Rapid discoveries in genomics and neuroscience stop short of the doctor’s office
Simon K. was worried that his mind had started to splinter. The 19-year-old engineering student was having trouble keeping up with his courses and had taken to drinking himself into a stupor on weekends. A typical description of sophomore year — except for the voices. Every once in a while, Simon thought he heard people calling his name, even when nobody was around. That couldn’t be good. His uncle had suffered his first psychotic break in college, and Simon wondered whether he, too, was tumbling toward schizophrenia.
His parents dragged him to a psychiatrist, who ordered a series of genetic tests and a brain MRI. When the results came in, Simon was relieved to find out that he had only four versions of the 319 genetic variations linked to schizophrenia, and that the brain scans revealed none of the telltale changes in his cortex that presage the onset of the disease. The chances that he would develop schizophrenia were quite small, he was told. Simon went back to school, feeling full of hope for his future.
If only this story were even remotely true. Although researchers know that schizophrenia has a strong genetic component, they have no clue how many genes contribute to the disorder or how such genes interact with environmental triggers. They also have yet to identify a reliable anatomical indicator of incipient schizophrenia that could be identified unequivocally with a brain scan.
The same can be said for anxiety disorders, autism, depression, post-traumatic stress disorder, and every other form of mental illness, aside from the few neurological conditions that generally strike people in their later years, such as Alzheimer’s disease and Huntington’s disease. Despite decades of work looking into the biology of brain disorders, clinicians still lack objective tests that could be used to diagnose the mental illnesses affecting more than 40 million Americans. Instead doctors resort to the age-old and imprecise practice of assessing symptoms and family histories to determine whether a patient has a particular disorder.
That’s not for lack of trying. Research has advanced rapidly, especially within the past decade, in understanding the genetics and anatomical signals of mental diseases. Scientists have developed new ways of peering into the brain and are retrieving some intriguing hints about the way disorders disrupt normal neural circuitry. For example, scientists at the University of California at Los Angeles have reported that people with schizophrenia lose part of their cortical gray matter and white matter, particularly in regions associated with planning and memory. If that patterns holds up in subsequent studies, it could eventually help in diagnosing and treating the disease.
“What’s exciting now is that we have some of the tools to answer these questions,” says Thomas R. Insel, director of the National Institute of Mental Health. But “we’re not there yet,” he warns. “It’s unfortunate that we just haven’t been able to move fast enough to get the kind of clearer picture that we need to provide more validity to the diagnostic scheme.”
The lack of progress will become all too evident in 2012, when the American Psychiatric Association is to publish the next edition of the mental-health bible known as the Diagnostic and Statistical Manual of Mental Disorders. The current version, the Fourth Edition, Text Revision (or DSM-IV-TR), was released in 2000. Since even before that 992-page tome landed on clinicians’ desks, the association has been working to produce the updated version, which is now taking shape.
The DSM plays a crucial role in determining how clinicians treat mental disorders, what kinds of treatments health providers will pay for, and how researchers study those diseases. By providing a set of criteria for diagnosing each particular mental disorder, the DSM unifies the mental-health field, enabling doctors, social workers, nurses, scientists, and others to consistently distinguish the different categories of disease.
The current DSM, however, has grown feeble with age. Some of its troubles stem from past authors’ predilections for treating each disorder separately, as distinct from all others and from normal behavior. But clinicians and researchers know that many patients suffer from multiple disorders, some of which may stem from similar roots in the brain. Moreover, the line between normal and diseased states is often fuzzy, as in anxiety and depression, suggesting a graded continuum between mental health and disease, rather than a simple either/or state. The sharp boundaries between diseases in the DSM has led in some cases to the proliferation of the NOS diagnosis, an appendix meaning “Not Otherwise Specified.” Physicians resort to that label when patients do not match the strict criteria for certain disorders — for example, in many cases of autism.
The developers of the forthcoming DSM-V have heard the complaints and are striving to incorporate a more “dimensional” approach, looking at relationships among disorders and between disorders and normal behavior. “Having a greater dimensional strategy for syndromes rather than yes/no gives you much greater power from a research standpoint and also from a clinical standpoint,” says Darrel A. Regier, director of research at the American Psychiatric Association and vice chair of the DSM-V task force. For example, he says, 50 percent of people with major depression also suffer from anxiety disorder and tend to fare worse. So the developers of the DSM-V are exploring ways to group those disorders and others that often cluster together.
Some of the tools of neuroscience might help in that process, says Regier. Genetic, imaging, and pharmaceutical studies have suggested that many disorders involve the neurotransmitter serotonin, which carries electrochemical signals from one nerve cell to another. This chemical is the target of drugs called selective serotonin reuptake inhibitors, such as fluoxetine hydrochloride, which is sold as Prozac. The serotonin connection may reveal groupings among related disorders or how one apparent disorder is actually multiple syndromes.
While the advances in research could help authors cluster diseases, the neuroscience data are not likely to influence how the disorders are actually defined in the DSM-V, or what treatments are prescribed. That failure may come as a shock to the public, which for several years
has been lapping up popular reports about advances in brain-imaging, genetics, animal-brain studies, pharmaceutical research, and other elements of neuroscience. Stories concerning new findings about the brain regularly appear on the cover of popular magazines such as National Geographic, Newsweek, and Time. In November the annual meeting of the Society for Neuroscience, in Washington, generated dozens of articles about the brain science of addiction, love, sex, and sleep, to name but a few topics.
The expectations for advances are not necessarily misplaced. The society itself was established in 1969, and membership has mushroomed from 500 to nearly 40,000, making neuroscience one of the fastest-growing areas in science. New brain-sensing tools have come on line, and the genomics revolution has provided extraordinary insights into the way that genes build the architecture of the brain.
“From a public-health point of view, physicians, patients, and their families should be frustrated with the slow progress of applying the findings of genetics and neuroscience to improve diagnosis and treatment,” says Steven E. Hyman, provost of Harvard University and a former director of the National Institute of Mental Health. But the expectations may be unrealistic, he says. “Looked at in another way, the brain is the most challenging object on which humanity has ever focused its science. It’s not surprising that the progress is slow and hard won.”
“The doctor in me is extremely impatient, and the scientist in me tells me this is very challenging stuff that will require, unfortunately, more time.”
Neuroscientists say their field is still quite young. Although studies into the biology and function of the brain reach back into the 19th century, the field truly emerged only in the past three decades, spurred in part by new technologies that could probe the working brain without having to invade the skull. In the 1970s, scientists developed positron emission tomography, or PET, a method to monitor the movement of radioactive tracers as they travel through the blood to active regions of the brain. The same period saw the birth of magnetic-resonance imaging, or MRI, which can map structures in the brain by reorienting the spin of atomic nuclei. Then, in the 1990s, researchers developed so-called functional MRI technology, or fMRI, to track blood flow in the brain. That advance has significantly broadened opportunities to study the brain in action because machines capable of performing fMRI are relatively common, and the new technology does not require the radioactive tracers used in PET scans.
On another front, researchers working with animals have developed models for studying disorders such as alcohol addiction and anxiety. Together with the genomics revolution, the multipronged approach of neuroscience has yielded results that could eventually pay off in diagnoses and treatments of mental disorders.
For an idea of how the field is developing, consider the work of Tyrone D. Cannon, a professor of psychology and psychiatry at the University of California at Los Angeles. For two decades he has investigated the genetics of schizophrenia. He has married those studies with MRI scans of people who suffer from the disorder and of their close relatives.
The combined approach has started to show how schizophrenia disrupts specific brain circuits as it takes hold. Cannon has captured the disease’s progress by starting with people who exhibit precursor symptoms that sometimes lead to schizophrenia. For example, some adolescents and young adults hear voices that they know are imagined and are not overwhelming. About 35 percent of those with such psychotic-like symptoms progress over two years to full-scale psychosis, at which point the voices become severely damaging and appear to come from real people, says Cannon. By scanning subjects at different stages, his group has found that people who eventually develop schizophrenia lose gray matter in the prefrontal cortex and parts of the temporal lobe, toward the front and sides of the brain.
In other studies, he has looked at the development of white matter, the myelin insulation that builds up around the axons of neurons, which carry signals from one cell to another. Such insulation increases the speed of the electrical transmissions among different parts of the brain. Normal adolescents acquire more myelin over time, allowing disparate brain centers to communicate more effectively, like several different processors working together in one computer.
Using relatively new imaging techniques that can detect myelin, Cannon’s team has studied people at various stages on the spectrum between prepsychosis and schizophrenia. His preliminary results indicate that people with schizophrenia do not grow as much myelin as those who remain healthy. The diseased brains have weaker connections and cannot work as efficiently as normal brains, he says.
As a next step in the research, Cannon will track individual patients over time, to see whether their brain development matches the reduced myelin pattern observed in the snapshots of different people. “We need that information before we can have conclusive things to say,” he says.
The hope is that the imaging might eventually provide clinicians with a tool to reliably predict whether an adolescent who hears voices will progress to develop schizophrenia. Right now there is not much that doctors could do with such a test, except to closely follow patients at risk of schizophrenia and ensure that they remain engaged with their families, peers, and schools. But pharmaceutical companies are seeking to develop drugs that could promote the growth of connections impaired in schizophrenia, says Cannon.
Those imaging tools won’t be ready in time for the DSM-V. “Maybe in the next five to 10 years, imaging may become useful diagnostically, and in that same time frame, genomics will reach the same level of importance,” says Cannon. Although the pace may feel too slow for patients and their families, he says, “we’ve learned more in the past five years than in the preceding 50 years about schizophrenia. It’s a period of rapid discovery, and it can’t be too far down the road before these discoveries will yield real changes in how patients are experiencing this disorder.”
Looking toward the horizon, researchers point to post-traumatic stress disorder, or PTSD, as another example of progress in neuroscience that might ultimately help patients. Much of the recent work has involved using fMRI to probe which circuits of the brain are particularly associated with that illness.
In one study, pictures of happy faces and fearful faces were shown to Vietnam veterans and firefighters, some of whom had PTSD. Brain scans taken during those sessions revealed that the people suffering from the disorder had exaggerated activity in the amygdala, a deep-seated area associated with the recognition and response to fear. At the same time, the PTSD subjects showed less activity in the medial prefrontal cortex, a part of the brain thought to exert control over the amygdala and circuits that sense trouble.
Lisa M. Shin, an associate professor of psychology at Tufts University who led that study, is following up by testing whether brain scans can help predict how patients will respond to therapy. In a small preliminary experiment, she has found that PTSD sufferers with the most activity in the medial prefrontal cortex when shown fearful faces responded the best to a treatment called exposure therapy, which consists of getting patients to relive the traumatic experience. People who had more-subdued reactions in that region did not do so well with that type of therapy.
“I think we’re getting there,” says Shin. “We’ve identified the right circuits, and those circuits are telling us something important clinically.”
Other fMRI experiments have shown that genetic differences play a role in determining the strength of the fear circuitry. These studies have focused on the 5-HTT gene, which regulates how much serotonin is available to ferry information between cells. The gene comes in a long and a short form.
In a study of healthy subjects, conducted at the National Institute of Mental Health, people who had the short version of the serotonin-transporter gene displayed more activity in the amygdala when they saw fearful or threatening faces. So that genetic variant, which is present in more than half of the American population, apparently causes the brain to construct a highly reactive threat center.
On its own, however, the short variation does not doom a person to a life of illness. In an influential study from 2003, a team led by researchers at King’s College London found that people with the short variation were more prone to depression and suicidal thoughts, but only if they had suffered major stressful events within the past five years. So the long variation of the transporter gene appeared to offer a protective effect, while the short version made people less resilient to crises.
Like others involved in this type of work, Shin says her findings regarding PTSD are not yet ready for the clinic — or for the DSM. But she says the research has progressed at an amazing pace: “I would hope that neuroimaging at some point would be helpful with diagnosis and treatment.”
When that day comes, it could allow doctors to intervene before illnesses take their toll. For now, clinicians do not recognize disorders like schizophrenia, depression, and autism until the symptoms are obvious. “The problem we have in this field is we make the diagnosis fairly late in the game,” says Insel, of the National Institute of Mental Health.
Psychiatry now stands where cardiology was a generation ago, when doctors diagnosed heart disease after a person suffered a heart attack. The key, says Insel, is to develop the knowledge and tools that will enable clinicians to spot trouble early, before the neural circuitry of the brain gets jumbled by disease.
Richard Monastersky is a senior writer at The Chronicle.
http://chronicle.com Section: The Chronicle Review Volume 55, Issue 15, Page B7