Duke Autism Center trial: no benefit of cord blood for autism

Dawson Kurtzberg cord blood for autism
Drs. Geraldine Dawson and Joanne Kurtzberg, some of the leading proponents of the idea of cord blood for autism.

A solid new, blinded placebo-controlled Phase II study from the Duke Autism Center clearly shows no benefit of cord blood for autism.

Surprisingly, the team has said they are encouraged by the results. As a result, they apparently will continue this kind of work anyway despite their data. In contrast, I see the data as a probable stop sign.

Few have been as upbeat over the years as this Duke team about the idea of cord blood for autism spectrum disorder. I have been one of the voices arguing on the other side against this idea for many reasons.

Their team is led by Drs. Joanne Kurtzberg and Geraldine Dawson. For background you can see a past YouTube video below of Kurtzberg talking about their autism research and the trial in the paper. I have followed her stem cell research for more than 15 years. The field overall has great respect for her as a stem cell researcher. So do I.

I don’t know Dr. Dawson’s work as well, but she seems to have an outstanding track record as a researcher too.

You can see my 2013 interview with Kurtzberg on The Niche in two parts here and here.

Let’s take a look at their newly published study.

Duke Autism Center Data: no efficacy of cord blood for autism

The paper in the Journal of Pediatrics by Dawson and Kurtzberg’s team presents data from their randomized, blinded, and placebo-controlled clinical trial. Overall, the trial found no significant efficacy of cord blood for autism.

This is a uniquely solid study in this area so in my view there’s not much room for hedging bets like “well, maybe trying things a little different will work” anymore. The key data in the paper are in Table II (screenshot below) showing across the board that there is no benefit of cord blood for these patients.

Note the huge p-values. We’re not talking about p=0.06 here or anything close to that.

The study missed both the primary and all of the secondary outcomes.

Dawson, et al Cord blood for autism paper Journal of Pediatrics 2020 Table II
Dawson, et al. cord blood for autism paper in the Journal of Pediatrics 2020. Screenshot of Table II.

While the paper has a neutral title, “A Phase II Randomized Clinical Trial of the Safety and Efficacy of Intravenous Umbilical Cord Blood Infusion for Treatment of Children with Autism Spectrum Disorder”, the conclusion in the abstract at first seems far more direct:

“Overall, a single infusion of CB was not associated with improved socialization skills or reduced autism symptoms.”

That’s clear, right? I think so.

Is more research justified?

Unfortunately, the paper’s abstract ends, “More research is warranted to determine whether CB infusion is an effective treatment for some children with ASD.”

But is it warranted? Why?

And if you proceed, what do you do differently? More carefully select and assign your trial participants? Won’t that risk introducing bias?

Try more infusions per patient? If so, where do you get enough cells? Purify just the stem cells from the cord blood and grow them in culture for infusion? What are the risks to that?

cord blood stem cells for autism
A Sac Bee article on a trial for cord blood stem cells for autism, which also found no benefit.

The simplest explanation for these new negative data and other past reports suggesting no benefit is that this approach is just not going to work.

The Duke Autism Center researchers looked for other reasons why cord blood still might work despite their study not showing it.

One main possible reason they suggest in the paper for why they didn’t find benefit was some imbalance. Participants with intellectual disability (ID) were imbalanced in representation in the autologous and allogeneic groups. Removing participants with ID from the analysis after the fact made some small possible trends show up, but notably still not significantly.

The paper emphasizes how hard it is to do this kind of trial and oddly enough that’s a possible reason to keep doing more trials. Can they learn from this one to do better trials of the same kind in the future? Maybe, but I’m not convinced it’s worth it. Further, would it make good sense to do a whole new series of cord blood trials just on ASD kids who have no ID?

This new study along with other data should also probably be the end of the road for the related idea of umbilical cord blood stem cells for autism, but it likely won’t be.

Why have I been so concerned about this path?

Why have I been skeptical of stem cells and specifically about cord blood for autism for so long? You can see some of my thinking in the many past posts I’ve done here on The Niche, listed below over a period of 8 years:

The key reasons for my skepticism

I have condensed down the major reasons I’ve been so skeptical from what I’ve written before:

  • “There still is no compelling evidence that I know of that stem cells or cord blood can help autism and that they would be definitely safe in kids, although the safety is starting to look much better overall as of 2020.
  • The most common route of administration, IV into the bloodstream, doesn’t make much sense to me. Do a meaningful number of cells even get into the brain that way?
  • If cells don’t get into the brain, you’re assuming they do something systemically that helps the autistic brain? Some folks believe that there is a definite autoimmune component to autism and that flooding the bloodstream with stem cells might reduce autoimmunity, but there’s little hard evidence for either of these ideas.
  • If the cells do get into the brain, what do the cells that get into the brain do that is helpful and where do they do it inside the brain? Which brings me to the next question.
  • What causes autism spectrum disorder? The field doesn’t really know so how can you treat it if you don’t understand it?
  • Under that umbrella “spectrum” term there are many manifestations and probably many causes so stem cells (and cord blood more generally) are unlikely to be an autism panacea.
  • Is there any evidence that once autism manifests that it is reversible?”

Stem cell clinics, cord products, and autism

Another big reason for concern here overall that was true 8 years ago and still is the case now in 2020 is that for-profit stem cell clinics around the world including here in the US and in Panama keep selling the idea of cord blood and umbilical cord stem cells for autism. They are literally selling this idea to make money. The clinics like to point to the Duke program and media about it as support for what they are doing too.

The most prominent clinic of his kind is The Stem Cell Institute in Panama led by Neil Riordan. They published a paper last year suggesting safety but without evidence of benefit of lab-grown umbilical cord stem cells for autism. I had a variety of reasons for concern about the publication of that article, which I wrote about here.

The funding for the Duke Autism Center trial

This kind of Phase II trial is very expensive to run.

The funding for the Duke research, as reported in the paper, apparently comes almost entirely from a philanthropic organization called the Marcus Foundation. I didn’t see any indication of NIH funding.  Different media stories over the years have listed the Duke funding from Marcus ranging anywhere from $15 million to tens of millions of dollars. I’m trying to learn more about the Duke autism research-Marcus connection, which will be the focus of another post.

Looking ahead

The Duke Autism Center apparently will keep forging ahead on this path. Remarkably, in a news piece about their new publication the team actually seemed somewhat upbeat about continuing:

“Overall, we are encouraged by these initial results and plan to build upon them in future studies of cellular therapies in children with autism,” Kurtzberg says.

I realize that parents of children with autism and their families need hope, but it’s got to be solid hope. If you keep going down the same road when even your own data don’t support doing that, is that providing real or false hope?

What would it take in terms of data for you to decide to stop and maybe try something else?

12 thoughts on “Duke Autism Center trial: no benefit of cord blood for autism”

  1. Just like the Wright brothers built their success in aviation based on multiple earlier failures, so too will Dr.Kurtzburg on her phase 2 data. As a physician and a father of a beautiful autistic child, I believe it is imperative we continue the work. Failure forces an evaluation to reformulate a study in design and content which potentially can lead to a successful outcome. I for one am not discouraged by the data but inspired by it. If we abandon an approach just because of an initial failure, we may have missed an amazing treatment arm for a condition that is so pervasive globally. If the Wright brothers just quit and gave up on the idea of flying and didn’t learn and modify their approach from the failures of Otto Lilienthal, we may have never known air travel. #StayOptimistic #KeepTrying #NeverQuit

  2. Paul,
    Expanded UC MSCs [large numbers] seem to be a better option than injecting very few MSC’s found in umbilical cord blood. Theoretically, we should be seeing better results.

  3. The report does have significant findings even though they may not have met their primary and secondary end-points.

    For example:

    “However, the mean 6-month change in VABS Socialization SS was larger in the CB group compared with the placebo group (3.13 vs 1.98) and met the threshold for a minimally clinically important difference.”


    “for children without ID, clinician CGI ratings indicated that children treated with allogeneic CB, but not those treated with autologous CB, showed improvement over placebo. Children without ID also showed significantly greater improvement in a pre-specified secondary outcome measure of VABS Communication SS when treated with CB compared with placebo.”

    I would have thought this is worth further research.

    1. @Wayne, I guess it depends on how one defines “Significant”. Statistically there really weren’t significant findings and it’s not necessarily kosher to start playing around with the data by eliminating certain groups of participants from the analysis.

  4. sairam atluri MD

    Great post Paul. The study was floored from the beginning as there are very few mesenchymal stem cells in umbilical cord. I am not surprised at the poor outcome.

    1. Ross Macdonald

      Yes Dr Atluri, maybe the problem is not with the concept but with the QUALITY and CONSISTENCY of the product??

    2. Just to be clear, the initial studies cited cord blood monocytes at the potential effector cells, not MSCs from cord blood. Ongoing research is including expanded cord tissue MSC’s, with some potency testing of said cells prior to infusion.

      1. @Matt,
        I’ve seen all kinds of proposed mechanisms by this team over the years. Early on they claimed the cells got into the brain in meaningful numbers, but they’ve shifted their proposed mechanism it seems to be an indirect immune system effect, perhaps independent on brain penetrance of the cells. The actual key cell type supposed involved, as you pointed out, has also changed over the years.

Leave a Reply