Disease correlates of rim lesions on quantitative susceptibility mapping in multiple sclerosis

Abstract

Quantitative susceptibility mapping (QSM), an imaging technique sensitive to brain iron, has been used to detect paramagnetic rims of iron-laden active microglia and macrophages in a subset of multiple sclerosis (MS) lesions, known as rim+ lesions, that are consistent with chronic active lesions. Because of the potential impact of rim+ lesions on disease progression and tissue damage, investigating their influence on disability and neurodegeneration is critical to establish the impact of these lesions on the disease course. This study aimed to explore the relationship between chronic active rim+ lesions, identified as having a hyperintense rim on QSM, and both clinical disability and imaging measures of neurodegeneration in patients with MS. The patient cohort was composed of 159 relapsing-remitting multiple sclerosis patients. The Expanded Disability Status Scale (EDSS) and Brief International Cognitive Assessment for Multiple Sclerosis, which includes both the Symbol Digit Modalities Test and California Verbal Learning Test-II, were used to assess clinical disability. Cortical thickness and thalamic volume were evaluated as imaging measures of neurodegeneration. A total of 4469 MS lesions were identified, of which 171 QSM rim+ (3.8%) lesions were identified among 57 patients (35.8%). In a multivariate regression model, as the overall total lesion burden increased, patients with at least one rim+ lesion on QSM performed worse on both physical disability and cognitive assessments, specifically the Symbol Digit Modalities Test (p = 0.010), California Verbal Learning Test-II (p = 0.030), and EDSS (p = 0.001). In a separate univariate regression model, controlling for age (p < 0.001) and having at least one rim+ lesion was related to more cortical thinning (p = 0.03) in younger patients (< 45 years). Lower thalamic volume was associated with older patients (p = 0.038) and larger total lesion burden (p < 0.001); however, the association did not remain significant with rim+ lesions (p = 0.10). Our findings demonstrate a novel observation that chronic active lesions, as identified on QSM, modify the impact of lesion burden on clinical disability in MS patients. These results support further exploration of rim+ lesions for therapeutic targeting in MS to reduce disability and subsequent neurodegeneration.

Figure 1

Examples of QSM rim+ lesions. Panels (A) and (B) represent slices of imaging for two separate patients. Within each panel, FLAIR is located on the left, and the corresponding slice of QSM is located on the right. In both panels (A) and (B), the lesion on FLAIR is located within an orange circle, while the corresponding lesion on QSM is located within a blue circle. These examples represent lesions with a complete rim on QSM.

Figure 2

Distribution of disability scores comparing patients with zero (0: red) rim+ lesions versus patients with at least one rim+ lesion (1+: blue). Dash lines represent the mean for each group. Patients with at least one rim+ lesion performed worse on average SDMT (A), CVLT-II (B), BVMT-R (C) and EDSS (D).

Figure 3

Association of rim+ lesions with disability measures after accounting for clinical and imaging covariates. Patients with both higher total lesion volume on FLAIR and at least one rim+ lesion performed worse on SDMT (A), CVLT-II (B), and EDSS (D) as demonstrated by 95% confidence intervals of the difference of mean disability scores between patients with zero rim+ lesions minus patients with at least one rim+ lesion. The 95% confidence intervals of the means are model-based estimates. The final model included current treatment duration, age, gender, total lesion volume on FLAIR, and the number of rim+ lesions (statistically significant covariates). Performance on BVMT-R (C) did not reach clinical significance in the model.