Accuracy of MRI criteria for dissemination in space for the diagnosis of multiple sclerosis in patients with clinically isolated syndromes

The MRI Barkhof-Tintore criteria have proved to be highly specific for predicting conversion to clinically definite multiple sclerosis in patients with clinically isolated syndromes (CIS), but lacked an optimal sensitivity. In order to improve the accuracy of early multiple sclerosis diagnosis, new imaging criteria have been proposed by Swanton et al. We aimed to evaluate the accuracy of both MRI criteria for dissemination in space to predict conversion from CIS to clinically definite multiple sclerosis. We studied 79 CIS patients with baseline MRI performed within the first 3 months after onset. The sensitivity and specificity of both MRI criteria to predict conversion to clinically definite multiple sclerosis were analysed. The time to develop clinically definite multiple sclerosis from CIS onset, according to each imaging criteria, was studied by Kaplan-Meier survival curves. The overall conversion rate was 75.7% with a median follow-up of 57 months. Barkhof-Tintore's criteria showed a sensitivity of 71.9% and a specificity of 77.2%. Swanton's criteria had a sensitivity of 91.2% and a specificity of 68.1%. Both MRI criteria identified CIS patients with higher risk and faster conversion to clinically definite multiple sclerosis. Swanton's criteria are simpler and more sensitive than Barkhof-Tintore's criteria, with a slight decrease in specificity. These results reinforce their use in multiple sclerosis diagnosis.