To determine the efficacy and tolerability of Azathioprine in patients with NMOSD using systematic review of relevant studies. 

Major healthcare databases (CENTRAL, MEDLINE, Embase, Scopus, LILACS, clinicalTrials.gov website, and the HERDIN Database of the Philippines) were searched from inception to January 2018. 

From a total of 274 records, 18 studies with various study designs were included. Azathioprine + Prednisone (AZA+Pd) regimen may be inferior to Rituximab (RTX) in terms of annualized relapse rate (P<0.001), risk for relapse (P=0.02), relapse-free rate (P=0.03), and disability reduction measured by expanded disability status scale (P<0.001). AZA+Pd treatment may be more beneficial than Interferon-beta particularly in relapse risk (P=0.001). On the other hand, no significant difference was noted when AZA+Pd was compared to Mycophenalate Mofetil + Prednisone (MMF+Pd) (P=0.08) and to Cyclophosphamide + Azathioprine + Prednisone (CYP+AZA+Pd) (P=0.45) regimens in terms of relapse risk. Gastrointestinal and hematologic adverse events associated with Azathioprine use, although majority were minor, were considered frequent and may significantly lead to drug discontinuity and inadherence. 

In patients with NMOSD, Azathioprine at 2 to 3 mg/kg/day for at least 6 months in combination with oral corticosteroids tapered to the lowest effective dose may be administered as an empiric regimen for the prevention of relapse and reduction of neurologic disability.  Current evidence suggest that this regimen may be inferior to RTX in terms of relapse prevention and disability reduction. In terms of relapse risk, AZA+Pd regimen was significantly favored than interferon-beta but yielded no statistical difference when compared to MMF+Pd and to CYP+AZA+Pd regimens. Azathioprine was associated with relatively frequent, minor gastrointestinal and hematologic adverse events.