Wednesday, February 14, 2007

Is Ranolazine Safe?

Yesterday a colleague of mine returned from a drug company sponsored symposium for Ranexa (generic name, ranolazine) made by CV Therapeutics (CVTX) of Palo Alto, CA.

He was mad.

Real mad.

He thought he’d been sold a bill of goods.

You see he’s a cardiologist who reads and pays attention to all things new in an attempt to offer leading-edge innovations to his patients. He attended this meeting to inquire about little known safety data about the drug ranolazine (Ranexa).

In January 2006, the FDA approved ranolazine, a piperazine derivative, as a new drug for patients with angina (chest pain) refractory to conventional medical therapy. It was hailed as the first new drug to help patients with angina in the past 10 years. One of its mechanisms of actions is to inactivate slowly inactivating component of the late inward sodium channel which may reduce the deleterious effects of calcium overload and intracellular sodium that accompany cells of the heart when they lack oxygen. But the drug also prolongs QT interval, an important determinant of potential lethal arrhythmias in patients with diseased hearts from coronary blockages.

To serve as a bit of background, QT prolongation came to light as a cause of mortality in a much earlier drug trial called the Coronary Artery Arrhythmia Suppression (CAST) trial. This trial tested the safety of drugs that suppressed premature ventricular beats well but also prolonged the QT interval – and was shown to have an increased mortality of patients in the drug treatment arm as compared to placebo. Many, many other drugs have never made it to market (or have been pulled from the market, witness the withdrawl of terfanidine (Seldane)) because of QT prolongation effects and pro-arrhythmia.

The company had a "hired gun" expert speaker to present their data from the trials presented to the FDA called the MARISA, CARISA, and ERICA trials. The data from the clinical trials looked great. Ranexa was clearly the one and only QT-prolonging drug out there that was safe for patients with chronic refractory chest pain. At least so it seemed. But there were troubling issues about these trials.

First, there was the usual conflict-of-interest issues. Most of the "investigators" of the MARISA and CARISA trials were either paid grant recipients, consultant/speakers, or on CV Therapeutics' payroll and owned stock in the company. Worse still, the longest follow-up period of any of these trials was 12 weeks.

Concerns about QT prolongation were not lost on the US Food and Drug Administration, even in 2004. In the report from the 100th Cardiovascular and Renal Drugs Advisory Committee Meeting, ranolazine was reviewed and “the principle safety issue with the drug was related to QT prolongation.” The committee tabled approval but wanted long-term efficacy data in females, elderly patients, and African Americans.

So the presenting "doctor expert" was asked a question. He was asked how any drug could have been approved with only 12 weeks of clinical data upon which to base safety concerns – especially in a drug known to prolong the QT interval of such sick patients. My colleague was told by the respected physician speaker that night that "no data" exists to suggest the drug is anything but safe in this patient population.

But some data do exist on the FDA's website. In their slides, CV Therapeutics claimed that the efficacy and safety of ranolizine demonstrated “No evidence for an adverse effect of ranolazine on survival.” (See slide 27 of the FDA’s Powerpoint presentation by Dr. Thomas Wolff). But then we look at Slide 26 – the slide just before this that includes 4-month follow-up data of deaths found by “pharmacovigelence” after the study completion -- here’s what the slide says:

Incidence per Patient Yr of Death and Sudden Death
Cause of DeathTherapyEventsFollow-up,Pt yrIncidence(%) (95% CI)
Sudden Death, death due to ventricular fibrillation/tachycardia or cardiac arrestPlacebo2169.31.17
(0.14, 4.18)
(0.36, 5.05)
(1.76. 3.21)
All causePlacebo3169.31.76
(0.36, 5.05)
(2.13, 3.68)

There it is: 4-month safety data that demonstrates an increased cardiovascular and all-cause mortality incidence in folks on ranolazine. Yet the drug was approved by the FDA in January 2006.

Can this be true? Was this slide not presented to the FDA reviewers? Did they not see this?

Well it seems they did. I called Michael Sweeney, MD, Vice President of Medical Affairs at CV Therapeutics, and he was kind enough to return my call. He acknowledged the event findings in the slide above, but because the numbers fell well within the 95% confidence intervals, they were not considered significant. Over a 10:1 1.5:1 safety difference was not significant? Ah, the power of statistic-speak.

Now the company has embarked on a large 6500-patient, randomized, double-blind, placebo-controlled multinational clinical trial to evaluate the drug as treatment for adjunctive treatment in acute coronary syndromes (Heart attack and unstable angina) called the MERLIN-TIMI 36 trial (MERLIN is a pneumonic for Metabolic Efficiency with Ranolazine for Less Ischemia in Non-ST-elevation acute coronary syndromes) in an attempt to extend its clinical indications. This study results are to be released at 8:50 am Central Standard Time on March 27,2007 as a late-breaking trial at the American College of Cardiology Meeting in New Orleans, LA. The trial will evaluate safety and efficacy of ranolazine after 730 cardiovascular events and 310 deaths have occurred during the study period. It is important to realize that the study’s Data Safety Monitoring Board met when one half of the patients had their pre-determined adverse event and elected to continue the study (good news for CV Therapeutics). But given the company’s own follow-up data, it begs a question:

Given the former FDA data, has the MERLIN genie already been let out of the bottle regarding the safety of this drug in patients with chronic ischemic coronary disease?

There are long stock-holders who are painting a rosy picture of this drug’s chances, particularly since this individual notes that other large institutional investors are increasing their positions in CV Therapeutics.

So go ahead and take your chances. Make a guess how ranolazine will fair. But I'm waiting until the MERLIN TIMI-36 data are presented in March before any of my patients with chronic ischemic coronary disease are placed on this drug.


18:16PM Link fixed to FDA slides

Clarification: Several commentors to this post (see attached) have correctly pointed out the calculation for safety risk represents a 1.5:1 rather than 10:1 difference as I posted earlier. I regret the error.


Anonymous said...

Dr. Fisher, thanks for your thoughtful post. With respect, I believe you have misunderstood the data.

For placebo patients, there were 2 deaths from VF in 169 patient-years of follow-up. This is 0.012 VF events per patient-year.

For Ranexa, there were 21 deaths from VF in 1848 patient years. This is a rate of 0.011 VF events per patient year.

There is no statistically significant difference between these two rates.

The large difference in patient years of follow up between placebo and Ranexa reflects the fact that at the end of a randomized clinical trial, there is commonly an open-label phase. All patients are offered the opportunity to continue the drug and are then monitored long term.

I believe the Ranexa safety database is now about 15,000 patient years. There has never been a case of torsades.

I am a cardiologist. I have no business relationship with CV Therapeutics, and I have received no grant support, consultant fees, speaker fees, or other payment from them. I have a long position in CVTX securities.

DrWes said...

Thanks for your clarification. I genuinely hope that we find that the low incidence rates are confirmed in the MERLIN trial. However, we must realize that this drug is used on our sickest of coronary patients (those with active angina refractory to medical therapy) with a high baseline mortality rate. While Torsades may not have been identified in the database to date, I would argue that that database may be an incomplete one. Torsades can be decidedly difficult to detect in patients taking medications due to its transient nature. Death is a hard endpoint less likely to be argued about. Could you speak about incidence of deaths on the database in the patients treated?

While I understand the issues at hand, the FDA's approval of this drug before mortality data are evaluated, especially in such a high-risk population, remains a puzzle to me.

BTW, I also have no relationship with the company and hold no financial interest - just the interest of my patients.

Anonymous said...

I agree that as a treatment for chronic angina, Ranexa is used in a very sick - and very torsades-prone - patient group.

Ranexa was approved on risk-benefit grounds and I believe the decision was reasonable. Chronic exertional angina is a debilitating disease, available therapeutic options were not always effective, and there was a substantial safety database that did not suggest problems. The degree of QT prolongation was very modest - 6 msec per 1000 ng per ml.

That said, I deeply respect colleagues such as yourself who prefer to wait for clinical outcomes data before prescribing the drug. It is a very reasonable and principled position to take. Such data will soon be available.

View from the Trekant said...

I won't be an early adopter of this medication either.

I'm not that impressed with the clinical improvement for one and agree that there are still QT concerns.

I also don't think the patient population represented in the last study represented those who are being put on the drug in clinical practice.

Just to really stir the pot: low-dose long acting MS contin is a very effective anti-anginal for those who have exhausted EVERY rehab and revascularization option and are on maxed-out triple therapy. MS lowers ventricular filling pressures, improves exercise tolerance and reduces anxiety as well as pain.

I'm talking about maybe one patient a year - I haven't had one yet who did not improve dramatically when we finally came down to this. Of course, we'll never see a randomized trial of this medication for this indication, however.

(And, of course, MS comes with serious consequences if mismanaged!)

This is not meant to be so much an advertisement for MS contin as a comment on how a cheaper therapy will never get the studies and support that the NEW patent-protected product does.

Anonymous said...

I just wanted to make a few points regarding your analysis.
1- The mortality rate was higher as you point out, but not 10 fold. The RR was about 1.5. Look at the much larger follow up in the ranolazine group.
2- This data included older trials on older formulations and from dose finding studies. The data from CARISA and MARISA- although they represent an even smaller data set- are a fairer data set to look at. They are shown on the next FDA slide- and don't show any increased risk.
I personally prescribe Ran- and have had some great clinical responses. I certainly share some of your concerns, but find the fact that MERLIN wasn't stopped early to be pretty reassuring.
I guess we will find out what the right answer is at the ACC.

Sam G said...


Note that Reuters has picked up on your blog.

I've pointed out to you, in the most collegial and polite terms, that your statements of fact are incorrect.

You acknowledged that fact, but did not correct the errors in your blog.

Meanwhile Reuters, a news agency committed to propagated falsehoods whenever it serves their agenda, has of course picked up on your errors and propagated them worldwide.

Wes, please. In the interest of truth, the only thing that matters. Correct your blog.

DrWes said...


Thank you. Your point is well-taken. A correction has been posted.

Although I am unaware if you were Anony 5:57, I would like to respond to one other issue made by that commentor:

"There has never been a case of torsades."

How, really, can one differentiate ventricular fibrillation from "sustained" torsades? It is a minor point, but an important one nonetheless. The safety data collected were all performed in relatively healthy population (ambulatory - able to walk on a treadmill). Ranexa is metabolized in the liver almost exclusively and several common drugs (diltiazem, some macrolide antibiotics) increase its drug levels since they compete for the same metabolic pathway in the liver. To my knowledge, no data regarding this drug's metabolism in congestive heart failure (CHF) patients exists. As you know, liver function is greatly diminished in heart failure patients. It would seem logical to this cardiologist that knowledge of this fact is critical to our understanding of the safety of Ranexa, particularly in such a high-risk population for the development of CHF. Unfortunately, even TIMI-36 will not test this population, but at least its results will be the first look at survival differences in one cardiovascular group and the randomized nature of the trial will eliminate the potential for selection bias.

Anonymous said...

Actually, it seems that the drug has been studied in NYHA class 1-4 CHF- per the label- and that there was no effect on pharmacokinetics.
Seems like you might want to check your facts a little better. Reading the FDA approved label would be a good place to start or you could try googling ranolazine and CHF as I just did.

DrWes said...

Anony 10:35 -

Thanks for pointing this out. But "pharmacokinetic trials" are not mortality trials, are they? You are correct when the package labeling suggests no dose adjustment is required (based on dosing studies) for patients with NYHA Class I-IV heart failure.

But there are other (counter-intuitive) CONTRAINDICATIONS to dosing Ranexa ALSO on its label: "Ranexa is contraindicated in patients:

With pre-existing QT prolongation
With hepatic impairment (Child-Pugh Classes A [mild], B [moderate] or C [severe]) (see CLINICAL PHARMACOLOGY, Hepatic Insufficiency, and Electrocardiographic Effects)
On QT prolonging drugs
On potent and moderately potent CYP3A inhibitors, including diltiazem


QT Prolongation
Ranolazine has been shown to prolong the QTc interval in a dose-related manner. While the clinical significance of the QTc prolongation in the case of ranolazine is unknown, other drugs with this potential have been associated with torsades de pointes-type arrhythmias and sudden death.

With repeat dosing, the mean effect on QTc of ranolazine 1000 mg b.i.d., at Tmax, is about 6 msec. However, in 5% of the population the prolongation of QTc is 15 msec. Age, weight, gender, race, heart rate, CHF NYHA Class I to IV, and diabetes have no significant effect on the relationship between ranolazine plasma level and increase in QTc. The relationship between ranolazine levels and QTc remains linear over a concentration range up to 4-fold greater than the concentrations produced by 1000 mg b.i.d., and is not affected by changes in heart rate. Doses greater than 1000 mg b.i.d. should not be used.

There are no studies examining the effects of ranolazine in patients with pre-existing QT prolongation or receiving other QT prolonging drugs. Because of possible additive effects on the QT interval, ranolazine should be avoided in patients with known QT prolongation (including congenital long QT syndrome, uncorrected hypokalemia), known history of ventricular tachycardia and in patients receiving drugs that prolong the QTc interval, such as Class Ia (e.g., quinidine) and Class III (e.g., dofetilide, sotalol) antiarrhythmics, and antipsychotics (e.g., thioridazine, ziprasidone).

Because the QTc-prolonging effect is increased approximately 3-fold in patients with hepatic dysfunction, ranolazine is contraindicated in patients with mild, moderate or severe liver disease (see Special Populations and Hepatic Insufficiency)."

I believe Anony, that I have checked my facts. I am quite aware that many very smart and learned individuals have made great efforts to assure the public's safety regarding this drug. But FDA approval letters are FDA approval letters, not prospective, randomized, placebo-controlled mortality trials. Many a drug has been approved by the FDA only to be later recalled.

So please, sir or madam, could we wait to see the TIMI-36 trial results? That's my only point here.

Sam G said...

No one could possibly object to a cardiologist who would prefer to wait for MERLIN results. It is a fair, reasonable and conservative course of action. I respect you for it.

Sam G (yes I was anon 557 - I have an unbelievably hard time signing in)

Anonymous said...

Looks like the top line Merlin results are out. Didn't show any benefit for ACS, but no trend towards increased mortality or arrythmias.
I guess you can use the drug now.

Anonymous said...

Wondering what your thoughts are on this now?


Is it your opinion that this drug is still not safe and doesnt work? I would love to hear any stories of your patients that have tried it. Not for arrhythmias, but for ischemia.