Fluoroquinolones and the risk of aortic aneurysm or dissection: A population-based propensity score-matched German cohort study.

Abstract

Objective: To investigate the risk of aortic aneurysm or dissection associated with fluoroquinolone (FQ) prescription compared to macrolides in German routine health care data in order to replicate the recent study (Pharmacotherapy 2023;43:883) extending the results by contributing evidence for six additional broad-spectrum antibiotic classes as active comparators.

Design: Cohort study in active comparator new user design comparing FQ with macrolides, tetracyclines, penicillins with extended spectrum, penicillins and beta-lactamase inhibitor combinations, second- and third-generation cephalosporins, sulfonamide and trimethoprim combinations, and lincosamides.

Setting: German statutory health insurance, the "Allgemeine Ortskrankenkasse" (AOK), January 2013 to December 2019.

Participants: Adults with at least one new prescription fill for FQ or active comparator antibiotics. New users were defined as individuals without antibiotic prescription fills, aortic aneurysm or dissection diagnoses, and hospitalization within 365 days prior to the cohort entry date. Users of FQ and active comparators were matched by nearest neighbor 1:1 propensity score matching.

Main outcome measures: Incident inpatient aortic aneurysm or dissection was observed within a 60-day risk window. In sensitivity analyses, an extended risk window of 90 days was applied, and specific FQ agents, dosages, and diagnoses were stratified.

Results: FQ episodes were associated with an increased risk for aortic aneurysm or dissection compared to macrolides (aHR = 1.52 [1.33; 1.74]), which replicates the risk estimate of Garg et al. (aHR = 1.34 [1.17; 1.54]). This association was robust in a 90-day risk window and for ciprofloxacin, levofloxacin, and moxifloxacin. Moxifloxacin comprised the greatest risk of aortic aneurysm or dissection compared to macrolides (aHR = 2.13 [1.64; 2.77]). Moreover, we observed similar associations when comparing FQ to tetracyclines, penicillins with extended spectrum, cephalosporins, and lincosamides (aHR = 1.86 [1.54; 2.24], aHR = 1.45 [1.28; 1.65], aHR = 1.23 [1.10; 1.37], and aHR = 1.73 [1.43; 2.11]), respectively.

Conclusion: In a German cohort study, FQ use was associated with a 52% increased risk for aortic aneurysm or dissection within 60 days compared with macrolide use. The risk of FQ-associated aortic aneurysm or dissection compared to macrolides can be replicated in German routine health care data. Extending the analysis, we provided new insights that the effect size may depend on the chosen AC.