← Back

Methotrexate in MG

A randomized controlled trial of methotrexate for patients with generalized myasthenia gravis

Year of Publication: 2016

Authors: Mamatha Pasnoor, Jianghua He, Laura Herbelin, ..., The Methotrexate in MG Investigators of the Muscle Study Group

Journal: Neurology

Citation: Neurology 2016;87:57-64

Link: https://clinicaltrials.gov/ct2/show/NCT00814138

Clinical Question

Does methotrexate provide a steroid-sparing benefit compared to placebo in patients with AChR-antibody positive generalized myasthenia gravis who are on stable prednisone therapy?

Bottom Line

Methotrexate 20 mg weekly did not demonstrate a significant steroid-sparing effect compared to placebo over 12 months of treatment. Although the direction of change in most outcomes favored methotrexate, no differences reached statistical significance. The study was limited by greater-than-anticipated variability in prednisone AUDTC, relatively low baseline prednisone doses, and possibly insufficient duration. MTX was well-tolerated with no serious drug-related adverse events.

        Major Points
        Primary endpoint (month 4-12 prednisone AUDTC) showed no significant difference: MTX 2,997 mg vs placebo 3,485 mg (difference -488 mg, 95% CI -2,443 to 1,467, p=0.26)
Average daily prednisone dose decreased in both groups, with no significant between-group difference
All secondary MG outcome measures (QMG, MGC, MMT, MG-ADL, MG-QOL) showed numerical trends favoring MTX but none reached statistical significance
MGC change of -3.3 (MTX vs placebo) would be considered clinically meaningful but did not reach statistical significance (p=0.09)
More dropouts in placebo group (7/25) vs MTX group (1/25), primarily due to worsening symptoms
Sensitivity analyses using LOCF and worst-value-carried-forward showed significance for QMG (p<0.01) and MGC (p<0.01)
MTX was well-tolerated; no serious MTX-related adverse events despite more GI symptoms and infections
Study highlights challenges of MG trials: recruitment difficulties, improvement on prednisone alone, outcome measure variability

      

Design

Study Type: Multicenter, randomized, double-blind, placebo-controlled trial

Randomization: 1

Blinding: Double-blind; investigators, evaluators, and participants were blinded to treatment allocation; drug and placebo were overencapsulated

Enrollment Period: April 2009 to December 2013 (screening); trial conducted April 2009 to September 2014

Follow-up Duration: 12 months

Centers: 19

Countries: United States, Canada

Sample Size: 50

Analysis: Intention-to-treat; multiple imputation for missing data (assumed missing at random); Wilcoxon rank sum test for primary outcome; two-sample t-test or Wilcoxon for secondary outcomes; Bonferroni correction for secondary outcomes (α=0.01); sensitivity analyses with LOCF and worst-value-carried-forward; STATA 13.1

Inclusion Criteria

Age ≥18 years
Diagnosis of generalized MG (MGFA Class II, III, or IV)
Positive acetylcholine receptor antibodies
Stable dose of ≥10 mg/day prednisone (or equivalent alternate-day dosing) for at least 30 days prior to enrollment

Exclusion Criteria

MTX use for MG within last 2 years
Thymoma or tumor
Active infection
Interstitial lung disease
Thymectomy within previous 3 months
Steroid-sparing immunosuppressive drugs within 60 days prior to screening
Contraindication to MTX (daily NSAID use, renal or hepatic disease)

Arms

Field	Methotrexate	Control
Intervention	Oral methotrexate 2.5 mg tablets, overencapsulated, titrated from 4 tablets/week (10 mg) increasing by 2 tablets every 2 weeks to 8 tablets weekly (20 mg); standardized prednisone taper protocol starting at month 3 if clinical improvement per MGFA post-intervention status guidelines	Matching placebo tablets, overencapsulated, same titration schedule as MTX; standardized prednisone taper protocol starting at month 3 if clinical improvement per MGFA post-intervention status guidelines
Duration	12 months	12 months

Outcomes

Outcome	Type	Control	Intervention	P-value
Prednisone area under the dose-time curve (AUDTC) from months 4 to 12, reflecting cumulative prednisone exposure over 9 months	Primary	3,484.7 mg (95% CI 2,151.8-4,817.5)	2,996.6 mg (95% CI 1,495.9-4,497.3)	0.26
12-month QMG change	Secondary	+0.3 (95% CI -1.8 to 2.4)	-1.4 (95% CI -2.9 to 0.1)	0.29
12-month MMT change	Secondary	-3.3 (95% CI -6.6 to 0.1)	-5.5 (95% CI -7.4 to -3.8)	0.28
12-month MG-QOL change (median)	Secondary	-3.7 (95% CI -8.4 to 1.0)	-4.6 (95% CI -9.1 to -0.1)	0.82
12-month MG-ADL change	Secondary	+0.26 (95% CI -0.9 to 1.5)	-1.2 (95% CI -2.3 to -0.5)	0.21
12-month MGC change	Secondary	-1.3 (95% CI -3.7 to 1.1)	-4.6 (95% CI -6.4 to -2.7)	0.09
Treatment failures (prednisone increased)	Secondary	15/25 (60%)	11/25 (44%)	0.26
Study completion	Secondary	18/25 (72%)	24/25 (96%)
Total adverse events	Adverse	161 events	176 events
Pain (nonspecific)	Adverse	18.0% of events (56% of patients)	20.6% of events (52% of patients)
Gastrointestinal	Adverse	9.9% of events (44% of patients)	16.0% of events (60% of patients)
Infection	Adverse	10.6% of events (28% of patients)	16.0% of events (52% of patients)
Elevated LFTs	Adverse	7.5% of events (16% of patients)	4.6% of events (12% of patients)
Pulmonary	Adverse	7.5% of events (28% of patients)	7.4% of events (36% of patients)
Muscle weakness/fatigue	Adverse	5.0% of events (20% of patients)	7.4% of events (24% of patients)
Serious MTX-related adverse events	Adverse	NA	0
Death	Adverse	1 (stroke)	0

Subgroup Analysis

Post hoc sensitivity analyses using different imputation methods (LOCF, worst-value-carried-forward) showed significant differences favoring MTX for QMG (p<0.01 with both methods) and MGC (p<0.01 with worst-value method). Exploratory analysis showed 7/8 dropouts had baseline prednisone <30 mg/day, explaining why imputation approaches did not affect prednisone AUDTC results.

Criticisms

Study may have been underpowered due to greater-than-anticipated variability in prednisone AUDTC
Participants were on relatively low baseline prednisone doses (median 20 mg/day), limiting ability to detect further reductions
12-month duration may have been too short; prior azathioprine study required >15 months to detect benefit
Sample size based on data from azathioprine study that had key design differences (immunosuppressant-naive patients, higher prednisone doses)
Broad inclusion criteria may have enriched for older patients with milder, late-onset MG
Male predominance (70% overall) and older age (median 67 years) may limit generalizability
Stratification by age at diagnosis was not performed but may be important for future studies
Missing data imputation method was not prespecified; results differed with alternative approaches
MTX dose of 20 mg weekly may not have produced effective immunosuppression in all participants
Clinically meaningful changes in MGC not reflected in statistical significance raises questions about optimal outcome measures
Unequal dropout rates (1 MTX vs 7 placebo) may indicate selection bias in completers

Funding

Food and Drug Administration Orphan Products Division grant RO1 FD 003538; University of Kansas Medical Center Clinical and Translational Science Awards (CTSA) grants UL1 RR 033179/UL1 TR 000001; NCATS grant KL2TR000119

Based on: Methotrexate in MG (Neurology, 2016)

Authors: Mamatha Pasnoor, Jianghua He, Laura Herbelin, ..., The Methotrexate in MG Investigators of the Muscle Study Group

Citation: Neurology 2016;87:57-64

Content summarized and formatted by NeuroTrials.ai.

Methotrexate in MG

(2016)

Objective

To determine the steroid-sparing effect of methotrexate in patients with symptomatic generalized myasthenia gravis

Study Summary

• Methotrexate 20 mg weekly showed no significant steroid-sparing benefit compared to placebo over 12 months (prednisone AUDTC difference -488 mg, p=0.26)
• Both groups improved on prednisone alone; no significant differences in secondary MG outcome measures
• MTX was well-tolerated with no serious drug-related adverse events

Intervention

Oral methotrexate 20 mg weekly (titrated over 4 weeks) vs matching placebo, with standardized prednisone taper protocol

Inclusion Criteria

Age ≥18 years, generalized AChR-Ab+ MG (MGFA Class II-IV), stable prednisone dose ≥10 mg/day for ≥30 days

Study Design

Arms: Methotrexate 20 mg weekly vs Placebo

Patients per Arm: MTX: 25, Placebo: 25

Outcome

• Primary: Month 4-12 prednisone AUDTC - MTX 2,997 mg vs placebo 3,485 mg (p=0.26)
• 12-month MGC change: MTX -4.6 vs placebo -1.3 (p=0.09, clinically meaningful but not significant)
• Treatment failures (prednisone increased): MTX 11 vs placebo 15 (p=0.26)

Bottom Line

Major Points

Primary endpoint (month 4-12 prednisone AUDTC) showed no significant difference: MTX 2,997 mg vs placebo 3,485 mg (difference -488 mg, 95% CI -2,443 to 1,467, p=0.26)
Average daily prednisone dose decreased in both groups, with no significant between-group difference
All secondary MG outcome measures (QMG, MGC, MMT, MG-ADL, MG-QOL) showed numerical trends favoring MTX but none reached statistical significance
MGC change of -3.3 (MTX vs placebo) would be considered clinically meaningful but did not reach statistical significance (p=0.09)
More dropouts in placebo group (7/25) vs MTX group (1/25), primarily due to worsening symptoms
Sensitivity analyses using LOCF and worst-value-carried-forward showed significance for QMG (p<0.01) and MGC (p<0.01)
MTX was well-tolerated; no serious MTX-related adverse events despite more GI symptoms and infections
Study highlights challenges of MG trials: recruitment difficulties, improvement on prednisone alone, outcome measure variability

Study Design

Study Type: Multicenter, randomized, double-blind, placebo-controlled trial
Randomization: Yes
Blinding: Double-blind; investigators, evaluators, and participants were blinded to treatment allocation; drug and placebo were overencapsulated
Sample Size: 50
Follow-up: 12 months
Centers: 19
Countries: United States, Canada

Primary Outcome

Definition: Prednisone area under the dose-time curve (AUDTC) from months 4 to 12, reflecting cumulative prednisone exposure over 9 months

Control	Intervention	HR/OR	P-value
3,484.7 mg (95% CI 2,151.8-4,817.5)	2,996.6 mg (95% CI 1,495.9-4,497.3)	- (-2,443.4 to 1,467.3)	0.26

Limitations & Criticisms

Study may have been underpowered due to greater-than-anticipated variability in prednisone AUDTC
Participants were on relatively low baseline prednisone doses (median 20 mg/day), limiting ability to detect further reductions
12-month duration may have been too short; prior azathioprine study required >15 months to detect benefit
Sample size based on data from azathioprine study that had key design differences (immunosuppressant-naive patients, higher prednisone doses)
Broad inclusion criteria may have enriched for older patients with milder, late-onset MG
Male predominance (70% overall) and older age (median 67 years) may limit generalizability
Stratification by age at diagnosis was not performed but may be important for future studies
Missing data imputation method was not prespecified; results differed with alternative approaches
MTX dose of 20 mg weekly may not have produced effective immunosuppression in all participants
Clinically meaningful changes in MGC not reflected in statistical significance raises questions about optimal outcome measures
Unequal dropout rates (1 MTX vs 7 placebo) may indicate selection bias in completers

Citation

Neurology 2016;87:57-64

View Original Publication →

Methotrexate in MG

A randomized controlled trial of methotrexate for patients with generalized myasthenia gravis

Clinical Question

Bottom Line

Major Points

Design

Inclusion Criteria

Exclusion Criteria

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about Methotrexate in MG