If you are interested in studying psychopathology, schizophrenia is the most severe of all psychiatric disorders. It affects the way people think, their motivation, their perceptions, and the way they relate to other people. It fills up more hospital beds outside of anything other than aging problems. The symptoms are so unusual—people hear voices, they have delusions—and it seems to me that that’s something you really want to understand and try to treat.

My own interest dates back to graduate school at Harvard, where I had a wonderful mentor, Professor Philip Holzman. He was investigating eye tracking—the way people follow objects with their eyes. He observed that tracking objects was abnormal in schizophrenia and he sought to understand why an anomaly like eye tracking was present in schizophrenia—he wanted to know exactly how this seemingly unrelated observation was related to schizophrenia.

This idea caught my attention as I had never thought about anomalies that may be seemingly unrelated to an illness but which might be more important in understanding the etiology of that illness than some of the more striking clinical symptoms like hallucinations or delusions.

“It was very exciting to be doing neuroimaging studies in schizophrenia, and to be a part of this new research, knowing that we were standing on the edge of something that was new and really promising.”

Symptoms like eye tracking may not be core features of schizophrenia, but Phil brought genetic aspects of the disorder to light—he found eye-tracking abnormalities in schizophrenia, albeit not in all patients with schizophrenia, but, even more importantly, he observed eye-tracking abnormalities in some of the family members of patients with schizophrenia.

The other thing that’s interesting about schizophrenia is that you’re probably not seeing one disorder, but many, and it is likely due, at least in part, to the effects of multiple genes—with weak effects, which, taken together, lead to schizophrenia. I think also the complicated nature of the disorder is what drew me to want to understand it.

My dissertation was on formal thought disorder, and I examined how patients with schizophrenia express themselves—they have peculiar word usages, and they make up words, and yet it was not at all obvious why this phenomenon was so common in schizophrenia. We even found that some formal thought disorder was also observed in the relatives of patients with schizophrenia. And again, the question that came to mind was, what is it that resulted in this person developing schizophrenia, but not his/her relative? What also was striking to me is that thought disorder had to be related to the brain. So I wanted to get closer to the brain.

There was, in fact, a long history of looking at the brain in schizophrenia, dating back to when Alzheimer discovered senile plaques in the brains of patients who ended up with this dementia named for him. Early investigators interested in schizophrenia were impressed with brain findings in dementia, and thought that they might also discover that the brain was also affected in schizophrenia.

Researchers in the late 19^th century became really disappointed, however, because the findings in schizophrenia weren’t there. But, I think the problem here is that they were looking for big alterations, when, in fact, any alterations in the brains of patients with schizophrenia are likely to be small, and subtle by comparison to disorders such as Alzheimer’s disease. And if you’re looking for small and subtle brain alterations then you can not use crude measurement tools—you have to be really precise, and this necessitates very refined and sophisticated tools to detect small differences in the brain.

There was a researcher by the name of Plum, who in a 1972 paper stated that schizophrenia was the "graveyard of neuropathologists." What he meant by this statement is that if you were going to spend your career working in the area of schizophrenia brain research, it would be a wasteland, because you would not get anywhere—this statement was made at a time when we still did not have the equipment available to make such an exploration fruitful. Many investigators had even moved away from looking at the brain, because if there were no good tools, there would be no good findings.

Then in 1976, there was a renewed interest in trying to look at the brain in schizophrenia—this interest was rekindled by a CT study done by Eve Johnstone and colleagues in the UK. This paper was important because it got researchers thinking about the brain and its connection to schizophrenia again. That was the seminal paradigm shift—the first CT study that said, "Hey, we actually can find abnormalities in the brain."

To get back on track though, my work on thought disorder with Philip Holzman led me to want to get closer to the brain in trying to understand schizophrenia. Consequently, when I finished my Ph.D. in ’84 at Harvard, I began a post-doctoral fellowship with my next mentor, Bob McCarley, in the Clinical Research Training Program at Harvard Medical School.

During my two-year fellowship with Bob, I conducted a study investigating the relationship between formal thought disorder, and an event related potential known as the P300. Event related potentials (ERP) are EEG activity in response to a stimulation controlled by the experimenter. The P300 occurs about 300 milliseconds following a discrepant or novel event in the environment. It is thought to be an index of the brain's ability to detect relevant from irrelevant stimuli. It’s the brain’s way of saying "hey, that’s unusual," and tells you that there’s something to which you need to pay attention.

As patients with schizophrenia have problems with differentiating relevant from irrelevant information in the environment, I thought it would be of interest to evaluate this phenomenon in conjunction with measures of formal thought disorder, delusions, and hallucinations. These studies led to the finding that patients with schizophrenia show a decrease in the P300 that is evident in the left temporal region of the brain, a region thought to be important in language functioning. This finding has since been replicated in our own research as well as by other laboratories.

In 1987 I was involved in a CT study with Bob McCarley and Marjorie LeMay at Brigham and Women’s Hospital, and I also talked to Ferenc Jolesz, who is a professor in the Department of Radiology, and I said, "I want to put in for a K award (NIH Career Development Award) to come here and study the brain in schizophrenia using MRI." He was very nice and said I could come but thought I would be wasting my time. I said, "OK, but can I come?" and he agreed. At the time, he had the same view of wondering why anyone would look at the brain and its role in schizophrenia—if there were anything to find, it would have been found already.

In ’88 we started the MRI work. Early on, the spatial resolution of MRI was only about 1 cm thick, and there were gaps between slices — so you couldn’t view the whole brain. It wasn’t until later, with our 1992 NEJM paper (Shenton ME, et al., "Abnormalities of the left temporal lobe and thought disorder in schizophrenia. A quantitative magnetic resonance imaging study," NEJM 327[9]: 604-12, 27 August 1992), that we were able to look at the entire brain through a series of 1.5mm slices.

It was very exciting to be doing neuroimaging studies in schizophrenia, and to be a part of this new research, knowing that we were standing on the edge of something that was new and really promising. And the result was that we really did find brain alterations in schizophrenia. And most exciting, we were able to confirm early speculations about brain abnormalities in schizophrenia in a way that was previously not possible, but became possible because of the new imaging tools available to explore the brain.

Dr. Martha Shenton's most-cited paper with 411 cites to date: Shenton ME, et al., "A review of MRI findings in schizophrenia," Schizophr. Res. 49(1-2): 1-52, 15 April 2001.  Source: Essential Science Indicators. RELATED INFORMATION: Dr. Martha Shenton was interviewed in our first analysis of Schizophrenia in July 2001. Dr. Shenton of Harvard Medical School talked about the influences and experiences that have shaped her career in neuroscience. Read the interview.  Read an interview with coauthor Dr. Robert McCarley about the paper above from the Special Topic of Schizophrenia.

Schizophrenia is such a terrible clinical disorder and we were all pleased that our work was published in a journal that would be read by people who were treating schizophrenia. We got the letter of acceptance on my mother’s birthday, which would have pleased her—she had passed away a few years prior, from Alzheimer’s disease.

The excitement we felt in those early years when we began MRI studies is very similar to the excitement generated from new work we are doing that involves diffusion tensor imaging studies in schizophrenia. There are no guideposts for the research, no standard operating procedures, no conventions, so to speak. We try things out, and find some things that work, and some that don’t, and have that "wow!" moment. To me, what I find most exciting is working in an area where two fields meet, which is also where new discoveries are made.

This new work is being done in collaboration with Marek Kubicki, a neuroradiologist and physicist by training; Sylvain Bouix, Carl-Fredrik Westin, Marc Niethammer, and Yogesh Rathi, computer scientists; Bob McCarley, Jim Levitt, Chandlee Dickey, and Ron Guerrera, psychiatrists; Dean Salisbury, Margaret Niznikiewicz, and Paul Nestor, psychologists; Ron Kikinis and Ferenc Jolesz, neuroradiologists; Zora Kikinis, a biochemist and geneticist; and visiting fellows including Moto Nakamura, a psychiatrist from Japan, Khang-Uk Lee, a psychiatrist from Korea, Jung Su Oh, a biomedical engineer from Korea, Toshiro Kawashima, a psychiatrist from Japan, Jennifer Fitzsimmons, a medical school graduate from Spain, and Lucas Torres, a medical school graduate from Argentina. We are also beginning studies evaluating ERP measures of interhemispheric transfer and diffusion tensor imaging with Kevin Spencer.

Sure, this work was done with Yoshio Hirayasu, Bob McCarley, Dean Salisbury, as well as others listed in the paper. This 1998 paper is important because you’re seeing real change early on in the illness, which means this is a point in time where you’d really want to intervene, and try to prevent these changes. If you look at patients with chronic schizophrenia at two points in time, you don’t see these changes. From this study we observed that brain alterations in the temporal lobe are observed at first hospitalization for schizophrenia, but are not observed in psychotic patients with a first-episode of affective disorder, or in normal controls.

Our lab is developing and applying new imaging tools to study schizophrenic brains, though we also are continuing work using structural MR, functional MR, and diffusion MR techniques to understand cognitive, clinical, and abnormalities in schizophrenia. We also have a number of young investigators with expertise in neuroscience, psychiatry, psychology, neuroradiology, physics, engineering, and computer science who are actively working on developing and applying new imaging tools to understand brain networks such as language, memory, and attentional networks in the brain.

Some of these individuals will be meeting and talking with you today about their work, although Marek Kubicki, one of my main collaborators, is not here today to discuss his work. He is very much involved in all of the imaging studies done in our lab.

Sylvain Bouix: Schizophrenia is a syndrome that affects the brain, but is not limited to one area. We need structural and functional pictures of the brain to get a complete picture of what is abnormal. We now have different ways of looking at the brain, and thanks to MRI we can perform volumetric studies and shape analyses. Our focus has been on regions involved in language processing, as well as language expression and auditory hallucinations, among the most common symptoms of schizophrenia.

We use a computer program 3D Slicer, to look at 3D images of the brain and automatically label the various structures for volume measurement. This is not easy to achieve, as it would be like taking a picture of a scene outside the window and wanting the computer to be able to tell the difference between the buildings, trees, cars, etc. It’s difficult to develop, and for some of the more complex regions, like the caudate, we still need to label them by hand. Nevertheless the refinement has come a long way over the last 15 years.

We are also performing shape analyses. Using a 3D shape representation technique, we can evaluate shape deformations in schizophrenic brains. We concentrate on where the volume loss is observed, although other parts of the brain will also be affected. We’ve done these studies with the caudate nucleus and amygdala-hippocampal complex, but we are certainly interested in investigating shape deformations in other structures as well, as these may be associated with neurodevelopmental influences in the brain.

Marc Niethammer: As recently as 10 years ago, researchers were unable to investigate white matter in detail from MR images, and mainly focused on gray matter. But now we’re doing something called diffusion tensor imaging (DTI), which allows us to look at the directionality of white matter, and the integrity of the white matter bundles in the brain.

These types of studies provide much more information than we had previously, and, of course, are more complicated to analyze and require more complex tools. DTI measures the diffusion of water in the brain. Though it cannot go down as far as to a single axon, it can provide the orientation of the macroscopic structure of fiber tracts, which allow us to make topographic maps of white matter, which helps to give an indication of global white matter behavior.

“We need structural and functional pictures of the brain to get a complete picture of what is abnormal.” Sylvain Bouix (left) “Combining structural MRI and DTI scans will provide a more complete picture than we’ve ever had before.” Marc Niethammer (right)

DTI has uses other than schizophrenic studies. It can also be used to study cases of brain trauma, tumors, or in preparation for brain surgery.

Another research strategy we are taking involves velocardiofacial syndrome (VCFS). A high percentage of VCFS patients also develop schizophrenia. These patients also have a known chromosomal deletion, on chromosome 22. With these patients, we can combine imaging and genetics by having a much more genetically homogeneous study population. Combining structural MRI and DTI scans will provide a more complete picture than we’ve ever had before.

Sylvain Bouix: One final type of study we are performing is functional MRI (fMRI)—we take the images during controlled experimental tasks, which we hope will help link anatomical abnormalities in the brain with clinical and psychological symptoms.

Shenton: We’re now part of a big NIMH grant with Bob McCarley, the principal investigator (PI), as well as Jill Goldstein, Larry Seidman, Jean Frazier, Tracey Petreshen, and Wilson Woo, called the Boston CIDAR. I’m PI for the imaging core, as well as PI on a special project on white-matter progression, using DTI. We’re going to be looking at prodromes (patients at high risk, before they become ill, or do not), first-episode patients, as well as different time points in the first episode, and also chronic patients. We’ll be collecting gene data from the patients as well, and this will be evaluated in conjunction with imaging and cognitive and behavioral data.

We’ll be able to look at endstage, and by including people who are at risk for developing schizophrenia (prodrome period), we will be able to evaluate people who go on to develop the schizophrenia and those who do not. That’s going to be important, because the individuals who go on to develop schizophrenia might be very different from those who don’t.

Also, if you find abnormalities early on, even before the first psychotic episode of illness, you’re going to be looking at something that’s probably much more inherently part of the disorder itself. And if you see things that change over time, even in a short time interval, it really indicates that this area is where you’re going to want to put your money and effort in terms of treatment and prevention strategies.

Schizophrenia is about thirds: a third of patients stay the same, a third get better, and a third get worse. Well, a third get "better," but they may not ever go back to baseline.

In 2001, I wrote a review (Shenton ME, et al., "A review of MRI findings in schizophrenia," Schizophrenia Research 49:1-52, 15 April 2001) that addresses this issue. Just to give you an idea of how far this research has come, in 2001 you could write a comprehensive review article, but you couldn’t do that today—there’s too much material now; you’d need a whole book.

Imaging genomics are going to be an area to watch—trying to get closer to the gene, with imaging. We know so much more about the brain now than we used to know, and we are just beginning to understand genes. Imaging and associations with genes will likely be important as will delineating further attentional, memory, and language networks in the brain, how they are affected in schizophrenia, and how they are associated with genes.

I also think that more individualized treatment may be an area that will burgeon in the near future. For example, once you know which brain region(s) are affected in a given patient, it may be possible to tailor treatment to that individual. There may, for example, be some patients for whom attention is more of an issue, and for others it could be something else—we might be able to find a way of delineating for each patient what their greatest deficits are, and what areas of the brain are affected most, and then you might be able to tailor new treatments based on new drugs developed specifically for that individual patient. You might be able to learn to block a gene’s effect(s) or begin to know where to intervene in terms of cognitive remediation.

Other future areas of research are likely in prevention, and in a focus on the prodrome period. You can diagnose a disease, but diagnosis sometimes doesn’t help in terms of treatment as, for example, in Alzheimer’s disease (AD). There are now clear diagnostic indicators of AD, and we know it is characterized by amyloid plaques and neural fiber tangles. But by the time you diagnose AD, there’s not much you can do in terms of treatment. But what if, early on, you could intervene?

For instance, there’s a researcher at Brigham and Women’s Hospital, Lee Goldstein, who is looking at AD plaques by looking at them in the lens of the eye. The eye’s lens metabolizes much more slowly than the brain, so the brain could be clearing out plaques, but the lens is slower, and as such, might not be clearing things out as quickly. So if you look at the lens of the eye and pick up amyloid plaques a decade or two before they’re seen in the brain, combined with perhaps information about an underlying predisposition to developing AD, such as ApoE4, or a slightly smaller hippocampus, and you could identify this decades before the onset of AD, then intervention might be more feasible.

Similarly, to intervene early in the course of schizophrenia will likely be an important future area of research. What if you could get in early to try to prevent the cascade of events that might follow? These are I think important areas for future research.

In terms of rehabilitation of chronic patients, I think that we need to impress upon people that the brain is more plastic than anyone thought. If there are deficits, we may be able to teach patients specific techniques to get around some of these deficits. With respect to early in the course of illness, I think more of a focus is needed here, prior to what is likely progressive changes in at least a subgroup of patients with schizophrenia.

With the imaging studies, we need to put structural, diffusion, and functional MRI together in a multimodal fashion, in the same patient—to really understand what’s going on in the brain using all the imaging tools we have. This is an area our group has been pushing forward on.

I don’t know how people answer questions like this! I think I would do what I’m already doing, but on a larger scale. I would add looking at postmortem brains to validate the DTI studies, and I would have a genetics group to work more closely with. It would be nice to get people from pharma involved as well.

We really do want to put all these pieces together—genetics, imaging, postmortem, pharma—and say, "Listen, this is a complex disorder, so let’s try to approach it from all these different areas." I would want to try to answer the same questions using all the tools we have available to us. This is truly a devastating disorder and we need to make it more tractable.

Martha E. Shenton, Ph.D.
Director, Psychiatry Neuroimaging Laboratory
Brigham & Women’s Hospital
Harvard Medical School
Boston, MA, USA
and
Clinical Neuroscience Division
Laboratory of Neuroscience
VA Boston Healthcare System
Brockton MA

Sylvain Bouix, Ph.D., Investigator and Instructor
Marc Niethammer, Ph.D., Investigator and Instructor
Psychiatry Neuroimaging Laboratory
Brigham & Women’s Hospital
Harvard Medical School
Boston, MA, USA