Is a trial of stem cell therapy in autism scientifically and ethically justified?

Houston, we have a problem.

Oh, wait. I’m not talking about Stanislaw Burzynski this time. But we do still have a problem, and it’s a problem that resembles the Burzynski problem I recently discussed. Specifically, it’s a problem of unethical clinical trials somehow winning approval from institutional review boards (IRBs). In academia, IRBs are basically ethics boards whose purpose is to protect the human subjects who agree to take part in clinical trials and other research from harm and from being subjected to experimental therapies in which the risk-benefit ratio isn’t sufficiently favorable. In other words, they are there to make sure that clinical trials are conducted with clinical equipoise. They have to approve clinical trials before they can begin and monitor them as they are performed, following up on adverse events and sometimes deciding to shut them down if there appears to be too much risk and/or harm. In the U.S., all institutions that receive federal funds (specifically, that have a Department of Health and Human Services-approved Federal Wide Assurance agreement) typically must agree to abide by the regulations governing human subjects research codified in what is known as the Common Rule. Similarly, any private entity (such as a pharmaceutical company) seeking FDA approval for its drug or device have to register with the FDA and abide by the Common Rule, whose most important set of rules mandate IRB approval and monitoring of the research. Some states also mandate that all human subjects research carried out within their borders, regardless of funding source, must abide by the Common Rule.

With all the laws, regulations, and rules that I have to abide by whenever I become involved in human subjects research, it never ceases to amaze me how some researchers can get away with the sorts of studies that they get away with. Think Mark and David Geier, who created a faux research institute and packed its IRB with their cronies to oversee their “studies” in which they subjected autistic children to chemical castration. Think Mary Holland. Think Stanislaw Burzynski. Think of all the other ethically dubious human subjects research that somehow manages to be approved and carried out in spite of laws and rules that, for those of us in academia, have become ever more onerous over the years.

Then ask yourself how this study could ever have been approved by any IRB, anywhere:

Families with autistic children must navigate a condition where questions outnumber the answers, and therapies remain sparse and largely ineffective. A clinical trial being conducted by the Sutter Neuroscience Institute in Sacramento, California to address this situation began recruiting participants today for a highly experimental stem cell therapy for autism. The institute plans to find 30 autistic children between ages 2 and 7 with cord blood banked at the privately-run Cord Blood Registry, located about 100 miles west of the institute.

I’ve written about stem cell quackery for autism many times. For instance, Kent Heckenlively hit his parents up for $15,000 to take his daughter to a highly dubious stem cell clinic in Costa Rica, where doctors injected “stem cells” (who knows what they really were) into her cerebrospinal fluid via a lumbar puncture. More recently, I noticed another autism “stem cell therapy” that was neither a therapy nor involved any actual stem cells and another child subjected to what are claimed to be stem cells by a quack clinic.

Is Sutter Neuroscience Institute another one? Certainly its website looks respectable enough, but then so do the websites of pretty much every dubious clinic offering unscientific therapies that I’ve ever seen. How spiffy a website appears tells us nothing, although the content of the website does give some clues, particularly the page Innovative medical and alternative therapies for autism spectrum disorders. There, you will find recommendations such as using vitamin C and vitamin B6 to treat autism, to “uncover food sensitivities” (which is a huge red flag for me), and to “try massage therapy” (which doesn’t strike me as a particularly good idea for at least some autistic children, given the sensory issues so many of them have. To be fair, though, the Institute does state unequivocally that the scientific evidence does not support the claim that vaccines cause autism and that children should get all their childhood vaccines. The website also correctly points out that chelation is dangerous and doesn’t work for autism; that facilitated communication also doesn’t work; and that there is no convincing evidence for any supplements having a positive effect on autistic symptoms (which makes me wonder why the webpage I mentioned before recommended vitamin C and B6).

So why is this place, which seems (mostly) reasonable, sponsoring a clinical trial using stem cells to treat autism? The title of the clinical trial can be found easily on its website, and you know what I have to do whenever I see that. Yes, indeed, it’s off to, where I first found the precursor trial mentioned in the Scientific American article:

Already one other clinical trial, with 37 total participants between ages 3 and 12 years old, has been completed in China. The researchers affiliated with Beike Biotechnology in Shenzhen, the firm that sponsored the study, have not yet published any papers from that the trial, which used stem cells from donated cord blood. Mexican researchers are currently recruiting kids for yet another type of autism stem cell trial that will harvest cells from the participant’s fat tissue.

The Chinese trial can be found on as well. Its rationale is:

Autism is one of those disorders in Autism spectrum disorders (ASD), which characterized by social interaction abnormalities, impaired verbal and non-verbal communication, and repetitive, obsessive behavior, while the therapeutic effect of current treatments remains limited progress. Neural hypoperfusion and immune deregulation are the two key pathologies associated with Autism. Human umbilical cord mesenchymal stem cells (hUC-MSCs) and human cord blood mononuclear cells (hCB-MNCs) have been shown to have the ability to modulate the immune response and enhance angiogenesis, suggesting the novel and promising therapeutic strategy. In this study, the safety and efficacy of hUC-MSCs and hCB-MNCs transplantation will be evaluated in patients with Autism.

This study was carried out overseas in china by a non-American company; so who knows what sort of ethics oversight it had. Its hypothesis is also dubious at best, but I’d rather discuss that in the context of the current clinical trial, whose entry informs us:

Evaluate the efficacy of one infusion of stem cells from autologous umbilical cord blood in patients with autism over six months after infusion as measured by changes in expressive and receptive language.

Also demonstrate improved behavior, learning, and changes in Serum tumor necrosis factor alpha (TNF-α), tumor necrosis factor beta (TNF-β), interleukin 1-alpha (IL-1α), interleukin 1-beta (IL-1β), interleukin 6 (IL-6), interleukin 10 (IL-10), and interleukin 13 (IL-13).

Basically, this is a single-center, randomized, placebo-controlled, crossover outpatient study with 15 subjects. Subjects will receive one infusion of autologous umbilical cord blood (AUCB) containing stem cells or an infusion of placebo (saline). After the 24-week follow-up, the groups will cross over so that patients who initially received AUCB will receive placebo and patients who received placebo at baseline will receive the cord blood. Looking at it, right off the bat I wonder why the investigators consider it so important to do a crossover design. Then I realized. It’s obvious. No parents would sign up for it if they didn’t know that their children would get stem cells; so the Institute accommodated them.

So what’s the problem with these studies? What why do I consider them to be unethical? It comes down to prior plausibility and preclinical evidence. According to the Helsinki Declaration, for a study to be ethical, there has to be solid preclinical evidence, in the form of in vitro and animal models, that strongly support the biological plausibility of the therapy proposed. What, I wonder, is the preclinical evidence supporting the use of autologous cord blood for autism? The introduction to the Chinese study tells us that the investigators are basing their rationale on the ability of human umbilical cord mesenchymal stem cells and mononuclear cells to “modulate the immune system.” Specifically, these cells have been tested in some autoimmune diseases. In essence, they appear to slow down the immune system, inducing Treg expression and decreasing the activation of T cells. In essence, these cells are immunosuppressive.

Which brings us to the problem.

The rationale for using these cells is not that they’re going to home to the brain and regenerate or repair parts of the brain in order to make them “normal” again. They are clearly being administered for their immunosuppressive/immunomodulatory activity, as evidenced by all the biomarkers being measured after their administration. They’re all cytokines and other immune molecules. The investigators wouldn’t measure all these cytokines if that’s not what they were interested in. For this therapy to have a chance of working, there must be a major immune component that is causative of autism. Although a variety of immune changes and abnormalities have been reported in autistic children, thus far, none of them have been convincingly shown to be causative. They could well be an epiphenomenon. Certainly, the evidence for immune dysfunction in autism is not sufficiently strong to justify something as invasive as performing autologous stem cell infusions, particularly when it’s not yet well understood what it is that autologous umbilical cord blood stem cells and monocytes do when reinfused.

Think of it this way: Do you think that the evidence implicating a hyperactive immune system is strong enough to justify treating autistic children with prednisone? Cyclosporine? Other immunosuppressive drugs? If not, then why would anyone advocate using autologous stem cells, which appear to be immunosuppressive? Why on earth would an institute like the Sutter Neuroscience Institute carry out such a trial based on low prior probability? What sort of preclinical evidence did they have to justify this trial? The scientists in the article who say that the likelihood of a positive result from this trial is low are, if anything, too optimistic. The likelihood of a positive result is almost homeopathically low. When it comes to clinical equipoise, this trial looks to me as though it’s all risk with too little prospect of benefit to be justifiable without a lot more clinical evidence.

I really wonder what the IRB is that approved and is overseeing this trial.