How the FDA Got the Markingson Case Wrong
During the years after Dan Markingson committed suicide in an industry-sponsored drug study at the University of Minnesota, his mother looked for justice anywhere she could find it. Mary Weiss wrote countless letters and e-mails; she filed a lawsuit against the university and the study sponsor; along with her friend, Mike Howard, she sent out complaints to any agency or government body that might possibly take an interest, no matter how tenuous its connection to the protection of research subjects. Most of these efforts went nowhere.
That is no surprise; there is simply very little protection for human subjects in privately-sponsored clinical trials. More surprising is the way that Ms. Weiss’ desperate efforts at justice are now being used against her. As calls to investigate her son’s suicide have escalated, University of Minnesota officials have repeatedly refused to look into the case, arguing that the university has already been exonerated by other agencies, especially the FDA.
It is true that the FDA investigated the case, in response to a complaint filed by Mike Howard. But the inspection report, issued in 2005, hardly qualifies as an exoneration. In fact, the report is a very puzzling document: occasionally misleading, remarkably uncritical, and often simply baffling.
When I was investigating Markingson’s suicide for an article in Mother Jones, FDA officials refused to answer my questions about the report. But if University of Minnesota officials plan to use the FDA report as a shield against criticism, the report is worth a very careful look.
First, though, a brief review of the case. Dan Markingson was a 26 year-old man from St. Paul and a recent graduate of the University of Michigan who began showing signs of mental illness in the summer of 2003. His thoughts became paranoid and delusional, and he became convinced that he was part of a vast cult, which was calling on him to murder people, including his mother. On November 12, 2003, Markingson was taken to Fairview Hospital in Minneapolis, where he was seen by Dr. Stephen Olson, a psychiatrist at the University of Minnesota. Olson recommended involuntary commitment, and a court agreed. Later, despite objections by Ms. Weiss, Olson recruited Markingson into a clinical trial of antipsychotic drugs.
The clinical trial, which was known as the CAFÉ study (an acronym for Comparison of Atypicals in First-Episode Schizophrenia), was sponsored by AstraZeneca, the manufacturer of Seroquel (quetiapine), and managed by Quintiles, a Contract Research Organization. The CAFÉ study compared three different atypical antipsychotic drugs, each of which was already being marketed: Zyprexa (olanzapine), Risperdal (risperidone) and Seroquel (quetiapine.) The University of Minnesota was one of 26 sites for the double-blinded trial, which was to last one year.
After Markingson was enrolled in the CAFÉ study, he spent about two weeks in Fairview Hospital before being discharged against his mother’s wishes to a halfway house. Over the next five months Ms. Weiss repeatedly expressed her concerns about her son’s medical condition, especially his increasing agitation and rage. Her warnings were largely ignored. Finally, in desperation, she warned the study coordinator that her son might kill himself.
On May 7, 2004, Markingson mutilated himself with a box cutter so violently that he nearly decapitated himself. His body was found by halfway house workers in the shower, along with a suicide note that said, “I went through this experience smiling.” Markingson had been taking Seroquel.
Was Markingson in any position to give informed consent for this study? After all, he had just been involuntarily committed – not just because he was psychotic, but because he was potentially homicidal. The Dakota County court had issued a legal order requiring Markingson to follow the treatment recommendations of his psychiatrist. If Markingson refused those recommendations, he would be confined to a mental institution. So when his psychiatrist recommended a clinical trial, it is no surprise that Markingson consented. But was that consent valid?
Yes, it was, according to the FDA inspector, who does not even consider the possibility that Markingson’s consent might have been influenced by the threat of involuntary commitment. One might have thought that a commitment order would violate the Common Rule, which requires researchers to minimize the “the possibility of coercion or undue influence” when recruiting subjects for research. Certainly the Minnesota legislature thought so in 2009, when it banned researchers from recruiting most psychiatric patients into a drug trial if they are under a commitment order.
But the FDA inspector concluded otherwise, writing, “There was nothing different about this subject than the others enrolled to indicate that he couldn’t provide voluntary informed consent” (p. 9).
So much for coercion. What about Markingson’s decision-making capacity? He was, after all, being recruited into a trial designed specifically for patients experiencing an acute psychotic episode. The FDA inspector summarily dismisses the suggestion that his psychosis might be a problem, noting that the study coordinator, Jeanne Kenney, had certified Markingson as competent to consent. In fact, the inspector implies that such an assessment was not even ethically necessary, pointing out that it was “neither required nor mentioned in the study protocol” (p. 9).
What the FDA inspector failed to note is that during the period leading up to his enrollment, Markingson had been repeatedly judged incapable of consenting to neuroleptic drugs. On November 14, 2003 Dr. Olson signed a commitment document stating that Markingson “lacks the ability to make decisions regarding such treatment.” On November 17, a pre-petition screening team recommended commitment, noting Markingson’s bizarre beliefs and his refusal to acknowledge his mental illness. On November 19, a court-appointed clinical psychologist confirmed those assessments, writing that Markingson “is believed not to have the capacity to make decisions regarding neuroleptic medication.” Yet on November 21, when Markingson was asked to consent to the CAFÉ study, this assessment of his mental state was reversed and he was judged competent.
Is it possible that a psychotic patient’s mental capacity could have improved so dramatically in two days? Yes. But there are some good reasons to doubt this happened.
First, the final competence assessment was made not by an independent party, but by the study coordinator for the CAFÉ study, Jean Kenney, whose job it was to recruit subjects for the study. This is hardly an impartial, disinterested assessment. Second, Kenney was a social worker, not a psychiatrist or psychologist trained to make competence assessments. Third, even after Mr. Markingson had been judged competent to consent to the CAFÉ study, his involuntary commitment order was not lifted. This suggests that his mental state had not changed quite so dramatically. And finally, one of the most persistent features of Markingson’s psychosis was a lack of insight into his condition. Markingson did not believe he had a mental illness. How could he be competent to consent to a study comparing treatments for his mental illness when he would not even acknowledge that he was mentally ill?
What the FDA inspection does address, indirectly but in some detail, is another controversy at the heart of the case: should Olson have dropped Markingson from the CAFE study? The answer to this question might seem obvious. It is hard to imagine an outcome much worse than the one that resulted from leaving him in. And nobody disputes that Mary Weiss believed her son was doing very poorly in the CAFÉ study, and that she warned the CAFÉ study team that she thought he might kill himself. Yet the obvious question goes unasked. Why didn’t Olson take these warnings seriously?
Instead, the FDA inspector goes to great lengths to refute the allegation that Markingson’s condition worsened in the study. Not an easy job to do for a subject who has stabbed himself to death, but there are notes to review, charts to examine, and clinical rating scales to evaluate. That record shows a subject who, by the most generous interpretation possible, failed to improve over a period of five and a half months; whose own psychiatrist acknowledged so little improvement that his stay of commitment could not be lifted; whose mother was convinced that his condition was spiraling dangerously downward; and whose life ended in a grisly suicide. The FDA’s conclusion? The record “does not appear to indicate a significant decline or deterioration” (page 10).
The FDA inspector dismisses the suggestion that Olson violated the court order for Markingson to get treatment by recruiting him into the CAFÉ study. Part of her rationale appears to be that Markingson’s caseworker, David Pettit, approved of this enrollment. The FDA inspector did not actually interview Pettit to confirm this, but she cites a chart entry where Pettit apparently expressed support.
What she fails to note is that Pettit was not even assigned to the case until three days after Markingson signed the consent form for the CAFÉ study. What is more, there is no evidence that the judge who issued the commitment order ever even knew about the CAFÉ study. In his deposition, Olson admitted that he did not inform the court that he had recruited Markingson into a clinical trial (pp. 78-79.)
Of crucial importance here is whether Markingson’s care was compromised by enrollment in the CAFÉ study. The FDA inspector appears to believe not. Summarizing Olson’s remarks uncritically, she cites the advantages of being in a clinical trial rather than getting standard therapy (p. 8). She does not mention the disadvantages.
If Markingson had not been enrolled in a double-blinded study, for example, both he and Olson would have known what drug he was being given. Olson would have had the freedom to change Markingson to another drug when it was clear that he was not improving. Also, he would have been able to use drug combinations not permitted by the protocol.
Almost as alarming as what the inspection report includes is what it leaves out. Most trials of antipsychotic drugs exclude subjects who are at risk of suicide or violence, to minimize the possibility that they will harm themselves or others in the trial. The CAFÉ study excluded subjects at risk of suicide, but for reasons that remain unclear, it did not exclude subjects at risk of violence, and the University of Minnesota IRB did not object. If the exclusion criteria had excluded potentially violent subjects, Markingson could not have been enrolled in the trial; he had been involuntarily committed precisely because he was potentially homicidal. In fact, at the time of his enrollment, there seems to have been broad agreement that he was at high risk of acting out his delusions.
In the years since the FDA report was issued, more alarming revelations about the trial have emerged. In May 2008, Paul Tosto and Jeremy Olson published a series of articles in the St. Paul Pioneer Press that revealed, among other things, financial conflicts of interest involving Olson and his co-investigator, Dr. Charles Schulz, and as well as the development of a locked unit for severely psychotic patients in Fairview Hospital, where patients are apparently targeted for recruitment. In 2010, AstraZeneca agreed to pay $520 million to settle two federal investigations and two whistleblower lawsuits alleging that it had marketed Seroquel illegally and concealed its health risks. One of those investigations involved physicians who had conducted clinical trials. As I reported in Mother Jones in September 2010, the documents unsealed in that case implicated Dr. Charles Schulz, the Chair of the Department of Psychiatry and a co-investigator for the CAFÉ study. Last month, City Pages journalist Andy Mannix reported even more evidence that Schulz was involved in manipulating trial results for AstraZeneca.
The response of University of Minnesota officials to these reports has been remarkably consistent. The suicide was tragic, they say, but the university has already been cleared of any blame. That uniform message has been issued by the Office of the General Counsel, the Vice-President for Research, the Communications Office, and the Dean of the Medical School. In December 2010, eight faculty members associated with the Center for Bioethics at the University of Minnesota wrote a public letter to the Board of Regents requesting an independent investigation. That letter was followed by a second letter from the Faculty for the Renewal of Public Education. On February 7, 2011, the Board of Regents refused that request. It cited, among other reasons, the fact that the FDA had already reviewed the case.
Carl Elliott, a Hastings Center Fellow, is a professor at the Center for Bioethics at the University of Minnesota. His most recent book is White Coat, Black Hat: Adventures on the Dark Side of Medicine.
Published on: March 2, 2011
Published in: Clinical Trials and Human Subjects Research
Receive Forum Updates
Hastings Bioethics Forum publishes commentaries from a range of perspectives on timely issues in bioethics.Subscribe