Azathioprine: tolerability, efficacy, and predictors of benefit in neuromyelitis optica

To evaluate the efficacy, tolerability, optimal dosing, and monitoring of azathioprine in patients with neuromyelitis optica (NMO).

This was a chart review and telephone follow-up study of 99 patients with NMO spectrum of disorders (NMOSD) treated with azathioprine (1994-2009). NMOSD were NMO (2006 diagnostic criteria) or partial NMO forms (NMO-immunoglobulin G seropositive). Wilcoxon signed rank test was used to compare pretreatment and postinitiation of azathioprine (posttreatment) annualized relapse rates (ARR), Expanded Disability Status Scale (EDSS) score, and visual acuity outcome. Linear regression was used to assess the effects of various factors on ARR change and disability.

The median duration of NMOSD symptoms prior to initiation of azathioprine was 2 years (range 1-27); 79 patients were women. Eighty-six patients had NMO and 13 limited NMO versions, including transverse myelitis in 8 and optic neuritis in 5. Median posttreatment follow-up was 22 months. Thirty-eight patients discontinued drug (side effects, 22; no efficacy, 13; lymphoma, 3). Among 70 patients with >12 months follow-up, 48 received ≥2.0 mg/kg/day (ARR: pretreatment, 2.20; posttreatment, 0.52); 22 received <2.0 mg/kg/day (ARR: pretreatment, 2.09; posttreatment, 0.82); 52 received concomitant prednisone (ARR: pretreatment, 2.20; posttreatment, 0.89) and 18 did not (ARR: pretreatment, 1.54; posttreatment, 0.23); p < 0.0001 for each comparison. EDSS was stable or improved despite ongoing attacks in 22 patients (31%). Twenty-six patients tolerated azathioprine and were relapse-free (37%, median follow-up 24 months; range 12-151). Mean corpuscular volume increase influenced ARR change (p = 0.049).

Azathioprine is generally effective and well-tolerated. Early initiation, adequate dosing, and hematologic parameter monitoring may optimize efficacy.

This study provides Class IV evidence that azathioprine is effective for reducing relapse rates and improving EDSS and visual acuity scores in patients with NMO spectrum of disorders.