Bipolar disorder

Researchers are developing a promising new tool to help in the diagnosis of bipolar disorder, a complex condition affecting 2.3 million Americans.

At a recent meeting of the Radiological Society of North America, scientists at Mayo Clinic reported on the breakthrough that utilizes magnetic resonance spectroscopic imaging (MRSI) to identify unique patterns of metabolic activity in the brains of bipolar patients.

Characterized by extremes of emotional highs and lows, bipolar disorder is currently diagnosed largely on the basis of symptoms and family history (see page 26), but the condition often remains undiagnosed, or misdiagnosed, for years. Sadly, a large percentage of young people experiencing the first severe depression of bipolar disorder commit suicide. Armed with better tools, mental health professionals can more accurately diagnose the problem and tailor appropriate treatments.

"The psychiatric community clearly needs a tool to help diagnose bipolar disorder. We are hopeful that very high-field MR spectroscopy will prove helpful by identifying metabolic markers of the disease," explained Dr. John D. Port, an assistant professor of radiology and a consultant at the Mayo Clinic. "Bipolar disorder is challenging to diagnose because individuals can cover up the symptoms of the illness or may recognize only their depression, not the manic phase of the disorder. It's also important to be able to distinguish bipolar disorder from major depression because a mistaken diagnosis can result in the wrong therapy and unstable moods for years."

The Post interviewed Dr. Port about the new technology and its potential use in bipolar disorder.

Post: Could you tell us about the use of MRSI and how it might be useful in diagnosis of bipolar disorder?

Dr. Port: The main problem with imaging a psychiatric patient is that typically the images of the patient's brain anatomy are normal, even though the person sitting on your table is hallucinating, manic, or obviously suffering from a mental disorder. In that regard, conventional neuroimaging has failed psychiatry, and is of limited use in helping with diagnoses.

Now, magnetic resonance spectroscopy and MRSI techniques have been around for 20 years but have never really caught on because they take a long time to perform and are hard to do well. When I came to Mayo, I looked though the literature and read every paper where researchers had used spectroscopy. It is pretty well established that psychiatric diseases are a result of chemical imbalance, and spectroscopy seems to be the way to look at chemical imbalances, but there were many problems in the literature with the techniques used. The patients, for example, in studies are often taking many different medications. The questions arise when looking at the chemicals in the brain: Are the differences you see because of the medication, or are the brains really different?

After reviewing the entire literature, I identified problems and created this new technique, which pulls together pieces of techniques to create something rather unique. We used MRSI done on a double-strength 3T magnet as our basic technique, then utilized something called tissue volume correction to compare the absolute concentrations of the chemicals in the brains, rather than just ratios. We also recruited as pure a sample of bipolar subjects as possible, requiring that they be drug-free.

Post: What chemicals in the brain do you measure?

Dr. Port: We look at many--NAA, choline, and creatine are the big three, but we also look at glutamate, glutamine, myoinositol, and lactate, among others. When beginning the study, I threw away any assumptions about what I was going to see. Many people have written papers about the theoretical circuitry behind schizophrenia, for example, and assign different functions to different areas of the brain. They hypothesize that if a given area breaks, then that defect will, for example, cause symptoms of mania or depression. Many previous spectroscopy studies have started with this idea and investigated different areas thought to be abnormal.

The problem with that technique is that if you have an abnormality somewhere else, you are not going to find the difference. We call this a sampling error. By sampling large numbers of voxels (little cubes of tissue in the brain noninvasively sampled through an MRI scanner) all over the brain at the same time, I looked at areas not involved in the theoretical circuits. By doing so, we discovered an area not previously thought to be involved in bipolar disease in the right parietal white matter of the brain, where I discovered abnormalities between people with different types of bipolar disease. Specifically, I looked at severe cases of bipolar 1, bipolar 2, and bipolar not otherwise specified (NOS). We wanted clear-cut cases.

Post: How did you discover the abnormalities?

Dr. Port: Every single scan was read by board-certified radiologists as normal, and once again, the research supported that there is no gross structural differences between a bipolar brain and a normal brain (although there have been a few papers finding small anatomical differences in groups of bipolar patients). We then looked at 14 separate regions in each subject to see if we could detect chemical differences. After a statistical analysis on the metabolite concentrations in these 588 voxels, I found that in two brain regions, specifically in the right frontal white matter and right lentiform nucleus, four different metabolites were statistically different between normal control patients and bipolar patients.

In our study, we found that in right frontal white matter, myoinositol levels are increased in bipolars relative to normals, and in the right lentiform nucleus, NAA, glutamate+glutamine (the levels are so low these compounds are often considered together), and creatine are all decreased in bipolars relative to normals.

Post: How can you ascertain these differences from the scans?

Dr. Port: When we are doing anatomical imaging in our institution, our images are black and gray. We estimate the pathology based on the different anatomical patterns that we see. With the spectroscopy, I generate a metabolite map, in which I colorcode the metabolites by concentration. The higher the concentration, the brighter the metabolite appears on my metabolite map. We then use the concentrations of metabolites from these maps to do our statistical tests.

Post: How many people were enrolled in the study?

Dr. Port: Even though a small sample with 21 people, it turned out to be one of the largest samples ever studied; these studies are very hard to do. It took me two years to find the appropriate people, so our study was particularly rigorous because we had drug- and medication-naive bipolar patients. In recruiting patients, we had to find newly diagnosed individuals who stopped taking medications months before, or people who were not on medications.

Post: Were you able to differentiate between the different bipolar groups?

Dr. Port: Yes. Looking at bipolar patients by diagnosis, we found that we could separate bipolar 1, bipolar 2, and bipolar NOS by looking at the right parietal white matter.

Post: Are these findings specific and sensitive enough to use for a clinical diagnosis?

Dr. Port: No, they are not and this is a key point. This is a preliminary study, evaluating a new technique. We need more work to refine the tool where it can diagnose individual patients. We are going to be working on this in the next two to three years.

Post: Prior to your research, has there been any objective diagnostic test for bipolar disorder?

Dr. Port: No, and that is the big problem. The diagnosis of bipolar disease is purely a clinical diagnosis, and it can take years or decades for people to receive an accurate diagnosis of bipolar disease. The tragedy is that oftentimes patients will go to their doctors in the depressed phase of the disease, and they will be given antidepressants because it looks like depression. The antidepressants can trigger a manic episode, so medicating patients with bipolar disorder is difficult.

Last year, there were 30,000 suicides among depressed people, who may never have had the diagnosis of bipolar because they had not yet experienced the manic phase of the disease. Clearly, if we could develop an imaging tool (or any tool, for that matter) to objectively diagnose bipolar disease, much of this morbidity and mortality could be avoided. I am a neuroradiologist. I am not a psychiatrist, but I work closely with two wonderful psychiatrists, and that is our goal--to develop a tool that we can use to assist psychiatrists in making a diagnosis.

Post: It certainly could be useful, especially for the large amount of young people who struggle with their first bout of depression as adolescents.

Dr. Port: Adolescent depression and adolescent suicide are horrible. I also am conducting a number of small studies looking at adolescent bipolar disease and depression. We are using proton and phosphorus spectroscopy in adolescents with bipolar disease, and we have focused on one tiny brain region thought to be involved in the disorder.

Post: When will the test be available?

Dr. Port: I hope that in the next two to three years, we can get the data needed to ascertain high sensitivity and specificity numbers. The goal is to be able to refine this technique where I can use it as a lab test. Presently, we are the only center on the planet that is doing spectroscopy this way.

Post: How can the technology assist in diagnosis?

Dr. Port: In the near future, I hope we can offer a tool at any medical center with a 3T magnet and the right software so that if a patient walks into a doctor's office and is severely depressed, that family physician will send them for a psychiatric MRSI and get help. In other words, the psychiatric MRSI will support the diagnosis of either bipolar or depression, then the physician can either refer them to a psychiatrist or treat them in his office. I would like to create for the first time in history an objective clinical tool that psychiatrists can use, like a blood or an imaging test, that will help diagnose psychiatric disease.

Post: If people are interested in undergoing the test at Mayo or learning more about the progress, how can they do that?

Dr. Port: My e-mail is port.john@mayo.edu, and I would be glad to let people know where we are in our work.

COPYRIGHT 2005 Saturday Evening Post Society
COPYRIGHT 2005 Gale Group