Prediction of Antidepressant Treatment Response Using Magnetic Resonance Imaging (MRI)

Christine DeLorenzo, Ph.D.

State University of New York (SUNY), Stony Brook

Funded in September, 2015: $100000 for 3 years
LAY SUMMARY . ABSTRACT . BIOGRAPHY .

LAY SUMMARY

back to top

Using new powerful MRI sequence to predict whether depressed patients will respond to SSRI treatment

Clinicians soon may be able to use a type of MRI imaging, called diffusion spectrum imaging, to predict whether an individual patient with depression is likely to respond to treatment with the class of antidepressants called SSRIs (selective serotonin reuptake inhibitors).        

Clinicians need to choose the best treatment option for a depressed patient at the outset, since it takes a month or more for a treatment to demonstrate effectiveness. Selection is a trial and error process, though, because: there are several classes of antidepressants; each has its own method of action in the brain; and clinicians currently have no brain “biomarker” to guide them in determining what mode of action is indicated for an individual patient. SSRIs are the most frequently prescribed antidepressant class. Yet a striking two–thirds of people with depression treated with SSRIs do not achieve treatment remission.            

One type of MRI imaging, called diffusion tensor imaging (DTI) has shown some evidence of being able to indicate whether an individual patient is likely to respond to SSRI treatment. Investigators now plan to use the even more powerful diffusion spectrum imaging (DSI) MRI technique to see if it can substantially improve the capacity to accurately predict which patients are likely to benefit from SSRI treatment.  

SSRIs act by blocking brain cells from taking up serotonin. This electrochemical “transmitter” is involved in mood regulation. Serotonin neurons primarily originate in a small brain region called the “raphe nucleus.” From there, the cells’ axons (communication cables) connect to other regions of the brain including the amygdala, which is implicated in depression. Bundles of these communication cables are the brain’s “white matter” connecting the raphe nucleus to the amygdala and they can be imaged. Diffusion imaging indicates whether axons are healthy (well insulated, running in parallel).  In the investigators’ prior studies, DTI imaging has indicated that patients’ white matter health prior to SSRI treatment is correlated with patients’ subsequent response to SSRIs—the healthier the white matter tracts, the better the SSRI response. But so far, imaging results of white matter health have explained less than half of the variability in patients’ treatment response.

The investigators hypothesize that they can substantially increase predictive accuracy by using the more powerful DSI imaging technique.  DSI is able to detect white matter fibers that cross each other (rather than running in parallel) and provides increased resolution of white matter, especially in areas surrounding the small raphe nucleus. They will enroll 100 patients with depression, half of whom will be treated with SSRIs while the other half will receive placebo. Patients will undergo DSI imaging prior to treatment, and followed for eight weeks to assess depression remission or recurrence. If investigators find that patients with good white matter health respond effectively to SSRI treatment while patients with poor white matter health do not, they will have preliminary evidence that white matter health is a biomarker for predicting that SSRI treatment will be effective.          

Significance: DSI imaging may provide the first marker in the brain that can be used to predict the likely effectiveness of SSRI treatment for depression on an individual level. 

 

 

ABSTRACT

back to top

Depression affects ~121 million people worldwide and is projected to be the second largest contributor of global disease burden by 2020. Monotherapy with selective serotonin reuptake inhibitors (SSRIs) is the most widely used treatment for major depressive disorder (MDD). However, on average, SSRIs require four to six weeks for onset of action, and two-thirds of patients on SSRIs fail to achieve remission. Although there are alternative treatments, clinicians currently do not have the tools to select a treatment plan based on an individual’s likelihood of remission. If such tools were developed, they could significantly reduce the morbidity and mortality resulting from ineffective treatment trials. Consequently, there is a critical need to identify markers predictive of an individual’s SSRI treatment outcome. Due to the role of the serotonergic system in the pathophysiology of depression and the fact that the most common SSRIs target the serotonin transporter, an index of serotonergic health may provide this prognostic indicator. In a preliminary study performed by our group, we measured the health of white matter connections between the raphe (a region of the brain from which most serotonergic neurons originate) and the amygdala (a region of the brain modulated by serotonin and implicated in depression) in depressed subjects. The location and health of these white matter tracts were quantified using diffusion MRI, a technique that detects water diffusion throughout the brain. Since water molecules preferentially diffuse along healthy, well-myelinated and coherent axons, diffusion MRI can be used to determine trajectories of neuronal bundles, as well as provide an index of the health of the identified neuronal (white matter) tracts, called fractional anisotropy (FA). Following this imaging, subjects received eight weeks of treatment with an SSRI. Correlation between average FA in the diffusion MRI-defined tracts and improvement in depression symptoms after SSRI treatment was then calculated. The results indicated that there is a significant correlation between average FA in the raphe-amygdala tracts and subsequent depression improvement. This suggests that reduced health and/or number of serotonergic fibers terminating at the amygdala are associated with a blunted SSRI response. Moreover, since these differences were not observed in tracts terminating at the hippocampus (raphe-hippocampus tracts), the result is regionally specific. Linear regression analysis indicated that 42% of the variance in SSRI treatment response was explained by pretreatment raphe-amygdala FA. Based on sensitivity and specificity estimates, the clinical utility index was estimated. At 0.64, this marker is defined as “good.” Though there is room for improvement, this very exciting initial finding indicates the strong potential of this image-based marker to yield the first pretreatment marker of SSRI effectiveness on an individual level. Further, recent improvements in diffusion imaging could provide the necessary enhancements for high clinical utility. Therefore, we propose a follow-up study on 100 subjects with the following additions: (1) Taking advantage of Stony Brook’s state-of-the-art imaging facilities, we will use the highest resolution diffusion MRI imaging available, Diffusion Spectrum Imaging, a diffusion MRI sequence designed to better handle the complex crossing structure of fibers. This will provide increased resolution especially in the area surrounding the small raphe. (2) We will compare treatment prediction for SSRIs versus placebo to confirm the prediction is SSRI-specific. Regardless of outcome, this study will provide insight into the pathophysiology of MDD and mechanisms of SSRI action. By design, the study also provides the optimal conditions for identifying the first clinically useful measure of treatment efficacy on an individual level. As such, this image-based marker could significantly improve current clinical practice and the lives of those suffering from depression.

KEYWORDS


Anatomy: Serotonin
White matter
Conditions: Depression
Technology: Antidepressants
Diffusion tensor imaging
MRI