Fluoroquinolones and Risk of Aortic Aneurysm or Dissection in Patients With Congenital Aortic Disease and Marfan Syndrome

Abstract

Background: Fluoroquinolone use can be associated with an increased risk of aortic aneurysm (AA) or aortic dissection (AD). The US Food and Drug Administration recently warned against fluoroquinolone use for high-risk patients, such as those with Marfan syndrome. However, the association between fluoroquinolone use and AA/AD risk was unknown in these high-risk patients and therefore it was studied in this work.

Methods and Results: Data were collected from a national database between 2000 and 2017 for 550 patients with AA/AD and any congenital aortic disease (mean age 41.5 years; 415 with Marfan syndrome). A case cross-over study was conducted to compare the risk of aortic events (AA/AD) associated with fluoroquinolone and amoxicillin use between the hazard period (from −60 days to −1 day) and a randomly selected reference period (−180 to −121 days; −240 to −181 days; and –300 to –241 days). Compared to the reference period without fluoroquinolone use, fluoroquinolone use during the hazard period was not associated with a greater risk of AA/AD (1.09% vs. 1.09%; odds ratio [OR] 1.000; 95% confidence interval [CI] 0.32–3.10), AA (OR 0.67; 95% CI 0.11–3.99), or AD (OR 1.33; 95% CI 0.30–5.96) in patients with congenital aortic disease or Marfan syndrome. This lack of association was maintained in subgroup analysis, including Marfan syndrome or not, age (≤50 vs. >50 years) and sex.

Conclusions: Fluoroquinolone use was not associated with an increased risk of AA/AD in patients with congenital aortic disease, including Marfan syndrome. More evidence is required for a fluoroquinolone pharmacovigilance plan in these patients.

Aortic aneurysm (AA) and aortic dissection (AD) are acute life-threatening conditions; therefore, identifying risk factors for AA and AD is crucial, particularly in high-risk populations.¹ Marfan syndrome is an autosomal-dominant connective tissue disorder that can lead to aortic enlargement, AA, and AD.^2–4 A recent national database study from Taiwan revealed that approximately 10% of patients with Marfan syndrome developed AD during the follow-up 12 years, and half the patients who developed AD experienced sudden death before the age of 40 years.⁵

Fluoroquinolones are among the most commonly used antibiotics in the US, accounting for 5% of total antibiotic use.⁶ Fluoroquinolones are widely used because they have few major complications and are well tolerated by patients; however, they may damage connective tissue collagen and jeopardize the integrity of the extracellular matrix in the vascular wall. Studies have demonstrated that fluoroquinolones significantly increase the risk of AA/AD.^7–10 In Taiwan, fluoroquinolone use has increased from 3.6% in 2000 to 11.5% in 2013.¹¹ We recently demonstrated that fluoroquinolone exposure in patients with AA/AD was associated with a higher risk of later all-cause mortality, aortic-related death, and later aortic surgery.¹¹ LeMaire et al also demonstrated that ciprofloxacin, a type of fluoroquinolone, accelerates aortic root enlargement and increases the incidence of AD and aortic rupture in Marfan mice.¹²

The US Food and Drug Administration (FDA) recently warned against fluoroquinolone use for high-risk patients, such as those with Marfan syndrome or a known aortic disease.¹³ However, this warning is based on inferences from studies designed for the general population instead of specific high-risk groups. No human-level evidence is available as to whether fluoroquinolones increase the risk of AA/AD in patients with Marfan syndrome or other congenital aortopathies. Accordingly, we conducted a population-based cohort study to investigate whether fluoroquinolone use increases the risk of AA/AD in patients with Marfan syndrome and other congenital aortopathies.^14,15

Methods

Data Source

We used data from the Taiwan National Health Insurance Research Database (NHIRD), maintained by the Health and Welfare Data Science Center. In Taiwan, the National Health Insurance (NHI) program is mandatory and covers >99% of the Taiwanese population (approximately 24 million people by the end of 2017). Moreover, the Taiwanese government is the single payer of insurance claims; therefore, the completeness of follow-up is very high. The Taiwan NHI launched in March 1995, and the data are updated annually with a time lag of approximately 2–3 years. Data across different subdatabases of the NHIRD are linked using deidentified civil identification numbers of insured individuals. Before December 31, 2015, diseases were diagnosed on the basis of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnostic codes; since January 1, 2016, diseases have been diagnosed on the basis of both ICD-9-CM and ICD-10-CM codes. Details of the NHI program and NHIRD have been reported previously.^16–18

The present study was approved by the Institutional Review Board of Chang Gung Memorial Hospital (Reference no. 202002205B0).

Study Cohort

From the NHIRD, we selected the data of patients who were admitted to an emergency department (ED) or hospitalized with a primary discharge diagnosis of AA/AD between January 1, 2000, and December 31, 2017; we included patients who had any congenital aortic disease, including bicuspid aortic valve, Turner’s syndrome, Ehlers-Danlos syndrome, and Marfan syndrome. The timing of diagnoses for congenital aortic disease could be before, during, or after the index date. If a patient had 2 aortic events during the study period, the first event of aortic disease was selected as the index admission. The day of AA/AD admission (ED or hospitalization) was assigned as the index date. We excluded patients with missing demographic data, a diagnosis of AA/AD before 2000, and no diagnostic code for any congenital aortic disease.

Study Design

On the basis of a previous study,¹⁰ we used a unidirectional case cross-over design without selecting controls from an external population. A case cross-over design has several advantages, such as avoiding within-person time-invariant confounding (i.e., genes) and possible selection biases of the external controls.¹⁹ Each patient contributed 1 hazard period (from −60 days to −1 day), 1 washout period (−120 to −61 days), and 3 reference periods (−180 to −121 days; −240 to −181 days; and −300 to −241 days). The risk of outcomes was compared for the same individual between the hazard period and one of the randomly selected reference periods (Figure 1B). The hazard period was set as 60 days preceding the admission date (ED or hospitalization) for AA/AD event based on previous reports investigating the association between fluoroquinolone use and AA/AD risk; however, these previous studies examined the short-term effect of fluoroquinolone use, rather than long-term effects.^7,9,10 Because prescription refills for medications for chronic illnesses last for a maximum length of 2–3 months according to the Taiwan NHI reimbursement policy, we set a washout period to clearly discriminate prescriptions between the hazard and reference periods. To avoid unexpected systematic biases, we randomly selected 1 period from the 3 reference periods.¹⁰ The drug of interest was fluoroquinolone. Amoxicillin was selected as the negative control because of its lack of effect on aortic events.^11,20,21 Several definitions of “drug exposure” were analyzed: (1) a prescription of ≥3 days in the outpatient department; (2) a prescription of ≥3 days in the outpatient department or during hospitalization; and (3) a prescription of ≥3 days or any injectable form in the outpatient department or during hospitalization.

Figure 1.

(A) Flowchart of patient selection. AA, aortic aneurysm; AD, aortic dissection. (B) The hazard, washout, and reference periods used in the present study. (C) Association between fluoroquinolone use and the risk of AA or AD in patients with AA or AD. CI, confidence interval; OPD, outpatient department; OR, odds ratio.

Covariates

The following covariates were included: patient demographics (age, sex, residence urbanization level, and monthly income), medical utilization status (number of outpatient visits, emergency room visits, and hospitalizations in the previous year), and comorbid conditions (hypertension, diabetes, dyslipidemia, heart failure, coronary arterial disease, myocardial infarction, peripheral arterial disease, atrial fibrillation, stroke, chronic kidney disease or dialysis, chronic obstructive pulmonary disease, asthma, and heart valve disease). The occurrence of heart failure, myocardial infarction, and stroke was verified using inpatient diagnoses prior to the index date, which can be tracked back to the inception of the Taiwan NHI program (March 1995). The other chronic comorbidities were identified if they were present in outpatient diagnoses at least twice or if they appeared in any inpatient diagnoses in the previous year. The presence of aortic surgery (i.e., open repair or stent surgery) during the index admission was captured using the Taiwan NHI reimbursement codes from the claims data. The ICD diagnostic codes are provided in the supplements (Supplementary Table 1).

Statistical Analysis

A conditional logistic regression model was used to compare the risk of AA/AD events between the hazard period and a randomly selected reference period for individual patients. The use of fluoroquinolone and amoxicillin was analyzed separately in 2 models. The analysis was conducted first on the whole AA/AD cohort, followed by the subset with congenital aortic disease, and finally on he subset with Marfan syndrome. The analysis was further stratified by aortic pathology (AA vs. AD). A subgroup analysis was also conducted in the AA/AD cohort with any congenital aortic disease, stratified by the presence or absence of Marfan syndrome, age group (≤50 vs. >50 years), and sex. Drug exposure was based on oral or injectable administration in the outpatient department or during hospital admission. Statistical significance was set at 2-sided P<0.05. All statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Results

Patient Characteristics

Initially, 40,090 patients admitted to the ED or hospitalized due to AA/AD were identified. The baseline characteristics of these patients are summarized in Supplementary Table 2. Of these patients, 550 had congenital aortic disease (415 [75.5%] with Marfan syndrome) and were considered for analysis (Figure 1A). The mean (±SD) age of these patients was 41.5±14.2 years, and only 28.2% of patients were older than 50 years (Table 1). The male-to-female ratio was approximately 2 : 1 (66% vs. 34%). Hypertension was the most prevalent comorbidity (33.3%), followed by coronary artery disease (14.4%), and peripheral artery disease (11.1%). The prevalence of the remaining comorbidities was <10%. Among the 550 patients, 63.8% had AD and 36.2% had AA. More than half (n=321; 58.4%) of the patients underwent aortic surgery at the index admission.

Table 1. Baseline Characteristics of Patients With Congenital Aortic Disease and Marfan Syndrome

	Congenital aortic disease (n=550)	Marfan syndrome (n=415)
Age (years)	41.5±14.2	37.0±11.6
Age >50 years	155 (28.2)	61 (14.7)
Male sex	363 (66.0)	265 (63.9)
Residence urbanization level
Low: Remote area	41 (7.5)	27 (6.5)
Moderate: Township	137 (24.9)	106 (25.5)
High: City	223 (40.6)	156 (37.6)
Very high: Metropolis	149 (27.1)	126 (30.4)
Monthly income
Tertile 1	191 (34.7)	153 (36.9)
Tertile 2	193 (35.1)	148 (35.7)
Tertile 3	166 (30.2)	114 (27.5)
Medical utilization in the previous year
No. outpatient visits	10.5 [5.0–18.0]	10.0 [5.0–17.0]
No. patients visiting emergency departments	239 (43.5)	179 (43.1)
No. patients hospitalization	103 (18.7)	68 (16.4)
Comorbid conditions
Hypertension	183 (33.3)	107 (25.8)
Diabetes	23 (4.2)	13 (3.1)
Dyslipidemia	41 (7.5)	19 (4.6)
Heart failure	26 (4.7)	12 (2.9)
Coronary artery disease	79 (14.4)	46 (11.1)
Myocardial infarction	4 (0.7)	4 (1.0)
Peripheral artery disease	61 (11.1)	34 (8.2)
Atrial fibrillation	15 (2.7)	8 (1.9)
Stroke	16 (2.9)	9 (2.2)
Chronic kidney disease or dialysis	24 (4.4)	7 (1.7)
Chronic obstructive pulmonary disease	13 (2.4)	7 (1.7)
Asthma	18 (3.3)	12 (2.9)
Heart valve disease	20 (3.6)	15 (3.6)
Aortic pathology
Aortic aneurysm	199 (36.2)	132 (31.8)
Aortic dissection	351 (63.8)	283 (68.2)
Aortic surgery during the index admission	321 (58.4)	241 (58.1)

Data are presented as frequency (percentage), the mean±SD, or the median [interquartile range].

Compared with all patients with congenital aortic disease, those with Marfan syndrome were younger (37 vs. 41.5 years) and had a lower prevalence of comorbidities (Table 1). The mean (±SD) age of patients with Marfan syndrome was 37.0±11.6 years, and only 14.7% of the patients were older than 50 years.

Association Between the Fluoroquinolone/Amoxicillin Use and the Risk of AA/AD

In the entire AA/AD population (n=34,995), the risk of AA/AD was significantly greater in the hazard period than in the reference period for the use of oral drugs in outpatient departments (odds ratio [OR] 1.20; 95% confidence interval [CI] 1.06–1.35; Figure 1C). The results were similar for the various definitions of drug exposure: oral drugs in the outpatient department or during hospitalization (OR 1.28; 95% CI 1.14–1.43) and oral drugs or injection in the outpatient department or during hospitalization (OR 1.41; 95% CI 1.28–1.57).

After restricting the AA/AD cohort to those with congenital aortic disease, we observed that the fluoroquinolone exposure rates were identical between the hazard and reference periods (1.09% vs. 1.09%) for the use of oral drugs in the outpatient department (Figure 2). The risk of AA/AD in the hazard period was comparable to that in the reference period (OR 1.000; 95% CI 0.32–3.10). The amoxicillin exposure rates in the hazard and reference periods were 3.45% and 4.36%, respectively, and amoxicillin use in the hazard period was not significantly associated with a greater risk of AA/AD (OR 0.79; 95% CI 0.43–1.45). The results were consistent for the other definitions of drug exposure. The results obtained for the Marfan syndrome subset were consistent with those obtained for patients with any congenital aortic disease (Figure 3).

Figure 2.

Association between fluoroquinolone use and the risk of aortic aneurysm or dissection in patients with congenital aortic disease. CI, confidence interval; OPD, outpatient department; OR, odds ratio.

Figure 3.

Association between fluoroquinolone use and the risk of aortic aneurysm or dissection in patients with Marfan syndrome. CI, confidence interval; OPD, outpatient department; OR, odds ratio.

Subgroup Analysis

The use of oral fluoroquinolones in outpatient departments during the hazard period was not associated with a greater risk of AA (OR 0.67; 95% CI 0.11–3.99) or AD (OR 1.33; 95% CI 0.30–5.96), and similar results were noted for oral amoxicillin use (AA: OR 0.57 [95% CI 0.17–1.95]; AD: OR 0.88 [95% CI 0.44–1.77]). The results were consistent for the various definitions of drug exposure. The results obtained for the cohort with Marfan syndrome were consistent with those obtained for patients with congenital aortic disease (Table 2).

Table 2. Association Between Fluoroquinolone/Amoxicillin Use and the Risk of Different Aortic Pathologies in Patients With Congenital Aortic Disease and Marfan Syndrome

Cohort/drug/exposure	Aortic aneurysm		Aortic dissection
	OR (95% CI)	P value	OR (95% CI)	P value
Congenital aortic disease
Fluoroquinolone
Oral drug in OPD	0.67 (0.11–3.99)	0.657	1.33 (0.30–5.96)	0.706
Oral drug	0.50 (0.09–2.73)	0.424	0.80 (0.22–2.98)	0.739
Oral and injected drug	1.25 (0.34–4.66)	0.739	0.83 (0.25–2.73)	0.763
Amoxicillin
Oral drug in OPD	0.57 (0.17–1.95)	0.372	0.88 (0.44–1.77)	0.724
Oral drug	0.57 (0.17–1.95)	0.372	0.88 (0.44–1.77)	0.724
Oral and injected drug	0.57 (0.17–1.95)	0.372	0.88 (0.44–1.77)	0.724
Marfan syndrome
Fluoroquinolone
Oral drug in OPD	0.67 (0.11–3.99)	0.657	1.000 (0.14–7.10)	1.000
Oral drug	0.67 (0.11–3.99)	0.657	0.50 (0.09–2.73)	0.424
Oral and injected drug	0.67 (0.11–3.99)	0.657	0.75 (0.17–3.35)	0.706
Amoxicillin
Oral drug in OPD	0.50 (0.05–5.51)	0.571	1.000 (0.46–2.16)	1.000
Oral drug	0.33 (0.04–3.21)	0.341	1.000 (0.46–2.16)	1.000
Oral and injected drug	0.33 (0.04–3.21)	0.341	1.000 (0.46–2.16)	1.000

CI, confidence interval; OPD, outpatient department; OR, odds ratio.

Furthermore, stratification by Marfan syndrome, age, and sex revealed that neither fluoroquinolone nor amoxicillin use in the hazard period was associated with an increased risk of AA/AD in any of the subgroups (Figure 4).

Figure 4.

Subgroup analysis (Marfan syndrome, age ≤50 or >50 years, and sex) of the association between the risk of aortic aneurysm or aortic dissection and (A) fluoroquinolone use and (B) amoxicillin use in patients with congenital aortic disease. CI, confidence interval; OR, odds ratio.

Discussion

Summary of Study and Findings

In this nationwide cohort study, we evaluated the effect of fluoroquinolones on the risk of aortic events (AA or AD) in patients with AA/AD with or without congenital aortic disease, including Marfan syndrome. According to our review of the relevant literature, this is the first study to evaluate this association in patients with congenital aortic disease. The results reveal that fluoroquinolone use (oral or injectable) increased the risk of aortic events in all patients with AA/AD (n=34,995) but not in the subgroup of patients with any congenital aortic disease (n=550) or specifically those with Marfan syndrome (n=415), even after stratification for age (≤50 vs. >50 years), sex, and the presence of Marfan syndrome.

Fluoroquinolone Use and Adverse Effects in Patients With AA/AD

As fluoroquinolone use increases, so does the reporting of uncommon adverse events, including torsade de pointes, hepatotoxic effects, and connective tissue disorders.^22,23 Frankel et al reported that approximately 20% of inpatients with AA/AD and 14% of inpatients with Marfan syndrome received fluoroquinolone therapy.²⁴

Studies conducted in Taiwan and Sweden using nationwide databases have reported that fluoroquinolone exposure in patients with AA/AD was associated with an increased risk of aortic events within 60 days following such exposure.^7,9 In 2018, Lee et al demonstrated that the risk of fluoroquinolones on AA/AD was modified by the duration of fluoroquinolone use and the length of the hazard period.¹⁰ Furthermore, a US study determined that fluoroquinolone use was associated with an increased incidence of AA formation in adults, even after adjustment for age, sex, and comorbidities.²⁵ Our previous study based on the NHIRD demonstrated a significantly increased risk of all-cause mortality, aortic-related death, and later aortic surgery in patients with AA/AD taking fluoroquinolones.¹¹ Fluoroquinolone use is associated with a relatively high risk of connective tissue disorder in the cardiovascular system, including the aorta and heart valves. A study using the FDA’s adverse reporting system database revealed that current or previous use of oral fluoroquinolones increased the risk of aortic and mitral regurgitation.²⁶ Two recent meta-analyses have reported a moderate level of evidence that the risk of AA/AD was significantly higher in all individuals using fluoroquinolones (OR 1.56–2.26) than in those using other antibiotics and/or those not using antibiotics;^27,28 this suggests that a cautious approach must be adopted towards fluoroquinolone use in not only high-risk individuals, but also the general population.

A few scholars have assessed previous research by using a robust design and performing a meticulous analysis. They have raised concerns regarding the findings of the assessed research, highlighting potential problems with bias and confounding factors that engender positive associations; accordingly, they have stressed the importance of including adequate comparator groups in pharmacoepidemiologic designs.^29,30 Gopalakrishnan et al²⁹ reported that fluoroquinolone use was associated with an increase in the relative risk (but not absolute risk) of AA/AD compared with the use of azithromycin or amoxicillin (but not trimethoprim/sulfamethoxazole) in patients with pneumonia; however, similar findings were not observed in those with urinary tract infection. Furthermore, Gopalakrishnan et al²⁹ reported that the association between fluoroquinolone use and amoxicillin use was no longer significant when baseline aortic imaging was considered. In another study, Dong et al³⁰ reported that fluoroquinolone use was not associated with an increased risk of aortic adverse events in inpatients with severe infection compared with the use of amoxicillin–clavulanate or ampicillin–sulbactam or with the use of extended-spectrum cephalosporin in patients with relevant indicated infections are UTI or pneumonia. However, the inclusion of inpatients with severe infection may bias the association between infection and aortic adverse events; this is because extensive imaging studies are typically conducted during index admission for diagnosis. In a systematic review and meta-analysis, Ribeiro et al³¹ attempted to control for infectious state, a confounding factor, by conducting a comparative analysis between the risk of AA/AD with fluoroquinolone use and the risk of AA/AD with β-lactam antibiotic use. In that study, the authors still found a diminished but still relatively high risk of AA/AD (OR 2.26 for no antibiotic use vs. 1.56 for β-lactam use).³¹ A 2021 meta-analysis stated that the risk of AA/AD with fluoroquinolone use was similar to that with other broad-spectrum antibiotics.³² The authors suggested that the increase in risk may thus be related to disease severity, and not the antibiotics themselves.³²

Use and Safety of Fluoroquinolones in Congenital Aortic Disease and Marfan Syndrome

Despite the FDA warning, no studies have focused on fluoroquinolone use in patients with congenital aortic disease, such as Marfan syndrome; this is possibly due to the lack of comprehensive datasets and insufficient sample sizes for rare diseases. Studies have also tended to exclude patients with cardiovascular comorbidities, including Marfan syndrome.¹⁰ LeMaire et al¹² conducted a comprehensive study involving a mouse model of Marfan syndrome. They observed that the administration of ciprofloxacin, a commonly used fluoroquinolone, led to a significantly increased predilection for aortic dissection and rupture in male mice through increases in aortic metalloproteinase levels, cell injury, and elastic fiber fragmentation; the resulting collagen breakdown directly affected aortic integrity.¹² However, our data indicated that fluoroquinolone use did not increase the risk of adverse aortic events in patients with congenital aortic disease, including Marfan syndrome. Our findings suggest that suspecting adverse aortic events in these patients is unnecessary; nevertheless, additional larger-scale investigations must be conducted to confirm our findings.

Whether age affects the effect of fluoroquinolones on the risk of aortic diseases requires further investigation. Fluoroquinolones have a high risk of adverse effects in older adults. In 2018, an FDA warning stated that the high risk of AA/AD in patients taking fluoroquinolones was mediated by factors such as known cardiovascular disease, high blood pressure, genetic vascular diseases, and old age. All previous studies^7–11 have used populations aged ≥65 years. Newton et al concluded that fluoroquinolone use increased the risk of abdominal AA only in adults aged ≥35 years and not in those aged 18–34 years.²⁵ Moreover, a study from Taiwan indicated no significant differences in collagen-associated adverse effects (gastrointestinal perforation) between fluoroquinolones users and non-users in a pediatric population.³³ The authors of that study also noted that most previous studies demonstrating a higher risk of collagen-associated adverse effects included adult patients (aged >18 years) and proposed that fluoroquinolones may not be as harmful in the pediatric population. However, in their cohort, no case of aortic event was reported.³³ The present study revealed that fluoroquinolone use did not significantly increase the risk of AA/AD in a very young population with Marfan syndrome; the mean age of the patients in this population was 37 years, and only 14% of patients were older than 50 years. A recent national database study from Taiwan demonstrated that approximately 10% of patients with Marfan syndrome developed AD during the follow-up 12 years, and half the patients who developed AD experienced sudden death before the age of 40.⁵ More robust studies are warranted to determine whether age influences the relationship between fluoroquinolone use and AA/AD overall, as well as in specific patient groups.

Study Limitations

Our study has several limitations. First, we used a retrospective observational design, which has inherent biases, including a lack of determination of causality. Moreover, although adjustments were made for known confounders, namely age, sex, and Marfan syndrome, other factors may have affected the results, such as computed tomography, blood pressure, body mass index, tobacco use, congenital collagen disorders, ongoing infection, and family history of aortic disease.³¹ Future studies should attempt to control for as many potential confounders as possible. Second, ICD-9-CM codes were used to identify all diagnoses, which may be subject to misdiagnosis and coding errors. However, the accuracy of the diagnostic codes for AA/AD and other major cardiovascular diseases in the NHIRD has been validated.^28,34 Third, the NHIRD does not include data regarding adherence to prescriptions, and we used records of prescription refills as a proxy for regular drug intake; accordingly, exposure to fluoroquinolone may have been overestimated. However, in Taiwan, patients with chronic diseases can refill prescriptions issued by physicians every month for 3 months;³⁵ therefore, any exposure misclassification would be rare and random, having a limited effect, if any, on our results. Fourth, we did not evaluate whether the pharmacological interaction of fluoroquinolones with other drugs, including steroids, affected the risk of aortic events. Fifth, in our cohort, most patients with congenital aortic disease had Marfan syndrome (75.5%); this prevented a robust assessment for patients with other types of aortopathy. Finally, our cohort comprised only Taiwanese patients, thus limiting the generalizability of our results to other ethnicities or health-care systems. Despite these limitations, we believe that our results based on a nationwide database provide a valuable contribution to the knowledge gap in this field, especially considering the rarity of cases. However, future studies, including animal studies, should explore the underlying mechanisms to help better characterize and even predict the effects of fluoroquinolones on variations in congenital aortic and other diseases, and thus support the development of future prevention or treatment strategies.

Conclusions

In patients with congenital aortic disease, including Marfan syndrome, the use of fluoroquinolones (oral or injectable) did not significantly increase the risk of AA/AD, even when stratified by age or sex. Therefore, more evidence is required for the FDA to implement its pharmacovigilance plan on fluoroquinolones in these patients, and the assessment of risk should be individualized. Further mechanistic and experimental research is warranted to verify and even elucidate the mechanisms underlying the association between fluoroquinolone use and congenital aortopathy in order to make definitive decisions regarding patient safety.

Acknowledgments

This study was based on data from the NHIRD provided by the NHI Administration, Ministry of Health and Welfare of Taiwan, and managed by the National Health Research Institutes of Taiwan. However, the interpretation and conclusions contained in this paper are those of the authors only. The authors thank and acknowledge the support of the Maintenance Project of the Center for Big Data Analytics and Statistics (Grant CLRPG3D0049) at Chang Gung Memorial Hospital for statistical consultation and data analysis. The authors also thank Alfred Hsing-Fen Lin and Zoe Ya-Jhu Syu, Raising Statistics Consultant Inc., for their assistance with the statistical analysis. The authors thank Wallace Academic Editing for editing the manuscript.

Sources of Funding

This work was supported by grants from the Chang Gung Memorial Hospital, Taiwan (CORPG3M0371, CFRPG3M0011, CMRPG3L0101, CMRPG3L0102, BMRPD95 [to S.-W.C.] and CIRPG3L0022 [to P.-H.C.]). This work was also supported by grants from the Ministry of Science and Technology (MOST-110-2314-B-182A-114 [to S.-W.C.] and MOST-111-2314-B-182A-013-MY3 [to P.-H.C.]).

Disclosures

The authors have no conflicts of interest to declare.

IRB Information

This study was approved by the Institutional Review Board of Chang Gung Memorial Hospital (Reference no. 202002205B0).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-22-0682

References

• 1.   Nienaber CA, Clough RE. Management of acute aortic dissection. Lancet 2015; 385: 800–811.
• 2.   Akutsu K. Etiology of aortic dissection. Gen Thorac Cardiovasc Surg 2019; 67: 271–276.
• 3.   Aoki H, Majima R, Hashimoto Y, Hirakata S, Ohno-Urabe S. Ying and yang of Stat3 in pathogenesis of aortic dissection. J Cardiol 2021; 77: 471–474.
• 4.   Izumi C, Eishi K, Ashihara K, Arita T, Otsuji Y, Kunihara T, et al. JCS/JSCS/JATS/JSVS 2020 guidelines on the management of valvular heart disease. Circ J 2020; 84: 2037–2119.
• 5.   Chiu HH, Wu MH, Chen HC, Kao FY, Huang SK. Epidemiological profile of Marfan syndrome in a general population: A national database study. Mayo Clin Proc 2014; 89: 34–42.
• 6.   Linder JA, Huang ES, Steinman MA, Gonzales R, Stafford RS. Fluoroquinolone prescribing in the United States: 1995 to 2002. Am J Med 2005; 118: 259–268.
• 7.   Lee CC, Lee MT, Chen YS, Lee SH, Chen YS, Chen SC, et al. Risk of aortic dissection and aortic aneurysm in patients taking oral fluoroquinolone. JAMA Intern Med 2015; 175: 1839–1847.
• 8.   Daneman N, Lu H, Redelmeier DA. Fluoroquinolones and collagen associated severe adverse events: A longitudinal cohort study. BMJ Open 2015; 5: e010077.
• 9.   Pasternak B, Inghammar M, Svanström H. Fluoroquinolone use and risk of aortic aneurysm and dissection: Nationwide cohort study. BMJ 2018; 360: k678.
• 10.   Lee CC, Lee MG, Hsieh R , Porta L, Lee WC, Lee SH, et al. Oral fluoroquinolone and the risk of aortic dissection. J Am Coll Cardiol 2018; 72: 1369–1378.
• 11.   Chen SW, Chan YH, Chien-Chia Wu V, Cheng YT, Chen DY, Lin CP, et al. Effects of fluoroquinolones on outcomes of patients with aortic dissection or aneurysm. J Am Coll Cardiol 2021; 77: 1875–1887.
• 12.   LeMaire SA, Zhang L, Zhang NS, Luo W, Barrish JP, Zhang Q, et al. Ciprofloxacin accelerates aortic enlargement and promotes dissection and rupture in Marfan mice. J Thorac Cardiovasc Surg 2020; 163: e215–e226.
• 13.   Bennett AC, Bennett CL, Witherspoon BJ, Knopf KB. An evaluation of reports of ciprofloxacin, levofloxacin, and moxifloxacin-association neuropsychiatric toxicities, long-term disability, and aortic aneurysms/dissections disseminated by the Food and Drug Administration and the European Medicines Agency. Expert Opin Drug Saf 2019; 18: 1055–1063.
• 14.   Upchurch GR Jr. Evidence against the use of fluoroquinolones in patients with aortic pathology. JAMA Surg 2018; 153: e181988.
• 15.   Singh S, Nautiyal A. Do fluoroquinolones increase the risk of aortic aneurysms and aortic dissection? J Am Coll Cardiol 2018; 72: 1379–1381.
• 16.   Hsieh CY, Su CC, Shao SC, Sung SF, Lin SJ, Kao Yang YH, et al. Taiwan’s National Health Insurance Research Database: Past and future. Clin Epidemiol 2019; 11: 349–358.
• 17.   Lin LY, Warren-Gash C, Smeeth L, Chen PC. Data resource profile: The National Health Insurance Research Database (NHIRD). Epidemiol Health 2018; 40: e2018062.
• 18.   Hsing AW, Ioannidis JP. Nationwide population science: Lessons from the Taiwan National Health Insurance Research Database. JAMA Intern Med 2015; 175: 1527–1529.
• 19.   Maclure M. The case-crossover design: A method for studying transient effects on the risk of acute events. Am J Epidemiol 1991; 133: 144–153.
• 20.   Lipsitch M, Tchetgen Tchetgen E, Cohen T. Negative controls: A tool for detecting confounding and bias in observational studies. Epidemiology 2010; 21: 383–388.
• 21.   Arnold BF, Ercumen A, Benjamin-Chung J, Colford JM Jr. Brief report: Negative controls to detect selection bias and measurement bias in epidemiologic studies. Epidemiology 2016; 27: 637–641.
• 22.   Etminan M, Forooghian F, Brophy JM, Bird ST, Maberley D. Oral fluoroquinolones and the risk of retinal detachment. JAMA 2012; 307: 1414–1419.
• 23.   van der Linden PD, Sturkenboom MC, Herings RM, Leufkens HG, Stricker BH. Fluoroquinolones and risk of Achilles tendon disorders: Case-control study. BMJ 2002; 324: 1306–1307.
• 24.   Frankel WC, Trautner BW, Spiegelman A, Grigoryan L, LeMaire SA. Patients at risk for aortic rupture often exposed to fluoroquinolones during hospitalization. Antimicrob Agents Chemother 2019; 63: e01712–e01718.
• 25.   Newton ER, Akerman AW, Strassle PD, Kibbe MR. Association of fluoroquinolone use with short-term risk of development of aortic aneurysm. JAMA Surg 2021; 156: 264–272.
• 26.   Etminan M, Sodhi M, Ganjizadeh-Zavareh S, Carleton B, Kezouh A, Brophy JM. Oral fluoroquinolones and risk of mitral and aortic regurgitation. J Am Coll Cardiol 2019; 74: 1444–1450.
• 27.   Dai XC, Yang XX, Ma L, Tang GM, Pan YY, Hu HL. Relationship between fluoroquinolones and the risk of aortic diseases: A meta-analysis of observational studies. BMC Cardiovasc Disord 2020; 20: 49.
• 28.   Chen SW, Lu YA, Lee CC, Chou AH, Wu VC, Chang SW, et al. Long-term outcomes after extracorporeal membrane oxygenation in patients with dialysis-requiring acute kidney injury: A cohort study. PLoS One 2019; 14: e0212352.
• 29.   Gopalakrishnan C, Bykov K, Fischer MA, Connolly JG, Gagne JJ, Fralick M. Association of fluoroquinolones with the risk of aortic aneurysm or aortic dissection. JAMA Intern Med 2020; 180: 1596–1605.
• 30.   Dong YH, Chang CH, Wang JL, Wu LC, Lin JW, Toh S. Association of infections and use of fluoroquinolones with the risk of aortic aneurysm or aortic dissection. JAMA Intern Med 2020; 180: 1587–1595.
• 31.   Ribeiro NV, Melo RG, Guerra NC, Nobre Â, Fernandes RM, Pedro LM, et al. Fluoroquinolones are associated with increased risk of aortic aneurysm or dissection: Systematic review and meta-analysis. Semin Thorac Cardiovasc Surg 2021; 33: 907–918.
• 32.   Lai CC, Wang YH, Chen KH, Chen CH, Wang CY. The association between the risk of aortic aneurysm/aortic dissection and the use of fluroquinolones: A systematic review and meta-analysis. Antibiotics 2021; 10: 697.
• 33.   Yu PH, Hu CF, Liu JW, Chung CH, Chen YC, Sun CA, et al. The incidence of collagen-associated adverse events in pediatric population with the use of fluoroquinolones: A nationwide cohort study in Taiwan. BMC Pediatr 2020; 20: 6, doi:10.1186/s12887-020-1962-0.
• 34.   Cheng CL, Lee CH, Chen PS, Li YH, Lin SJ, Yang YH. Validation of acute myocardial infarction cases in the national health insurance research database in Taiwan. J Epidemiol 2014; 24: 500–507.
• 35.   Wang TC, Ku PJ, Lu HL, Hsu KC, Trezise D, Wang HY. Association between potentially inappropriate medication use and chronic diseases in the elderly. Int J Environ Res Public Health 2019; 16: 2189.