To investigate the relationship between iron staining and magnetic resonance (MR) imaging measurements in postmortem subjects with multiple sclerosis (MS).

Institutional ethical approval was obtained, and informed consent was obtained from the subjects and/or their families. Four MR imaging methods based on transverse relaxation (T2 weighting, R2 mapping, and R2* mapping) and phase imaging were performed by using a 4.7-T system in three in situ postmortem patients with MS less than 28 hours after death and in one in vivo patient 1 year before death. Iron staining with the Perls iron reaction was performed after brain extraction. Region-of-interest measurements from six subcortical gray matter structures were obtained from MR imaging and then correlated with corresponding locations on photographs of iron-stained pathologic slices by using a separate linear least-squares regression in each subject. Iron status of white matter lesions, as determined by staining, was compared with appearance on MR images.

R2* mapping had the highest intrasubject correlations with iron in subcortical gray matter (R² = 0.857, 0.628, and 0.685; all P < .001), while R2 mapping (R² = 0.807, 0.615, 0.628, and 0.489; P < .001 and P = .001, .034, and .001, respectively), phase imaging (R² = 0.672, 0.441, 0.596, 0.548; all P ≤ .001), and T2-weighted imaging (R² = 0.463, 0.582, 0.650, and 0.551; all P < .001) had lower but still strong correlations. Within lesions, hypointense areas on phase images did not always represent iron. A hyperintense rim surrounding lesions on R2* maps was only present with iron staining, yet not all iron-staining lesions had R2* rim hyperintensity.

All four MR imaging methods had significant linear correlations with iron and could potentially be used to determine iron status of subcortical gray matter structures in MS, with R2* mapping being preferred. A reliable method of determining iron status within MS lesions was not established.

© RSNA, 2013