A Cautionary Note on the Use of Transcraneal Magnetic Stimulation in Autism

For many years now my group has used Transcraneal Magnetic Stimulation (TMS) to treat autism. We had the distinction of having the first published article on the subject, and have treated more patients in clinical trials than anybody else in the world. Our idea was judged as meritorious by the NIMH and I received a EUREKA award for the same.  It may therefore be of interest to know that contrary to what I was told and advised by both colleagues and our university officials I never pursued taking a patent on the idea.Since it was my original idea and I had collected data regarding both safety and efficacy this seemed to many a feasible idea. The reason for not pursuing a patent was my desire to make the method available to all patients without them having to pay a surcharge based on a patent. We also did not want to stifle the progress of research in this area.  At least on my part, I have never charged any of our patients for the TMS treatment or any associated interventions.

TMS action

An interesting side note to my story is that I had made my ideas and data on TMS and autism the object of several academic lectures before pursuing their publication. Not long after one of these lectures I received a call from a business developer from another institution. He was in-charge of commercializing the efforts of their autism center and told me about some of their ventures. To say the least many of his efforts were quite naïve. What really sparked my attention was his insistence on acquiring information regarding my use of TMS in autism. After a long sales pitch and more than one telephone conversation it was obvious that he wanted to patent my idea, but did not have any studies on the subject! I was asked to partner with his commercial venture (belonging to another academic institution), which I refused to do. The chutzpah of the individual!

The basic premise for the use of TMS in autism spectrum disorders is that patients have deficits within the modular organization (also called minicolumns) of their cerebral cortex. Under normal circumstances signal processing within these cortical modules is kept within the center of any given module by a group of surrounding inhibitory cells. According to some investigators these inhibitory cells provide a “strong vertical flow of inhibition” or a “shower curtain of inhibition”. In autism, a faulty inhibitory surround allows for the signal to escape and excite surrounding minicolumns. The hyperexcitable cortex of autistic individuals usually works under a “rich get richer scheme”, that is, a relentless cascade of amplification that makes it difficult to differentiate signal from noise. The resultant hyperexcitability also lowers the threshold for seizures to become manifest. It is not surprising that by puberty about one third of autistic individuals have had at least two seizures and comply with criteria to diagnose epilepsy.

figure 2 July 2010 without MET

Figure: Schematic illustration of minicolumns in neurotypicals (at top) and autistic (at bottom) individuals. Minicolumns are thinner and some of the inhibitory cells (labelled in red) are missing. Illustration from Vaccarino Frontiers in Neurogenesis.


Figure: The surrounding shower curtain of inhibition is faulty in autism (left panel).  It is hypothesized that using low frequency TMS helps strengthen the inhibitory surround of minicolumns.

More recent investigators have proposed other mechanisms for the action of TMS in autism. Most prominent among these is the putative effect of TMS on mirror neurons. It has never been explained why TMS would have a specific effect on mirror neurons. Still, if there was any credible evidence as to the role of mirror neurons in autism then we would have to consider this hypothesis. However, the theory of mirror neurons in autism has been shattered from a clinical, eletrophysiological and neuroimaging standpoint (see http://bit.ly/17wSldl ).

Transcraneal Magnetic Stimulation at low frequencies is inhibitory to the cerebral cortex. Most equipment commercially available emphasizes higher frequencies for the treatment of depression. The need in depression is to stimulate the cortex, the opposite of what is required in autism. It is difficult to use commercially available equipment and use it for the lower frequencies needed in autism.

TMS depression

At present the use of TMS in autism spectrum disorders has not been approved by the FDA.  This means that its use remains experimental or as it is called “off-label”.  According to the FDA off-label refers to the use of drugs and/or devices for other than their approved use. Many insurance companies won’t reimburse the cost of using a device when the same is off-label. Data for research regarding off-label use of a devise can be collected from patients but it requires an investigational device exemption (IDE).

Clinical evaluation of devices that have not been cleared for marketing requires:

  • an IDE approved by an institutional review board (IRB). If the study involves a significant risk device, the IDE must also be approved by FDA; (In our case the use of TMS has been approved and supervised by our IRB)
  • informed consent from all patients;
  • labeling for investigational use only
  • monitoring of the study and;
  • required records and reports.

Note: Unless exempt the IDE must be obtained before the research is initiated. Without an IDE the equipment per se can’t be shipped legally.

Since the FDA regulates not only the use but the also the marketing approval of drugs and medical devices, these products can only be advertised for the specific use(s) the FDA has cleared. This means that active promotion (including brochures) of devices for off-label use with the intent of making money goes against FDA regulations.

In a similar way the FDA has stated that a sponsor, investigator, or any person acting for or on behalf of a sponsor or investigator shall not commercialize an investigational device by charging the subjects or investigators for a device a price larger than that necessary to recover costs of manufacture, research, development, and handling (see http://bit.ly/15MGEm0 ).

In summary, the use of TMS in autism spectrum disorders remains experimental and has not been approved by the FDA for this purpose. Its use for research should be approved by an Internal Review Board (IRB) or equivalent organization that is able to supervise the ongoing clinical trial.  In the meanwhile the distribution of material for the purpose of advertising its use or commercializing the effort is prohibited by the FDA.

Addendum 9/24/2013: It should be very apparent that I am very protective about my patients as well as the proper use of TMS.  Unfortunately based on my work other physicians have taken the opportunity to establish private ventures commercializing the technique for ASD. This is premature and certainly runs into problems with federal regulations in this country. A similar and complementary opinion can be found at http://bit.ly/17zXXpv ).

11 responses to “A Cautionary Note on the Use of Transcraneal Magnetic Stimulation in Autism

  1. Hi Manny, the nerve of some people hey! I know you and your research… the ethics and the commitment to our best… these are always upper most in your thinking… I admire this in you. Having gone through rTMS with you and the team … I know there is value in this. It’s also interesting that motivation seems to play a role in broadening attention… all good stuff.


  2. Hi Manuel Casanova –

    A while ago I spent some time reading one of your blog posts reviewing some of the evidence of perturbed neuronal migration in autism. I went through the references and read the ones that were open source; one thing I took away from them was that while the findings of altered migration were frequently found, there wasn’t much of a pattern to the changes. In other words, structure consistent with altered migration was present in many cases, but it was altered in different areas of the brain depending on the case.

    But thinking about the underlying mechanism of TMS has me wondering a bit more; the minicolumn structure that you hope to affect with TMS is a pretty specific compared to the findings of general heterotopias / dysplasia observed in the studies you referenced previously.

    Is my problem reconciling this a function of my primitive understanding, differences in study methodology/capabilities in visualizing minicolumns, something else, or some of all of the above?

    Thanks for any insight.

    – pD


  3. The problem with applying TMS seems to be we don’t know exactly what areas are affected in autism and I though the magnet only goes over a certain area of the brain and not the entire brain. Also, would TMS just affect the cerebral cortex and not the deeper nuclei, e.g. the deep cerebellar nucleui or maybe limbic structures beneath the cortex? In my novel, as you remember the research subject undergoes TMS of the pars opercularis but there may not be enough evidence, as you said that this is an area involved in the etiology of autism.


  4. We selected to stimulate the dorsolateral prefrontal cortex (DLPC) because of its role in executive functions and initial neuropathological findings on minicolumns in autism. The DLPC may be the best connected area of the whole brain. It is regarded as a major connection hub of the brain. In this regard we decided to pursue what is called a cascading principle, that is, by correcting the function of the DLPC others would be corrected given the area’s connectivity to them. In this regard we are counting with the plasticity of the brain to correct itself.

    As you suggest TMS does not penetrate all that much, only a few centimeters. This allows for stimulation of the crest of the gyri, but it is doubtful whether the depth of the sulci are in any way affected directly affected.

    Thanks for the comment. Loved your novel. Have recommended it whenever I can.


  5. thanks for the praise of my novel, but one other thought. According to your logic and the cascading effect, then it could be argued that stimulation of mirror neuron areas with TMS would be beneficial for autism since the insula connects the pars opercularis to the limbic system which some post mortem studies have found to be abnormal in autism. Unless of course the pars opercularis does not have the same connections as the dorsolateral prefontal cortex and would not cascade as easily to other areas of the brain or perhaps there is something else i’m missing.


  6. The insula is buried and far removed from the surface of the hemispheres. TMS would not be able to stimulate the same. The penetration of the magnetic field for TMS is just a couple of centimeters and you have to take away the bone (skull). Also the insula has limited connections and there are many areas of the brain that have been involved in autism. The dorsolateral prefrontal cortex would be able to connect to all of them, the insular would connect to a limited number of them. Best regards.


  7. Hi Manny, I found this post intriguing because you mention the “hyperexcitable cortex” of ASD children. You may remember from the Baltimore think tank my suspicion that autistic behavior was due to lack of cortical inhibition because ammonia accumulation shifts metabolism/blood flow from cortex to subcortex.

    Is a “hyperexcitable” cortex compatible with lack of cortical inhibition? Do the seizures these children have always originate in the cortex, or can they arise in the subcortex?



    Peter Good Autism Studies http://www.autismstudies.info


  8. Individuals with autism spectrum disorders do have a hyperexcitable cortex. This fact has even given rise to a theory of an excitatory-inhibitory imbalance in autism. We spoke about the same soon after publishing our minicolumnopathy article in Neurology back in 2002. I would agree with you that it appears to be due to lack of or siminished inhibition. We claim that this can explain the defect in surroung inhibition of minicolumns that we have described potmortem. Other researchers have used vibrotactile stimulation or EEG to describe the same thing.
    The seizures in autism are multifocal which makes them difficult to treat. If they occur very early on, they may be widely distributed and often give rise to a particular pattern (West syndrome). Individual cases may have one preferred foci. However, their origin in the majority of cases suggests a cortical abnormality.


  9. I really appreciate this post and your commitment to protecting individuals with autism and their families from ill intended folks. Thanks for the work you do.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s