Comparison of B-cell Depletion and Natalizumab for Treatment of Multiple Sclerosis: A Semi-supervised Causal Analysis
Dominic DiSanto1, Weijing Tang2, Wen Zhu3, Tianxi Cai1, Zongqi Xia3
1Department of Biostatistics, Harvard University, 2Department of Statistics, Carnegie Mellon University, 3Department of Neurology, University of Pittsburgh
Objective:
We compared BCD and NTZ in managing MS patient-reported disability progression using registry-linked electronic healthcare record (EHR) data.
Background:
B-cell depleting (BCD) agents (ocrelizumab, ofatumumab, rituximab) and natalizumab (NTZ) are both higher efficacy disease-modifying therapies (DMTs) for multiple sclerosis (MS). However, there is no randomized clinical trial and only limited real-world evidence comparing these drug classes.
Design/Methods:
The study population included MS patients in the UPMC EHR cohort (2004-2022), including a subset enrolled in a clinic-based MS registry (Pittsburgh, PA), which provided the gold-standard outcome labels. To estimate the comparative treatment effect, we applied a robust and efficient semi-supervised approach to enable analysis of all patients (not just those in the registry) and comprehensively adjusted for confounders that included not only a priori clinical and demographic features but also a high-dimensional set of codified and narrative EHR features selected by knowledge graph. While gold-standard longitudinal disability outcomes were available in registry patients, we imputed the baseline disability status before treatment initiation and the disability progression status after treatment initiation for non-registry patients.
Results:
In this MS cohort (n=1,738, Age=46±13 years, Non-Hispanic White=86.71%), there was no significant difference between BCD (n=1,245, 71.63%) and NTZ (n=495, 28.37%) in mitigating sustained worsening (ATE=0.012, 95% CI [-0.123, 0.102], p=0.803) or promoting sustained improvement (ATE=-0.0823, 95% CI [-0.194, 0.022], p=0.172) of patient-reported disability. Sensitivity analyses using a 2-year window after treatment initiation confirmed no difference in sustained worsening (ATE=-0.0141, 95% CI [-0.107, 0.080], p=0.895) or sustained improvement (ATE=-0.0713, 95% CI [-0.263, 0.026], p=0.245) between BCD and NTZ. In power analysis, our semi-supervised approach provided greater statistical power than the standard approach of using gold-standard data alone.
Conclusions:
This real-world comparative effectiveness analysis based on a novel semi-supervised approach found no difference between BCD and NTZ in managing MS disability progression.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.