3.3 Clinical Relevance of Brain Macromolecular Proton Fraction Changes in MS

Friday, May 31, 2013: 4:30 PM
Vasily L Yarnykh, PhD , Radiology, University of Washington, Seattle, WA
James D Bowen, MD , Swedish Medical Center, Seattle, WA
Alexey A Samsonov, PhD , Radiology, University of Wisconsin, Madison, WI
Pavle Repovic, MD, PhD , Swedish Medical Center, Seattle, WA
Angeli Mayadev, MD , Swedish Medical Center, Seattle, WA
Bart P Keogh, MD, PhD , Swedish Medical Center, Seattle, WA
Beena Gangadharan, PhD , Swedish Medical Center, Seattle, WA
Hunter R Underhill, MD, PhD , Radiology, University of Washington, Seattle, WA
Kenneth R Maravilla, MD , Radiology, University of Washington, Seattle, WA
Lily K JungHenson, MD , Swedish Medical Center, Seattle, WA


Background: Macromolecular proton fraction (MPF) is a key biophysical parameter determining magnetization transfer (MT) between water and macromolecules in tissues. MPF has recently attracted significant interest as a quantitative MRI biomarker of demyelination in MS. A recently published fast single-point MPF mapping method enables clinical applications of this technology with high image quality, whole-brain coverage, and a reasonable scan time.

Objectives: Establish associations of MPF in normal appearing brain tissues (NABT) and MS lesions with clinical disability and disease course.

Methods: Images were acquired on a 3 Tesla MRI scanner from 14 healthy controls (HC), 19 relapsing-remitting (RRMS), and 11 secondary-progressive (SPMS) MS patients. MPF mapping protocol included a gradient-echo sequence with MT saturation, a reference sequence without saturation, three-point variable flip-angle T1 mapping, and fast B0 and B1 field-mapping sequences. Whole-brain three-dimensional MPF maps were obtained with 1.5x1.5x4 mm3 spatial resolution and 15 min total acquisition time. MPF maps were segmented into normal appearing white matter (NAWM), gray matter (NAGM), and lesions. T2-weighted FLAIR images were used to define lesions. Mean MPF values in each tissue class were compared between subject groups using independent t-test. Associations between imaging variables and clinical disability scores (EDSS and MSFC) were assessed using Pearson’s correlation coefficient (r). Data are reported as mean±SD.

Results: MPF monotonically decreased from HC to RRMS and from RRMS to SPMS in both NAWM (13.5±0.4% vs. 12.6±0.6% vs. 11.8±0.8%; P<0.01 for all comparisons) and NAGM (7.4±0.3% vs. 7.0±0.3% vs. 6.3±0.4%; P<0.001 for all comparisons). MPF in lesions was significantly lower in SPMS compared to RRMS (7.4±1.0% vs. 8.5±0.8%; P<0.01), whereas lesion volume (LV) was larger in SPMS (21.2±17.7 ml vs. 7.4±6.5 ml, P<0.01). MPF in NAGM demonstrated the strongest correlations with disability (r=-0.70 for EDSS, r=0.81 for MSFC, P<0.001) followed by MPF in NAWM (r=-0.56, P<0.01 for EDSS; r=0.72, P<0.001 for MSFC). The weakest correlations were observed for MPF in lesions (r=-0.42, P<0.05 for EDSS; r=0.50, P<0.01 for MSFC), similar to LV (r=0.42, P<0.05 for EDSS; r=-0.57, P<0.001 for MSFC).

Conclusions: Pathological MPF changes in NABT and lesions are consistent with progressive demyelination and associated with MS severity. MPF values in NABT demonstrate a higher clinical significance than MPF in lesions and LV. MPF in NAGM appears the strongest predictor of disability. This finding emphasizes a critical role of gray matter demyelination in MS. Our data establish clinical relevance of the fast and robust MPF mapping technology applied in this study. This new quantitative imaging method provides a promising approach for quantitative assessment of myelin damage and repair in both white and gray matter for various MS clinical studies.