The cross-sectional, single-center study enrolled patients with clinically isolated syndrome (CIS), MS, and conditions which can mimic MS on MRI. Dr. Clarke and her coinvestigators looked at the frequency of lesions with the central vein sign, and with iron rings, in all patients.
An additional aim of the study was to compare how experienced and inexperienced raters fared in their identification of both central veins and iron rings in 25 scans randomly chosen from within the study population. Inter-rater reliability between experienced and inexperienced raters was assessed as good, with little difference between experience levels in detecting iron rings and central veins, said Dr. Clarke.
Criteria used for central vein determination were those established by the North American Imaging in MS initiative, said Dr. Clarke: The vein needs to be seen entering and/or exiting the lesion, and the vein must course through the lesion’s center. If lesions are confluent, each of the larger lesion’s “fingers” must be assessed individually.
Iron rings appear as a hypointense area rimming the lesion’s edge; for the study, an iron ring was considered present if it could be seen fully or partially encircling a lesion, and if the ring was visible on at least two slices.
The study enrolled 103 patients with relapsing-remitting MS, 49 with progressive MS, 112 with CIS, and 35 non-MS patients; about 60% of this latter group had either autoimmune or vascular disease.
The fewest white matter lesions – a median of 4 per patient - were seen in the CIS group, while the progressive MS and non-MS group each had a median of 7 lesions, and the relapsing-remitting MS group had a median of 10 lesions.
In all, 2,617 lesions were analyzed, and 1,352 were assessed as having the central vein sign. Patients with MS or CIS had central vein sign in more than 50% of their lesions, while the non-MS patients had fewer than 20% central vein–positive lesions. In CIS and MS patients, central vein–positive lesions occurred more frequently in the periventricular and subcortical regions, compared with other brain regions (P less than .001).
Iron rings were detected in 392 lesions; none of the non-MS patients had iron ring–positive lesions. In terms of the brain regions where iron rings were most likely to be seen, said Dr. Clarke, “Over half of all iron ring-positive lesions were periventricular.” This finding was statistically significant as well (P less than .001). At least one lesion with an iron ring was seen in 59% of relapsing-remitting MS patients, 39% of progressive MS patients, and 48% of CIS patients.
In terms of patient characteristics, men were 40% more likely to have iron ring–positive lesions, and patients with relapsing-remitting MS were 50% more likely than were patients with CIS to have iron rings. Iron rings became 3% less likely for each additional year of age, as well (P less than .01 for all comparisons).
“Our results show that iron ring numbers peak in relapsing-remitting MS and decrease with longer disease duration,” Dr. Clarke and colleagues reported.
Dr. Clarke acknowledged several limitations of the study, including its single-center and retrospective nature, as well as the relatively low numbers of non-MS patients and patients with progressive MS. She and her colleagues are planning larger studies using 5-year follow-up data, she said.
Dr. Clarke is an ECTRIMS-MAGNIMS fellow and reported a speaker honorarium from Novartis.
SOURCE: Clarke M et al. ECTRIMS 2019. Abstract 108.