On the front page of the Washington Post last week was a headline geared to spark the ire of those who have a congenital animus toward pharmaceutical companies--"Comparison of Schizophrenia Drugs Often Favors Firm Funding Study".
Resident Scholar Sally Satel
The study, published in the February issue of the American Journal of Psychiatry by German and American psychiatrists, reviewed 30 drug trials of second-generation antipsychotics, also called atypicals, published between 1966 and 2004. Five drug companies were responsible for underwriting these 30 trials.
Aside from its headline, the Post article was well done and revealed important lessons from the study. For instance:
- From a clinical standpoint today, it doesn't matter much which
drug wins a horse race.
- The company studies provide useful information, not pabulum.
- Most important, we need more research.
Let's look at these points one by one.
In the real world of practice trial results are less important than you think. In modern psychiatry, there are no first-line drugs. Years ago, by contrast, lithium was the pharmaceutical of choice for maintenance treatment of the manic symptoms of bipolar illness. Today, the FDA has approved several mood–stabilizers (anticonvulsants) for treatment and suppression of mania, not to mention off–label drugs. Take posttraumatic stress disorder. Today, only two SSRI–type drugs are FDA approved for PTSD but they are not dazzlingly effective. In fact, because the condition varies so much from patient to patient, other medications that are not approved for the condition (e.g., low-dose antipsychotics, anticonvulsants and benzodiazepines) can be very helpful in certain individuals.
Why do doctors struggle so much to find effective treatments? The state of the art in prescribing effective antipsychotic drugs is still pretty poor and, as a result, doctors are often forced to take a trial and error approach. When starting a depressed patient on an antidepressant or an antipsychotic, doctors typically choose the drug they are most skilled in using. This is a perfectly reasonable choice, given our rather slim knowledge about how to tailor medications for each patient.
The largest and most recent government–funded studies tend to be enormously humbling for physicians and companies alike. Data from the new STAR*D Study (Sequenced Treatment Alternatives to Relieve Depression) show that only 25-33% of subjects achieved full remission with antidepressants. (Granted, a higher percentage of patients had partial relief, and sometimes that is enough to improve daily functioning significantly. Still, higher rates of full-remission are surely desirable.)
Another study, CATIE (Clinical Antipsychotic Trials of Intervention Effectiveness Study) found that three-fourths of all subjects asked to stop their antipsychotic, either because it was not helping or because side-effects were unacceptable to them. The CATIE study lasted an unprecedented duration (18 months) and enrolled an unprecedented number of subjects (nearly 1500), thus permitting a fuller picture than most other clinical trials.
For the record, psychiatrists do have some guideposts in selecting drugs. In my own practice, I rely on four rules of thumb for using (or avoiding) a specific drug. I use a particular treatment (1) if the patient had a good response to it during a prior treatment episode, and (2) if a close relative benefited from that drug (aka: the pharmacogenetic rationale).
I avoid treatments if (1) the patient says he would find a side-effect totally unacceptable, such as weight gain. If the medication poses a sizeable risk of causing a medical complication, such as worsening hypotension, diabetes or high cholesterol, I still might use it but watch the patient very carefully. I also avoid treatments if (2) a drug is very expensive--and the patient has to pay out of pocket. In that case, I use a first generation, generic antipsychotic and follow for signs of tardive dyskinesia, a movement disorder that can emerge with long-term use.
Company funded studies still provide useful information. It is no mystery (or conspiracy) why studies comparing the same drug can come out differently. In short, they are conducted using different designs, different statistical tests, and/or different outcome measures. Anyone can read the published studies and see for themselves how they differ. There is transparency here, but the psychiatrist must take time to read the entire study, not just the abstract.
Some factors that can favor one or another drug are: the doses used (e.g., a low-dose of a clearly superior drug will make it seem lame when compared to a standard dose of an inferior one), the duration of the study (i.e., trials may be only four weeks long but different drugs may have different times to maximum effectiveness), and outcome measures (i.e., reduction of psychosis but not behavioral withdrawal, or vice-versa).
One can be forgiven for wondering whether some trials have been conveniently engineered to make the sponsor look better--for example, when a low dose of a competitor's drug is used or when the researcher doesn't seem to look too hard for side effects or doesn't report all of them. On the other hand, often researchers truly don't know to look for a novel side effect or do not know what the optimal duration of study should be. Why? Because we need to know more!
We need more research. This is the main conclusion physicians, patients, and the media should take away from February's American Journal of Psychiatry study. The main limitation of so many clinical trails is that the data they yield don't help the clinician with a specific patient. The kinds of information psychiatrists need are how to home in on a medication that will be best for a particular patient. Switching around medications and dealing with bad side-effects take large tolls on patients. Not only do they cause suffering or vast inconvenience, these adverse events problems can cause poor compliance or outright refusal to try any other drug. It is demoralizing to a patient to continue to be sick and to go from one drug to the next; he starts to fear he will never get better. And, of course, it is a problem, though a routine one, for the physician.
Legitimate concerns about clinical trials abound. For instance, how do the findings of clinical trials compare to patients that physicians see everyday (who are, generally speaking, more likely to have other mental and physical problems)? Clinical trials typically enroll subjects who have classic symptom constellations and are medically relatively healthy. As well, such subjects are able to give informed consent; very psychotic patients often cannot. Also, do differences in drugs justify differences in cost and how do those differences vary across patient subtypes? And, of vast importance, what drugs really help the patient get better, not in just alleviating symptoms, but in terms of day-to-day functioning?
Pharmaceutical companies may not be interested in these questions. Either the National Institutes of Health or some other government or non-profit entity could conduct these trials. Currently, only NIH can sponsor studies large enough to involve a sufficiently large number of patients that the results yield more clinically-relevant information. That's a major weakness of company-sponsored trials, no matter how well done. An additional virtue of non-industry funding is that the specter of conflict of interest is erased.
Schizophrenia is one of the most mysterious and vexing diseases in psychiatry. The most afflicted often give up their lives to it. They are desolately lonely and tormented by hallucinations and paranoia. Neuroscientists have made tangible progress but the etiology or etiologies of the disease (the most popular theory is that gene mutations lead to abnormal development of neural circuits, particularly in the context of external stress, but a number of respected researchers have even suggested a viral cause) remain murky. This makes it difficult for physiology to inform drug development.
There is no question that some patients have Lazarus-like recoveries when the right medication is found. The search, though, can be a time-consuming pharmaceutical odyssey. Worse, however, is that not enough patients get excellent results. Clearly, we need more evidence about the medications we have now and even better drugs in the future.
We shouldn't become cynical about the companies pursuing these cures. Competition between competing therapies broadens our understanding of mental illness and can help researchers chart a course towards the next generation of treatments. In a sea of uncertainty, that's the best we can hope for.
Dr. Sally Satel is a resident scholar at AEI.
 Some psychiatrists consider clozapine (now off patent) to be a first line drug for schizophrenia. It has been clearly demonstrated to be superior to other atypical antipsychotics but is not very popular because of the necessary blood monitoring. Also, because it is generic, there is no pharmaceutical advertising behind it. See Tamminga, C. Practical Treatment Information for Schizophrenia American Journal of Psychiatry 163:563-565, 2006. Of interest, now lithium is making a comeback. A number of new studies have shown that it has a significant suicide-prevention effect, see Tondo L., Hennen J., Baldessarini R.J Lower suicide risk with long-term lithium treatment in major affective illness: a meta-analysis. Acta Psychiatrica Scandinavica, vol. 104, Number 3, September 2001, pp. 163-172(10); Cipriani A, Pretty H, Hawton K, Geddes JR. Lithium in the prevention of suicidal behavior and all-cause mortality in patients with mood disorders: a systematic review of randomized trials. Am J Psychiatry. 2005 Oct;162(10):1805-19; Kessing LV, Sondergard L, Kvist K, Andersen PK. Suicide risk in patients treated with lithium. Arch Gen Psychiatry. 2005 Aug;62(8):860-6;Muller-Oerlinghausen B, Felber W, Berghofer A, Lauterbach E, Ahrens B.The impact of lithium long-term medication on suicidal behavior and mortality of bipolar patients. Arch Suicide Res. 2005;9(3):307-19.
 Rush AJ, Trivedi MH, Wisniewski SR, Stewart JW, Nierenberg AA, Thase ME, Ritz L, Biggs MM, Warden D, Luther JF, Shores-Wilson K, Niederehe G, Fava M; STAR*D Study Team. Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression. N Engl J Med. 2006 Mar 23;354(12):1305-7;Rubinow DR Treatment strategies after SSRI failure--good news and bad news. N Engl J Med. 2006 Mar 23;354(12):1305-7.
 Lieberman JA, Stroup TS, McEvoy JP, et. al. Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Investigators. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia.N Engl J Med. 2005 Sep 22;353(12):1209-23.
 And, still, patients on first generation, or typical, antipsychotics were not followed long enough to track emergence of tardive dyskinesia, the major aversive side-effect from long term use.
 Other suggestions: perhaps the FDA could require a new medication be included in at least one large, long-term trial funded by the company and designed and conducted by an independent group at the expense of that company or jointly supported by NIH. Alternatively, one could have the FDA require that as part of Phase IV (post-FDA approval) that companies would be obligated to contribute to CATIE-like studies to see how effective medications are when used they way clinicians use them for the patients that they treat. Since these would be head-to-head studies, several companies would contribute and this would combat any bias that could be alleged when a single company pays.