Technology-enabled comprehensive characterization of multiple sclerosis in clinical practice

Abstract

BackgroundObjective and longitudinal measurements of disability in patients with multiple sclerosis (MS) are desired in order to monitor disease status and response to disease-modifying and symptomatic therapies. Technology-enabled comprehensive assessment of MS patients, including neuroperformance tests (NPTs), patient-reported outcome measures (PROMs), and MRI, is incorporated into clinical care at our center. The relationships of each NPT with PROMs and MRI measures in a real-world setting are incompletely studied, particularly in larger datasets. ObjectivesTo demonstrate the utility of comprehensive neurological assessment and determine the association between NPTs, PROMs, and quantitative MRI measures in a large MS clinical cohort. MethodsNPTs (processing speed [PST], contrast sensitivity [CST], manual dexterity [MDT], and walking speed [WST]) and physical disability-related PROMs (Quality of Life in Neurological Disorders [Neuro-QoL], Patient Determined Disease Steps [PDDS], and Patient-Reported Outcomes Measurement Information System Global-10 [PROMIS-10] physical) were collected as part of routine clinical care. Fully-automated MRI analysis calculated T2-lesion volume (T2LV), whole brain fraction (WBF), thalamic volume (TV), and cervical spinal cord cross-sectional area (CA) for brain MRIs completed within 3 months of a clinic visit during which NPTs and PROMs were assessed. Spearman's rank correlation coefficients evaluated the cross-sectional associations of NPTs with PROMs and MRI measures. Linear regression was utilized to determine which combination of clinical characteristics, patient demographics, MRI measures, and PROMs best cross-sectionally explained each NPT result. Results997 unique patients (age 47.7 ± 11.4 years, 71.8% female) who underwent assessments over a 2-year period were included. Correlations among NPTs and PROMs were moderate. PST correlations were strongest for Neuro-QoL upper extremity (NQ-UE) (Spearman's rho = 0.43) and lower extremity (NQ-LE) (0.47). CST correlations were strongest for NQ-UE (0.33), NQ-LE (0.36), and PDDS (-0.31). MDT correlations were strongest for NQ-UE (-0.53), NQ-LE (-0.54), and PDDS (0.53). WST correlations were strongest for PDDS (0.64) and NQ-LE (-0.65). NPTs also had moderate correlations with MRI metrics, the strongest of which were observed with PST (with T2LV (-0.44) and WBF (0.49)). Spearman's rho for other NPT-MRI correlations ranged from 0.23 to 0.36. Linear regression identified age, disease duration, PROMIS-10 physical, NQ-UE, NQ-LE, T2LV and WBF as significant cross-sectional explanatory variables for PST (adjusted R2=0.46). For CST, significant variables included age and NQ-LE (adjusted R2 = 0.30). For MDT, significant variables included PDDS, PROMIS-10 physical, NQ-UE, NQ-LE, T2LV, and WBF (adjusted R2=0.37). For WST, significant variables included sex, PDDS, NQ-LE, T2LV, and CA (adjusted R2=0.39). ConclusionsImpaired performance on NPTs correlated with worse physical disability-related PROMs and MRI disease severity, but the strongest cross-sectional explanatory variables for each NPT component varied. This study supports the use of comprehensive, objective quantification of MS status in clinical and research settings. Future longitudinal analyses can determine predictors of treatment response and disability worsening.