Wednesday, September 19, 2007

An NIH Hearing About the Death of a Patient in a Gene Therapy Trial

We previously discussed, most recently here, the unfortunate and unexplained death of young patient after an injection of an experimental gene-therapy agent. A US National Institutes of Health (NIH) committee, the Recombinant DNA Advisory Committee, just conducted a hearing about the case, which received considerable media coverage. The coverage did illuminate a few more important details, although much about the case remains mysterious.

To summarize the case thus far, a young woman received a dose of an adeno-associated virus carrying a gene that was meant to have a local immune suppression effect, tgAAC94, made by Targeted Genetics, directly into her joint as part of a phase I study meant to assess the safety of the treatment. Soon after the treatment, she became ill, then very ill, and ultimately died in an intensive care unit. Whether the treatment had anything to do with her illness and death were initially unclear. Questions were raised about the conduct of the trial, including
  • whether the patient was lead to believe she might benefit from it, even though the study was meant only as an initial, Phase I, safety assessment;
  • whether she was influenced to participate because her rheumatologist was an investigator, and he administered the informed consent form, without disclosing he was paid to participate in the trial;
  • whether she found the consent form understandable;
  • what institutional review board (IRB) reviewed the trial; and
  • whether Targeted Genetics unduly delayed reporting the patient's condition to the US Food and Drug Administration (FDA)?

The key points from the hearing, as reported by the New York Times and Washington Post, were:

  • The patient died apparently from a massive histoplasmosis (fungal) infection, possibly complicated by major internal bleeding
  • The patient was also taking "Humira, a drug similar to Enbrel that has also been linked to infections, including histoplasmosis. She was also on two other immune-suppressing drugs." [Times]
  • The genetically engineered AAV had spread to her liver and spleen, although the President of Targeted Genetics said the the levels of the virus were "tiny"
  • The patient did not appear disabled or seriously ill from arthritis before entering the trial, her "husband, Rob, told the panel that his wife was only mildly affected by her arthritis and was not properly warned of the experiment's risks." [Post]
  • The panel was concerned "about the way his wife was signed up for the study. They noted that, although no regulation prohibits the practice, it can be ethically problematic when a patient's personal physician recruits the patient.... Patients routinely overestimate the chance that an experimental treatment will help them, and this 'therapeutic misconception' is even more likely when the recruiter is the patient's personal doctor...." Also, her "physician also signed her up immediately, rather than having her take the materials home to consider.... And there was no mention in the informed consent document that her doctor was being paid by Targeted Genetics for every patient he enlisted." [Post]

In my humble opinion, the main achievement of this meeting was affording a potentially important case more publicity. To me, the major questions remain: what relationship, if any, was there between the experimental therapy and the patient's illness and death; would changes in the conduct of experiments like this, particularly involving informed consent, lower the probability of future tragedies?

The larger question is whether companies and other organizations should be sponsoring and controlling research on human beings when the companies and organizations have financial (or ideological) stakes in the outcome of the studies?

Targeted Genetics employees seemed to be taking a persistently optimistic view about the role of their treatment in this case. Note the comments above of the President minimizing the importance of admittedly preliminary data showing that the AAV had spread beyond the joint into which it was injected. Furthermore, in a Seattle Times report, a company spokesperson also asserted that the virus was "unlikely" to have been related to the patient's death, and that the data so far was "very positive." Of course, when one's job and company's financial future depend on a particular interpretation of the science, one is likely to support that interpretation.

ADDENDUM (20 September, 2007) - The consent form for this study is apparently available on the web here. See first comment below. It appears to raise more questions than it answers. Stay tuned.

1 comment:

Anonymous said...

You mentioned the consent form - if you haven't seen it, it's posted here: http://www.circare.org/consents/consentform_13G01_20060508.pdf

As to your question, "would changes in the conduct of experiments like this, particularly involving informed consent, lower the probability of future tragedies?"

Perhaps it might, but it's more likely that sponsors and the regulator could decrease the likelihood of future tragedies by selecting site investigators who are qualified experts. Like the regulations say.