John DeWitt Port MD, Presenter: Nothing to Disclose

Sencan Solay Unal MD, Abstract Co-Author: Nothing to Disclose

Mandie J Maroney-Smith, Abstract Co-Author: Nothing to Disclose

Joel Patrick Felmlee PhD, Abstract Co-Author: Nothing to Disclose

David A Mrazek MD, Abstract Co-Author: Nothing to Disclose

To test the hypothesis that magnetic resonance spectroscopic imaging (MRSI) can detect metabolic markers of bipolar disease

A prospective single-blinded case-control study design was used. Drug- and medication-naive bipolar patients without other underlying medical conditions were recruited from the inpatient and outpatient psychiatric services at Mayo Clinic. Age-, sex- and handedness-matched normal control subjects were recruited from the general population. Diagnosis (bipolar or normal) was confirmed by psychological testing. Subjects underwent mood assessment and MRSI scanning on the same day. MRSI was performed on a GE 3T longbore scanner using PRESS-MRSI (TE=30, voxel size 1.25x1.25x1.3cm = 2.02cc); two slices were obtained, one through the basal ganglia and one through the anterior cingulate gyrus. MRSI metabolite data were quantified using LCModel; this quantitative data was overlaid onto anatomical images and regions of interest selected for further statistical analysis. Metabolite ratios were used for the analysis.

Nineteen bipolar subjects (12 females, 7 males) were matched with nineteen normal control subjects. The average age for each cohort was 32 years. Of the bipolar subjects, 7 were diagnosed as bipolar I, 8 as bipolar II, and 4 as bipolar NOS; 13 were inpatients and 6 were outpatients. Four regions of interest showed statistically significant differences between bipolar and normal subjects. Right frontal white matter mI/Cr was significantly increased (p<0.039, Wilcoxon signed rank), and right parietal white matter mI/Cr was significantly decreased (p<0.004) in bipolar subjects relative to normal controls. Right lentiform Cho/Cr was significantly increased (p<0.034) in bipolar subjects relative to normal controls. Left occipital gray matter NAA/Cr was significantly decreased (p<0.043), and mI/Cr was significantly increased (p<0.031) in bipolar subjects relative to normal controls.

Proton MRSI techniques using very high field MR scanners may prove helpful for the diagnosis of bipolar disease.

Port, J, Unal, S, Maroney-Smith, M, Felmlee, J, Mrazek, D, Imaging Markers of Bipolar Disease: Evaluation of Proton Magnetic Resonance Spectroscopic Imaging at 3T.  Radiological Society of North America 2004 Scientific Assembly and Annual Meeting, November 28 - December 3, 2004 ,Chicago IL. http://archive.rsna.org/2004/4411606.html