ADAPT
Safety, efficacy, and tolerability of efgartigimod in patients with generalised myasthenia gravis (ADAPT): a multicentre, randomised, placebo-controlled, phase 3 trial
Clinical Question
Is efgartigimod, a human IgG1 antibody Fc fragment that reduces pathogenic IgG autoantibody levels by blocking the neonatal Fc receptor (FcRn), safe and effective in patients with generalized myasthenia gravis?
Study Overview
Objective
To assess the safety and efficacy of efgartigimod, a human IgG1 antibody Fc fragment engineered for increased affinity to FcRn, in patients with generalized myasthenia gravis
Study Summary
- 68% of AChR-Ab+ patients on efgartigimod were MG-ADL responders vs 30% placebo (OR 4.95, P<0.0001)
- 63% were QMG responders vs 14% placebo (OR 10.84, P<0.0001)
- 40% achieved minimal symptom expression (MG-ADL 0-1) vs 11% placebo
Intervention
Efgartigimod 10 mg/kg IV infusion, four infusions per cycle (one per week), cycles repeated based on clinical response no sooner than 8 weeks after previous cycle initiation
Patients per Arm
Efgartigimod: 84, Placebo: 83
Bottom Line
Efgartigimod was well tolerated and highly efficacious in patients with generalized myasthenia gravis. In AChR-antibody positive patients, 68% achieved MG-ADL response (≥2-point improvement sustained ≥4 weeks) compared to 30% with placebo (OR 4.95, 95% CI 2.21-11.53, P<0.0001). The drug produced rapid onset of action (84% of responders improved by week 2), durable benefit, and reproducible efficacy across treatment cycles. 40% of patients achieved minimal symptom expression. The individualized, cycle-based dosing approach based on clinical response represents a novel paradigm for MG treatment.
Major Points
- First phase 3 trial of an FcRn antagonist in generalized myasthenia gravis
- Largest clinical trial in generalized MG to date (167 patients), including both AChR-Ab positive and negative patients
- Primary endpoint met: 68% MG-ADL responders vs 30% placebo in AChR-Ab+ patients (OR 4.95, 95% CI 2.21-11.53, P<0.0001)
- QMG responder rate 63% vs 14% (OR 10.84, P<0.0001) - highly significant
- 40% of efgartigimod patients achieved minimal symptom expression (MG-ADL 0-1) vs 11% placebo
- Rapid onset: 84% of responders had first improvement by week 2
- Mean maximum IgG reduction 61.3%; AChR-Ab reduction 57.6% at week 4
- Response reproducible across cycles: of 44 cycle 1 MG-ADL responders, 32 qualified for retreatment and 29 (90%) responded again in cycle 2
- Duration of response: 34% maintained response ≥12 weeks
- Well tolerated: AEs similar to placebo; most mild to moderate; no deaths
Design
Study Type: Multicenter, randomized, double-blind, placebo-controlled, phase 3 trial
Randomization: 1
Blinding: Double-blind: patients, investigators, study personnel, clinic staff, and funder all masked to treatment assignments; randomization performed centrally using interactive response technology (web and voice systems)
Enrollment Period: September 5, 2018 to November 26, 2019
Follow-up Duration: 26-week treatment period; open-label extension (ADAPT+) ongoing
Centers: 56
Countries: Japan, United States, Canada, Multiple European countries (14 countries in Europe and North America total)
Sample Size: 167
Analysis: Modified intention-to-treat (mITT) for efficacy; all randomized patients with valid baseline and ≥1 post-baseline MG-ADL assessment; primary endpoint tested by two-sided exact logistic regression with baseline score as covariate and stratification factors (AChR-Ab status, NSIST use, Japanese nationality); hierarchical testing of secondary endpoints at 5% two-sided significance; Kaplan-Meier for time-to-event; ANCOVA for proportion of time with improvement; SAS version 9.2
Inclusion Criteria
- Age ≥18 years
- Generalized myasthenia gravis (with or without AChR antibodies)
- MGFA class II to IV
- MG-ADL score ≥5 with >50% due to non-ocular symptoms
- Diagnosis supported by abnormal neuromuscular transmission tests, positive edrophonium chloride test, or improvement with acetylcholinesterase inhibitors
- Stable dose of ≥1 treatment for MG (AChE inhibitors, corticosteroids, or NSISTs) before screening and throughout trial
Exclusion Criteria
- Rituximab or eculizumab within 6 months before screening
- Thymectomy within 3 months
- IVIG or plasma exchange within 1 month of screening
- Active hepatitis B
- Seropositive for hepatitis C
- Seropositive for HIV with low CD4 count
- Serum IgG levels <6 g/L at screening
- Pregnant
Arms
| Field | Control | Efgartigimod |
|---|---|---|
| Intervention | Matching placebo IV infusion, four infusions per cycle (one per week), cycles repeated based on clinical response when MG-ADL score ≥5 (>50% non-ocular) and loss of clinically meaningful improvement, no sooner than 8 weeks after previous cycle initiation; maximum 3 cycles in 26-week study | Efgartigimod 10 mg/kg IV infusion, four infusions per cycle (one per week), individualized dosing based on clinical response - subsequent cycles administered when MG-ADL score ≥5 (>50% non-ocular) and loss of clinically meaningful improvement (if previous responder), no sooner than 8 weeks after previous cycle initiation; maximum 3 cycles in 26-week study |
| Duration | 26 weeks | 26 weeks |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Proportion of AChR-Ab positive patients who were MG-ADL responders in cycle 1; MG-ADL responder defined as ≥2-point improvement sustained for ≥4 consecutive weeks, with first improvement occurring by week 4 | Primary | 19/64 (30%) | 44/65 (68%) | 4.95 | <0.0001 |
| QMG responder in cycle 1 (AChR-Ab+ population) - ≥3-point improvement sustained ≥4 weeks | Secondary | 9/64 (14%) | 41/65 (63%) | OR 10.84 (95% CI 4.18-31.20) | <0.0001 |
| MG-ADL responder in cycle 1 (all patients) | Secondary | 31/83 (37%) | 57/84 (68%) | OR 3.70 (95% CI 1.85-7.58) | <0.0001 |
| Percentage of time with ≥2-point MG-ADL improvement up to day 126 (AChR-Ab+ population) | Secondary | 26.6% | 48.7% | 0.0001 | |
| Median time from day 28 until no clinically meaningful improvement (AChR-Ab+ population) | Secondary | 8 days (IQR 1-57) | 35 days (IQR 18-71) | 0.26 (log-rank); 0.013 (Wilcoxon post-hoc) | |
| Early MG-ADL responder - first improvement by week 2 (AChR-Ab+ population) | Secondary | 16/64 (25%) | 37/65 (57%) | Not assessed (hierarchical testing stopped) | |
| Minimal symptom expression (MG-ADL 0-1) in cycle 1 (AChR-Ab+ population) | Secondary | 7/63 (11%) | 26/65 (40%) | <0.0001 | |
| MG-ADL responder in cycle 2 | Secondary | 11/43 (26%) | 36/51 (71%) | ||
| Cycle 1 MG-ADL responders (who qualified for retreatment) who also responded in cycle 2 | Secondary | N/A | 29/32 (90%) | ||
| Cycle 1 non-responders who became MG-ADL responders in cycle 2 | Secondary | N/A | 7/19 (37%) | ||
| MG-ADL responder in cycle 3 | Secondary | N/A | 6/7 (86%) | ||
| Duration of MG-ADL responder status - 6-7 weeks | Secondary | N/A | 14/44 (32%) | ||
| Duration of MG-ADL responder status - 8-11 weeks | Secondary | N/A | 10/44 (23%) | ||
| Duration of MG-ADL responder status - ≥12 weeks | Secondary | N/A | 15/44 (34%) | ||
| Mean maximum IgG reduction (week 4, AChR-Ab+ patients) | Secondary | N/A | 61.3% (SD 0.9) | ||
| Mean maximum AChR-Ab reduction (week 4) | Secondary | N/A | 57.6% (SD 1.3) | ||
| IgG1 subtype reduction | Secondary | N/A | 68% (SD 1.0) | ||
| IgG4 subtype reduction | Secondary | N/A | 52% (SD 1.7) | ||
| AChR-Ab negative: MG-ADL responders in cycle 1 | Secondary | 12/19 (63%) | 13/19 (68%) | ||
| AChR-Ab negative: Both MG-ADL and QMG responders in cycle 1 (post-hoc) | Secondary | 4/19 (21%) | 9/19 (47%) | ||
| MuSK-Ab positive: MG-ADL responders in cycle 1 | Secondary | 3/3 (100%) | 3/3 (100%) | ||
| Any adverse event | Adverse | 70/83 (84%) | 65/84 (77%) | ||
| Any serious adverse event | Adverse | 7/83 (8%) | 4/84 (5%) | ||
| AE leading to discontinuation | Adverse | 3/83 (4%) | 3/84 (4%) | ||
| Any infection | Adverse | 31/83 (37%) | 39/84 (46%) | ||
| Infusion-related reaction | Adverse | 8/83 (10%) | 3/84 (4%) | ||
| Headache | Adverse | 23/83 (28%) | 24/84 (29%) | ||
| Nasopharyngitis | Adverse | 15/83 (18%) | 10/84 (12%) | ||
| Nausea | Adverse | 9/83 (11%) | 7/84 (8%) | ||
| Diarrhoea | Adverse | 9/83 (11%) | 6/84 (7%) | ||
| Upper respiratory tract infection | Adverse | 4/83 (5%) | 9/84 (11%) | ||
| Urinary tract infection | Adverse | 4/83 (5%) | 8/84 (10%) | ||
| Severe AEs | Adverse | 8/83 (10%) | 9/84 (11%) | ||
| Deaths | Adverse | 0 | 0 | ||
| SAEs in efgartigimod group | Adverse | N/A | Thrombocytosis (1), rectal adenocarcinoma (1), MG worsening (1), depression (1) | ||
| SAEs in placebo group | Adverse | Myocardial ischaemia (1), atrial fibrillation (1), spinal ligament ossification (1), upper respiratory infection (1), spinal compression fracture (1), MG worsening (1), MG crisis (1) | N/A |
Subgroup Analysis
Predefined exploratory analyses showed no efficacy differences based on gender, age, or baseline MG-ADL. Concomitant use of NSISTs did not affect efficacy: 72% (18/25) of AChR-Ab+ patients NOT on NSISTs were MG-ADL responders. In patients with previous thymectomy, 60% (27/45) were responders vs 85% (17/20) without previous thymectomy - suggesting thymectomy status did not favor efgartigimod (more patients in efgartigimod group had prior thymectomy). All 6 MuSK-Ab positive patients (3 per group) were MG-ADL responders. AChR-Ab negative patients showed similar MG-ADL response rates (68% vs 63%) but a post-hoc analysis of combined MG-ADL and QMG responders showed 47% vs 21% favoring efgartigimod.
Criticisms
- Sponsored by argenx with funder involvement in study design, data collection, analysis, interpretation, and writing
- Three authors are full-time argenx employees
- Limited 26-week follow-up; long-term safety and efficacy require ongoing open-label extension data
- Trial not powered for AChR-Ab negative subgroup analysis (only 38 patients)
- Imbalanced thymectomy history between groups (70% vs 43%) though this appeared to favor placebo
- High placebo response rate in AChR-Ab negative patients (63% MG-ADL responders) limits conclusions in this subgroup
- Retreatment criteria requiring return to <2-point improvement from baseline may not reflect real-world practice
- Median cycle duration only 10 weeks, limiting understanding of optimal dosing intervals
- Log-rank test for time to loss of improvement was not significant (P=0.26); only post-hoc Wilcoxon test was significant
- No head-to-head comparison with other treatments (IVIG, plasma exchange, eculizumab)
- Infection rates numerically higher in efgartigimod group (46% vs 37%) though mostly mild-moderate
- Only 6 MuSK-positive patients enrolled - insufficient to assess efficacy in this subgroup
Funding
argenx
Based on: ADAPT (The Lancet Neurology, 2021)
Authors: James F Howard Jr, Vera Bril, Tuan Vu, ..., ADAPT Investigator Study Group
Citation: Lancet Neurol. 2021;20(7):526-536
Content summarized and formatted by NeuroTrials.ai.