Jennifer Margulis and her daughter Leone Francesca

Jennifer Margulis, Ph.D., is a senior fellow at the Schuster Institute for Investigative Journalism. A former Fulbright Fellow and an award-winning journalist, she has worked on a child survival campaign in Niger, West Africa, and championed the rights of rescued child slaves in Pakistan. Research for her fifth book, The Business of Baby: What Doctors Don’t Tell You, What Corporations Try to Sell You, and How to Put Your Pregnancy, Childbirth, and Baby Before Their Bottom Line, was partially made possible by the Fund for Investigative Journalism.

The following is a telephone interview conducted by Jennifer Margulis. I have taken the liberty of posting it in my blog while retaining her wording and introduction.A previous blog goes into more details and provides references https://corticalchauvinism.wordpress.com/2013/01/30/the-need-for-further-studies-on-the-safety-of-prenatal-ultrasound/

Jennifer Margulis (JM): I interviewed Dr. Manuel Casanova, M.D., the Gottfried and Gisela Kolb Endowed Chair in Outpatient Psychiatry and a Professor of Anatomical Sciences and Neurobiology at the University of Louisville, in October 2011 and spoke to him several times afterwards. Since 1996, Dr. Casanova has been involved in autism research. His particular concentration has been on how autistic brains differ from the neurotypical brains and what factors might be triggering the rise in neurological disorders in children in America and so many other countries in the industrialized world. For the past five years Dr. Casanova, a preeminent neuroscientist in his field, editor and peer-reviewer for several highly respected scientific journals, co-author of over 200 peer-reviewed papers, close to 70 book chapters, and a meticulous researcher, has been investigating whether or not ultrasound exposure could be one of the factors triggering neurological disorders. His research is not easy for the non-scientist to understand and I found myself interrupting him several times during our talks to ask him for clarification. Below is most of the transcript from our first long interview.

Manuel Casanova (MC): Initially I had to understand what was wrong with the brain before I could try to attempt to identify risk factors. So a lot of our initial research was based on neuropathology–trying to study the brain and what is abnormally expressed within the same in patients with autism. Once I understood the pathology, I could look at potential interventions as well as the causes.

JM: What are the most important things that you now understand about the pathology of the brains in people with autism?

MC: First of all, I approach everything from the standpoint of a neurologist, as a clinician. When diagnosing a condition, a neurologist first tries to localize the pathology in the brain. Autism is a condition of the cerebral cortex. The language defects, socialization, and other deficits found in autism are best localized within the cerebral cortex (the outer rim of grey matter surrounding the brain). What we found is that the units that provide for information processing within this part of the brain, in the cortex, which are called minicolumns, are abnormal.

JM: I don’t know what minicolumns are.

MC:From the standpoint of systems theory, let’s say you have a system, which in this case I’ll say a car. That car is composed of different modules; it has tires, a transmission, alternator, exhaust, etc. It’s only when you put everything together that you have the emergence of a property that wasn’t there in the individual modules, and that’s locomotion—a nice drive. Okay?

JM: You’re saying there are all these different aspects you can isolate, if you are talking about a car or the brain. But when you take each aspect, whatever it’s doing individually, when you put it together it might do something you don’t expect.

MC: Not predictable, necessarily. If you had a tire by itself and you have never seen a car, you wouldn’t really know what its ultimate purpose was.

Within the brain we have multiple modules that we call minicolumns and depending on how you tie them together you have the emergence of multiple properties, that’s the way we have face recognition, joint attention, that we are able to process visual perception and things of that sort.

JM: Is a minicolumn actually a structure in the brain?

MC: The concept of minicolumns is derived from a very famous neuroscientist named Vernon B Mountcastle, who was an electrophysiologist. His work was to impale the cortex with microelectrodes and then see the recordings of the cells that he had impaled. He noticed that when he impaled the cortex perpendicular to the surface of the cortex and went straight down through the gray matter (like a birthday candle on a cake), every time he impaled a neuron, they all shared the same properties: they all had the same terminal fields for sensory perception, and once you stimulated a cell, all reacted together. They were thus reacting as a unit. Because they were acting together, as a unit, he decided to give them a name: “mini” because they are microscopic, they span a very small amount of tissue, like 25 microns to 60 microns, and “columnar” because if he impaled them tangentially all of those properties would disappear. You can only have columns of cells that share the same properties vertically through the extent of the gray matter.

So those are the minicolumns, they are a module of information processing, and depending on how you connect them, it is how you get the emergence of higher cognitive functions: facial recognition, joint attention, theory of mind, you name it.

JM: What is joint attention?

MC: Joint attention is when, for example, you call attention to a child that he should be observing something. If I am with a child and all of a sudden I notice something within the room, I gaze in that direction. Because the child sees me, he will gaze his attention to what I am looking at. It is attention shared between more than one individual. It can be done reflexively. I can turn my face, look somewhere, and because I did so the child will imitate me. That quality appears to be abnormal in autistic children.

In terms of autism, schizophrenia, bipolar disorder, practically any psychiatric condition–there wasn’t a lot known in regards to the pathological substratum of these conditions.

If you examine the microscopy of the cortex in patients with autism, for example, most people would believe that it’s completely normal. I realized one thing. That we had the wrong paradigm in looking at the cortex. Thomas S. Kuhn, a historian of science, a philosopher of science, he wrote the book, History of Scientific Revolution, and he said that, “Science advances by introducing a new perspective at looking at an old problem.” It’s not by incremental degrees but actually by introducing an aha moment, a eureka moment, a paradigm shift. I thought we needed something similar in terms of neuropathology for autism.

We have a very neuron-o-centric approach to neuropathology. We put the neurons and cells at the center of our diagnostic universe. We call things ‘abnormal’ primarily because cells are lost or diminish in size or stain differently. What happens if the pathology escapes that level of resolution? We need a paradigm shift in neuropathology of psychiatric conditions. One that looks at modules, at circuitry, rather than single cells. We therefore decided to study circuitry within the brain of patients with psychiatric conditions, primarily autism.

What we found is that the minicolumns that provide for information processing within the brain in autistic individuals are abnormal: there are more of them in autistic individuals AND they are constructed abnormally.

A lot of this research could not have been done before because in order to analyze the minicolumn, their numbers and size, you really need the eyes of a computer. The amount of reduction that we noticed in a minicolumn was like 10 – 12 percent. That’s something that the human eye cannot discern.

A minicolumn is a network of interconnected cells, maybe 80 to 100 of them, and then all of their connections and their projections.

JM: Is it possible to know where it starts and where it ends?

MC: Yes, but then you have to divide the cortex into Lamina. And then you would probably say that they extend all the way from Lamina 2 to Lamina 6.

JM: Why would there be more minicolumns instead of less, that seems counterintuitive to me?

MC: That’s in part where ultrasound comes into play. May I give you a small introduction to the ultrasound bit?

JM: Of course.

MC: Let’s talk about sound waves. A sound wave is a wave that conveys a certain amount of energy and when it impinges on your tympanic membrane, it makes it vibrate, and it activates certain mechanisms that allow you to hear.

Ultrasound is a similar type of wave, whose energy deforms cell membranes. There are certain cells within the body called mechano-sensitive that have a proclivity for being deformed, a vulnerability. Their membranes are sensitive to deformation.

Ultrasound, this energy wave, preferentially affects those cells. When it deforms the membrane of the cell, it activates mechanisms that have to do with cell growth and with cell divisions. These cells tend to be the faster dividing cells within the body. Many of these mechano-sensitive cells are stem cells. We know that this actually happens. There have been many studies.

You should be aware, for example, that the FDA has approved the use of ultrasound for bone fractures because it accelerates cell division, it accelerates healing of the bone. So within the brain there is a nidus, a conglomerate, of stem cells in and around cavities which are called the ventricles. These stem cells usually divide and migrate from the ventricular wall (this cavity in the center of the brain) all the way to the surface of the brain where they form the cortex, the gray matter that we were talking about before.

JM: That’s part of the process of human gestation, that brain cells migrate?

MC: It occurs during gestation. When they migrate to the cortex, they follow a scaffolding, and they acquire a vertical orientation within the gray matter and from there they form minicolumns. These cell divisions actually occur within the germinal zone (the germ cells surrounding the ventricles) at a higher rate in autistic individuals, providing for more migration of cells to the cortex and more minicolumns.

JM: That’s a hypothesis or that has been shown to be true?

MC: That’s our studies, we have several, one was an international study, that included people from Germany, from the Netherlands, we were all blinded to the results, I provided the analysis. The study was published in Acta Neuropathologica. It has all been reproduced and reported within the literature. Within the brain of autistic individuals there appears to be something that promotes the division of these stem cells that surround the ventricles to divide supernumerary-wise, at a time when they shouldn’t be dividing. That division provides for daughter cells that migrate to the cortex and acquire a vertical arrangement that we call minicolumns. There are more minicolumns in the brains of autistic children than in normal but it all stems from the fact that something impinged on the germinal cells and caused them to divide.

JM: Caused them to divide more than they should have?

MC: Minicolumns are compartmentalized. You have a central or core compartment that is provided by this radial cell migration to the cortex, and these are excitatory cells. If they divide at a proper time, these cells migrate to the cortex where they mature in synchrony with another type of cell that migrates tangentially, these are inhibitory cells. So normally you have a radial migration of cells to the cortex which is primarily excitatory and that’s coupled in a very fine ballet-like fashion with inhibitory cells that are moving tangentially through the cortex. It has been said that these inhibitory cells provide a shower curtain of inhibition to the minicolumn. You know that a shower curtain keeps water inside of the bathtub. If you have a defect in the shower curtain of inhibition, water will spill out of the bathtub. If the radial migration is not coupled with the tangential migration of inhibitory cells then the minicolumns would have a faulty “shower curtain” of inhibition and information would no longer be kept within the core of the minicolumn, it would be able to suffuse to adjacent minicolumns and have an overall amplification affect. And actually the cortex of autism individuals is hyper-excitable and they suffer from multi-focal seizures. One third of autistic individuals have suffered from at least two seizures by the time they reach puberty. What we are proposing is that something impinges on the germinal cells, causes them to divide at a time when they should not divide. Cells migrate to the cortex but because it is at an anomalous time they are not synchronized with inhibitory cells, so there is an excitatory-inhibitory imbalance.

JM: When you say something “impinges” on the germinal cells, what you’re saying is something disrupts them, something changes them, moves them?

MC: It makes them divide when they shouldn’t.

JM: OK. In the normal migratory pathway of brain development, when we have cells that are migrating, they are also coupled with inhibitory cells. I’m not sure I understand this. Those cells are coming along?

MC: The cells meet them there in the cortex during the migration of both. They develop together.

JM: When we get there there’s this welcoming committee of cells that are going to keep us in place and keep us from spilling out to a part of the brain where we don’t want to be.

MC: Exactly. But if you force those cells to migrate when they aren’t supposed to migrate, there’s no welcoming committee.

MC: We have shown this to be true. We have gone as far as to validate the significance of the same. Facts are only as valuable as you can apply to the patients. We have used the findings to predict many things about autism that were unknown before and to explain clinical symptoms of the condition that are well known. This hypothesis has been validated scientifically in terms of its explanatory and predictive powers.

JM: OK. So now let’s go back to ultrasounds.

MC: The other thing that you should know is that there are many things that may actually cause the germinal cells to divide abnormally. Some viral infections, like cytomegalovirus and rubella, target these cells. Genetic conditions like tuberous sclerosis or Ehler Danlos interfere with their divisions. Extreme prematurity may cause hemorrhages and destruction of many germinal cells surrounding the ventricles. All of these conditions, not surprisingly, express in a significant proportion of cases an autism phenotype.

JM: You mentioned several things in one of the papers I read. You said maternal infection, seizures, drugs…

MC: The main thing is that if it was only the effect of genetics, since all the germ cells have the same genetic component, you would probably see abnormal migration everywhere throughout the cortex. It’s only when you have an exogenous factor, like X-rays, that you see abnormal development and migration in the cortex in different areas of the brain in different organisms. In terms of autism, you can actually see there is abnormal migration of cells to the cortex. This group of cells cluster and come to rest where they shouldn’t, even before reaching the gray matter. They actually stop their migration within the white matter. So you have a small island of gray within the white matter (“heterotopias” in medical lingo). In autism we find those islands of gray within the frontal lobe, but in the next patient it will be within the occipital lobe. The next patient may exhibit them within the cerebellum. It’s almost like every single patient is different. And that goes along with environmental migration and abnormalities. That’s usually seen when something from the environment is the precipitant for those germinal cells to divide, and because it impinges differently on the germinal cells of different patients, you are going to have a different spectrum in terms of pathology. But if these were only the effects of the genetics, since all of the germ cells have the same genetic component, you would see the same abnormal migration everywhere throughout the cortex. It’s only when you have an exogenous factor, like X-rays, you see. In ultrasound, for example, the way I approach the fetus with my probe may actually vary from exam to exam. So the amount of energy that is impinging at any time in the germinal cell layer of the brain will vary.

Those were my initial thoughts about why ultrasound could be of significance in terms of autism. Then the more you examine ultrasound and its epidemiology, the more proof you can draw of the same as a risk factor for autism. Populations within the United States that don’t use as much ultrasound are at a lower risk, like the Amish. The Somalis, where autism is practically unknown in their native countries, when they migrate to developed countries, they acquire a higher risk. This is not a property of being in this country. Somalis also acquired a higher risk for autism if they move to other developed nations. Obviously, in these new surroundings, they are receiving more ultrasounds as they are considered a higher risk pregnancy. Other risk factors for autism like older parental age, multiparity (like having twins or triplets) and obese pregnant women require more ultrasound studies.

Many people believe that because having a child with autism confers a greater risk for having a sibling similarly diagnosed, that this is a genetic condition. However, this could also be explained by ultrasound, as the mothers tend to go to the same OB-GYN practitioners. Ultrasound may or may not be a risk factor. That is something that research will tell us in the future. And I am really not advocating to stop using ultrasound, it is a valuable tool, but rather for practitioners to adhere to existing safety regulations.

Studies say that one third of all practitioners do not adhere to safety regulations. One safety regulation is to never perform an ultrasound study during the first trimester in a non-risk pregnancy. No more than two in a low-risk pregnancy. Right now one third of practitioners are using ultrasound during the first trimester and they don’t see anything wrong with that. Alarmingly, about 40 percent of ultrasound equipment presently in use is defective. One study from around 2009 took 700 machines from about seven different companies and they found defective transducers, the probe that you use to apply the ultrasound was defective in 20 to 70 percent of the equipment depending on manufacturer (the median was 40 percent.) Furthermore, the end users do not know what they are doing. When asked to define the thermal index, the mechanical index, they did NOT know where to find it on the machine in order to provide an index of danger to the patient. Right now because it has been deregulated since 1993, people have been using ultrasound with a 7-fold to 8-fold increased energy without having done the proper safety studies. Now we have a whole industry of 4-D ultrasounds, of fetal colorized reconstructions of babies using ultrasounds. Now we have patients that buy their own equipment through Amazon and eBay, they are providing reviews for other patients, it’s not the physicians. They are buying it themselves. They really do not know the dangers in what they are buying.

Ultrasound is actually also used in fetal heart rate monitoring. Those are also being bought in a non-restricted fashion by the patients themselves. We should voice some caution about what is happening within our society. There have been many committees and organizations that have called for safety regulations, and they are not being followed as of present.

We need more studies. Ultrasounds are being done without regards to the safety of the patients. I have been to many ultrasounds. The techs, who are well mannered and well-educated, at the end of the ultrasound session will ask the patients, “Wouldn’t you like to see more?” “Would you like me to take more pictures?” They have no idea that they should be getting in and out within a certain defined period of time.

[4/16/13 Addendum: I have often been asked whether the increased incidence of autism among siblings points to a genetic condition. I say that other factors, besides genetics may be help explain the fact, e.g., maybe women tend to go to the same gynecologist for all of their pregnancies. As I said in my interview, the use of ultrasound may also help explain some of the other risk factors for autism, e.g. old maternal age, obese pregnant woman, multiparity (twins, triplets). All of these conditions usually require additional ultrasounds or exams with higher energy levels. Retrospetive or epidemiological studies will add little assurance to the safety of ultrasound as its practice varies tremendously among end users. This is one place where animal studies could be of use but the NIH has shown little interest in funding this type of study. A petition by David Blake to study the health risks associated with ultrasound and to promote more research on the subject can be found at: https://www.change.org/petitions/health-risks-of-prenatal-ultrasound-the-urgent-need-for-more-research-and-regulation ]