dl). Drug was discontinued and the abnormalities started to resolve.

 

Deaths

There were reports of 21 additional deaths for this reporting period.

Most of the causes of the newly reported deaths were cardiovascular in nature.

Deaths

Subject ID	Dose SR bid /duration (days)	Cause of death
3032/129/129 1082	1000 mg/1107	Liver carcinoma and sepsis
3032/505/505 1510+	1000 mg/1281	Non-Hodgkin’s lymphoma
3032/506/506 1449	1000 mg/1159	Myocardial infarction
3034/141/141 8124	1000 mg/374	Elevated LFTs, jaundice, sepsis with biliary obstruction and probable cholangio-carcinoma
3034/182/182 9079	750 mg/740	Sudden death
3034/183/183 8354	750 mg/489	MI
3034/190/190 8006	1000 mg/864	Sudden death
3034/204/204 7021*	1000 mg/744	GI carcinoma, MI, UTI
3034/224/224 7300	1000 mg/650	Sudden death
3034/493/493 9516	1000 mg/83^	Collapse, prostatic carcinoma
3034/502/502 8117	750 mg/634	Sudden death
3034/512/512 8243	500 mg/666	MI
3034/519/519 9211	1000 mg/344	Angina, MI
3034/525/525 8258	750 mg/324	Sudden death
3034/706/706 8645	750 mg/178	Pulmonary embolism, endometrial carcinoma
3034/707/707 9604	1000 mg/176	Osteosarcoma
3034/710/710 9608	750 mg/202	MI
3034/712/712 7613	500 mg/117	Sudden death, Myocardial ischemia
3034/718/718 7643	1000 mg/46	Acute coronary syndrome
3034/721/721 9636	750 mg/192	Sudden death, hemopericardium
3034/728/728 9686	1000 mg/49	MI

+died after cut off date of 31 Oct 2002

*previous submitted, material updated

^ drug discontinued; patient died 44 days later

 

Survival curve

The updated version is shown below.

 

Discontinuations for adverse events

An additional 30 patients were discontinued from treatment because of an adverse event, an increase of 0.8% compared to the original NDA.

 

 

The adverse events most commonly resulting in discontinuation are shown below.

 

This is similar to what was previously reported.

 

ECG

There is no new information about the effect of ranolazine on QT interval prolongation and T wave morphology changes. The summary of QTc outliers is shown below.

 

There is a dose related increase in the incidence rate of QTc outliers.

Protocol CVT 3031

 

A Double-Blind, Placebo-Controlled, 4-Period Cross Over, Multiple-Dose Study of Ranolazine SR as Monotherapy for Chronic Stable Angina Pectoris at Doses of 500 mg bid, 1000 mg bid, and 1500 mg bid.

 

The primary objective of this study was to determine the effect of ranolazine SR monotherapy compared to placebo in patients with chronic stable angina on exercise treadmill test duration at the time of trough ranolazine plasma levels (12 hours post dose) when given at the following doses 500 mg bid, 1000 mg bid and 1500 mg bid. There was no washout phase between doses.

 

Demographics

The majority of subjects were male, mean age was about 64 years, and most were white. Approximately half of the subjects were at least 65 years of age.

 

Study completion

 

There were 3 (1.7%) placebo subjects discontinued prematurely compared to 6 (3.3%) of the ranolazine 500 mg, 1 (0.5%) in the ranolazine 1000 mg, and 13 (7.0%) in the ranolazine 1500 mg groups.

Most of the discontinuations were the result of an adverse event.

 

 

All adverse events

 

Because of the cross over design of the study, it is very difficult to associate adverse event and dose. The following The table below shows the most common adverse events using the first period analysis (the analysis counting an AE in the period in which it occurred unless it worsened).

 

 

 

 

 

All of the reported events except angina pectoris were more frequent in the ranolazine 1500 mg group compare to lower dose groups and placebo.

 

Serious safety

 

Death

There was one death, the cause reported as ventricular fibrillation (sudden death), and the patient was on study drug for 17 days. He was taking ranolazine 500 mg at the time of death and in the previous weeks had received ranolazine 1000 mg and placebo in that order.

 

Serious adverse events

There were 12 patients who reported adverse events. The details of these events are shown below.

 

 

 

 

 

 

Serious events were reported more often in the ranolazine 1500 mg group (7/191, 4%), compared to placebo (1/191, <1%), ranolazine 500 mg (3/191, 2%), and ranolazine 1000 mg (1/191, <1%). The events reported in the 1500 mg group include allergic reaction (post flu shot), accidental injury (occupation related), postural hypotension, angina (2), syncope, and dizziness (and headache and vertigo).

 

 

 

Withdrawals for adverse events

There were 15 subjects who withdrew from the study because of an adverse event.

 

 

 

 

Of the 15 patients who withdrew, 10 (5%) were receiving ranolazine 1500 mg, 1 (<1%) was receiving 1000 mg, 2 (1%) were receiving 500 mg, and 2 (1%) were receiving placebo.  The 10 subjects on the highest ranolazine dose withdrew because of increased salivation (and nausea, paresthesia, thirst, urine abnormality, vomiting), hematuria (and nausea), abnormal vision (and constipation, hypesthesia, nausea, vomiting), abnormal gait (and confusion, constipation, headache, myasthenia, twitching), asthenia (and dizziness), syncope (and CHF), angina, dizziness (and headache, nausea), BUN increased (and dizziness, headache, vertigo).

 

QT/QTc intervals

The frequency of QTc interval changes are shown below.

There were 9 subjects with 15 ECGs that matched the criteria of > 60 msec from baseline to values to >500 msec.

 

 

Pt number/dose	Baseline QTc (msec)	Highest QTc (msec)	Change (msec)
1231001/500 mg	461	543	82
1411161/500 mg	482	583	101
1411161/1500 mg	482	561	79
1491198/1000 mg	460	528	68
1561230/placebo	465	527	62
1561230/1500 mg	438	535	97
5011437/1000 mg	426	503	77
5011441/1500 mg	438	558	120
5071428/1500 mg	433	522	89
5101417/500 mg	446	539	93
5251309/500 mg	506	571	65

 

 

T wave morphology

 

The frequency of notched T waves is shown below by treatment group at peak and trough concentrations.

 

% of subjects with notched T waves

	Placebo	Ranol 500	Ranol 1000	Ranol 1500
peak	2%	1%	3%	6%
trough	<1%	<1%	5%	5%

 

More notched T waves were reported in the Ranolazine 1000 mg and 1500 mg doses than in the placebo and ranolazine 500 mg dose groups.

 

Vital signs

 

The table below shows the mean blood pressure and heart rate at rest.

 

At doses less than 1500 mg, there were no statistically significant differences between ranolazine and placebo for changes in blood pressure. There were significant decreases for the 1500 mg dose at peak drug concentrations (reductions were in the order of 2-3 mmHg[27]).

 

Heart rate reductions were significant for the 1500 mg dose at both peak and trough drug concentrations. Mean reductions in heart rate were 3 bmp or less.

Protocol CVT 3033

 

A Double-Blind, Randomized, Stratified, Placebo-Controlled, Parallel Study of Ranolazine SR at Doses of 750 mg Twice a Day and 1000 mg Twice a Day in Combination with Other Anti-Anginal Medications in Patients with Chronic Stable Angina Pectoris.

 

The primary objective of this study was to determine the effect of ranolazine SR at doses of 750 mg twice a day and 1000 mg twice a day compared to placebo on symptom-limited treadmill exercise over 12 week treatment period. Study patients had chronic stable angina and were receiving a stable dose of a single concomitant anti-anginal medication (diltiazem 180 mg once a day in a formulation intended for once-a-day dosing, atenolol 50 mg once a day, or amlodipine 5 mg once a day).

 

A 2-day rebound assessment for possible increase in anginal events, as measured by exercise treadmill test duration, was included following discontinuation of ranolazine SR at doses of 750 mg twice a day or 1000 mg twice a day compared to patients who were maintained on placebo during a 12 week treatment period.

 

Demographics

Most patients were male, around 64 years of age, and almost completely white. The groups were similar in these characteristics.

 

About one fifth of the population had unstable angina at least 2 months prior to study enrollment, about 30% had congestive heart failure (NYHA class I or II), and more than half had a prior MI. The groups were well balanced with rare exception.

 

Study completion

 

No. and (percent) of patients

	Placebo N=269	Ran 750 mg N=279	Ran 1000 mg N=275
Completed trial	243 (90.3)	250 (89.6)	238 (86.5)
Early withdrawal	26 (9.7)	29 (10.4)	37 (13.5)
For AE	16 (5.9)	22 (7.9)	25 (9.1)
Death	2 (0.7)	2 (0.7)	1 (0.4)
Elective withdrawal	4 (1.5)	1 (0.4)	5 (1.8)
Other^	7 (2.9)	6 (2.2)	7 (2.5)

^ includes non compliance, lost to follow up

table 12B, Table 1.4.1 study report

 

More patients failed to complete the study in the high dose ranolazine group (13.5%) compared to low dose ranolazine (10.4%) and placebo (9.7%). The main reason for discontinuation for all treatment groups was an adverse event. The high dose group had more discontinuations for adverse events (8.7%) compared to low dose (7.2%) and placebo (4.8%). The percent of reported deaths were similar across treatment groups.

 

Withdrawals by concomitant anti-anginal medication.

           

Percent of patients who withdrew early:  placebo subtracted

	Ran 750 mg			Ran 1000 mg
	Dilt N=74	Aten N=119	Amlo N=86	Dilt N=69	Aten N=117	Amlo N=89
Early withdrawal	3.3	5	-7.6	4.5	9.4	-4.4
For AE	6.5	5.1	-5.1	8.8	6.9	-4.0
Death	0	-0.8	1.2	-1.4	-0.8	1.1
Elective withdrawal	0	-1.7	-1.2	0	0.9	-0.1

 

There were more withdrawals for adverse events by patients taking diltiazem in all treatment groups including placebo (table 1.4.3).

 

N.B. Diltiazem (180-360 mg) in previous studies (CVT 3012, RAN0121, and RAN068) was shown to increase ranolazine average steady-state plasma concentrations of 1.5-2.4 fold.

 

Individual adverse events

Adverse events reported by at least 4 subjects in at least 1 of the ranolazine groups and more than the placebo group are shown below.

 

No. and (percent) of patients

Adverse event	Placebo N=269	Ran 750 mg N=279	% Pl subtracted	Ran 1000 mg N=275	% Pl subtracted
Any event	71 (26.4)	87 (31.2)	4.8	90 (32.7)	6.6
Constipation	2 (0.7)	18 (6.5)	5.8	20 (7.3)	6.6
Dizziness	5 (1.9)	10 (3.6)	1.7	19 (6.9)	5.0
Nausea	2 (0.7)	9 (3.2)	2.5	14 (5.1)	4.4
Asthenia	6 (2.2)	5 (1.8)	-0.4	13 (4.7)	2.5
Syncope	0	0	0	5 (1.8)	1.8
Abdominal pain	2 (0.7)	2 (0.7)	0	7 (2.5)	1.8
Sweating	0	3 (1.1)	1.1	4 (1.5)	1.5
Vomiting	1 (0.4)	2 (0.7)	0.3	4 (1.5)	1.1
Diabetes mellitus	0	5 (1.8)	1.8	2 (0.7)	0.7
Headache	4 (1.5)	7 (2.5)	1.0	6 (2.2)	0.7
Myocardial infarct	0	4 (1.4)	1.4	1 (0.4)	0.4
Dyspepsia	4 (1.5)	7 (2.5)	1.0	5 (1.8)	0.3
Dyspnea	4 (1.5)	5 (1.8)	0.3	1 (0.4)	-1.1

Table 3.0.1

 

Constipation was the most frequently reported adverse event for patients randomized to ranolazine with the 1000 mg dose group reporting more (6.6%) than the 750 mg dose (5.8%). Dizziness was the next most frequently reported event with the 1000 mg dose group reporting more (5.0%) than the 750 mg dose group (1.7%). Syncope was reported by 5 subjects, all from the 1000 mg dose group and 4 of these 5 subjects were taking concomitant diltiazem.

 

Serious safety

 

The table below shows the number and percent of patients how reported early withdrawal for an adverse event, a serious adverse event, and/or death.

 

No. and (percent) of patients

	Placebo N=269	Ran 750 mg N=279	Placebo subtracted (%)	Ran 1000 mg N=275	Placebo subtracted (%)
Death	3 (1.1)	2 (0.7)	-0.4	1 (0.4)	-0.7
Serious adverse event	15 (5.6)	20 (7.2)	1.6	19 (6.9)	1.3
Early withdrawal for AE	16 (5.9)	22 (7.9)	2.0	25 (9.1)	3.2

Table 3.0.0

 

There were 6 reported deaths with half occurring in the placebo group (3 placebo, 2 ranolazine 750 mg and 1 ranolazine 1000 mg).

 

There were more reported serious events in the ranolazine 1000 mg group (6.9%) and ranolazine 750 mg (7.2%), compared to placebo (5.6%). In addition, compared to placebo, there were more reported withdrawals for adverse event in the ranolazine 1000 mg group (9.1%) and the ranolazine 750 mg group (7.9%).

 

 

Deaths

There were 6 deaths and they are listed in the table below.

Patient ID	Treatment group/background med	Duration of treatment (days)	Cause of death
177/9027	Ranol 750 mg/amlodipine	33	Acute MI
704/7600	Ranol 750 mg/diltiazem	18	Sudden death
706/9575	Ranol 1000 mg/amlodipine	13	Sudden death
710/7631	Placebo diltiazem	18	Sudden death
717/8668	Placebo/atenolol	6	Acute coronary insufficiency
751/9386	Placebo/amlodipine	83	Cardiac arrest

Tables 3.2.0 and 3.2.1

 

There were 3 deaths reported in the placebo group, 2 in the ranolazine 750 mg group and 1 in the ranolazine 1000 mg group. The causes of deaths are not unusual for this type of patient population.

 

Serious adverse events

The numbers and percents of patients reporting a serious event are shown below by event.

 

Cardiovascular events combined were reported by ranolazine subjects more often than placebo (3.7%) compared to the ranolazine groups (5.4% and 4.7% for 750 mg and 1000 mg, respectively). 

 

Serious adverse events reported by 3 or more subjects in 1 or both of the ranolazine groups and reported by more ranolazine subjects than placebo subjects include myocardial infarct and syncope. On the other hand, angina pectoris was reported more often by placebo patients.  No particular event, other than syncope (all in ranolazine 1000 mg), was convincingly reported more often in the ranolazine group compared to placebo.

 

The patients taking concomitant diltiazem with ranolazine had more serious adverse events compared to patients taking atenolol or amlodipine (table 3.0.6).

 

Withdrawals for adverse events

Adverse events leading to discontinuation reported for more than one subject included nausea, headache, asthenia, dyspnea, dyspepsia, palpitations, syncope, angina pectoris, dizziness, atrial fibrillation, myocardial infarction, constipation, myasthenia, abdominal pain, tinnitus, vomiting, coronary artery disorder, myocardial ischemia and sudden death.

 

The table below shows the adverse event leading to discontinuation in 3 or more ranolazine subjects and more in the ranolazine than the placebo subjects are shown below.

 

No. and (percent) who discontinued for adverse event

	Placebo N=269	Ran 750 mg N=279	Ran 1000 mg N=275	Total ranol N=554	% Placebo subtracted
Total discont	16 (5.9)	22 (7.9)	25 (9.1)	47 (8.5)	2.6
Nausea/N&V/vomiting	0	2 (0.7)	6 (2.2)	8 (2.9)	2.9
Dizziness	1 (0.4)	2 (0.7)	7 (2.5)	9 (1.6)	1.2
Constipation	0	2 (0.7)	2 (0.7)	4 (0.7)	0.7
Asthenia	0	1 (0.4)	3 (1.1)	4 (0.7)	0.7
Syncope	0	0	3 (1.1)	3 (0.5)	0.5
Headache	0	1 (0.4)	2 (0.7)	3 (0.5)	0.5
MI	0	2 (0.7)	1 (0.4)	3 (0.5)	0.5
Myocardial ischemia	1 (0.4)	2 (0.7)	1 (0.4)	3 (0.5)	0.1

Attachment 6 dated 3-18-03

 

Nausea (with the addition of nausea & vomiting and vomiting) was the leading adverse event resulting in discontinuation for the ranolazine group (2.9%), compared to placebo (0). Dizziness was the next most cited adverse event followed by  constipation and asthenia. Syncope was reported only by the high dose group.

 

Patients were more likely to discontinue study early because of an adverse event if they were receiving diltiazem and ranolazine (table 3.0.6).

 

 

Syncope

There were patients reporting serious syncope/dizziness and they are listed in the table below.

Patient ID	Treatment group/background med	Days on drug at time of event	Comments
174/7012	Ranolazine 1000/diltiazem	3	Lab tests and ECG were reported as normal. Withdrawn from study
530/7169	Ranolazine 1000/diltiazem	9	Experienced dizziness on day 7 followed by syncope 2 days later
549/7219	Ranolazine 1000/diltiazem	5	Witnessed event with loss of consciousness and “jerking movements” and “snorting breathing” Pulse not palpable, systolic BP 88 mmHg. Remained on drug.
562/8071	Ranolazine 1000/atenolol	70	This 86 year old female was hospitalized for syncope reported as mild. Withdrawn from drug.
569/7065	Ranolazine 1000/diltiazem	17 and 33	Treatment interrupted
177/7406	Ranolazine 1000/diltiazem	18	Nausea and vomiting followed by syncope. Withdrawn from study

Appendix 14.7.2

 

There were 5 patients reporting syncope as an adverse event. Of the 5, 4 were receiving ranolazine 1000 mg plus diltiazem and 1 was receiving ranolazine 1000 mg plus atenolol.

 

 

Clinical Laboratory

 

Hematology

The mean changes from baseline at the last double blind visit are shown below by treatment group.

 

 

There were decreases in hemoglobin, hematocrit, and RBCs in all three treatment groups but more so in the ranolazine groups.

 

Mean changes from baseline at endpoint and (SE)

Parameter	Means
	Placebo	Ranol 750 mg	Ranol 1000 mg
Hemoglobin g/dl	-0.11 (0.06)	-0.55 (0.06)	-0.56 (0.05)
Hematocrit %	0.1 (0.2)	-1.0 (0.2)	-1.2 (0.2)
RBC (106/uL)	-0.03 (0.02)	-0.25 (0.02)	-0.3 (0.02)

Table 12I

 

Shift changes for the 3 hematology parameters are shown below.

 

 

 

There were more subjects in the ranolazine groups compared to placebo who were normal at baseline and became abnormally low at endpoint.

 

Clinical chemistry

 

The mean changes from baseline at the last double blind visit are shown below by treatment group.

 

 

 

 

 

 

Nothing seems alarming.

 

 

Resting ECG

Standard supine 12-lead ECGs were obtained at Screening (Visit 1), at trough at Visits 2-6 (or early withdrawal), and peak at Visits 2, 3,and 5,  and whenever clinically indicated. The ECG was to be inspected by the investigator to ensure patient safety. The QT interval was to be examined for evidence of prolongation. Any other new clinically significant ECG findings appearing during treatment with study medication was to be discussed with a study monitor to determine whether the patient should continue in the study.

 

Official reading of each ECG for analysis was measured on the supine rest electrocardiogram and performed by the ECG core laboratory. The patient was to be withdrawn from the study and monitored to ensure the QTc returned to baseline if, at any point during the study, the QTc interval widened to 130% of its duration at Visit 2 and was longer than 500 msec.

 

In the ECG safety population, key parameters from the centrally coded electrocardiogram (e.g., corrected QT interval, T wave amplitude, T wave notching) were to be analyzed using analysis of variance for continuous measures or Cochran-Mantel-Haenszel tests for categorical measures.

 

ECG Interval changes

Mean ECG changes from baseline at week 12 at peak (4 hrs + 0.5 after dosing) drug concentrations are shown below.

 

 

Heart rate was essentially unchanged by ranolazine. Mean changes in QT and QTc intervals were significantly longer (p<0.001) in the ranolazine 750 mg and ranolazine 1000mg groups compared to placebo.

 

T wave amplitude decreased significantly (p<0.001) from baseline in both ranolazine groups compared to placebo.

 

The table below shows the QT interval changes at weeks 2 and 12 comparing the ranolazine groups to placebo using ANCOVA model.

Mean differences of ranolazine versus placebo for QT interval at peak were 7.5 msec at week 2 and 11.2 msec at week 12 for ranolazine 750 mg and 8.6 msec for ranolazine 1000 mg at week 2 and 11.7 msec at week 12. Both doses at both time points were statistically significantly different from placebo (p<0.001).

 

The upper limits of the 95% confidence interval for change at peak at week 12 were 15.7 msec for the 750 mg dose and 16.2 msec for the 1000 mg dose.

 

The table below shows the number and percent of patients with selected QT changes from baseline, by treatment group.

 

No. and (percent) of patient with selected QT changes

QT changes from baseline msec	Placebo N=257		Ranolazine 750 mg N=271		Ranolazine 1000 mg N=255
	Week 2	Week 12	Week 2	Week 12	Week 2	Week 12
0-<30	123 (48)	102 (42)	125 (46)	121 (48)	134 (53)	112 (47)
30-<60	9 (4)	12 (5)	29 (11)	33 (13)	26 (10)	34 (14)
>60	4 (2)	5 (2)	12 (4)	13 (5)	9 (4)	12 (5)

Attachment 3 dated 3-18-03

 

The percents of patients with prolonged QT intervals at Weeks 2 and 12 are at least twice as high in the ranolazine groups compared to placebo. There is little difference between the ranolazine groups.

 

Trough changes

Mean ECG changes from baseline at week 12 at trough drug concentrations are shown below.

 

As with peak effects, the mean changes at week 12 for QT and QTc intervals at trough were statistically significantly greater compared to placebo (8.5 msec and 4.5 msec for ranolazine 750 mg, respectively, and 10.0 msec and 7.7 msec for ranolazine 1000mg, respectively). There was little effect on heart rate, but the effect on T wave amplitude was significantly different from placebo for both treatment groups.

 

The table below shows the number and percent of patients with selected QTc changes from baseline, by treatment group, at weeks 2, 6 and 12.

 

 

 

 

No. and (percent) of patient with selected QTc changes

	Placebo		Ranolazine 750 mg		Ranolazine 1000 mg
QT changes from baseline msec	Week 6	Week 12	Week 6	Week 12	Week 6	Week 12
<0	141 (56)	134 (54)	102 (38)	112 (44)	89 (35)	78 (33)
0-<30	110 (43)	112 (45)	156 (59)	134 (52)	154 (61)	152 (63)
30-<60	1 (0)	0	4 (2)	8 (3)	4 (2)	6 (3)

Table 3.7.10.6.1

 

The incidence rates for QTc changes less than 0 msec at weeks 6 and 12 were higher for placebo than for the 2 ranolazine groups. For changes greater than 0 msec, incidence rates greater than 0 msec at both weeks were higher for the ranolazine groups than placebo groups.

 

T wave morphology

 

Changes in T wave morphology by drug group are shown below.

 

The numbers of patients with notched T waves were greater in the ranolazine groups.

 

No of patients with notched T waves

	Placebo		Ranolazine 750 mg		Ranolazine 1000 mg
	peak	trough	peak	trough	peak	trough
Week 2	1	1	11^	7	5#	8
Week 6	-	0	-	4	-	6
Week 12	0	1	3	2	8	4

^2 patients also had notched T waves at baseline/screening

# 1 patient also had notched T waves at baseline/screening

 

Subject 218/7250 (ranolazine 1000 mg) was withdrawn on day 12 because of an increase of >25% QTc, asthenia, nausea, and dizziness. Events started on day 1 of study drug dosing.