Health professionals are warned to beware of drugs that trigger a heart disorder that can lead to sudden death
Health professionals are warned to beware of drugs that can trigger a heart disorder that can lead to sudden death
In July of 2000, Propulsid (cisapride, Janssen Pharmaceutica) became the latest of several high-profile drug withdrawals in recent years, joining Seldane (terfenadine, Aventis), Hismanal (astemizole, Janssen), Raxar (grepafloxacin, GlaxoSmithKline), and Posicor (mibefradil, Hoffmann-La Roche). Despite prominent changes to product labeling, educational programs, and multiple "Dear Doctor" letters, Janssen Pharmaceutica was forced to remove cisapride from the general market.
As most pharmacists know by now, cisapride can cause a disorder called long QT syndrome (LQTS). When combined with certain other drugs, such as erythromycin, fluconazole, or quinidine, the chances of developing LQTS greatly increase. According to reports collected by Janssen, at least 80 patients have died as a result of cisapride-associated LQTS.
Some experts believe that number is grossly inaccurate and estimate thousands of deaths may have been caused by cisapride. It's impossible to know for certain. "Often [LQTS] is not recognized, and often it is not reported," said Raymond Woosley, M.D., Ph.D., chairman of the department of pharmacology at Georgetown University.
Cisapride has been on the market since 1993, before many clinicians were even aware of the syndrome. It is quite possible that many patients had cardiac complications that were inadvertently attributed to some other cause. After death, no evidence of QT prolongation remains. Other cases may have slipped through the cracks because the Food & Drug Administration's reporting system for adverse events is voluntary.
"Cisapride was taken off the market not because it was such a dangerous drug if used as directed, but ... because the community couldn't use it as directed," said Woosley. The physician was one of several experts recruited by the Sudden Arrhythmia Death Syndromes (SADS) Foundation for a recent LQTS seminar in New York. "So this is an effective drug that is now no longer available to patients not because of its toxicity but because the use of the drug caused toxicity."
Since 1991, SADS has been getting the word out about LQTS. While some of the most dangerous drugs have been removed from the market, there are still many others that can prolong the QT interval. The problem is not going away. On the contrary, it will become worse as more QT-prolonging drugs are approved.
In order to understand LQTS, pharmacists first need to know what the QT interval is. No need to dust off those old physiology books, though. G. Michael Vincent, M.D., the founder and president of SADS, explained the basics at the SADS seminar.
The QT interval refers to electrocardiogram readings. Each wave on the ECG is designated by a letter: P, Q, R, S, and T. The Q-wave is the beginning of the electrical discharge of the ventricles. The T-wave represents the repolarization of the heart. The time elapsed between the Q-wave and the T-wave is the QT interval. So, the QT interval represents the time it takes for the heart to recharge after the ventricles have discharged.
During the recharging or repolarizing phase, the heart muscle is most vulnerable to stimuli that can provoke arrhythmias, according to Joe Slechta, Pharm.D., cardiology clinical specialist at Wesley Medical Center in Wichita, Kan. When the QT interval is prolonged, a type of ventricular tachycardia called torsade de pointes (TdP) can occur. The likelihood of TdP increases with the length of the QT interval.
Victims of TdP often lose consciousness until the arrhythmia corrects itself. TdP does not always resolve on its own, though. Instead, it may progress to ventricular fibrillation and, if not corrected, sudden death. "You're not going to have much warning with that, unfortunately," cautioned Slechta.
So how long is too long? Because the QT interval varies in length according to heart rate and so on, it is not a set unit. "Unfortunately, there isn't a single number that divides a long QT interval from a normal QT interval," said Vincent.
Because of the variability of the interval, researchers correct for heart rate to simplify comparison of QT values. One of the most common correction techniques is Bazett's formula. By dividing the length of the QT interval by the average interval between QRS complexes, a value called QTc (QT corrected) can be determined (QTc = QT/(RR)0.5. Different values for QTc are produced by the 25 or so other formulas.
Vincent explained that when experts are citing a measurement for the QT interval, they are most likely referring to QTc, and most often they have used Bazett's formula. Normal values for QTc range from 0.38 to 0.44 seconds. Some LQTS experts believe that an interval of .45 or greater is prolonged. Women tend to have longer QT intervals, though, so .46 or greater may be more applicable for them.
Long QT syndrome, then, is a disorder in which there is an electrical abnormality in the heart. Because electrical charges are generated by ion channels in cardiac muscle cells, these channels are the source of the abnormality. This condition can be inherited or acquired. In the inherited form, mutations of certain genes produce potassium channels that slow the efflux of potassium from the cells, which prolongs the QT interval.
The acquired form of LQTS can be caused by several conditions or events, such as stroke, heart disease, and/or protein/electrolyte disturbances (see Table 1 for a list of risk factors). But the condition is most often drug-induced. Although the cause of the acquired disorder is quite different, the effects are the same: The flow of potassium ions out of the potassium channels is slowed, and the QT interval is prolonged.
The congenital form of LQTS is relatively rare. Vincent estimated that about 50,000 people in the United States might have the mutated genes that cause LQTS. Acquired LQTS is much more common. But because relatively few patients are identified, it is difficult to know exactly how many people are affected. Some experts estimate 1% to 6% of patients taking QT-prolonging drugs actually acquire LQTS. Others put the number at 12%. While this is still a small percentage, it may represent millions of people.
The big question for many researchers is, Why do only certain people on QT-prolonging drugs acquire LQTS? One possible answer is that the syndrome is severely underdiagnosed. For example, erythromycin is known to prolong the QT interval, but rarely are patients on the oral form of the drug given ECG tests (intravenous infusions are more likely to cause LQTS). Because many cases of LQTS are asymptomatic, there is no way of knowing if patients have acquired LQTS.
Another explanation is that people who acquire LQTS may actually have a mild form of the inherited disorder, which only manifests itself when aggravated by QT-prolonging drugs.
Identifying patients with LQTS can be difficult because the two symptoms are syncope and death. Unless patients have had fainting episodes, there is often no sign of the disorder before their sudden and unexpected deaths. Patients with the hereditary form are a bit easier to diagnose because they usually have family members with the disorder, and genetic testing can be done. However, according to Vincent, one-third to one-half of these patients never experience any symptoms.
Practitioners must therefore rely on thorough evaluation of all patients for risk factors for LQTS and order ECGs accordingly. Heart disease and aging may reduce the number of working potassium channels, increasing the risk for LQTS. Electrolyte abnormalities, particularly hypokalemia and hypomagnesemia, reduce the available ions for efflux, setting the stage for prolonged QT intervals. Gender is also a factor. Women are far more susceptible to the disorder than are men. All of these patients are at risk for drug-induced LQTS.
Several drugs can cause LQTS by acting directly on potassium channels. These agents inhibit potassium efflux, prolonging the QT interval, and they do so in varying degrees. Some drugs, such as quinidine, are much more likely to prolong the QT interval than others. Administering QT-prolonging drugs to at-risk patients greatly increases the incidence of LQTS. The higher the doses of these drugs, the greater the risk. Hepatic or renal insufficiency, then, can worsen the situation by reducing metabolism and excretion of the drugs.
"The danger of drug-induced QT prolongation has now become of major concern to a host of different groups," said Arthur Moss, M.D., director of the International Long QT Syndrome Registry. The FDA, physicians, pharmacists, drug manufacturers, lawyers, and the public have become aware of the extent of the problem. Both the Pharmaceutical Research & Manufacturers of America and the FDA have established task forces to assess drug-induced LQTS.
The National Institutes of Health is funding a registry to document drug-induced arrhythmias and to catalogue DNA samples from those patients. "We hope then to identify the genes associated with drug-induced arrhythmia," said Woosley.
Woosley has created and maintained a list of QT-prolonging drugs on the Web. The list can be found at www.qtdrugs.org and at Georgetown University's Center for Education on Research and Therapeutics (Georgetown CERT) at www.torsades.org. The list (see Table 2) includes more than 50 drugs, in a variety of drug classes, from amiodarone (Cordarone, Wyeth-Ayerst) to zolmitriptan (Zomig, AstraZeneca).
|Drug (brand name)||Drug (brand name)|
Arsenic trioxide (Trisenox)
Doxepin (Sinequan, Zonalon)
Erythromycin (E.E.S., Erythrocin)
Fluoxetine (Prozac, Sarafem)
Pentamidine (Pentam, NebuPent)
Procainamide (Procan, Pronestyl)
Quinidine (Cardioquin, Quinaglute)
While 50 is a formidable number of drugs to track, consider the number of potential drug interactions that could involve any of these drugs. The QT-prolonging drugs can have additive effects when given together, and any drug that inhibits metabolism or excretion of these drugs can increase the risk of TdP. So, then, the problem is considerably more complex.
As herbal remedies become increasingly popular, interactions with these drugs become more of a concern. Many patients forget or are reluctant to tell their R.Ph.s and doctors that they are taking herbals. And even less is known about the effects and interactions of herbals than those of drugs. Some herbals have an additive effect with antiarrhythmics. Quite a few others lower potassium levels.
Any patient with one or more of the risk factors for LQTS is more vulnerable to drug-induced LQTS. Not only must R.Ph.s be aware of the 50-plus products and their associated interactions, but they must also know the conditions that can interact with these drugs. Only then can at-risk patients be identified and monitored.
As pharmacists take on more and more responsibility for drug therapy, we become more liable for any consequences. According to Woosley, physicians are not generally very aware of LQTS. Pharmacists will be relied upon to pick up the slack. "This is a tremendous challenge for pharmacists, just as it is for the physicians," said Vincent.
Vincent felt that patient profiles and drug-interaction software would be the answer and envisioned a worldwide system that could be accessed by any pharmacist. For now, though, he suggested patients use one pharmacy exclusively in order to get optimal benefits from drug-interaction software. "I think [the drug-interaction] program is the real salvation," he continued. "In today's busy world, what pharmacist has time to read the books and look at the drug interactions ... of each of these medicines?"
"There is the belief that computers make things better, and it ain't necessarily so," said Thomas Hazlet, Pharm.D., DrPH, assistant professor at the school of pharmacy at the University of Washington. Hazlet, along with Philip Hansten, Todd Lee, and John Horn, conducted a study of drug-interaction software. They evaluated the ability of nine different software programs, installed in a total of 516 pharmacies, to detect several drug interactions. The results of the study were published in the March/ April 2001 issue of the Journal of the American Pharmaceutical Association.
Several fictitious patient profiles were entered into each pharmacy's computer. Each profile contained at least one well-documented interaction, and each pharmacy received identical profiles. Fifteen interactions were deliberately entered into each computer, and only one of these was detected in all 516 pharmacies. Most notably, the QT-prolonging interactions (cisapride/erythromycin and cisapride/fluconazole) were not correctly identified by 8% of the pharmacies' programs.
If pharmacists are to be relied upon to prevent drug-induced LQTS, and in turn are relying upon inadequate software programs, how on earth are they going to solve the problem? Should pharmacists rush out to buy more liability insurance? Do they need to stay up nights memorizing lists of potentially fatal drug combinations? Not necessarily, said Hazlet.
"The point of the article should be interpreted carefully," said Hazlet. He noted that, depending on the source, there are 75 to 125 "really significant interactions," all of which are relatively rare. And, while some interactions were not accurately detected by the pharmacy software, the majority of them were. "The advice is that pharmacists need to understand the limits of the software ... and be sure the technicians aren't overriding it."
Another finding of the study was that while some chains had corporate-level control over the software, others gave their pharmacies the option of customizing it. Some pharmacies chose to reduce the sensitivity of the programs in order to eliminate inconsequential drug interactions. Unfortunately, "part of the problem is the result is not what you expect," said Hazlet. That is, decreasing the sensitivity of some programs may actually eliminate important warnings.
Hazlet's advice for pharmacists is to start pestering their employers to provide improved software. The employers could then encourage software vendors to come up with better products. "The marketplace has not put sufficient pressure on the software producers," he said. He also suggested that incentives for vendors and encouraging FDA involvement might speed up the process.
In the meantime, Jerry L. Bauman, Pharm.D., FACC, advised pharmacists to bone up on LQTS. "Pharmacists clearly have to know, No. 1, the drugs that cause it ... [and] they probably should know the risk factors." In addition to female gender, electrolyte abnormalities, and underlying heart disease, he added subarachnoid hemorrhage and severely slow bradycardia as risk factors. He also noted that LQTS usually occurs within the first several days of starting a QT-prolonging drug.
"Clearly, for the pharmacists, if they're in the community, it's more difficult," said Bauman, a professor in the departments of pharmacy practice and cardiology medicine at the University of Illinois at Chicago. While hospital pharmacists have access to patient charts and ECG readings, pharmacists in the community have fewer such resources.
Slechta agreed that monitoring for LQTS in the community setting can be tricky. "I think the best thing you can stress is just communication with the patient," he said. When presented with a prescription for a drug such as erythromycin or fluconazole, "a question about a patient's cardiac history is warranted."
Slechta also said R.Ph.s should make a regular habit of getting cardiac history for any patient taking a QT-prolonging drug and include this information in patient profiles. Check with patients when switching antiarrhythmics to see if their doctor recommended waiting between drugs. If the patient is unsure, check with the physician. Also keep an eye on patients undergoing aggressive diuretic therapy, which can lead to hypokalemia and hypomagnesemia. These patients will need to have proper electrolyte supplementation in order to decrease the risks of LQTS.
Be wary of Rxs written by physicians other than cardiologists for cardiac patients. For example, a psychiatrist may prescribe Thorazine or Prozac for a patient who is being treated for an arrhythmia. A call to the psychiatrist could prevent a fatality.
Although these interventions require more calls to prescribers, Slechta pointed out, "you get a better response from doctors when [the drug interaction] is life-threatening." He suggested pharmacists be ready with therapeutic alternatives just in case. He also recommended pharmacists check out the drug list at qtdrugs.org and keep an eye on the updates.
New drugs continue to come to market with warning labels about QT prolongation, and more are in development. With the FDA ready to pull the plug on any product that behaves like terfenadine or cisapride, drug companies have been forced to address the issue.
Although Janssen withdrew cisapride from the market, the company is still making the drug available to patients via a limited-access program. Adult patients with gastroesophageal reflux disease (GERD), gastroparesis, pseudo-obstruction, and severe chronic constipation may enter the program. Children and newborns that meet certain criteria may participate as well. All patients must have failed other therapies, have been properly diagnosed, and must receive several tests, including ECGs. The prescribers must either be gastroenterologists or primary care physicians consulting with gastroenterologists.
Pfizer has devised a different approach for its antiarrhythmic, Tikosyn (dofetilide). "In trials, it caused QT prolongation in a significant number of patientsenough to be a concern," said Slechta. Knowing the risk of TdP was fairly high, Pfizer chose to limit the drug's marketing. Only physicians who have completed a dofetilide educational program are allowed to prescribe it. Pfizer maintains a list of all qualified physicians for R.Ph.s to refer to before dispensing the drug.
Therapy with other antiarrhythmics must be discontinued for three half-lives before starting dofetilide. Amiodarone is the one exception to this rule. Because it has a long half-life, amiodarone must be discontinued for three months. A qualified physician and trained staff may then initiate dofetilide therapy, but only in a hospital. Patients must remain hospitalized for at least three days.
Pfizer has very specific instructions for dofetilide initiation. Baseline QTc values must be obtained, and if the interval is longer than 0.44 seconds, the drug is contraindicated (except in patients with ventricular conduction abnormalities). Because dofetilide is renally eliminated, creatinine clearance should be calculated and the dose adjusted accordingly. Monitoring for QT prolongation is to continue for three days. According to Slechta, that is the time period in which most problems occur.
Dofetilide, like other antiarrhythmics, interacts with several drugs. Concomitant use of cimetidine, verapamil, ketoconazole, and trimethoprim (alone or in combination with sulfamethoxazole) is contraindicated. Other drug interactions may occur with amiodarone, diltiazem, nefazodone, norfloxacin, quinine, zafirlukast, SSRIs, protease inhibitors, macrolides, azole antifungals, and others listed in the package insert.
As if all that weren't bad enough, if patients skip dofetilide for three days or more, the whole initiation procedure must be repeated. With dofetilide therapy being so complicated, it's a wonder Pfizer went to the trouble to market the drug at all. Slechta believes there may be a place for the drug, though. "I think its niche is probably going to be for people who can't tolerate amiodarone."
An unfortunate by-product of drug-induced LQTS is that the syndrome has become a marketing liability. Manufacturers are happy to point out that their competitor's product can cause a life-threatening arrhythmia, while their product does not.
Ziprasidone (Geodon), Pfizer's new antipsychotic, is the victim of such a ploy. Concern over QT-prolongation slowed the drug's approval by the FDA. Finally cleared in February, with a warning label about the risk of sudden death, ziprasidone now faces a barrage of negative advertisements from its competitors. Pfizer, in turn, has found the weaknesses of other antipsychotics and is attempting to turn the tables on its adversaries.
SADS advises all patients with symptomatic LQTS to receive treatment, usually with beta-blockers. Some patients respond better to treatment with potassium, mexilitine, or implanted devices such as pacemakers and defibrillators. For patients with drug-induced LQTS, the remedy is simply the discontinuation of the drug and the administration of supportive measures, if needed.
Naturally, SADS would like to prevent acquired LQTS altogether. The foundation has organized educational programs and seminars to alert health professionals to the disorder. Although SADS has been hard at work on this since 1991, many physicians still remain uninformed. "I think it is very difficult to convince many physicians that this is a problem because they haven't seen it in their practice that they know of ... because it is fairly rare," said Woosley.
In March, SADS released the results of a survey polling primary care physicians and psychiatrists about their knowledge of LQTS. One-fifth of the respondents admitted they knew very little about drug-induced LQTS. Fewer than half of them were aware that a history of cardiac problems is a risk factor for the disorder, and only 15% indicated that QT-prolonging drugs could induce LQTS. Furthermore, a substantial number (more than 30%) did not bother to find out what other drugs patients were on before prescribing drugs that may prolong the QT interval.
Clearly, LQTS education for health professionals needs to continue. Other tactics may also be needed. Researchers are hard at work looking for a reliable, relatively inexpensive screening test for QT prolongation. The idea is that by using computer molecular models, drug companies can determine whether a new compound is likely to block potassium channels and potentially cause LQTS. Otherwise, researchers will have to continue with the more expensive preclinical and clinical studies.
The FDA is concerned that molecular screening might unnecessarily eliminate useful drugs. The fear is that a drug that might have substantial benefit may never be developed because a screening program found it blocked potassium channels. While this blockage is a strong indication that QT prolongation could occur, it does not measure the propensity for the drug to induce LQTS.
Bauman agreed: "Just because a drug prolongs QT does not mean that it causes life-threatening arrhythmias." He added, though, that most QT-prolonging drugs do increase the frequency of TdP. "I don't know of a drug that lengthens QT that doesn't cause torsade de pointes." The distinction is that the likelihood of arrhythmias varies greatly among drugs. Amiodarone is known to cause TdP, but only rarely. Dofetilide, on the other hand, has a much higher incidence of TdP.
"Europe has tackled this issue a little bit more than the States," said Bauman. He participated in a working group of cardiologists and other experts that recommended every compound be screened for QT prolongation. Their guidelines were published in the European Heart Journal. He suggested that promising new compounds be screened in vitro and again in clinical studies before a decision is made about QT prolongation.
While current methods are able to identify some new drugs that prolong QT in a small number of patients, they do not always accurately predict the prevalence of drug-induced LQTS once the drugs hit the market. Terfenadine is a good example of this. The drug's hazards were identified only after years of use by millions of patients. The small number of patients admitted into clinical studies is not always representative of the general population. Also, many studies exclude patients with known arrhythmias and other risk factors for LQTS.
Even if QT prolongation problems are discovered, the information is not making it to the prescribers and dispensers. "The capacity to alert various components [of the healthcare system] is a miserable failure," said Hazlet. Although "Dear Health Professional" letters outline important changes in product labeling and warnings, doctors and pharmacists are not getting the message.
Part of the problem, said Hazlet, is the limited capacity of manufacturers and the FDA to inform. Current information dissemination methods simply are not working. "I think it needs a different look." Until then, he asserted, R.Ph.s need to assume responsibility for educating themselves and keeping up-to-date on developments.
Jillene Lewis. KO'ING QT SYNDROME. Drug Topics 2001;13:53.