PREVENT
Eculizumab in Aquaporin-4–Positive Neuromyelitis Optica Spectrum Disorder (Prevention of Relapses in Neuromyelitis Optica)
Year of Publication: 2019
Authors: Sean J. Pittock, Achim Berthele, Kazuo Fujihara, ..., Dean M. Wingerchuk
Journal: The New England Journal of Medicine
Citation: N Engl J Med 2019;381:614-625
PDF: Clinical Question
Does eculizumab, a terminal complement inhibitor, reduce the risk of relapse compared with placebo in patients with AQP4-IgG-positive neuromyelitis optica spectrum disorder?
Study Overview
Objective
To evaluate the efficacy and safety of eculizumab, a terminal complement inhibitor, in reducing relapse risk in patients with AQP4-IgG-positive neuromyelitis optica spectrum disorder
Study Summary
- Eculizumab reduced adjudicated relapse risk by 94% (3% vs 43%, HR 0.06, p<0.001)
- Adjudicated ARR: 0.02 vs 0.35 (rate ratio 0.04, p<0.001)
- No significant difference in EDSS change (-0.18 vs +0.12); trial stopped after 23/24 prespecified relapses
Intervention
Eculizumab 900 mg IV weekly x 4 weeks, then 1200 mg IV every 2 weeks vs placebo; 76% continued stable-dose immunosuppressive therapy
Patients per Arm
96 (eculizumab), 47 (placebo)
Bottom Line
In this phase 3, randomized, double-blind, placebo-controlled trial, eculizumab dramatically reduced the risk of adjudicated relapse by 94% compared with placebo (3% vs 43%, HR 0.06, p<0.001) in patients with AQP4-IgG-positive NMOSD. The adjudicated annualized relapse rate was 0.02 with eculizumab vs 0.35 with placebo. However, there was no significant between-group difference in EDSS disability change. The trial was stopped early after 23 of 24 prespecified relapses. Eculizumab was generally well tolerated; one patient died from pulmonary empyema, but no meningococcal infections occurred despite this known risk. This trial led to FDA approval of eculizumab for AQP4-IgG-positive NMOSD.
Major Points
- First phase 3, randomized, placebo-controlled trial of complement inhibition in NMOSD
- Dramatic 94% reduction in adjudicated relapse risk with eculizumab (HR 0.06; 95% CI 0.02-0.20; p<0.001)
- Only 3/96 (3%) patients on eculizumab had adjudicated relapse vs 20/47 (43%) on placebo
- Adjudicated ARR: 0.02 (eculizumab) vs 0.35 (placebo); rate ratio 0.04 (p<0.001)
- 76% of patients continued stable-dose immunosuppressive therapy during trial
- Efficacy maintained in subgroup not receiving concomitant immunosuppressive therapy (0% vs 54% relapse)
- No significant difference in EDSS change (-0.18 vs +0.12; 95% CI crossed zero) - hierarchical testing failed here
- Trial stopped early after 23/24 prespecified adjudicated relapses due to uncertainty in timing of final event
- Physician-determined relapses higher than adjudicated (sensitivity analysis: 15% vs 62%, HR 0.18, p<0.001)
- One death from pulmonary empyema (S. intermedius and P. micros; organisms not associated with complement deficiency)
- No meningococcal infections despite this being a known risk of complement inhibition
- Led to FDA approval of eculizumab (Soliris) for AQP4-IgG-positive NMOSD
Design
Study Type: Phase 3, multicenter, randomized, double-blind, placebo-controlled, time-to-event trial
Randomization: 1
Blinding: Double-blind. Patients, site staff, and sponsor representatives were unaware of trial-group assignments. Trial agents and treatment kits appeared identical. EDSS raters were separate from treating physicians and also blinded. Independent relapse adjudication committee (3 members) blinded to treatment assignment.
Enrollment Period: April 2014 to October 2017
Follow-up Duration: Time-to-event design; median not specified; trial terminated after 23/24 prespecified adjudicated relapses
Centers: 70
Countries: 18 countries
Sample Size: 143
Analysis: Modified intention-to-treat (all randomized who received ≥1 dose). Stratified log-rank test for primary endpoint. Stratified Cox proportional-hazards model for hazard ratio. Poisson regression for ARR. Fixed-sequence hierarchical testing procedure for multiplicity control. Sample size: 132 patients needed for 90% power to detect HR 0.24 at α=0.05 two-sided, assuming 80% relapse-free at 12 months for eculizumab vs 40% for placebo, with 10% dropout.
Inclusion Criteria
- Age ≥18 years
- Diagnosis of neuromyelitis optica (2006 criteria) or NMOSD (2007 criteria)
- AQP4-IgG seropositive status confirmed by cell-binding kit assay (Euroimmun) analyzed centrally
- History of at least two relapses during previous 12 months, or three relapses during previous 24 months with at least one in previous 12 months
- EDSS score ≤7
- Stable-dose immunosuppressive therapy permitted
Exclusion Criteria
- Treatment with mitoxantrone or rituximab during previous 3 months
- Receipt of intravenous immune globulin during previous 3 weeks
- Prednisone doses >20 mg/day or equivalent at screening
- Unresolved meningococcal disease
- Systemic bacterial or other infection considered clinically significant or not treated with appropriate antibiotics
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| Group Name | Placebo | Eculizumab |
| N | 47 | 96 |
| Female | 42 (89%) | 88 (92%) |
| Age at first receipt of trial agent (years), mean ± SD | 45.0 ± 13.29 | 43.9 ± 13.32 |
| Age at initial clinical presentation (years), mean ± SD | 38.5 ± 14.98 | 35.8 ± 14.03 |
| Diagnosis - Neuromyelitis optica | 38 (81%) | 69 (72%) |
| Diagnosis - NMOSD | 9 (19%) | 27 (28%) |
| Diagnosis - LETM | 5 (11%) | 14 (15%) |
| Diagnosis - Optic neuritis | 2 (4%) | 8 (8%) |
| Annualized relapse rate during previous 24 months, mean ± SD | 2.07 ± 1.04 | 1.94 ± 0.90 |
| Type of relapse previous 24 months - Optic neuritis | 22 (47%) | 58 (60%) |
| Type of relapse previous 24 months - Transverse myelitis | 42 (89%) | 74 (77%) |
| Type of relapse previous 24 months - Brain-stem symptoms | 15 (32%) | 18 (19%) |
| Type of relapse previous 24 months - Cerebral symptoms | 5 (11%) | 10 (10%) |
| Median EDSS score (range) | 4.00 (1.0-6.5) | 4.00 (1.0-7.0) |
| Immunosuppressive therapy at baseline - None | 13 (28%) | 21 (22%) |
| Immunosuppressive therapy at baseline - Glucocorticoids alone | 11 (23%) | 16 (17%) |
| Immunosuppressive therapy at baseline - Azathioprine ± glucocorticoids | 13 (28%) | 37 (39%) |
| Immunosuppressive therapy at baseline - Mycophenolate mofetil ± glucocorticoids | 8 (17%) | 17 (18%) |
| Immunosuppressive therapy at baseline - Other ± glucocorticoids | 2 (4%) | 5 (5%) |
| Previous rituximab treatment | 20 (43%) | 26 (27%) |
Arms
| Field | Control | Eculizumab |
|---|---|---|
| Intervention | Matching intravenous placebo administered over approximately 35 minutes (range 25-45 minutes) on the same schedule as eculizumab: weekly for first four doses, then every 2 weeks. Stable-dose concomitant immunosuppressive therapy (except rituximab) permitted. Meningococcal vaccination required before first dose. | Intravenous eculizumab 900 mg weekly for first four doses (days 1, 8, 15, 22), then maintenance 1200 mg every 2 weeks starting at week 4. Administered over approximately 35 minutes (range 25-45 minutes). Stable-dose concomitant immunosuppressive therapy (except rituximab) permitted. Meningococcal vaccination required before first dose. |
| Duration | Until relapse, discontinuation, or trial end (time-to-event design) | Until relapse, discontinuation, or trial end (time-to-event design) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| First adjudicated relapse. Relapse defined as: new onset or worsening of neurologic symptoms with objective change on examination persisting >24 hours, attributable to NMOSD rather than other causes, onset preceded by at least 30 days of clinical stability. Independent relapse adjudication committee (2 neurologists, 1 neuro-ophthalmologist) reviewed all physician-determined relapses. | Primary | 20/47 (43%) | 3/96 (3%) | 0.06 | <0.001 |
| Adjudicated annualized relapse rate | Secondary | 0.35 (95% CI 0.20-0.62) | 0.02 (95% CI 0.01-0.05) | Rate ratio 0.04 (95% CI 0.01-0.15) | <0.001 |
| Mean change from baseline in EDSS score | Secondary | 0.12 ± 0.95 | -0.18 ± 0.81 | Least-squares mean difference -0.29 (95% CI -0.59 to 0.01) | NA |
| Mean change from baseline in modified Rankin scale | Secondary | 0.09 ± 0.75 | -0.24 ± 0.72 | Least-squares mean difference -0.32 (95% CI -0.57 to -0.06) | NA (no inferences due to hierarchical failure) |
| Mean change from baseline in Hauser Ambulation Index | Secondary | 0.51 ± 1.61 | -0.39 ± 1.08 | Least-squares mean difference -0.87 (95% CI -1.32 to -0.42) | NA (no inferences due to hierarchical failure) |
| Mean change from baseline in EQ-5D-3L visual analogue scale | Secondary | 0.57 ± 16.39 | 5.42 ± 18.53 | Least-squares mean difference 6.43 (95% CI 0.63 to 12.23) | NA (no inferences due to hierarchical failure) |
| Mean change from baseline in EQ-5D-3L summary index | Secondary | -0.04 ± 0.21 | 0.05 ± 0.18 | Least-squares mean difference 0.09 (95% CI 0.02 to 0.15) | NA (no inferences due to hierarchical failure) |
| First physician-determined (non-adjudicated) relapse (sensitivity analysis; original primary endpoint) | Secondary | 29/47 (62%) | 14/96 (15%) | HR 0.18 (95% CI 0.10-0.34) | <0.001 |
| Adjudicated relapse in subgroup NOT receiving concomitant immunosuppressive therapy | Secondary | 7/13 (54%) | 0/21 (0%) | ||
| Adjudicated relapse in subgroup receiving concomitant immunosuppressive therapy (post hoc) | Secondary | 13/34 (38%) | 3/75 (4%) | ||
| Exposure (patient-years) | Adverse | 53 | 173 | ||
| Any adverse event (events/100 patient-years) | Adverse | 1127 | 745 | ||
| Any adverse event - patients | Adverse | 43/47 (91%) | 88/96 (92%) | ||
| Serious adverse events (events/100 patient-years, excluding NMOSD relapses) | Adverse | 55 | 27 | ||
| Serious adverse events - patients | Adverse | 13/47 (28%) | 25/96 (26%) | ||
| Upper respiratory tract infection (events/100 patient-years) | Adverse | 19 | 31 | ||
| Headache (events/100 patient-years) | Adverse | 38 | 55 | ||
| Nasopharyngitis - patients | Adverse | 9/47 (19%) | 20/96 (21%) | ||
| Urinary tract infection - patients | Adverse | 10/47 (21%) | 13/96 (14%) | ||
| Serious infections (events/100 patient-years) | Adverse | 15 | 8 | ||
| Death | Adverse | 0 | 1 (pulmonary empyema; S. intermedius and P. micros; probably related per investigator; patient had extensive pulmonary history and was active smoker) | ||
| Meningococcal infection | Adverse | 0 | 0 | ||
| Discontinuation due to adverse event | Adverse | 2 (4%) | 0 | ||
| Trial discontinuation (all reasons) | Adverse | 3 (6%) | 16 (17%) |
Subgroup Analysis
Prespecified subgroup analysis of patients not receiving concomitant immunosuppressive therapy showed consistent efficacy: 0/21 (0%) eculizumab vs 7/13 (54%) placebo had adjudicated relapse. Post hoc analysis of patients receiving concomitant immunosuppressive therapy also showed consistent efficacy: 3/75 (4%) eculizumab vs 13/34 (38%) placebo. Kaplan-Meier curves for both subgroups demonstrated treatment effect consistent with overall population. Of 45 physician-determined relapses, 24 (53%) were adjudicated as relapses; disagreements mostly due to adjudication committee finding insufficient objective neurologic change (18/21 disagreements). Most relapses were myelitis.
Criticisms
- Only AQP4-IgG-positive patients included - results cannot be extrapolated to seronegative NMOSD or MOG-antibody disease
- No significant between-group difference in EDSS change (hierarchical testing failed here) - long-term disability effects uncertain
- Trial design precluded follow-up beyond 6 weeks after single relapse - cannot assess sustained disability impact
- Trial stopped after 23/24 prespecified relapses, though retained at least 80% power
- Discordance between physician-determined and adjudicated relapses (53% agreement) - concern that adjudication was too strict
- Placebo control ethically questionable given known disease severity, though 76% received concomitant immunosuppressive therapy
- Rituximab excluded as concomitant therapy due to mechanistic incompatibility - cannot assess eculizumab vs rituximab
- High previous rituximab use (32%) may have affected baseline characteristics
- Higher withdrawal rate in eculizumab group (17% vs 6%) for reasons other than adverse events
- One death from pulmonary empyema in eculizumab group (though organisms not complement-related)
- 46 patients (32%) previously received rituximab but not within 3 months - potential carryover effects
Funding
Alexion Pharmaceuticals (designed trial in consultation with academic authors, provided trial agents, analyzed data, paid for writing assistance)
Based on: PREVENT (The New England Journal of Medicine, 2019)
Authors: Sean J. Pittock, Achim Berthele, Kazuo Fujihara, ..., Dean M. Wingerchuk
Citation: N Engl J Med 2019;381:614-625
Content summarized and formatted by NeuroTrials.ai.