ObjectiveTo assess whether it is feasible to establish specific cut-off values able to discriminate ‘physiological’ or ‘pathological’ brain volume rates in patients with multiple sclerosis (MS).MethodsThe study was based on the analysis of longitudinal MRI data sets of patients with MS (n=206, 87% relapsing–remitting, 7% secondary progressive and 6% primary progressive) and healthy controls (HC; n=35). Brain atrophy rates were computed over a mean follow-up of 7.5 years (range 1–12) for patients with MS and 6.3 years (range 1–12.5) for HC with the SIENA software and expressed as annualised per cent brain volume change (PBVC/y). A weighted (on the follow-up length) receiver operating characteristic analysis and the area under the curve (AUC) were used for statistics.ResultsThe weighted PBVC/y was −0.51±0.27% in patients with MS and −0.27±0.15% in HC (p<0.0001). There was a significant age-related difference in PBVC/y between HC older and younger than 35 years of age (p=0.02), but not in patients with MS (p=0.8). The cut-off of PBVC/y, as measured by SIENA that could maximise the accuracy in discriminating patients with MS from HC, was −0.37%, with 67% sensitivity and 80% specificity. According to the observed distribution, values of PBVC/y as measured by SIENA that could define a pathological range were above −0.52% with 95% specificity, above −0.46% with 90% specificity and above −0.40% with 80% specificity.ConclusionsOur evidence-based criteria provide values able to discriminate the presence or absence of ‘pathological’ brain volume loss in MS with high specificity. Such results could be of great value in a clinical setting, particularly in assessing treatment efficacy in MS.

No abstract

Background: Subclinical paroxysmal atrial fibrillation (AF) is one of the main occult causative mechanisms of embolic stroke of undetermined source (ESUS). Aim of this study was to identify AF predictors, and to develop a score to predict the probability of AF detection in ESUS. Methods: We retrospectively analyzed ESUS patients undergoing 2-week external electrocardiographic monitoring. Patients with and without AF detection were compared. On the basis of multivariate analysis, predictors of AF were identified and used to develop a predictive score, which was then compared with other existing literature scores. Results: Eighty-two patients, 48 females, mean age±SD 72±10 years, were included. In 36 patients (43.9%) AF was detected. The frequency of age 75 years or above and arterial hypertension, and the median CHA2DS2-VASc score were significantly higher in patients with AF compared with those without. National Institutes of Health Stroke Scale (NIHSS) score ≥8 was the only independent variable associated with AF detection. We derived the Empoli ESUS-AF (E2AF) score (NIHSS ≥8 5 points, arterial hypertension 3 points, age 75 years or above 2 points, age 65 to 74 years 1 point, history of coronary/peripheral artery disease 1 point, left atrial enlargement 1 point, posterior lesion 1 point, cortical or cortical-subcortical lesion 1 point), whose predictive power in detecting AF was good (area under the curve: 0.746, 95% confidence interval: 0.638-0.836) and higher than that of CHA2DS2-VASc and other scores. Conclusions: In our study NIHSS score ≥8 was the only independent predictor of post-ESUS-AF detection. The E2AF score appears to have a good predictive power for detecting AF. External validations are required.

No abstract