Author + information
- Sandy Maumus-Robert, PharmD, PhD∗ (, )@CHUBordeaux@1APariente@univbordeaux@DrugsSafe,
- Xavier Bérard, MD, PhD,
- Yohann Mansiaux, PhD,
- Pascale Tubert-Bitter, PhD,
- Stéphanie Debette, MD, PhD and
- Antoine Pariente, MD, PhD
- ↵∗University Bordeaux, Inserm, Bordeaux Population Health Research Center, Team Pharmacoepidemiology, UMR 1219, F-33000 Bordeaux, France
Fluoroquinolones have been associated with aortic aneurysm or dissection (1–4); the European Medicines Agency recently warned against this risk (5). We added to the existing body of evidence using: 1) a self-controlled design; and 2) amoxicillin as an active comparator for fluoroquinolones.
We performed case-time-control analyses using the French health insurance nationwide databases (SNIIR-AM) covering >60 million inhabitants. Cases were patients age 18 years or older with incident aortoiliac aneurysm or dissection who were diagnosed between July 1, 2010, and December 31, 2015; had been exposed to fluoroquinolones (or amoxicillin) during the 180 days before the date of outcome occurrence (i.e., index date); and were not hospitalized in the 180 days before the outcome (no information on in-hospital antibiotic exposure available from the database). Drug exposures were identified from out-hospital reimbursements, and outcomes from in-hospital diagnoses (International Classification of Diseases-10th Revision codes) and medical procedures. The main outcome of interest corresponded to aortoiliac ruptured aneurysm and/or dissection; it was extended to unruptured events for secondary analyses.
To estimate the association using each patient as his or her own control, frequency of exposure to fluoroquinolone was compared between a defined risk window (day 30 to day 1 before the outcome) and matched control windows (day 120 to day 91, day 150 to day 121, and day 180 to day 151 before the outcome). Trend for exposure along the period was controlled using a reference group of randomly selected subjects free of the event who were individually matched to cases regarding age and sex (up to 10 per case).
The conditional logistic model used considered as time-varying confounders the use of other antibiotics, corticosteroids, anticoagulant agents, antiplatelet agents, and antihypertensive drugs.
Association with amoxicillin use, considered a marker of a potential indication bias, was studied similarly. The association with fluoroquinolone corrected for this risk of indication bias was expressed in terms of amoxicillin-adjusted odds ratio (OR) for fluoroquinolones (ratio of OR: OR for fluoroquinolones divided by OR for amoxicillin) and its corresponding 95% confidence interval.
Sensitivity analyses were conducted: 1) extending the risk window (and corresponding control windows) to 60 or 90 days; and 2) extending the population to patients with history of hospitalization during the 180 days prior to index date. A complementary case-control analysis considering cases and ≤10 randomly selected control subjects matched on a disease risk score was conducted, which also explored the association with fluoroquinolone past use (day 61 to day 365 before the event).
Of the 5,946 subjects who presented with incident aortoiliac ruptured aneurysm or dissection during the study period, 86 (median age 70 years; interquartile range: 62 to 80 years; 36% women) had been exposed to fluoroquinolones during the prior 180 days, of which 36 had been exposed during the 30-day risk window preceding the event. Exposure to amoxicillin occurred in 316 cases (median age 67 years [interquartile range: 57 to 79 years]; 25.3% women), including 82 exposed during the risk window. The amoxicillin-adjusted OR-R for the association with aortoiliac aneurysmal rupture or dissection was estimated at 2.44 (95% confidence interval: 1.31 to 4.57) for fluoroquinolones; it was also significant when considering both ruptured and unruptured aneurysms or dissections (Figure 1). Consistent results were obtained from the sensitivity and complementary analyses; the latter highlighted no association with fluoroquinolone past use (data not shown).
Altogether, these results are consistent with those of the previous studies (1–4). However, contrary to a recently published study (3), we found no association with fluoroquinolone past use. This study did not use an active comparator (e.g., amoxicillin in ours) to account for a potential indication bias, which could explain this discrepancy.
Aside from this internal and external consistency, the study we herein report has several strengths, some relating to the data source (drug reimbursements and hospitalization diagnoses and procedures exhaustive recording), and others to the design used, which, to mitigate the risk of confounding or indication biases, combines the advantages of a self-controlled design (self-adjustment for all time-unvarying confounders) with those of time-depending adjustment and use of an active comparator.
In conclusion, our results confirm the existence of an increased risk of aortoiliac aneurysms or dissections in the month following fluoroquinolone initiation, which appears to be independent from the potential role of underlying infections.
Please note: The present study is part of the DRUGS-SAFE (Drugs Systematized Assessment in real-liFe Environment) research platform that is funded by the French Medicines Agency (Agence Nationale de Sécurité du Médicament et des Produits de Santé [ANSM]). This platform aims at providing an integrated system allowing the concomitant monitoring of drug use and safety in France. The funder had no role in the design and conduct of the studies; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication. This publication represents the views of the authors and does not necessarily represent the opinion of the French Medicines Agency. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2019 American College of Cardiology Foundation
- Daneman N.,
- Lu H.,
- Redelmeier D.A.
- Lee C.C.,
- Lee M.G.,
- Hsieh R.,
- et al.
- Lee C.C.,
- Lee M.T.,
- Chen Y.S.,
- et al.
- Pasternak B.,
- Inghammar M.,
- Svanstrom H.
- Pharmacovigilance Risk Assessment Committee (PRAC)