Monday, January 20, 2014

Cardiovascular and Diabetes Outcomes Among Those taking Novel Antipsychotics

I remember from medical school and the early days of my residency when the only medications available to treat psychosis were the neuroleptics.  Patients hated taking them: the high potency medicines like Haldol and Prolixin left people rigid; they had pill-rolling movements with their fingers, cogwheeling in their joints, and they walked liked zombies.  The lower potency medications like Mellaril left people drooling and sedated.  I once heard these medications described as like having molasses poured into your brain.  We'd cajole people in to taking them, and like all medications, there were some people who had no side effects and just got better, but that wasn't the usual.  I soon realized  that the best thing about Zyprexa (before it became evident that there were other problems which I will discuss below) is that most patients didn't mind taking them!  They didn't look like zombies, or obvious psychiatric patients, and they didn't complain about feeling badly.  It did feel like a revolutionary change for psychiatry.  This I remember.

So it was disappointing and distressing to realize that the novel anti-psychotics were associated with weight gain in some people, and sometimes a lot of weight was gained, diabetes, and lipid abnormalities.  Sometimes I wondered about things that things that no one was asking.  Were these problems reversible if the medications were stopped? I had a patient who became diabetic (in a rather dramatic fashion, I might add, with a glucose topping 700) and whose diabetes reversed when the medication was stopped. Was this true for everyone.  And if someone took a very low dose on an as-need basis for severe agitation and catastrophic distress -- something I see a fair amount of -- either instead of a benzodiazepine or in addition to one --  but only too a low dose for  a short period of time, were they subject to these problems?  Or what about the people with schizophrenia in their later years who were going along just fine, living independently, seeing family, going to church, looking quite well, and happy with their treatment, but who also had diabetes and cardiac risk factors?  I tried to change medications on a few of them to lower their risk, only the long-dormant schizophrenia became evident, and I decided to settle on a risk-benefit discussion and leave it to the patients.  Dose does matter?  And does the fact that lipid metabolism change mean that the patient is at greater risk for cardiovascular events, or is it just laboratory values we're discussing here? 

I pleased to see that I'm not the only one thinking about this and The Journal of Clinical Psychiatry, December 2013, published a study, "Cardiovascular Outcomes and Diabetes Mellitus Among Users of Second-Generation Antipsychotics" by Cirtome, et al.    I thought I'd summarize the finding for you.

This was a retrospective study done with data extracted from a database.  All patients who started a 2nd generation antipsychotic from 2003-2010, who were 18-64 years old were included in the study, and an index date was determined, the patient needed to fill at least 60 days worth of medicine after the index date.  The patient was followed until: 1. switch or discontinuation of the medicine  2. disenrollment from the health plan   3. end of the study, or 4. a cardiovascular event or procedure or the diagnosis of diabetes. Patients were followed on 4 medications: Risperdal, Zyprexa, Seroquel, and Geodon (no Latuda, Saphis, and the sample size for Clozapine was deemed insufficient) and the dose of the index drug was categorized as high, medium or low, though the range for Ability was given as 2-30 mg as medium and there is no smaller pill available than 2mg.  The study included 138,523 patients.  These divided to Risperdal: 22,357;  Zyprexa: 17,428; Seroquel:58,807; Geodon:6,753 and Abilify: 32,890.  In every cohort, the most common diagnosis was Major Depression, followed by Bipolar Disorder.  5.5% of patients or less in each cohort had a diagnosis of schizophrenia.  All patients had medium level dosing of the medications.   The diagnosis of diabetes required that the diagnosis be made after 45 days on the medication or the patients were excluded with the idea that it was likely a pre-existing condition.
    Results showed that a new diagnosis of diabetes was more common then a cardiovascular event.  Rates of diagnosis of diabetes per 1000 person-years: these number are subsets of the Abilify group matched against the comparator drug (here, I'm sad to report that I'm a bit lost in the pages of data):
            Abilify vs. comparator 19.8 vs 17.8; 
            Zyprexa: 19.7 vs 20.5
            Seroquel: 18.9 vs. 18.1
            Geodon: 19.8 vs 20.6
While previous studies have shown a decrease risk of diabetes with Abilfy and Geodon, this study did not find any significant difference between any of the medications for the risk of diabetes.  The authors speculated that perhaps those at high risk for diabetes were preferentially prescribed Abilify because of the findings of past studies, and if such preferential prescribing occurred, then the patients on Abilify would at higher risk and the lower rate that would be expected could be washed out.  It was speculation and I wasn't totally sold.  We'll leave it that there was no significant difference for causing diabetes between the medications.  The authors note also that there may have been confounding issues, such as race or smoking status, and these were not available from the claims data.  Also death rates could not be examined, because insurance claims are not filed on dead people. (I would have been interested in this). 

When the low dose cutoff of Abilify was raised from 2mg to 5mg, then the comparison with Zyprexa showed a "small but statistically significant reduction of risk with olanzapine" (olanzapine is the generic for  Zyprexa) .  Huh?  The risk of diabetes with 2-30mg of Abilify (considered low for addressing psychosis) was the same as Zyprexa, but when a higher dose of Abilify 5-30 mg was compared, then Zyprexa (presumably at any indicated dose) was less likely to cause diabetes.

Overall, there were no difference between drugs for diabetes and the expected decrease in diabetes with Abilify and Geodon was not found.

The risk of heart failure and stroke rose from 28% to 58% with some of the comparator drugs compared with Abilify.  These are low frequency events and "may be considered small in magnitude and given the low incidence rates, are not likely to be clinically meaningful  on an individual practice level..."

Compared to Abilify, there was an increased risk of stroke, heart failure, and any cardiovascular event with Seroquel.
Compared to Abilify, there was an increased risk of stroke and any cardiovascular event with Zyprexa and Risperdal.
Compared to Abilify, there was no increased risk with Geodon on any measure.

Compared to Abilify, there was an increased risk of stroke for Zyprexa