Considering patient clinical history impacts performance of machine learning models in predicting course of multiple sclerosis.

Multiple Sclerosis (MS) progresses at an unpredictable rate, but predictions on the disease course in each patient would be extremely useful to tailor therapy to the individual needs. We explore different machine learning (ML) approaches to predict whether a patient will shift from the initial Relapsing-Remitting (RR) to the Secondary Progressive (SP) form of the disease, using only "real world" data available in clinical routine. The clinical records of 1624 outpatients (207 in the SP phase) attending the MS service of Sant'Andrea hospital, Rome, Italy, were used. Predictions at 180, 360 or 720 days from the last visit were obtained considering either the data of the last available visit (Visit-Oriented setting), comparing four classical ML methods (Random Forest, Support Vector Machine, K-Nearest Neighbours and AdaBoost) or the whole clinical history of each patient (History-Oriented setting), using a Recurrent Neural Network model, specifically designed for historical data. Missing values were handled by removing either all clinical records presenting at least one missing parameter (Feature-saving approach) or the 3 clinical parameters which contained missing values (Record-saving approach). The performances of the classifiers were rated using common indicators, such as Recall (or Sensitivity) and Precision (or Positive predictive value). In the visit-oriented setting, the Record-saving approach yielded Recall values from 70% to 100%, but low Precision (5% to 10%), which however increased to 50% when considering only predictions for which the model returned a probability above a given "confidence threshold". For the History-oriented setting, both indicators increased as prediction time lengthened, reaching values of 67% (Recall) and 42% (Precision) at 720 days. We show how "real world" data can be effectively used to forecast the evolution of MS, leading to high Recall values and propose innovative approaches to improve Precision towards clinically useful values.

Autoimmune Encephalitis and CSF Anti-GluR3 Antibodies in an MS Patient after Alemtuzumab Treatment.

A 45-year-old Italian woman, affected by relapsing-remitting multiple sclerosis (RR-MS) starting from 2011, started treatment with alemtuzumab in July 2016. Nine months after the second infusion, she had an immune thrombocytopenic purpura (ITP) with complete recovery after steroid treatment. Three months after the ITP, the patient presented with transient aphasia, cognitive deficits, and focal epilepsy. Serial brain magnetic resonance imaging showed a pattern compatible with encephalitis. Autoantibodies to glutamate receptor 3 peptide A and B were detected in cerebrospinal fluid and serum, in the absence of any other diagnostic cues. After three courses of intravenous immunoglobulin (0.4 mg/kg/day for 5 days, 1 month apart), followed by boosters (0.4 mg/kg/day) every 4-6 weeks, her neurological status improved and is currently comparable with that preceding the encephalitis. Autoimmune complications of the central nervous system during alemtuzumab therapy are relatively rare: only one previous case of autoimmune encephalitis following alemtuzumab treatment has been reported to date.

The Contribution of Gut Barrier Changes to Multiple Sclerosis Pathophysiology.

The gut barrier consists of several components, including the mucus layer, made of mucins and anti-bacterial molecule, the epithelial cells, connected by tight junction proteins, and a mixed population of cells involved in the interplay with microbes, such as M cells, elongations of "antigen presenting cells" dwelling the lamina propria, intraepithelial lymphocytes and Paneth cells secreting anti-bacterial peptides. Recently, the influence of intestinal permeability (IP) changes on organs far from gut has been investigated, and IP changes in multiple sclerosis (MS) have been described. A related topic is the microbiota dysfunction that underpins the development of neuroinflammation in animal models and human diseases, including MS. It becomes now of interest to better understand the mechanisms through which IP changes contribute to pathophysiology of neuroinflammation. The following aspects seem of relevance: studies on other biomarkers of IP alterations; the relationship with known risk factors for MS development, such as vitamin D deficiency; the link between blood brain barrier and gut barrier breakdown; the effects of IP increase on microbial translocation and microglial activation; the parallel patterns of IP and neuroimmune changes in MS and neuropsychiatric disorders, that afflict a sizable proportion of patients with MS. We will also discuss the therapeutic implications of IP changes, considering the impact of MS-modifying therapies on gut barrier, as well as potential approaches to enhance or protect IP homeostasis.