To identify MS lesions characteristic on QSM and identify iron participation at different disease stages by comparing the lesions volume differences on QSM and T2w images.

Two patient cohorts were employed in this study. The first one includes MRI images of MS patients obtained from February to March in 2015 in our MS center. The second one includes patient data from 2011 to 2015. The first one is used to survey the lesion characteristics and the second one is used to track the lesion temporal behavior. Pre- and post-gadolinium T1-weighted (T1w) images, T2w images and a multi-echo gradient echo (GRE) image were included for each patient. QSM was reconstructed from GRE data using the Bayesian method ¹. All other images were co-registered to QSM images.

MS lesions were selected based on their hyperintensity on T2w. The susceptibility value of each lesion was labeled as hyperintense, isointense and hypointense using nearby white matter as reference. Lesions with hyperintense rim on QSM were noted as QSM-ring lesion. The lesion volume were measured manually in 3D on both T2w and QSM images and the volume ratio was calculated as (T2w volume – QSM volume)/T2w volume.

Lesion age was determined using gd-enhancement on T1w+ images as time 0. Four age groups were used (0, 0–6 months, 6–12 months and 12–24 months). Lesions susceptibilities were referenced to normal appearing white matter (NAWM).

In the first patient cohort, a total number of 376 T2w lesions were identified in 51 patients and only 2 lesions were enhancing on T1w+Gd. 272 (72.34%) were hyperintense on QSM and the other 104 (27.66%) including the 2 enhancing lesions were isointense. In the 272 hyperintense QSM lesions, 63 (23.16%) were QSM-ring lesions. Central veins were observed in 181 (66.54%) QSM hyperintense lesions. 142 (52.21%) QSM lesions had a larger volume on QSM images than on T2w images. Example images are showed in Figure 1. Histogram of the volume ratio frequency is showed in Figure 2.

In the second patient cohort, 133 enhancing lesions were found at baseline MRI in 60 MS patients. 32 lesions in 15 patients only had the baseline MRI with QSM (age 0) and 45 patients had follow-up QSM. In the 101 follow-up lesions, 39 lesions from 14 patients were found with age smaller than 6 months, 34 lesions from 16 patients with age ranged from 6 to 12 months and 28 lesions from 15 patients were aged 12 to 24 months.

The 32 lesions were of age 0. 24 of 32 (75%) were nodular enhancing and isointense on QSM. 8 were shell enhancing lesions and hyperintense on QSM with worse-defined boundary. Other lesions older than 0 were non-enhancing and hyperintense on QSM. Three lesions in the 6–12 months group and one lesion in 12–24 months group were ring-QSM lesions. Example image are showed in Figure 3.

The lesions susceptibilities increased with the lesion age, as shown in Figure 4. The volume ratio also increased with time, as shown in Figure 5.

In the survey study, heterogeneity in lesion patterns on QSM was observed. The lesions volumes between QSM and T2w may be different.

In the second longitudinal study, lesion susceptibility values keep increase within the first two years. It may be attributed to reduction in myelin content, clearance of myelin debris and accumulation of iron².

The lesions volume on QSM are first smaller than that on T2w, but after one year the lesions on QSM become larger than on T2w. QSM and T2w hyperintense both reflect demyelination. The lesion volume on QSM being bigger than that on T2w indicates the existence of iron associated with proinflammatory M1 type of microglia/macrophages^2,3.

QSM adds information on inflammatory activities of MS lesions. Longitudinal tracking of lesion evolution reveals lesion dynamics among heterogeneity. The lesion volume on QSM overtaking that on T2w around one year after lesion onset indicates the appearance of iron in active chronic lesions to promote inflammation.