In clinical research, the road to hell is paved with good intentions

I’ve written a lot about Stanislaw Burzynski and what I consider to be his unethical use and abuse of institutional review boards and clinical trials. Before that, I used to regularly write about Mark and David Geier and their unethical use and abuse of IRBs and clinical trials. In both cases, I lamented how they could use their own IRBs stacked with their own cronies to rubberstamp scientifically and ethically dubious studies. Mark and David Geier got away with it for years. Stanislaw Burzynski got away with it for decades and, apparently, is still getting away with it to some extent. (His IRB is chaired by an old Baylor crony of his from the 1970s, and he has been cited for numerous problems with his IRBs.) I’d like to contrast how their unethical research, in which the Geiers subjected autistic children to chemical castration with Lupron to decrease testosterone levels and allegedly make mercury easier to chelate (to them mercury was bound by testosterone, something that doesn’t happen under physiological conditions but requires organic solvents) and Stanislaw Burzynski administered an unproven cancer chemotherapy (antineoplastons) to hundreds of patients over the years and charged them for it, compares to a recent case in the news.

The case has been mentioned by PZ Myers. It happened at t involves the same sort of tumors that Stanislaw Burzynski claims to be able to cure, namely brain tumors. It happened at the University of California Davis (UC Davis) and involved two very prominent neurosurgeons there, a former head of the department Dr. J. Paul Muizelaar and Dr. Rudolph J. Schrot, who were found to have violated university’s faculty code of conduct with their experimental work. When you read this part of the story, you’ll shiver. At least, I did:

The procedure in question involved three patients described in documents only as Patients 1, 2 and 3 – two middle-aged women and one man who had a common enemy: glioblastoma.

For anyone unlucky enough to be diagnosed with the highly malignant brain tumor, the prognosis is dismal. Median survival is less than 15 months.

Muizelaar and Schrot called their novel approach “probiotic intracranial therapy,” or the introduction of live bowel bacteria, Enterobacter aerogenes, directly into their patients’ brains or bone flaps. The doctors theorized that an infection might stimulate the patients’ immune systems and prolong their lives.

The first patient lived about 5 1/2 weeks. The second survived another year, an outcome that buoyed the doctors and seemed to bolster their theory, they said.

The institutional trouble began in March 2011, when a newly diagnosed third patient developed sepsis, became unresponsive and died two weeks after being deliberately infected. The university’s first internal investigation soon followed.

As a cancer scientist and surgeon myself, my first thought here was: Who on earth thought this was a good idea? Let me put it this way. What Muizelaar and Schrot were doing was intentionally causing bacterial meningitis. In cancer patients! You know, patients whose immune systems are often compromised to begin with! To be honest, on the surface, Burzynski’s antineoplastons have more biological plausibility than this. Thirty years ago, I could say that they were at least drugs and therefore might have had some activity against cancer.

The concept that inducing an inflammatory reaction might help “prime” the immune system to counter cancer is not a new idea. In general, however, it’s not a good idea to use bacteria that could cause potentially life-threatening infections. The CNS is generally viewed to be an immune “privileged” site, meaning that the immune system doesn’t function the same way there as it does in the rest of the body. Immune privilege was first observed when it was noted that tissue grafts that would be rapidly rejected if grafted to the skin or other sites would survive a long time when grafted into the CNS. The original thought was that this meant the CNS was somehow cut off from the immune system by the blood-brain barrier, its lack of draining lymphatics, and the apparent inability of cells known as microglia to function as immune cells. However, recently this concept has been reevaluated in light of new data that show that the CNS is neither isolated nor passive in its interactions with the immune system, but that doesn’t change the observation that the immune system functions somewhat differently in the CNS. In any case, gut bacteria are in general not expected to be in the CNS, and bacterial meningitis can be really nasty. It’s also not that difficult to induce meningitis. Remember the outbreak of fungal meningitis from contaminated steroids injections last year?

Apparently, Muizelaar based his idea on an apparent observation in among neurosurgeons that postoperative infection was associated with better survival in patients with glioblastoma. However, the data supporting this particular association is, to put it kindly, mighty thin, a retrospective study. What Muizelaar apparently did was to place an enteric (gut) bacteria, Enterobacter aerogenes, into the open wound and bone flap after surgery in order to produce a wound infection. E. aerogenes tends not to be pathogenic, except in patients who have compromised immunity, such as the elderly, infants, and, yes, cancer patients.

It’s instructive to read some of the PDFs included with the article, such as the report of the UC Davis investigation of Dr. Muizelaar and of Dr. Schrot. It reveals a bit of the complexity of doing investigations like this and, more importantly, why Drs. Muizelaar and Schrot went too far and how they gamed the system. He started with a consultation to the Bioethics Consultation Committee, which, after a lengthy debate, concludes that the practice “may be seen as consistent with the customs and practices of medicine.” However, the committee thought that the procedure required more oversight than it could provide, and suggested that the protocol should be reviewed by the IRB. So Dr. Schrot submitted a full application to the IRB in which he admitted that further preclinical work was needed but argued that in this patient the benefit/risk ratio was favorable. On what evidence? Who knows? I don’t have the IRB application. Dr. Schrot also discussed the procedure with the Center for Biologics Evaluation and Research at the FDA and was told that “animal studies will be necessary prior to entering into the clinic with your proposed therapy.” In addition, he was told that “these animal studies must show not only safety, but also establish a reasonable proof of concept in order for this investigational therapy to be introduced into patients.” In other words, the FDA told them to do the preliminary work first.

In the meantime, Drs. Muizelaar and Schrot submitted an NIH grant application for the animal studies. It was not well received by the study section, which characterized it as having “low translational impact with seriously flawed methodology and statistical analysis.” The summary statement concluded, “There is absolutely no feasibility evidence in support of this project” and comments that it has a “low translational impact since it is very unlikely that any IRB or the FDA will allow the introduction of Enterobacter aerogenes directly into patients.”

Here’s where things get interesting. The FDA had told Dr. Muizelaar that he and Dr. Schrot could proceed with their “innovative treatment” if the bacteria were available locally. No doubt this was because the FDA is a federal agency and doesn’t have jurisdiction if everything happens within one state. (It’s the same issue with Stanislaw Burzynski, actually.) If it was not available locally and had to go through the investigational new drug (IND) process, then more animal work was required, as the CBER stated. It turns out that a graduate student had some Enterobacter aerogenes from ATCC, and that’s what was used on these patients.

What follows then is wrangling between Dr. Muizelaar and the university. Dr. Muizelaar went to the Chief Medical Officer Al Siefkin and told him that the IRB had told Dr. Schot that the proposed procedure wouldn’t be considered research. In any case, both gave the impression that they were “shopping” for approval, and, although probably not lying, being a bit selective in what they told various parties involved in the approval process. As the news report notes:

Among the university’s findings, detailed in reports, email correspondence, interviews and regulatory documents:

• A former top administrator said he believed the doctors went “shopping for approval” and were “doing all they could to get around” the formal approval process. A physician who served on the research oversight board said the surgeons seemed to be “gaming” the system to bypass any institutional rejection or skepticism of their plans.
• Muizelaar and Schrot did not complete a study on rats before they began treating the three human patients, despite an FDA directive in 2008 that “animal studies will be necessary prior to entering into the clinic with your proposed therapy,” according to an email to Schrot from an FDA official. When asked by a compliance investigator why animal trials were not done first, Schrot allegedly responded that such testing would take “10 years … his entire career,” one internal review states. The investigator found Schrot’s “eagerness to proceed” to be concerning and his actions “reckless.”
• One university investigator questioned whether the doctors ever would have been able to get federal approval of their concept. According to one report, Schrot’s grant application to the National Institutes of Health to study the bacterial treatment was rejected, with NIH commenting in 2009 that “(t)his is a very poorly developed scientific proposal that lacks feasibility.” The NIH thought it unlikely that university or federal regulators would allow the introduction of live bacteria directly into patients, the report states.

Ultimately, the procedures were done on three patients. Particularly disturbing is the consent forms that were drawn up, which basically said that there was no evidence that this would work but that the surgeons wanted to do it anyway. While technically that is accurate, one has to wonder how much the patients might have felt coercion and how free they felt to say no. After all, they all had horrible diseases with a median survival of less than 15 months expected. As this editorial points out, patients also weren’t told everything:

Despite the absence of any FDA approval or IRB review, the neurosurgeons proposed their treatment, to Terri Bradley. The consent form she and her daughter signed appears straightforward. It made clear that the infection the surgeons were deliberately introducing into Bradley’s brain “may be totally ineffective.” It then laid out the many serious risks, including “further neurological deterioration … paralysis, coma or death.” The women were told the procedure did not have FDA approval.
Faced with the alternatives presented and the near certainty of death, not surprisingly, the Bradleys signed.

Here’s what the Bradleys weren’t told. They were not told that the FDA hadn’t just failed to approve the procedure their physicians were pushing, federal officials had told the doctors specifically not to perform the procedure until they had conducted animal studies. The Bradleys were not told that the doctors had bypassed the university’s IRB process or that the IRB had previously told the doctors not to perform their “experimental” procedures on human patients.

In other words, the “informed consent” provided was every bit as much misinformed consent as what antivaccinationists advocate.

The end result was one patient living less than six weeks, another surviving two weeks, and the second patient, the “success story” living over a year. However, the family of that patient, Terri Bradley, didn’t view it as a “success.” Indeed, Mrs. Bradley required at least one operation to “clean out” her infected head wound and directly administer antibiotics. Indeed, her daughters report that the infected wound had begun to smell bad. True, this patient did apparently have some tumor shrinkage. As I read about the tumor shrinkage, thoughts of Stanislaw Burzynski came to mind, as did the observation that three patients are too few to say anything and that one sees longer-than-expected survival in cancer patients with homeopathy (i.e., treatment with nothing but water). In any case, despite the doctors’ thinking that Ms. Bradley had benefited, she did not:

“Patient 2 did not experience significant functional benefit,” investigators concluded, “and upon her death … recurrences of the tumors and chronic infection of the brain were identified.”

Shockingly, even after this almost certainly treatment-related death, Drs. Muizelaar and Schrott convened an ad hoc ethics committee and presented them with a proposal to proceed with five additional patients. According to this report, the ad hoc committee concurred with the proposal and recommended continuing clinical activities with the understanding that they would develop an IRB-approved protocol as soon as possible.

In the end, this case demonstrates just how much oversight is needed on human subjects research and just how disturbing what “brave maverick doctors” like Stanislaw Burzynski, Mark Geier, and, yes, Drs. Muizelaar and Schrot do can be. The problem, of course, is that the line between research and trying something “innovative” (or truly innovative without the scare quotes) is not always clear, nor is the line between research and making a “last ditch” effort to save a patient’s life. It also speaks to the inherent conflict between the desire of a doctor to save lives and the need to do rigorous research in order to determine whether what the doctor does actually does save lives. As a physician, it’s very easy to develop a tunnel vision and believe that your latest idea is something that will work, that will save lives, if only those unimaginative clods would see what you see. Also, in the service of the goal of saving lives, it becomes progressively easier to justify cutting ethical corners, as Drs. Muizelaar and Schrot appear to have done. It’s very easy to start to view the system designed to protect human subjects as obstacles to be overcome rather than the patients’ advocate.

You know how they say that the road to hell is paved with good intentions? I can’t think of a better example in medicine than a case like this.