The Background
The second generation of antipsychotic medications was launched by the arrival of clozapine in1990. Subsequent pharmacologic development produced a handful of similar follow-ups including risperidone, olanzapine, zispradone, aripiprazole and quetiapine. Their novel mechanism of action (increased antagonism at certain dopamine and serotonin receptors) combined with their reduction of certain side effects (movement disorders, especially tardive dyskinesia and hyperprolactinemia) gave these new agents a sense of promise and potential advantage. Later findings that several of these new atypicals might be specifically effective in mood disorders, particularly bipolar states, made their glow even stronger.
The bloom is now off the rose. The first corrective was the uncovering of metabolic problems associated with the use of these second generation antipsychotics. We now know that, as a whole, this group of medications can produce weight gain, increased cholesterol and triglycerides, raised blood sugar and associated diabetes. While these effects vary by particular medication and dose, the idea of a newer, cleaner class of antipsychotics without the burden of troublesome side-effects is no longer tenable.
The Story
It is with additional disappointment then that we report on the article by Ray et al in the January 15th (2009) issue of the New England Journal Medicine. In a large, retrospective study of Tennessee Medicaid enrollees, the authors compared the rates of sudden cardiac death (SCD) in three groups: those using conventional antipsychotics, those using atypicals and those using neither. They found significantly increased rates of SCD in those groups using either conventional or atypical antipsychotics relative to non-drug-using controls who were matched (in a questionable and novel manner) on a variety of sociodemographic and illness-related variables. (The validity of the matching method is critical to enable genuine drug-related differences to be discerned from patient variables that might predispose to increased rates of cardiovascular mortality, such as tobacco use, obesity or poverty which are all found in higher frequency in patients with persistent, severe mental illness).
More surprisingly, the rates of SCD were essentially similar in those patients using the conventional and the atypical drugs. So, not only were the 2nd generation agents found to be associated with increased rates of SCD, they did so to a degree comparable to that found with the older, conventional drugs.
Last, the increased risk was found to be dose-dependent with low doses (eg, less than or equal to 75mg/day of quetiapine, or 2 mg/day or risperidone or 7.5 mg/day of aripiprazole) conferring a incidence rate ratio of 1.6, relative to non-users. This incidence rate ratio increased to 2.86 for patients on high doses, which were defined as 3x or more than the maximum lower doses just specified.
Now, what do these numbers mean in real terms? The Ray study found a rate of SCD in the general population of 143 per 100,000 patient years or 0.0014 per year per single person. This rate increased two to three-fold, with higher doses of medication, to between 0.0029 to 0.0033% per year for a given individual. So, if you were to take an atypical medication for a year, you would have a risk of SCD of between 2 to 3.3 per thousand.
Our Reactions
As usual, we are now left trying to give the appropriate interpretation and attention to these new data. Specifically, to what degree should this new information alter our existing clinical practice? After getting over the initial shock and alarm, we have several reactions and suggestions.
First, though the data is worrisome, this is hardly the last word to be written on atypicals and cardiovascular risk. An earlier study by Liperoti in the 2005 Archives of Internal Medicine found no increased risk of ventricular arrhythmias or cardiac arrest in a case-controlled study of nursing home residents taking atypical antipsychotics (they did, however, confirm an approximately two-fold higher rate of risk in those residents taking first-generation medications). We expect that future studies will refine the emerging picture of risk associated with this class of medications.
Second, if the association between atypicals and SCD is replicated and found to hold water, it does not portend the immediate elimination of this class of medications. Even if this risk is verified, the absolute numbers involved are small enough that judicious use with appropriate cautions will enable ongoing use in those patients where the potential benefits outweigh the associated potential for harm.
A third is to sensitize us to predisposing clinical factors that increase cardiovascular risk in our patients and, thus, make them even greater targets for any rhythm-disrupting properties of antipsychotics. A partial list of such factors includes female sex, older age, the presence of dementia, pre-existing cardiovascular disease, a personal history of syncope, a family history of sudden cardiac death before the age of 40, electrolyte abnormalities, and the use of other drugs which can effect cardiac conduction such as stimulants and tricyclic antidepressants. See Zarate and Patel (Archives of General Psychiatry, 2001, 58, pp 1168-1171) and Abdelmalwa and Mitchell (Advances in Psychiatric Treatment 2006, 12(1), pp 35-44) for thorough reviews of this topic.
The Controversy
Having said all this, the bottom-line question remains: Will we now conduct ourselves differently with this new information? This topic has been the subject of several controversial and conflicting commentaries, each weighing in on the Ray et al study. The accompanying editorial in the New England Journal of Medicine advocated for reduced usage in vulnerable populations (the young and the elderly) and EKG monitoring both before and during treatment with these medications. The Carlat Report (www.thecarlatreport.com, a fantastic source of impartial information on psychiatric research) takes a roughly similar stand. An APA report prepared by Lieberman, Merrill and Parameswaran (www.omh.state.ny.us/omhweb/advisories/adult_antipsychotic_use_attachement.html) takes a stance that is more critical and questioning of the original study and argues that routine EKG monitoring may be premature. We refer readers to these commentaries to obtain a more direct picture of this debate.
Conclusion
Whether EKG monitoring becomes a standard of care or not, I am certainly more cognizant and aggressive in my screening for cardiovascular risk factors in patients in whom I am starting or continuing on antipsychotic medications. The Lieberman et al and NEJM editorial both provide guidelines for assessing risk and managing EKG readings (see also, Abdelmalwa and Mitchell, Advances in Psychiatric Treatment, 2006, 12, pp100-109, for a comprehensive review of monitoring and prevention recommendations).
Given the uncertainty of these findings and the relative ease of getting an EKG, I am trending towards routine monitoring.
How does the Ray et al study affect the status of atypicals? Were the promised benefits and claims of greater safety illusory? The answer, I think, pertains both to this class of medication specifically and to the larger issue of drug development and marketing. Because of the phenomenal costs involved in bringing a psychotropic to market, pharmaceutical companies promote new drug launches with marketing and supporting research designed to maximize usage. This is especially true in the post-approval phase. To a certain extent, we all greet new treatment developments with hope and anticipation. When exaggerated, this optimism can become excessive, leading to the type of irrational exuberance seen in the stock market rise prior to last year. When this type of exuberance takes hold and critical analysis is put aside, this almost ensures a subsequent market correction which will bring a product or a stock back to a more realistic place.
The recent studies about the metabolic and cardiac toxicities of 2nd generation antipsychotics are simply bringing these new medications back to the type of more realistic valuation they should have had and maintained from the start. Atypicals will need to take their place with all other psychotropic medications as drugs with distinct risks and potential efficacies. True progress will have been made if, in the future, we are a little more circumspect about the next class of agents that comes our way.