1 s2.0-s0002934317307180

Aortic Dissection and Aortic Aneurysms
Associated with Fluoroquinolones:
A Systematic Review and Meta-Analysis
Sonal Singh, MD, MPH,a
Amit Nautiyal, MDb
a
Department of Medicine, University of Massachusetts Medical School, Worcester, Mass; b
Department of Medicine, Albany Medical
College, Albany, NY.
ABSTRACT
BACKGROUND: Our objective was to evaluate the association between fluoroquinolone use and aortic dis-
section or aortic aneurysm in a systematic review and meta-analysis.
METHODS: We searched Medline, Embase, and Scopus from inception to February 15, 2017. We selected
controlled studies for inclusion if they reported data on aortic dissection and aortic aneurysm associated
with fluoroquinolones exposure versus no exposure. Data were extracted by 2 independent reviewers, with
disagreements resolved through further discussion. We assessed the quality of studies using the Newcastle–
Ottawa Scale for observational studies and the strength of evidence using the Grading of Recommendations
Assessment, Development, and Evaluation approach. The odds ratios (ORs) from observational studies were
pooled using the fixed-effect inverse variance method, and statistical heterogeneity was assessed using the
I2
statistic.
RESULTS: After a review of 714 citations, we included 2 observational studies in the meta-analysis.
Current use of fluoroquinolones was associated with a statistically significantly increased risk of aortic
dissection (OR, 2.79; 95% confidence interval [CI], 2.31-3.37; I2
= 0%) and aortic aneurysm (OR,
2.25; 95% CI, 2.03-2.49; I2
= 0%) in a fixed-effects meta-analysis. The unadjusted OR estimates and
sensitivity analysis using a random-effects model showed similar results. We rated the strength of
evidence to be of moderate quality. The number needed to treat to harm for aortic aneurysm for elderly
patients aged more than 65 years who were current users of fluoroquinolones was estimated to be 618
(95% CI, 518-749).
CONCLUSIONS: Evidence from a small number of studies suggests that exposure to fluoroquinolones
is consistently associated with a small but significantly increased risk of aortic dissection and aortic
aneurysm.
© 2017 Elsevier Inc. All rights reserved. • The American Journal of Medicine (2017) 130, 1449–1457
KEYWORDS: Aortic aneurysm; Aortic dissection; Fluoroquinolones; Meta-analysis; Systematic review
INTRODUCTION
Fluoroquinolones are one of the most widely used group of
antibiotics. They are used to treat a wide variety of infec-
tions, including urinary tract, respiratory, skin, bone, and joint
infections. They are routinely prescribed by internists, family
practitioners, specialists, subspecialists, and surgeons. There
have been several observational studies and their systematic
reviews linking fluoroquinolones to adverse effects on col-
lagenous structures, such as tendon rupture1
and retinal
detachment.2
The aorta is also rich in type 1 and type III collagen.
However, the effect of fluoroquinolones on the aorta has not
been systematically evaluated. Aortic aneurysms are degen-
erative disorders associated with sudden death due to aortic
rupture and aortic dissection. They may involve both the prox-
imal aorta (Stanford Type A) and the distal aorta beyond the
Funding: None.
Conflict of Interest: None.
Authorship: Both authors had access to the data and played a role in
writing this manuscript.
Requests for reprints should be addressed to Sonal Singh, MD, MPH,
University of Massachusetts Medical School, 55 Lake Ave North, Worces-
ter, MA 01655-0002.
E-mail address: Sonal.Singh@umassmemorial.org
CLINICAL RESEARCH STUDY
0002-9343/$ - see front matter © 2017 Elsevier Inc. All rights reserved.
https://doi.org/10.1016/j.amjmed.2017.06.029
Downloaded for rachmad sammulia (iwandrake@yahoo.co.id) at Universitas Muslim Indonesia from ClinicalKey.com by Elsevier on March 17, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.

aortic arch (Stanford Type B). Although aortic dissection and
aneurysm are relatively rare,3
they may have serious and life-
threatening consequences if not diagnosed and treated early.
In a Swedish population–based cohort study of more than
14,000 cases from 1987 to 2002, the incidence of thoracic
aortic disease (aneurysm and dissection) was noted to be 16.3
per 100,000 per year for men and
9.1 per 100,000 per year for
women.4
Another study from Berlin
reported a similar incidence of 11.9
cases per 100,000 inhabitants per
year from 2010 to 2014.5
Several
patients may die before reaching
the hospital. In a cohort of 153
patients with abdominal aortic an-
eurysm, the mortality ranged from
4.2% for elective cases of aneu-
rysm repair, to 16.7% for acute
cases, to 55.8% for ruptured cases.6
The known risk factors for aortic an-
eurysm development include age,
smoking, hypertension, diabetes,
and Marfan’s syndrome. More than
one third of patients with Marfan’s
syndrome may experience aortic dissection.7
Given the wide-
spread use of fluoroquinolones, it is important to fully evaluate
the risk of aortic dissection or aortic aneurysm associated with
fluoroquinolone use.
OBJECTIVES
Our objective was to determine the effect of fluoroquinolone
use on the risk of aortic dissection or aneurysm compared
with controls.
METHODS
Systematic Review Registration
We carried out a systematic review and meta-analysis ac-
cording to a prespecified protocol. The protocol was registered
at PROSPERO8
(Appendix 1, available online). We fol-
lowed the Preferred Reporting Items for Systematic Reviews
and Meta-Analyses guidelines for reporting of systematic
reviews and meta-analysis (Appendix 2, available online).9
Ethical Approval. We did not obtain ethical approval because
the study was a systematic review and meta-analysis of
summary data.
Eligibility Criteria. We selected controlled studies that re-
ported on aortic aneurysms or aortic dilation in patients
receiving any of the fluoroquinolones: besifloxacin,
ciprofloxacin, gatifloxacin, levofloxacin, moxifloxacin,
norfloxacin, ofloxacin, enoxacin, and pefloxacin. The con-
trols included unexposed patients or patients exposed to other
antibiotics. We included studies of a cohort or case-control
design that enrolled adult participants who had used
fluoroquinolones for any indication. The primary outcomes
of interest were aortic aneurysms and aortic dissection. El-
igible studies had to present 1 of the following: odds ratio
(OR), relative risk, hazard ratio, or sufficient raw data to enable
calculation of the OR. We excluded
animal studies, preclinical studies,
pharmacokinetic studies, pharma-
codynamics studies, narrative
reviews, case reports, and case
series. We excluded studies not rel-
evant to the study question, meeting
or conference abstracts, and studies
for which the full text was not
available.
Search Strategy. We systematical-
ly searched PubMed, Scopus, and
Embase from inception through
February 15, 2017. We used a com-
bination of the search terms for
these databases. The details of the
search strategy for each of the da-
tabases is shown (Appendix 3, available online). We also
searched the references of included articles to find addition-
al references. We did not use any language restrictions in the
search for articles. We also signed up with PubMed to receive
automated electronic notification of any new articles.
Data Extraction. Two reviewers (SS or AN) checked all titles
and abstracts for studies that could meet the inclusion crite-
ria. We retrieved full reports of these potentially eligible studies
for detailed assessment by 2 reviewers (SS and AN), who then
independently extracted information on the study design, lo-
cation of the study, specific fluoroquinolone use, dose and
duration of drug use, and follow-up and outcomes onto a
preformatted spreadsheet. We extracted data on participant
characteristics, confounders, unadjusted and adjusted rela-
tive risk, hazard ratio or OR, and 95% confidence intervals
(CIs). All discrepancies were resolved with agreement after
rechecking the source articles and further discussion among
the reviewers with full consensus before inclusion. When dif-
ferent timings and durations of fluoroquinolone use were
reported in the study, we prespecified that data would be
preferentially extracted from the participants with current
use.
Risk of Bias Assessment. We used an adapted version of the
Newcastle–Ottawa Scale for Quality Assessment for cohort
and case-control studies.10
We rated cohort studies on rep-
resentativeness, selection bias, ascertainment of exposure, and
outcome, ensuring comparability of the exposed and
nonexposed cohorts, follow-up length, and duration. We
rated case-control studies on case definitions, representative-
ness, selection and definitions of controls, comparability of
CLINICAL SIGNIFICANCE
• Fluoroquinolones are associated with a
significantly increased risk of aortic an-
eurysm (OR, 2.25; 95% CI, 2.03-2.49)
and aortic dissection (OR, 2.79; 95% CI,
2.31-3.37), although the absolute in-
crease in risk is modest.
• The number needed to treat to harm for
aortic aneurysms associated with
fluoroquinolones is 618 (95% CI, 518-
749) among an elderly population aged
more than 65 years.
1450 The American Journal of Medicine, Vol 130, No 12, December 2017

cases and controls, ascertainment of exposure, sampling of
cases and controls, and nonresponse rates. All items were
independently rated by both reviewers with full consensus
reached before inclusion.
Data Synthesis. We conducted both qualitative and quanti-
tative synthesis. We considered meta-analysis when 2 or more
studies could be pooled after evaluating the studies for clin-
ical and statistical heterogeneity. Clinical heterogeneity was
examined by evaluating for differences in study population,
intervention, or comparators. We used the I2
statistic to assess
statistical heterogeneity.11
We considered I2
of <50% as low
heterogeneity, I2
50% to 75% as moderate heterogeneity, and
I2
>75% as evidence of severe statistical heterogeneity. We
conducted fixed-effects meta-analysis using the inverse vari-
ance method for pooled ORs when the number of studies was
low (<3) and there was no evidence of statistical heteroge-
neity, and we used the random-effects model for sensitivity
analysis. We assumed similarity between the OR and other
relative measures, such as relative risk, rate ratios, or hazard
ratios because aortic dissection and aneurysms were rare
events.12
When possible, we aimed to pool adjusted ORs from
the primary studies; otherwise, we used the unadjusted ORs.
We planned to assess for publication bias using the Egger’s
test and examined whether all outcomes evaluated in the
methods were reported. All meta-analyses were conducted
in StatsDirect.13
Because the OR represents a relative measure of effect,
we also estimated an absolute measure of effect, the number
needed to treat to harm, and its 95% CI.14
The number needed
to treat to harm is the number of patients who need to be
treated with fluoroquinolones for 1 additional patient to have
an adverse event. The pooled ORs were used to calculate the
number needed to treat to harm using the Visual Rx soft-
ware to convert the ORs to number needed to treat to harm.15
The baseline risk was obtained from the unexposed group in
the population-based studies.
Analysis of Subgroups. We planned to conduct sensitivity
analysis to study the influence of study design, study quality,
and use of active versus untreated comparators.
Strength of Evidence Rating. We also rated the strength of
evidence by adapting the Grading of Recommendations As-
sessment, Development, and Evaluation approach using the
major domains of study limitations, imprecision, inconsis-
tency, indirectness, and publication bias.16
RESULTS
Study Selection
The Preferred Reporting Items for Systematic Reviews and
Meta-Analyses flow sheet for studies is shown in Figure 1.
A total of 438 citations were available for screening after
removal of duplicates. A total of 425 articles were removed
after reviews of titles and abstracts. We reviewed a total of
13 full-text articles. Two controlled observational studies were
included in the systematic review and meta-analysis.17,18
There
were no randomized controlled trials that evaluated this
outcome. The population-based incidence of acute aortic dis-
section and thoracic aortic aneurysm rupture is low and
unlikely to be fully captured in underpowered randomized
controlled trials.3,4
Study Design and Characteristics
The study design and characteristics are shown in Table 1.
The 2 studies included 1 propensity-matched, nested, case-
control study from the National Health Insurance Research
Database in Taiwan.17
This study included 1477 cases and
1,477,000 controls matched on age, sex, and index date.17
Another study was a propensity-matched population–based
longitudinal cohort study among 1.7 million elderly partici-
pants aged more than 65 years in Ontario, Canada, with a total
duration of follow-up of 22,380,515 patient-days (minimum
2 years; maximum 17 years).18
Adult participants aged more
than 18 years were sampled from the National Health Insur-
ance Research Database in the case-control study,17
and elderly
participants aged more than 65 years were sampled from the
Ontario Drug Database in the cohort study.18
Both studies
evaluated different types of fluoroquinolones, with the case-
control study from Taiwan evaluating some fluoroquinolones
that are unavailable in North America.17
Exposure
Exposure was defined using prescription databases in both
studies. The case-control study defined 3 categories of ex-
posure: 1) current exposure within 60 days of index date, 2)
recent exposure within 60 days up to 365 days from the index
date, and 3) any exposure up to 365 days from the index date.
The cohort study defined current exposure as exposure up to
30 days after fluoroquinolone use.18
The mean time to diag-
nosis of aortic aneurysm in the cohort study was 20 days.
Outcome
Both studies used a combination of administrative diagnos-
tic codes to identify aortic aneurysms and aortic dissection.
The primary outcome was the first occurrence of hospital-
ization for aortic aneurysms and aortic dissection in the case-
control study.17
Risk of Bias of Included Studies
The risk of bias assessment using the Newcastle–Ottawa Scale
is shown in Table 2. The included studies were broadly similar
in terms of ascertainment of drug use and outcomes; they relied
mainly on administrative codes and pharmacy claims data-
bases. They did not verify drug history directly with the
patients or check the validity of the prescriptions data source.
The cohort study reported a 99% concordance between the
Ontario Drug Database and the pharmacy chart review in pre-
vious studies.18
Both studies reported on outcome ascertainment
based on administrative codes; the case-control study
conducted independent validation of their algorithm for aortic
1451Singh and Nautiyal Fluoroquinolones and Aortic Dissection

dissection and aortic aneurysm in a tertiary medical center
and reported a positive predictive value of 92%.17
Both studies used a wide variety of variables to adjust for
potential confounders, including smoking, hypertension, di-
abetes, and cardiovascular risk factors. Both studies used
propensity score methods to adjust for confounding and re-
ported both unadjusted and propensity-matched adjusted risk
estimates.
Fixed-Effects Meta-Analysis of Aortic
Dissection and Aortic Aneurysm Among
Current Users
The results of the fixed-effects meta-analysis on aortic dis-
section and aortic aneurysm for current users of
fluoroquinolones versus nonusers for the propensity-adjusted
OR are shown in Figures 2 and 3, respectively. Compared
with nonusers, current use of fluoroquinolones was associ-
ated with a statistically significant increased odds of aortic
dissection (propensity-adjusted OR, 2.79; 95% CI, 2.31-
3.37; I2
= 0%) and aortic aneurysm (propensity-adjusted OR,
2.25; 95% CI, 2.03-2.49; I2
= 0%) in a fixed-effects meta-
analysis of 2 studies.17,18
The results of the fixed-effects meta-analysis on aortic dis-
section and aortic aneurysm for current users of
fluoroquinolones versus nonusers of fluoroquinolones for the
unadjusted OR are shown in Appendix 4 (available online)
and Appendix 5 (available online), respectively. The unad-
justed analysis also showed a statistically significantly
increased risk for aortic dissection (OR, 3.13; 95% CI, 2.59-
3.79; I2
= 0%) and aortic aneurysm (OR, 2.73; 95% CI, 2.54-
2.93; I2
= 0%).17,18
There was no evidence of statistical
heterogeneity for the pooled results for any of these outcomes.
Sensitivity Analysis
Sensitivity analysis using the random-effects model yielded
propensity and crude estimates for the risk of aortic dissection
or aortic aneurysm among current users of fluoroquinolones
versus nonusers that were similar in direction, magnitude, and
statistical significance to those from the fixed-effects model
(Appendix 6, available online).
Risk of Aortic Dissection and Aortic Aneurysm
Among Past Users and Any Users
Only 1 study reported on the reported on risk of aortic dissec-
tion and aortic aneurysm among past or any users compared with
Records identified through database
search Pubmed (n=44), Scopus
(n=335), Embase ( n=335)
(Total n = 714)
ScreeningIncludedEligibilityIdentification
Additional records identified
through other sources
(n =0 )
Records after duplicates removed
(n =438)
Records screened
(n =438 )
Records excluded
(n =425)
Case report (n=310)
Not relevant (n=109)
Review (n=6)
No original data (n=0)
Non-humans (n=0)
Full-text articles assessed
for eligibility
(n =13 )
Full-text articles excluded,
with reasons
(n =11 )
Case report (n=1)
Commentary (n=1)
Review articles (n=1)
No relevant data on
outcome (n=2)
Not relevant intervention
(n=1)
Not relevant data (n=1)
No original data (n=4)
Studies included in
qualitative synthesis
(n =2 )
Studies included in
quantitative synthesis
(meta-analysis)
(n =2 )
Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2009
flow diagram.

Table 1 A. Study Design and Characteristics of Controlled Observational Studies Reporting on Fluoroquinolones and Aortic Aneurysm and Aortic Dissection. B. Study Design and Char-
acteristics of Controlled Observational Studies Reporting on Fluoroquinolones and Aortic Aneurysm and Aortic Dissection
A
Study
Authors
Study Design and
Location
Follow-up
Duration, Mean Participants Interventions Outcomes Controls
Lee et al17
Propensity-matched
population-based
nested case control
in Taiwan
3613 Adults in the National
Health Insurance
Research Database
Taiwan 1998-2011
Ciprofloxacin, levofloxacin, ofloxacin,
sparfloxacin, norfloxacin, lomefloxacin,
moxifloxacin, gemifloxacin, exoxacin,
pefloxacin
1477 cases of first AA or AD requiring
hospitalization plus imaging with
echocardiography, magnetic
resonance imaging, or angiography
1,477,000 cases using
risk set from the
same population
A, continued
Study
Authors Adjustment Exposure Category Outcomes Crude RR, 95% CI
Adjusted RR,
95% CI
Propensity Score–
Adjusted RR, 95% CI
Lee et al17
Demographic
cardiovascular and
healthcare use
including indications
for fluoroquinolone use
and comorbidity score
Current within 60 d of
index date
Aortic aneurysm or dissection 2.93 (2.17-3.97) 2.28 (1.67-3.13) 2.43 (1.83-3.22)
Aortic aneurysm 2.98 (2.05-4.32) NA 2.36 (1.66-3.36)
Aortic dissection 2.93 (1.79-4.80) NA 2.55 (1.58-4.11)
Past: 60 to 365 d before
index date
Any use in prior 1 y of
index date
B
Study
Authors Study Design Study Location
Follow-up Duration,
Patient-Days Participants Interventions Outcomes
Daneman
et al18
Propensity-matched,
population-based
longitudinal cohort
Ontario, Canada 22,380,515 Inception cohort of
elderly aged >65 y
from 1997-2012
Ciprofloxacin, levofloxacin, ofloxacin,
norfloxacin, moxifloxacin in the
outpatient setting
Aortic aneurysms and
aortic dissection using
administrative codes
B, continued
Study
Authors Exposure Controls Adjustment Outcomes Crude HR (95% CI)
Propensity-Adjusted
HR (95% CI)
Daneman
et al18
Currently exposed during
fluoroquinolones use
and 30 d after
treatment
Positive and negative
tracer
Demographic, cardiovascular, and
healthcare use including indications
for fluoroquinolone use and
comorbidity score
Aortic aneurysm 2.72 (2.53-2.93) 2.24 (2.02-2.49)
Aortic dissection* 3.17 (2.58-3.90) 2.84 (2.32-3.50)
AA = aortic aneurysm; AD = aortic dissection; CI = confidence interval; HR = hazard ratio; NA = not available; RR = relative risk.
*Limited to emergency hospitalizations.
1453SinghandNautiyalFluoroquinolonesandAorticDissection
Downloadedforrachmadsammulia(iwandrake@yahoo.co.id)atUniversitasMuslimIndonesiafromClinicalKey.combyElsevieronMarch17,2018.
Forpersonaluseonly.Nootheruseswithoutpermission.Copyright©2018.ElsevierInc.Allrightsreserved.

Table 2 Risk of Bias Assessment of Controlled Observational Studies Reporting on Fluoroquinolones and Aortic Aneurysm and Aortic
Dissection Using the Newcastle–Ottawa Scale
Lee et al17
Selection
Adequate case definition Independent validation with positive predictive value of 92%;
more specific definition by combing with surgical procedures
Representative case Representative cases
Selection of controls Population based
Definition of controls Prevalent controls removed
Comparability
Comparability of controls Controls for demographic and 96 covariates; cases with higher
burden of cardiovascular disease and comorbidities
Exposure
Ascertainment of exposure Medical records of prescription fill
Same method for ascertainment of cases and controls Yes
Nonresponse rate Not reported
Daneman et al18
Selection
Representativeness of the exposed Limited to the elderly aged >65 y
Representativeness of the nonexposed Drawn from the same source
Ascertainment of exposure Medications in the Ontario database
Outcome of interest not present at inception Adequate in this inception cohort
Comparability
Study controls Adequate and used both positive tracer and negative tracer
Outcome
Assessment Administrative diagnosis
Adequate duration Adequate minimum was 2 y
Completeness of follow-up Complete
Figure 2 Fixed-effects meta-analysis of ﬂuoroquinolone use and aortic dissection (propensity-
matched OR).

nonusers.17
The propensity score estimates for past use (rel-
ative risk, 1.19; 95% CI, 0.85-1.66) and any use (relative risk,
1.37; 95% CI, 1.04-1.79) were attenuated in the direction of
those of current users but remained statistically significant
(Table 1).
Assessment of Publication Bias
We did not report on statistical tests for publication bias, such
as funnel plots, which are underpowered when the number
of studies is low.
Number Needed to Treat to Harm for Aortic
Aneurysm
The number needed to treat to harm for aortic aneurysm with
current use of fluoroquinolone was estimated to be 618 (95%
CI, 518-749) on the basis of the underlying risk in the un-
exposed cohort (0.13/100 patient-years) of elderly
participants.18
Our estimates of the number needed to treat
to harm are only applicable to similar patients receiving
fluoroquinolones for the same treatment duration as in the
included study.
Strength of Evidence Rating
We rated the strength of evidence to be of moderate quality
because of study limitations, directness of evidence, large mag-
nitude of treatment effect, consistency, and relative precision.
Discussion
Our systematic review and meta-analysis show that there is
significantly increased risk of aortic dissection or aortic an-
eurysm associated with current use of fluoroquinolones.
Although current fluoroquinolone use more than doubles the
risk of aortic dissection and aortic aneurysm, the absolute risk
is modest with the number needed to treat to harm at 618.
In addition, the quality of the evidence is moderate.
There are some hypothesized biological mechanisms by
which fluoroquinolones may accentuate the development of
aortic aneurysms (mean duration of use ≈20 days in the cohort
study). The use of fluoroquinolones has been known to have
adverse effects on other collagenous structures by inducing matrix
metalloproteinase, providing one potential mechanism for in-
creasing the risk of aortic aneurysms and aortic dissection. Matrix
metalloproteinases have been known to play a role in the patho-
genesis of aortic aneurysms, whereas matrix metalloproteinase
inhibitors may oppose the development of aneurysms.19
The pro-
teolytic activities of matrix metalloproteinases, particularly
matrix metalloproteinase-2 and matrix metalloproteinase-9,
have been shown to play a role in the induction of aortic an-
eurysms in mice20
and a role in the development of aortic
aneurysms in humans through degradation of the collagen
fibril.21
In another small study among humans, therapeutic con-
centrations of ciprofloxacin (5 ng/mL) significantly increased
total matrix metalloproteinase activity more than 2-fold in cul-
tured human aortic smooth muscle.22
Study Limitations
There are several limitations to our review, which primarily
reflect the limitations of underlying data. We did not find any
randomized controlled trials that evaluated these outcomes.
Our findings must be interpreted with caution given that only
studies were eligible for inclusion in the meta-analysis. We
were unable to conduct meta-regression according to expo-
sure dose and duration because of the absence of individual-
level data. Misclassification of exposure and outcome is always
possible in a database study. Although the estimates were
Figure 3 Fixed-effects meta-analysis of fluoroquinolone use and aortic aneurysm (propensity-
matched OR).

adjusted for a wide variety of confounders using propensity
score methods, residual confounding is always possible. The
estimates for the number needed to treat to harm are only ap-
plicable to an elderly population aged more than 65 years.
Our study has several strengths. We searched multiple da-
tabases according to a prespecified protocol, analyzed various
types of exposure, conducted both random- and fixed-
effects analyses, and rated the quality of studies and the
strength of evidence and provided both relative and abso-
lute risk estimates. Two reviewers were independently involved
in all stages of data-extraction study interpretation and analysis.
Although definitive proof of a causal association between
fluoroquinolone use and aortic dissection and aortic aneurysm
will require further replication in other observational study
designs and populations, several findings are noteworthy.23
We noted the large strength of the association, consistency
in increased risk in different study designs across diverse
settings, and evidence of temporality in prospective cohort
studies. The biological plausibility of the association, the
coherence between mechanistic and epidemiological find-
ings, and the analogous findings of associated tendon rupture
lend support to our findings.
Implications for Clinical Policy and Practice
Our findings have clinical and policy implications. The number
needed to treat to harm for those with aortic aneurysm should
be considered along with the beneficial and harmful effects
on other outcomes. The number needed to treat for tendon
rupture in the cohort study was estimated to be 178.18
The
number needed to treat for ciprofloxacin relative to
trimethoprim-sulfamethoxazole for treating acute uncompli-
cated pyelonephritis is estimated to be 10 from randomized
controlled trials.24
The benefit/harm ratio may be less of a
concern for patients with a low underlying risk of aortic dis-
section or aortic aneurysm. However, the risk of repetitive
fluoroquinolone exposure in older patients with multiple risks
needs more careful consideration. Future studies using in-
strumental variables or self-controlled case designs, which
control for time-invariant confounders, may provide addi-
tional information on the increased risk, because adequately
powered randomized controlled trials are impractical for such
rare outcomes.
CONCLUSIONS
Our findings suggest an increased relative risk of aortic dis-
section or aortic aneurysm associated with the use of
fluoroquinolones, although the absolute risk is modest and
evidence is of moderate quality. Clinicians should carefully
consider the risks of aortic dissection or aortic aneurysm as-
sociated with the use of fluoroquinolones, along with their
adverse effects on other outcomes and balance this against
their established benefits.
References
1. Stephenson AL, Wu W, Cortes D, Rochon PA. Tendon injury and
fluoroquinolone use: a systematic review. Drug Saf. 2013;36(9):
709-721. doi:10.1007/s40264-013-0089-8. [Epub 2013/07/28]; PMID
23888427.
2. Raguideau F, Lemaitre M, Dray-Spira R, Zureik M. Association between
oral fluoroquinolone use and retinal detachment. JAMA Ophthalmol.
2016;134(4):415-421. doi:10.1001/jamaophthalmol.2015.6205. [Epub
2016/03/12]; PMID 26967005.
3. Clouse WD, Hallett JW Jr, Schaff HV, et al. Acute aortic dissection:
population-based incidence compared with degenerative aortic aneu-
rysm rupture. Mayo Clin Proc. 2004;79(2):176-180. [Epub 2004/02/
13]; PMID 14959911.
4. Olsson C, Thelin S, Stahle E, Ekbom A, Granath F. Thoracic aortic an-
eurysm and dissection: increasing prevalence and improved outcomes
reported in a nationwide population-based study of more than 14,000
cases from 1987 to 2002. Circulation. 2006;114(24):2611-2618.
doi:10.1161/circulationaha.106.630400. [Epub 2006/12/06]; PMID
17145990.
5. Kurz SD, Falk V, Kempfert J, et al. Insight into the incidence of acute
aortic dissection in the German region of Berlin and Brandenburg. Int
J Cardiol. 2017;doi:10.1016/j.ijcard.2017.05.024. [Epub 2017/05/14];
PMID 28499667.
6. Campbell WB, Collin J, Morris PJ. The mortality of abdominal aortic
aneurysm. Ann R Coll Surg Engl. 1986;68(5):275-278. PMID 3789625.
PMCID: PMC2498338.
7. Groth KA, Stochholm K, Hove H, et al. Aortic events in a nationwide
Marfan syndrome cohort. Clin Res Cardiol. 2017;106(2):105-112.
doi:10.1007/s00392-016-1028-3. [Epub 2016/08/24]; PMID 27550511.
8. Singh S, Nautiyal A. Aortic dissection and aortic aneurysms associated
with fluoroquinolones: a systematic review and meta-analysis of ob-
servational studies; 2017. Available at: https://www.crd.york.ac.uk/
PROSPERO/display_record.asp?ID=CRD42017057426. Accessed June
4, 2017.
9. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items
for systematic reviews and meta-analyses: the PRISMA statement. PLoS
Med. 2009;6(7):e1000097. doi:10.1371/journal.pmed.1000097. [Epub
2009/07/22]; PMID 19621072. PMCID: PMC2707599.
10. Wells G, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS)
for assessing the quality of nonrandomised studies in meta-analyses
Ottawa. Available at: http://www.ohri.ca/programs/clinical_epidemiology/
oxford.asp. Accessed June 4, 2017.
11. Deeks JJHJ, Altman DG. Cochrane Handbook for Systematic Reviews
of Interventions. Chichester, UK: John Wiley & Sons, Ltd; 2008 Avail-
able at: http://onlinelibrary.wiley.com/doi/10.1002/9780470712184.ch9/
summary. Accessed March 2, 2017.
12. Davies HT, Crombie IK, Tavakoli M. When can odds ratios mislead?
BMJ. 1998;316(7136):989-991. [Epub 1998/04/29]; PMID 9550961.
PMCID: PMC1112884.
13. StatsDirect Ltd. StatsDirect statistical software. UK: StatsDirect Ltd;
2013 Available at: http://www.statsdirect.com. Accessed August 9, 2017.
14. Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful
measures of the consequences of treatment. N Engl J Med.
1988;318(26):1728-1733. doi:10.1056/nejm198806303182605. [Epub
1988/06/30]; PMID 3374545.
15. Cates C. Dr Chris Cates’ EBM Website Visual Rx; 2017. Available at:
http://www.nntonline.net/visualrx/. Accessed March 2, 2017.
16. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consen-
sus on rating quality of evidence and strength of recommendations. BMJ.
2008;336(7650):924-926. doi:10.1136/bmj.39489.470347.AD. [Epub
2008/04/26]; PMID 18436948. PMCID: PMC2335261.
17. Lee CC, Lee MT, Chen YS, et al. Risk of aortic dissection and aortic
aneurysm in patients taking oral fluoroquinolone. JAMA Intern Med.
2015;175(11):1839-1847. doi:10.1001/jamainternmed.2015.5389. [Epub
2015/10/06]; PMID 26436523.
18. Daneman N, Lu H, Redelmeier DA. Fluoroquinolones and collagen as-
sociated severe adverse events: a longitudinal cohort study. BMJ Open.
2015;5(11):e010077. doi:10.1136/bmjopen-2015-010077. [Epub 2015/
11/20]; PMID 26582407. PMCID: PMC4654346.
19. Amin M, Pushpakumar S, Muradashvili N, Kundu S, Tyagi SC, Sen
U. Regulation and involvement of matrix metalloproteinases in

vascular diseases. Front Biosci. 2016;21:89-118. [Epub 2015/12/29];
PMID 26709763.
20. Longo GM, Xiong W, Greiner TC, Zhao Y, Fiotti N, Baxter BT.
Matrix metalloproteinases 2 and 9 work in concert to produce
aortic aneurysms. J Clin Invest. 2002;110(5):625-632. doi:10
.1172/jci15334. [Epub 2002/09/05]; PMID 12208863. PMCID:
PMC151106.
21. Liapis CD, Paraskevas KI. The pivotal role of matrix metalloproteinases
in the development of human abdominal aortic aneurysms. Vasc Med.
2003;8(4):267-271. doi:10.1191/1358863x03vm504ra. [Epub 2004/05/
06]; PMID 15125488.
22. Wang DJA, Tajik JK. Ciproﬂoxacin enhances matrix metalloproteinases
activity in human aortic smooth muscle cells. J Patient Cent Res Rev.
2014;1:58-59.
23. Hill AB. The environment and disease: association or causation? Proc
R Soc Med. 1965;58(5):295-300. PMID PMC1898525.
24. Talan DA, Stamm WE, Hooton TM, et al. Comparison of ciproﬂoxacin
(7 days) and trimethoprim-sulfamethoxazole (14 days) for acute un-
complicated pyelonephritis in women: a randomized trial. JAMA.
2000;283(12):1583-1590. [Epub 2000/03/29]; PMID 10735395.
SUPPLEMENTARY DATA
Supplementary Material accompanying this article can be
found in the online version at doi:10.1016/j.amjmed
.2017.06.029.

APPENDIX 1
1457.e1 The American Journal of Medicine, Vol 130, No 12, December 2017

1457.e2Singh and Nautiyal Fluoroquinolones and Aortic Dissection

APPENDIX 2

APPENDIX 3
Aortic dissection and aortic aneurysms associated with fluoroquinolones: a systematic review
and meta-analysis of observational studies.
Search Strategy
Pubmed
(((("aorta"[MeSH Terms] OR "aorta"[All Fields] OR "aortic"[All Fields]) AND ("dissection"[MeSH Terms]
OR "dissection"[All Fields])) OR ("aortic aneurysm"[MeSH Terms] OR ("aortic"[All Fields] AND
"aneurysm"[All Fields]) OR "aortic aneurysm"[All Fields])) OR (("aorta"[MeSH Terms] OR "aorta"[All
Fields] OR "aortic"[All Fields]) AND ("dilatation"[MeSH Terms] OR "dilatation"[All Fields])))
AND
(((((((((((("fluoroquinolones"[MeSH Terms] OR "fluoroquinolones"[All Fields]) OR ("quinolones"[MeSH
Terms] OR "quinolones"[All Fields])) OR ("pefloxacin"[MeSH Terms] OR "pefloxacin"[All Fields])) OR
("pefloxacin"[MeSH Terms] OR "pefloxacin"[All Fields])) OR ("enoxacin"[MeSH Terms] OR "enoxacin"[All
Fields])) OR ("ofloxacin"[MeSH Terms] OR "ofloxacin"[All Fields] OR "levofloxacin"[MeSH Terms] OR
"levofloxacin"[All Fields])) OR ("norfloxacin"[MeSH Terms] OR "norfloxacin"[All Fields])) OR
("moxifloxacin"[Supplementary Concept] OR "moxifloxacin"[All Fields])) OR ("levofloxacin"[MeSH Terms]
OR "levofloxacin"[All Fields])) OR ("gatifloxacin"[Supplementary Concept] OR "gatifloxacin"[All Fields]))
OR ("ciprofloxacin"[MeSH Terms] OR "ciprofloxacin"[All Fields])) OR ("7-(3-aminohexahydro-1H-azepin-
1-yl)-8-chloro-1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-3-quinolinecarboxylic acid"[Supplementary
Concept] OR "7-(3-aminohexahydro-1H-azepin-1-yl)-8-chloro-1-cyclopropyl-6-fluoro-1,4-dihydro-4-oxo-
3-quinolinecarboxylic acid"[All Fields] OR "besifloxacin"[All Fields]))
SCOPUS
( TITLE-ABS-KEY ( aortic AND dilation ) OR TITLE-ABS-KEY ( aortic AND aneurysm ) OR TITLE-ABS-KEY (
aortic AND rupture ) AND TITLE-ABS-KEY ( besifloxacin ) OR TITLE-ABS-KEY ( ciprofloxacin ) OR TITLE-
ABS-KEY ( gatifloxacin ) OR TITLE-ABS-KEY ( levofloxacin ) OR TITLE-ABS-KEY ( moxifloxacin ) OR TITLE-
ABS-KEY ( norfloxacin ) OR TITLE-ABS-KEY ( ofloxacin ) OR TITLE-ABS-KEY ( enoxacin ) OR TITLE-ABS-
KEY ( pefloxacin ) OR TITLE-ABS-KEY ( quinolones ) OR TITLE-ABS-KEY ( fluoroquinolones ) )

EMBASE
besifloxacin OR 'ciprofloxacin'/exp OR
ciprofloxacin OR 'gatifloxacin'/exp OR
gatifloxacin OR 'levofloxacin'/exp OR
levofloxacin OR 'moxifloxacin'/exp OR
moxifloxacin OR 'norfloxacin'/exp OR norfloxacin
OR 'ofloxacin'/exp OR ofloxacin OR 'enoxacin'/exp
OR enoxacin OR 'pefloxacin'/exp OR pefloxacin OR
'quinolone derivative'/exp OR 'quinolone
derivative' OR 'fluoroquinolones'/exp OR
fluoroquinolones AND ('aorta aneurysm' OR 'aorta
dissection'/exp OR 'aorta dissection' OR 'aortic
dilatation'/exp OR 'aortic dilatation')
APPENDIX 4
Fixed effects meta-analysis of ﬂuoroquinolone use and aortic dissection (Unadjusted Odds Ratio).

APPENDIX 5
Fixed effects meta-analysis of ﬂuoroquinolone use and aortic aneurysm (Unadjusted Odds Ratio).
APPENDIX 6
Odds Ratio in Meta-Analysis of Aortic Aneurysm and Aortic Dissection with Current Fluoroquinolone Use
Unadjusted OR Estimate (95% CI); I2
% Propensity Score–Adjusted OR (95% CI); I2
%
Aortic aneurysm
Fixed-effects meta-analysis 2.73 (95% CI, 2.54-2.93); I2
= 0% 2.25 (95% CI, 2.03-2.49); I2
= 0%
Random-effects meta-analysis 2.73 (95% CI, 2.54-2.93); I2
= 0% 2.25 (95% CI, 2.03-2.49); I2
= 0%
Aortic dissection
Fixed-effects meta-analysis 3.13 (95% CI, 2.59-3.79); I2
= 0% 2.79 (95% CI, 2.31-3.37); I2
= 0%
Random-effects meta-analysis 3.13 (95% CI, 2.58-3.79); I2
= 0% 2.79 (95% CI, 2.31-3.37); I2
= 0%
CI = confidence interval; OR = odds ratio.

1 s2.0-s0002934317307180

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to 1 s2.0-s0002934317307180

Similar to 1 s2.0-s0002934317307180 (20)

Recently uploaded

Recently uploaded (20)

1 s2.0-s0002934317307180