To investigate the association between the use of two fluoroquinolones (ciprofloxacin and levofloxacin) and one macrolide (azithromycin) and myasthenia gravis (MG) exacerbations

Several commonly used antibiotics have been linked to MG exacerbations due to their potential impairment of the neuromuscular junction based on case reports. Practice guidelines recommend against their use in MG patients out of safety concerns. However the true associated risks of MG exacerbation remains unknown due to a lack of evidence from large scale retrospective or prospective studies.

Patients with confirmed MG seen at the Cleveland Clinic Neuromuscular Center between 2002-2022 and with documented use of ciprofloxacin, levofloxacin or azithromycin were included. The strength of association between antibiotic usage and MG exacerbation was scored using the Adverse Drug Reactions Probability Scale as definite, probable, possible or doubtful. Descriptive statistics were utilized to summarize patient and clinical characteristics. Mixed-effects model was constructed to model MG exacerbation status following antibiotic use.

Three hundred and sixty-six patients were included, with a total of 918 episodes of antibiotic usage (n=339 for ciprofloxacin, n=186 for levofloxacin and n=393 for azithromycin). The frequencies of MG exacerbation following antibiotic use were as follows: 17 (1.9%) for the whole group, 8 (2.4%) for ciprofloxacin, 2 (1.1%) for levofloxacin and 7 (1.8%) for azithromycin. History of MG-related hospitalization or ED visit within 6 months prior to antibiotic use was significantly associated with MG exacerbation (p=0.004). The strength of the association between antibiotic use and MG exacerbation was determined as definite in 0, probable in 3, possible in 13 and doubtful in 1 patient.

Analysis of a large group of MG patients reveal that the usage of ciprofloxacin, levofloxacin or azithromycin resulted in MG exacerbation in a minority of patients. MG patients with recent hospitalization or ED visit appear to be at higher risk.