Treatment Discontinuation in MOGAD
(2026)Objective
To determine relapse risk after treatment discontinuation in MOGAD and identify optimal treatment duration
Study Summary
• 39% of MOGAD patients relapsed after treatment discontinuation at median 5.4 months.
• Relapsing course was the strongest predictor (HR 1.95).
• Optimal duration: 10–18 months for monophasic, 20–30 months for relapsing MOGAD.
• Relapsing course was the strongest predictor (HR 1.95).
• Optimal duration: 10–18 months for monophasic, 20–30 months for relapsing MOGAD.
Intervention
Observation after treatment discontinuation (retrospective cohort)
Inclusion Criteria
MOGAD patients with documented treatment discontinuation events from Oxford specialist service
Study Design
Arms: Single observational cohort with treatment discontinuation events (190 patients, 236 intervals)
Patients per Arm: 190 patients, 236 discontinuation intervals
Outcome
39% relapsed at median 5.4 months. Relapsing course HR 1.95 (P=0.003). Optimal duration: 10–18 months (monophasic), 20–30 months (relapsing). MOG-Ab titer at discontinuation was NOT predictive.
Bottom Line
39% of MOGAD patients relapsed after treatment discontinuation at median 5.4 months. Relapsing course was the strongest predictor (HR 1.95). Optimal duration: 10–18 months for monophasic, 20–30 months for relapsing MOGAD.
Major Points
- Largest treatment discontinuation study in MOGAD to date: 190 patients, 236 treatment intervals from the Oxford MOGAD service.
- 39% relapsed after stopping treatment, with median time to relapse of 5.4 months — most relapses occur within the first year.
- Relapsing disease course was the strongest predictor of post-discontinuation relapse (HR 1.95, P=0.003).
- Optimal treatment duration: 10–18 months for monophasic onset, 20–30 months for relapsing course.
- MOG antibody titers at the time of discontinuation were NOT predictive of relapse — challenging the common practice of monitoring titers to guide decisions.
- Provides first evidence-based framework for treatment discontinuation in MOGAD.
Study Design
- Study Type
- Retrospective cohort study
- Sample Size
- 190
Limitations & Criticisms
- Retrospective design with inherent selection bias — patients discontinued for various reasons
- Single-center (Oxford) may limit generalizability to other populations
- Treatment decisions were clinician-driven, not protocolized — confounding by indication possible
- MOG-Ab assay methods varied over the study period
- Optimal duration recommendations derived from observational data — RCT confirmation needed