Functional MRI brings hope in psychiatric drug therapies

New research on the ability of functional MRI to assess the effects of antidepressants in psychiatric patients is part of a small but growing body of evidence showing the modality's potential in neuroscience drug discovery.

"Drug discovery data are promising in terms of future clinical and commercial applications. But there needs to be a lot more consolidation of the early academic results, and the technique must be road-tested quite comprehensively before you could have real-life application," said Prof. Edward Bullmore, a professor of psychiatry at Addenbrookes Hospital in Cambridge, U.K.

In a trial funded by GlaxoSmithKline, researchers examined patients with severe depression, correlating changes in facial expression with functional brain changes over the course of eight weeks of antidepressant use. At a Horizon Seminar held in October at Churchill College in Cambridge, Bullmore discussed the results, which had been published in September (Arch Gen Psychiatry 2004;61:877-889).

In the study, depressed patients had lower capacity for activation in the left amygdala, ventral striatum, and frontoparietal cortex and a negatively correlated increase of dynamic range in the prefrontal cortex. Patients who improved with the course of antidepressants showed a reduction of dynamic range in the pregenual cingulate cortex, ventral striatum, and cerebellum. The researchers concluded that fMRI may be a useful indicator of antidepressant treatment response.

A technique that could spot the winners and losers among hundreds of psychiatric drug compound contenders would have great clinical and commercial value. Moving a psychiatric drug through development stages and complex trials to regulatory approval can cost from $500 million to $800 million, said Dr. Richard Hargreaves, vice president of imaging at Merck Research Laboratories in West Point, Pennsylvania, U.S. Most research projects are conceptually flawed, and some psychiatric drugs that do make it through may fail, at great expense, once they reach the market.

"We have to get better at picking the concept and picking the molecule," he told delegates at the Cambridge seminar. "It is a sobering thought that most of the projects we run in industry are conceptually flawed. If we are going to choose a target, we are likely to choose the wrong one. If we choose the wrong one, we need to fail quickly and cheaply, so we have dollars left in the pot."

Demand for drugs to treat diseases of the central nervous system is likely to grow in the future, given the aging population and the expected rise in incidence of conditions such as Alzheimer's disease. CNS drugs have become better tolerated and patient compliance has improved over the last 25 years, but the range of treatments has not expanded greatly, according to Hargreaves.

"There has been an explosion in new neuroscientific methods that should make it easier to find drugs," Bullmore said. "Therapeutics have not yet been revolutionized by the neuroscientific developments of the last 20 years."

Functional MRI is one tool that could help industry take advantage of these developments. Unlike PET, MRI does not involve radiation and is therefore more practical for pharmacological studies, which tend to require repeated scanning. The body of evidence for fMRI in pharmacological studies is small, however; Bullmore reviewed the literature and found only about 50 studies in the field. But trial activity has been growing rapidly (Trends in Pharmacological Sciences 2004;25[7]:366-374), and studies indicate that fMRI could help predict treatment response and identify new compounds for drug development.

"You can see different effects on fMRI signals of other drugs-antipsychotics, benzodiazepines-every major class of pyschotropic drugs has been studied with fMRI. All show a change in brain activation," Bullmore said.

In the case of antidepressants, preliminary research is promising. Drug companies would like to see a study of the use of fMRI to chart functional changes in healthy volunteers or people with transient depression and the use of this information to predict therapeutic effects in very sick people.

Drugs currently must be tested in a large number of treatment-naive patients with certifiable severe depression, and as a result, patients are hard to find and trials are expensive and difficult to conduct.

"That would be the holy grail with MRI. It would be incredibly useful, changing the way clinicians manage patients and drug companies manage the drug development process," he said. "The jury is still out on whether functional MRI or any other surrogate marker can do that for a drug company. I think it could, but to be sure of that you would need to run additional experiments. The technique is quite new."