Ethical Considerations in Medication-Free Research With Schizophrenia Patients: An Expert Interview With William T. Carpenter, Jr., MD

October 25, 2006

Editor's Note:

In a recent article in Schizophrenia Bulletin, John R. Bola, PhD, of the University of Southern California, Los Angeles, asserted that there is insufficient evidence to conclude that short-term postponement of medication in early episodes results in long-term harm, and suggested reconsidering the categorical prohibition against medication-free research.

On behalf of Medscape, Jessica E. Gould, BA, interviewed William T. Carpenter, Jr., MD, Editor-in-Chief, Schizophrenia Bulletin, and Professor of Psychiatry and Pharmacology, University of Maryland School of Medicine, Baltimore, Maryland, about the ethical issues surrounding medication-free research.

Medscape: Please tell me and the readers about the article "Medication-Free Research in Early Episode Schizophrenia: Evidence of Long-term Harm.[1]" What issues did it raise?

Dr. Carpenter: The Bola article in Schizophrenia Bulletin was a meta-analysis evaluating how subjects in research studies do when they are off medication. The topic had the potential of informing clinicians about how their patients might fare off of prescribed drugs. A number of studies have observed that patients who are doing very well are often doing very well off of medication. The original Bola hypothesis was that some patients would do better if they were off medication; the paper that we published in the Schizophrenia Bulletin, however, limited the interpretation to whether certain patients might do well off medication. We did not think there were any data that would answer the question of whether they were doing well because they were off medication. So, the article related to whether it's possible to do off-medication research in schizophrenia safely, and the data would support the proposition that at least a subgroup of patients can do well with periods off medication.

Medscape: In which patients and circumstances might it be a good idea to go off medication?

Dr. Carpenter: In general, the sort of considerations include first of all, how long they are going to be off medication. If they have a problem being off medication, how rapidly could medication be restored and is it likely to be effective? Will there be appropriate clinical observation?

Then, there is the careful use of exclusion criteria in which you try to exclude patients who are likely to experience risk or harm from the study and include patients only through careful selection. In other words, they are selected because they are expected to do okay and because they can give competent and voluntary consent to procedures and participation.

The predictors of doing well off medication are the same as just doing well:

  • Better background;

  • Better social relationships;

  • Better work; and

  • Better everything at the time they became ill.

The ones who have done well without drugs are always those who are drawn from that better prognostic end of the schizophrenia spectrum.

It is really a combination of excluding those patients who would have the greatest risk and then maximizing the safety routines.

Medscape:How do you enhance safety?

Dr. Carpenter: The first step toward safety, again, is excluding people at greatest risk. The second step would be clinical observation to spot early signs of a person doing worse and then, if necessary, intervening. You might consider only doing off-medication research with inpatients who you observe continuously. If you're doing off-medication research with outpatients, you'd be more likely to see people once a week, make sure that they have phone numbers of people who they can call anytime day or night, and have a contact person if somebody doesn't show for an appointment. There would be a routine so that if there is any indication of exacerbation, the patient would quickly be remedicated.

Medscape: Could you give me and the readers a sense of some of the ethical considerations that are relevant to medication-free studies of patients with schizophrenia?

Dr. Carpenter: The ethics and principles for medication-free research would be the same as that which guides ethical oversight of any human subject research.

If research is deemed justified by the scientific merit, you must ensure that it is also justified by having adequate safeguards to minimize risk. That risk should be comparable in proportion to the importance of the experiment.

Medscape: Tell us about the issue of biological toxicity.

Dr. Carpenter: There was a hypothesis that Richard Wyatt put forward some years ago that psychosis might, in and of itself, be neurotoxic.[2,3] If true, the implication would be that the more you experience psychosis, the worse it is for your brain tissue. He was very clear that this was a speculation. He also was clear -- he and I had companion articles in Schizophrenia Bulletin some years ago[4,5] -- that not only was it a speculative hypothesis, but that he did not believe it was actually supported by postmortem examination or empirical evidence. Furthermore, he believed that it was not relevant to patients selected for off-medication research, which is time-limited and done under clinical controls. His concern was for persons with schizophrenia who might have prolonged periods of psychosis even with effective intervention. I think all clinicians would share this concern.

The issue was raised in a legal journal and probably other places as an unacceptable risk for medication-free research.[6] Wyatt[5] published the Schizophrenia Bulletin article to try to make it clear that the psychosis-is-neurotoxic hypothesis was speculative and that he did not consider off-medication research to be relevant, even if it were a correct hypothesis.

Medscape: What are the benefits of continuing off-medication research?

Dr. Carpenter: There are many interesting hypotheses that require off-medication study designs. One example from a few years ago is the hypothesis that excessive dopamine was causal of psychosis in schizophrenia. Technology developments enabled a direct test of this hypothesis with positron emission tomography (PET) assessment of the quantity of dopamine released following an amphetamine stimulation. Data would not be interpretable in subjects receiving dopamine antagonists. Attempting the study with medicated patients may be judged unethical because of questionable scientific merit. There are numerous study questions in which medicated subjects seriously weaken scientific merit, and testing an efficacy hypothesis for an experimental treatment is best done with a placebo comparison.

Another benefit is that if there's any chance that the treatment doesn't work or may be harmful, you expose fewer people in the study than if you compared an experimental drug against a standard antipsychotic drug. So it not only costs more, but if there were risks associated with it, you'll expose a lot more people to that risk.

There is a great deal of interest in how patterns of dopamine signaling relate to learning, reinforcement, reward, motivation, and basic cognitive mechanisms.

Many study types, from clinical trials to assessment of basic mechanisms, are designed to address these issues. If you look for subtle abnormalities in the dopamine signaling processes in patients who are on antipsychotic drugs, a lot of scientists would say, "Don't fund that project." If they are going to be on drugs that are blocking dopamine, don't ask questions about dopamine.

Medscape: What are the drawbacks of continuing off-medication research?

Dr. Carpenter: Many institutional review boards have a strongly negative approach toward off-medication research in schizophrenia. The cost of doing business has gotten extraordinarily high without there being any evidence that people are actually being harmed by participating in off-medication studies. People often react to an off-medication study as though all treatment is withdrawn, but other elements of treatment are generally enhanced and clinical "out" procedures facilitate rapid antipsychotic drug intervention. If you think that all treatment is stopped and that the result is severe relapses, then the ethical objections are evident. This ill-informed view can be very influential in a review committee.

Medscape: Given those fears, how has the psychiatric community responded to the Bola article?

Dr. Carpenter: The response from the field has been interesting. The New York Times article entitled "Revisiting Schizophrenia: Are Drugs Always Needed?" published on March 21, 2006, raised the possibility that it would be best to treat some patients without medication; that view alarmed many people.[7] Despite the sensational title, the article was focused on the question of whether some patients with schizophrenia could participate in off-medication research with reasonable risk and benefit considerations.

We invited a series of commentaries from leaders in the field to accompany the Bola article. Two commentaries of interest on the ethical questions were prepared by a leading bioethicist named Jonathan Moreno, PhD (Professor of Biomedical Ethics, University of Virginia, Charlottesville; Director, Center for Biomedical Ethics, and Senior Fellow Center for American Progress, Washington, DC),[8] and from a leading advocate for the mentally ill named James McNulty (President, The National Alliance for the Mentally Ill (NAMI) National Board of Directors, Arlington, Virginia).[9] Both of those commentaries point to potential problems in medication-free research. They also considered circumstances in which such research designs could be conducted ethically. This was very reassuring coming from 2 people known to be concerned with mental illness research ethics and widely respected as carefully reasoned critics. I think both of them would line up behind a set of assertions, such as:

  • To use patient participants who, in clinical care, would be recommended to be on their medication continuously for a research proposal that would merit off-medication, requires very good justification;

  • Subjects must be selected carefully;

  • Clinical care safeguards should be built;

  • It must be clear to the individuals that they are free to decide whether or not they want to participate; and

  • Participants must be able to make a voluntary and competent judgment in regard to participation.

I think that both Moreno and McNulty would emphasize the need to very carefully apply these ethical standards, and they would probably note that the field has not always been sufficiently careful in this regard.

I also received a call from a person at National Institute of Mental Health (NIMH) who was familiar with an earlier version of the Bola manuscript, which interpreted the data as some patients doing well because they were off medication. They reminded me that I had reviewed the paper and recommended rejection. So why was I now publishing it? This person and others were concerned with the way The New York Times article moved the discussion from the manuscript's focus on research ethics to whether some people have their best shot at recovery when being treated without drugs. As I mentioned before, Bola's data were not suitable for testing that therapeutic hypothesis. For the patients who have done well without medication, it is simply not known how they would have done with medication. There is, of course, much data that a group of schizophrenia patients randomly assigned to placebo will be more symptomatic than the comparison patients randomly assigned to antipsychotic drug treatment. What we don't know is whether, within those groups, there is a minority of individuals who would have done better if they had not received medication. If there is, we do not know how to identify who they are except for the fact that the best prognostic group will have the best course of illness regardless of treatment approach.

A number of people have asked me about the quote by Jeff Lieberman, MD (Professor, Department of Psychiatry, University of North Carolina, Chapel Hill, North Carolina), in the Times article; he was quoted as saying that not treating patients with medication "is a huge mistake.[7]" Dr. Lieberman is one of the most respected experts on the treatment of persons with schizophrenia. He sent me an email explaining that he had not read the article in Schizophrenia Bulletin at the time when he was interviewed by Mr. Carey of The New York Times. His remarks were a response to the reporter's interest in the advantages of treatment without medication. Dr. Lieberman had no problem with the article in Schizophrenia Bulletin, and shares the belief that off-medication research can be designed with adequate attention to safety.

Medscape: How might this picture change if research concerning the use of atypical antipsychotic medications were introduced into the mix?

Dr. Carpenter: The basic issues would be the same. The new drugs work via the same therapeutic mechanism, have very similar effects to old drugs, and differ in side effects. The exception is clozapine, which has modestly superior efficacy in patients not responding well to the other antipsychotic drugs. Clozapine is complicated when considering medication-free research because clozapine withdrawal has been associated with symptom exacerbation. So in terms of the safety issues, if you were withdrawing people from clozapine, you would have a greater risk of precipitating an exacerbation in the withdrawal period.

Medscape: Thinking ahead, what kind of research needs to happen to build on the Bola study?

Dr. Carpenter: There is extensive nonadherence to medication. Studies designed to help determine how to best treat patients who do not take drugs continuously would be very important.

A second issue is associated with present efforts to advance antipsychotic drug intervention to a much earlier stage of the first psychotic episode. We will necessarily end up treating more young people who may not have a recurrent illness and their diagnosis will be less clear. The risk for the drugs will not be justified if they do not have persevering psychotic symptoms or a relapsing psychotic illness. It will be important to have ways to take these young people off medication to determine whether they need it to prevent future episodes.

Medscape: Thank you very much for your time and your insight.

Dr. Carpenter: My pleasure. Thank you for the opportunity.

This interview is published in collaboration with NARSAD: The Mental Health Research Association, and is supported by an educational grant from Pfizer.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.