Background

Neuro-performance testing has been used extensively in MS clinical trials, resulting in a large literature on processing speed (Symbol Digit Modalities Test [SDMT]), manual dexterity (9-Hole Peg Test [9HPT]), and walking speed (25-foot walk [25FW]). Computer adapted versions were developed and validated, to support widespread use in clinical practice. The Multiple Sclerosis Performance Test (MSPT) includes a self-administered Processing Speed Test (PST), simulating SDMT; Manual Dexterity Test (MDT), simulating 9HPT; and Walking Speed Test (WST), simulating 25FW. MSPT is deployed within the Multiple Sclerosis Partners Advancing Technology and Health Solutions (MS PATHS) network. Optimal test thresholds associated with employment status in a real-world population have not been reported.

Objectives

To determine thresholds for standardized test scores of processing speed, manual dexterity, and walking speed as predictors of employment status in a real world MS population.

Methods

Neuroperformance testing was done as part of clinical visits in MS PATHS. Employment status was collected via standardized questionnaire. Patients aged 18 to 60 in the US were divided into a training set (n=3210) and a test set (n=1605). PST, MDT and WST benchmarks predicting unemployment at baseline and employment worsening at 2 years were identified as the test scores with the minimum p-value in logistic regression models adjusting for age, sex and education. Odds ratios representing the risk of unemployment or employment worsening were calculated based on the identified benchmarks.

Results

4815 of 9585 participants (50%) were employed full-time at baseline. In the training set benchmarks for unemployment were: PST ≤44 correct, OR (95% CI) 5.3 (4.7, 6.0); MDT >28.7 seconds, OR 7.2 (6.3, 8.1); and WST >8 seconds, OR 6.7 (5.8, 7.7). For patients employed at baseline, benchmarks for worsening employment status were: PST ≤44 correct, OR 4.3 (3.1, 6.0); MDT >24 seconds, OR 3.3 (2.3, 4.6); and WST >7.6 seconds, OR 6.4 (4.7, 8.8). Benchmarks were confirmed in the validation set.

Conclusions

Clinically relevant neuroperformance test benchmarks for predicting unemployment and employment worsening were identified in a training set and confirmed in a validation set using a large real world MS population. Future research will determine early risk factors for these benchmarks in order to identify potential employment preservation strategies.

Disclosures: MS PATHS is sponsored by Biogen

Background

Natalizumab extended interval dosing (EID) is associated with lower risk of progressive multifocal leukoencephalopathy than every-4-week standard interval dosing (SID). Independent real-world (RW) studies suggest that natalizumab effectiveness is maintained in patients who switch from SID to EID. MS PATHS (Multiple Sclerosis Partners Advancing Technology and Health Solutions) is a learning health system comprised of a collaborative network of healthcare institutions that provides access to RW clinical and MRI data collected using standardized acquisition protocols.

Objectives

Compare the effectiveness of natalizumab EID and SID using quantitative MRI metrics from highly standardized RW images in MS PATHS.

Methods

An MRI segment was defined as 2 MRI acquisitions and associated interval duration. MS PATHS patients with ≥1 MRI segment, ≥2 infusion cycles (infusion interval >21 days and ≤84 days), and complete covariate information were eligible. MRI segments with average infusion cycle (AIC) ≤35 days and >35 days were defined as SID and EID, respectively. For each MRI segment, new T2 lesions and changes in T2 lesion volume (T2LV) and brain parenchymal fraction (BPF) were compared for SID and EID using inverse probability weighting (IPW) with logistic regression and robust linear regression.

Results

IPW analysis included 327 SID patients (596 MRI segments) and 67 EID patients (85 MRI segments). The mean AIC for SID was 29.5 (standard deviation [SD] 1.8) days and 40.8 (4.9) days for EID. Mean MRI segment duration for SID and EID was 9.7 (SD 5.5) and 9.6 (5.3) months, respectively. Proportions of patients with no new T2 lesions were similar for the SID and EID groups (75.6% vs 75.3%; adjusted odds ratio for 0 vs ≥1 lesion=0.967 [95% CI 0.500, 1.871]; P=0.921). SID and EID patients did not differ significantly in adjusted T2LV change (−0.070 [95% CI −0.129, 0.111] mL vs 0.022 [−0.132, 0.180] mL; P=0.233) or BPF change (−0.1222% [95% CI −0.1524%, −0.0921%] vs −0.1442% [−0.2234%, −0.0649%]; P=0.617).

Conclusions

MRI activity was low in SID and EID MS PATHS patients, and there were no significant differences in MRI outcomes between the 2 groups. These results confirm and extend previous RW studies of natalizumab EID effectiveness. Study limitations include modest EID sample size and potential channeling bias. The ongoing phase 3b randomized trial NOVA (clinicaltrials.gov NCT03689972) will further evaluate the effectiveness of natalizumab EID versus SID.

MS PATHS is supported by Biogen.

Background

Background: Multiple sclerosis (MS) incidence in Hispanic Americans (HA) is increasing, highlighting the need to understand disease features and clinical course trends among this subpopulation.

Objectives

Objective: To compare demographic features and clinical characteristics of a large population of HA and non-Hispanic Caucasian Americans (NHCA) with MS.

Methods

Methods: MS PATHS is a network of MS Centers in the United States (n=7) and Europe (n=3) contributing standardized data acquired during routine care. US-based MS PATHS participants who self-reported as HA (irrespective of race) or as NHCA, and compared the groups according to demographic (sex, years of education, smoking status, BMI, employment, and insurance status), MS clinical (self-reported disability via Patient Determined Disease Steps [PDDS]), and neuro-performance (via the MS Performance Test (MSPT): walking, manual dexterity, and processing speed) features. Odds ratios and mean differences for PDDS and neuro-performance outcomes were adjusted for age, sex, disease duration and subtype, smoking status, BMI, insurance status, employment status, and years of education. Z-score is compared to a representative healthy population.

Results

Results: Compared to NHCA (n=9003), HA (n=609) had earlier MS symptom onset (mean 28.6y [SD:10.7y] vs 33.6y [11.3y]; p<0.001) and younger age at diagnosis (31.6y [10.9y] vs 36.6y [10.9y]; p<0.001). HA were more likely to have mild disability by the PDDS, compared to NHCA (OR 0.62, 95% CI [-0.89, -0.06], p=0.02). However, HA had worse performance on both manual dexterity times (z score: 0.31 [0.14, 0.47], p<0.001), and cognitive processing speed score (# correct: 0.37 [0.27-0.47], p<0.0001). 25-foot walking speed was not different between the groups (z score:0.09 [-0.23,0.41], p=0.56).

Conclusions

Conclusion: Using standardized data collection in this large MS sample, HA compared to NHCA patients were found to have younger age of onset and diagnosis and higher levels of cognitive and manual dexterity slowing. However, HA were less likely to rate themselves with severe disability on the PDDS. As the groups did not differ in walking speed, this may reflect the scale relatively weighting ambulation over other functions or other language/cultural differences.

Background

Understanding patient-reported changes in physical, mental, and social health after starting MS therapy is important in optimizing treatment.

Objectives

Assess changes in the Quality of Life in Neurological Disorders (Neuro-QoL; NQ) questionnaire after starting natalizumab (NAT) and compare to another high efficacy therapy - ocrelizumab (OCR).

Methods

T-scores of 12 NQ domains were obtained at routine visits in the MS Partners Advancing Technology and Health Solutions (MS PATHS) network. NQ scores from visits post NAT initiation were compared to last previous NQ (baseline, BL) to calculate the annualized rate of change and the likelihood of clinically meaningful change (≥5-point) in the overall cohort and in patients with abnormal BL NQ (T-score worse than 50; 36%-76% of the population). Subgroup analyses in NAT- and OCR-treated patients were conducted with multivariate mixed-effects regression models after propensity score weighting and adjustment for antidepressants, year and drug*year interaction.

Results

164 NAT patients were analyzed; mean (SD) follow-up was 6 (6) months and number of assessments was 2.3 (1.6). Significant improvements from pre-NAT BL were seen in 8 of 12 NQ domains. Patients with BL impairment had significant improvements in 10 NQ domains and higher rates of improvement compared to the overall cohort (p<0.05). In this subgroup, the largest number of patients with ≥5-point improvement was seen for positive affect and well-being (PAF) (43%), emotional and behavioral dyscontrol (EBD) (38%) and sleep disturbances (35%). In the subgroup of NAT (n=145)- and OCR (n=520)-treated patients, the annualized improvement rates were higher with NAT than with OCR, reaching statistical significance for PAF (p=0.02), sleep disturbances (p=0.003), and satisfaction with social roles and activities (SRA) (p=0.03). In patients with impaired BL NQ, significantly higher rates of improvement were seen with NAT than with OCR for EBD (p=0.01), participation in SRA (p=0.0001) and satisfaction with SRA (p=0.02). The percentage of patients with ≥5-point improvement was numerically higher with NAT than OCR for 9 of 12 NQ domains; differences in the likelihood of ≥5-point improvement were not significant.

Conclusions

NAT can lead to clinically meaningful improvements in mental and social health. Such improvements are unlikely to be primarily driven by expectation bias as their magnitude exceeded improvements with another high-efficacy therapy (OCR).