MGTX
Thymectomy Trial in Non-Thymomatous Myasthenia Gravis Patients Receiving Prednisone Therapy
Year of Publication: 2016
Authors: G.I. Wolfe, H.J. Kaminski, I.B. Aban, ..., for the MGTX Study Group
Journal: New England Journal of Medicine
Citation: N Engl J Med 2016;375:511-22. DOI: 10.1056/NEJMoa1602489
Link: https://www.nejm.org/doi/full/10.1056/NEJMoa1602489
PDF: https://www.nejm.org/doi/pdf/10.1056/NEJ...ticleTools=true
PDF: Clinical Question
Is extended transsternal thymectomy combined with a standardized prednisone protocol superior to prednisone alone for improving myasthenic weakness, reducing prednisone requirements, and enhancing quality of life over 3 years in patients with non-thymomatous generalized myasthenia gravis?
Bottom Line
Thymectomy significantly improved clinical outcomes over 3 years, with lower QMG scores (difference 2.85 points), 41% lower prednisone requirements, reduced need for azathioprine (17% vs 48%), and fewer hospitalizations for exacerbations (9% vs 37%). This landmark trial provides the first randomized evidence supporting thymectomy in non-thymomatous myasthenia gravis.
Major Points
- Time-weighted average QMG score was significantly lower in thymectomy group (6.15 vs 8.99; difference 2.85; p<0.001)
- Alternate-day prednisone requirement was 41% lower with thymectomy (32 mg vs 54 mg; p<0.001)
- Only 17% of thymectomy patients required azathioprine vs 48% in prednisone-only group (p<0.001)
- Hospitalizations for MG exacerbations: 9% thymectomy vs 37% prednisone-only (p<0.001)
- 67% of thymectomy patients achieved minimal manifestation status at 36 months vs 47% (p=0.03)
- Thymectomy group had fewer treatment-associated symptoms (p<0.001) and lower distress levels (p=0.003)
- No significant difference in treatment-associated complications between groups (p=0.73)
- First randomized, controlled trial to definitively establish benefit of thymectomy in non-thymomatous MG
Design
Study Type: Multicenter, international, randomized, single-blind (rater-blinded), controlled trial
Randomization: 1
Blinding: Single-blind (rater-blinded); participants wore high-collared obscuring shirts to conceal surgical incisions; separate blinded evaluators for efficacy assessments; unblinded neurologist managed clinical care until month 4
Enrollment Period: September 2006 to November 2012
Follow-up Duration: 36 months (3 years)
Centers: 36
Countries: United States, United Kingdom, Germany, Argentina, Canada, Chile, South Africa, Italy, Thailand, Australia, Brazil, Japan, Poland, Mexico, Netherlands
Sample Size: 126
Analysis: Intention-to-treat; two-stage primary analysis with 99.5% CI for QMG score difference followed by two-sided t-test for prednisone dose at α=0.05; time-weighted averages computed using trapezoidal method; multiple imputation methods for missing data
Inclusion Criteria
- Age 18-65 years (initially 18-60, expanded in October 2008)
- Generalized non-thymomatous myasthenia gravis
- Disease duration <5 years (initially <3 years, expanded in October 2008)
- Serum acetylcholine-receptor antibody level >1.00 nmol/L (or 0.50-0.99 nmol/L with confirmatory testing)
- MGFA clinical classification II-IV
- Taking appropriate anticholinesterase therapy with or without oral glucocorticoids
Exclusion Criteria
- Thymoma on CT or MRI of chest
- Previous thymectomy
- Immunotherapy other than prednisone
- Pregnancy or lactation
- Unwillingness to avoid pregnancy
- Contraindications to glucocorticoids
- Substantial medical illness precluding participation
Baseline Characteristics
| Characteristic | Prednisone Alone | Thymectomy + Prednisone |
|---|---|---|
| N | 60 | 66 |
| Female | 65% | 76% |
| Age (years) - Median (range) | 33 (18-64) | 32 (18-63) |
| Race - Asian | 7% | 9% |
| Race - Black | 10% | 11% |
| Race - Hispanic | 28% | 26% |
| Race - Non-Hispanic White | 50% | 47% |
| Race - Other | 5% | 8% |
| Pyridostigmine at enrollment | 93% | 91% |
| Glucocorticoid at enrollment | 78% | 74% |
| Previous IVIg | 22% | 18% |
| Previous plasma exchange | 12% | 14% |
| MGFA Class IIa | 42% | 38% |
| MGFA Class IIb | 23% | 27% |
| MGFA Class III | 33% | 32% |
| MGFA Class IV | 2% | 3% |
| Disease duration (years) - Median (range) | 1.14 (0.15-4.38) | 1.08 (0.02-4.41) |
| QMG score - Mean ± SD | 12.35 ± 4.90 | 11.40 ± 5.12 |
| Prednisone dose at baseline (mg) - Mean ± SD | 42.49 ± 23.52 | 43.43 ± 28.92 |
| Prednisone use at baseline | 78% | 74% |
Arms
| Field | Thymectomy + Prednisone | Control |
|---|---|---|
| Intervention | Extended transsternal thymectomy via median sternotomy with en bloc resection of all mediastinal thymic tissue, performed within 30 days of randomization, plus standardized alternate-day prednisone protocol starting at 10 mg and increasing to maximum 100 mg or 1.5 mg/kg alternate-day; prednisone tapered after achieving minimal manifestation status and QMG <14 | Standardized alternate-day prednisone protocol starting at 10 mg, increased in 10 mg steps to maximum 100 mg alternate-day or 1.5 mg/kg (up to 120 mg if needed); tapered by 10 mg every 2 weeks after achieving minimal manifestation status and QMG <14; azathioprine permitted after 12 months if minimal manifestation status not achieved |
| Duration | 36 months follow-up | 36 months follow-up |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Dual primary outcome: (1) Time-weighted average Quantitative Myasthenia Gravis (QMG) score over 3 years; (2) Time-weighted average required alternate-day prednisone dose over 3 years | Primary | QMG Score - Control: 8.99 ± 4.93 · QMG Score - Intervention: 6.15 ± 4.09 · QMG Score Difference: 2.85 (99.5% CI 0.47-5.22) · QMG P-value: <0.001 · Prednisone Dose - Control: 54 ± 29 mg alternate-day · Prednisone Dose - Intervention: 32 ± 23 mg alternate-day · Prednisone Dose Difference: 22 mg (95% CI 12-32) · Prednisone P-value: <0.001 | |||
| Time-weighted average MG-ADL score over 3 years | Secondary | 3.41 | 2.24 | 0.008 | |
| Azathioprine use | Secondary | 48% | 17% | <0.001 | |
| Minimal manifestation status at month 36 | Secondary | 47% | 67% | 0.03 | |
| Hospitalization for MG exacerbation | Secondary | 37% | 9% | <0.001 | |
| Treatment-associated complications (survey) | Secondary | No difference | No difference | 0.73 | |
| Treatment-associated symptoms (number of patients with symptoms) | Secondary | Higher | Lower | <0.001 | |
| Distress level related to symptoms | Secondary | Higher | Lower | 0.003 | |
| Cumulative hospital days | Secondary | 19.2 ± 24.5 | 8.4 ± 8.6 | 0.09 | |
| Any serious adverse event | Adverse | 55% (33/60 patients; 93 events) | 38% (25/66 patients; 48 events) | 0.05 | |
| Life-threatening event | Adverse | 12% (7/60) | 2% (1/66) | 0.03 | |
| Disability or incapacity | Adverse | 3% (2/60) | 12% (8/66) | 0.10 | |
| Death | Adverse | 2% (1/60) | 0% | 0.48 | |
| Complication due to thymectomy | Adverse | NA | 2% (1/66) | ||
| Hospitalization (any cause) | Adverse | 52% (31/60) | 23% (15/66) | <0.001 | |
| Hospitalization for MG exacerbation | Adverse | 37% (22/60) | 9% (6/66) | <0.001 | |
| Nervous system disorder hospitalization | Adverse | 37% (22/60) | 12% (8/66) | 0.001 | |
| Infection/infestation hospitalization | Adverse | 12% (7/60) | 6% (4/66) | 0.35 | |
Subgroup Analysis
Three prespecified subgroups analyzed: (1) Previous vs no previous glucocorticoid use: Among glucocorticoid-naive patients, significant difference in prednisone dose but not QMG; among those with prior exposure, both outcomes favored thymectomy. (2) Sex: Women showed significant between-group differences in both outcomes; men showed significant difference in prednisone dose only. (3) Age at disease onset (<40 vs ≥40 years): Both subgroups showed significant benefit for thymectomy on both outcomes. Tests for interaction were not significant in any subgroup analysis.
Criticisms
- Single-blind (rater-blinded) design without sham surgery control; not feasible to subject controls to sham sternotomy
- Pill counts may not precisely measure actual prednisone intake despite diary confirmation
- Extended transsternal thymectomy may not represent less-invasive approaches now commonly used
- Concern that minimally invasive techniques may leave ectopic thymic tissue affecting long-term outcomes
- 8 patients in thymectomy group declined surgery; 8 in prednisone group crossed over to thymectomy
- Prednisone doses at 36 months higher than routine clinical practice due to protocol requirements
- Unable to conclude about differential benefits in subgroups due to non-significant interaction tests
- Small numbers limit interpretation in glucocorticoid-naive and male subgroups
- Did not test less-invasive thymectomy approaches with similar effectiveness and better cosmesis
- Restrictive eligibility criteria (disease duration <5 years, age 18-65) limit generalizability
Funding
National Institute of Neurological Disorders and Stroke (grant U01 NS042685), Muscular Dystrophy Association, Myasthenia Gravis Foundation of America
Based on: MGTX (New England Journal of Medicine, 2016)
Authors: G.I. Wolfe, H.J. Kaminski, I.B. Aban, ..., for the MGTX Study Group
Citation: N Engl J Med 2016;375:511-22. DOI: 10.1056/NEJMoa1602489
Content summarized and formatted by NeuroTrials.ai.