Cognition beyond relapses: cognitive progression independent of relapse activity reflects early smoldering neurodegeneration in multiple sclerosis

Read full text
Authors: Stefano Ziccardi, Tom Fuchs, Luigi Martinelli, Teresa Maltempo, Francesco Guarnaccia, Francesco Crescenzo, Maddalena Guandalini, Gian Marco Schiavi, Albulena Bajrami, Agnese Tamanti, Valentina Camera, Damiano Marastoni
+2 more Ralph H. B. Benedict, Massimiliano Calabrese
Cited by: 3
References: 45

Abstract

BACKGROUND: Progression independent of relapse activity (PIRA) is frequent in relapsing-remitting MS patients (RRMS). While often studied in terms of physical disability, cognitive decline occurring without relapses, termed "cognitive PIRA", has also been reported. Cortical/deep gray matter atrophy reflect both clinical PIRA and cognition. This study investigated whether early global/region-specific brain atrophy predicts long-term cognitive decline due to PIRA in RRMS. METHODS: A retrospective cohort study was conducted on RRMS patients followed over time at the Verona MS Centre to evaluate the relationship between clinical/radiological data at diagnosis (T0) and 2 years after (T2) with cognitive PIRA after 20 years of MS. Patients underwent 1.5T MRI scans at T0 and T2 (regional cortical/subcortical volumes, cortical thickness, lesion burden) and multiple neuropsychological assessments at the end of follow-up. Significant cognitive decline was defined using reliable change indices and attributed PIRA if no relapses occurred within ±9 months the time between the assessments showing the decline. Random forest models evaluated early predictors of cognitive PIRA. RESULTS: 115 RRMS patients were followed for 19.1 ± 5.4 years. Fifty-seven (49.6 %) patients exhibited significant cognitive decline, most of whom (80.7 %) were classified as cognitive PIRA. Global cortical thickness emerged as the feature with the highest relative importance in Random Forest models. Baseline thalamic volume and longitudinal atrophy in the middle frontal gyrus ranked among the most informative regional features contributing to model performance. CONCLUSIONS: Cognitive PIRA accounted for most long-term cognitive decline in RRMS. In a predictive machine-learning framework, early measures of brain atrophy contributed most to the identification of individuals who later developed cognitive PIRA. These findings suggest that cognitive PIRA reflects MS silent disease activity and smoldering processes.

Fields of Study & Subjects

Multiple Sclerosis Research Studies Amyotrophic Lateral Sclerosis Research Systemic Sclerosis and Related Diseases

Publication details

Journal / source
Multiple Sclerosis and Related Disorders
Publisher
Elsevier BV
Volume
108
Pages
107080–107080
Publication date
16 February 2026
Language
EN