University of Melbourne
Department of Biomedical Engineering

Author Of 1 Presentation

Imaging Poster Presentation

P0626 - Quantitative Susceptibility Mapping at 7 Tesla detects ongoing active lesions in relapse-free RRMS patients (ID 1540)

Abstract

Background

Microglia are iron-rich cells, found surrounding multiple sclerosis (MS) lesions in areas of active inflammation. Quantitative Susceptibility Mapping (QSM) can detect this increased iron and thus could be a novel MRI biomarker for microglia-associated inflammation in the brain. The proportion of patients with active inflammation is currently unknown, as is the proportion of MS lesions seen on conventional MRI sequences that are active across patients. Ultra-high field MRI (7 Tesla +) provides superior signal to noise and susceptibility contrast making it the optimal method for detecting iron in MS lesions and tracking active inflammation.

Objectives

To compare the number of lesions with positive QSM signal indicating active inflammation with lesion size and number in patients with relapsing-remitting MS (RRMS) using 7T MRI.

Methods

21 people with RRMS (mean ± SD age = 42 ± 11 yrs; sex: 2m/19f; mean ± SD disease duration = 5.5 ± 3.2 yrs; all EDSS < 4; no relapses in previous 12 months) were scanned using MP2RAGE anatomical and multi-echo gradient echo sequences on a Siemens 7T MAGNETOM MRI scanner. MP2RAGE was used to identify lesions and then co-registered to QSM (calculated from gradient echo phase images using an in-house pipeline). The number of lesions with an average QSM value over 0 (QSM+), indicating the presence of iron associated with active inflammation, were compared to the total number and total volume (log10 transformed) of lesions across patients using linear regression.

Results

The number of lesions in patients ranged from 3 to 92 (mean ± SD = 33 ± 25) and volumes ranged from 26 to 14505 mm3 (mean ± SD = 2554 ± 3445 mm3). Across all patients, the average proportion of QSM+ lesions was 0.61 (95% CI = 0.50-0.72, R2=0.87, p<0.0001), and for each log10 cubic millimeter change in the lesion volume, there were an additional 15 QSM+ lesions (95% CI = 7.0-24, R2=0.43, p=0.0012). There were no associations between the proportion of QSM+ lesions and any disease or demographic variables.

Conclusions

Irrespective of disease severity or duration, the proportion of QSM+ lesions was highly consistent. Based on the assumption that QSM+ lesions are undergoing active inflammation, our results indicate that around ~60% of lesions in RRMS patients could be active.

Collapse