Eadie B, Etminan M, Mikelberg FS. Risk for Uveitis With Oral MoxifloxacinA Comparative Safety Study. JAMA Ophthalmol. 2015;133(1):81-84. doi:10.1001/jamaophthalmol.2014.3598
Numerous case reports have suggested that the recent use of moxifloxacin may lead to uveitis. This epidemiologic study is, to our knowledge, the first to explore the relationship between oral moxifloxacin hydrochloride use and uveitis.
We conducted a case-control study within a cohort of men aged 40 to 85 years followed up from January 2001 through December 2011. All cases with a new diagnosis of uveitis were identified. For each case, 10 control subjects were matched by age, cohort entry date, and follow-up duration. For identified uveitis cases, current use of moxifloxacin, levofloxacin, or ciprofloxacin hydrochloride was compared with nonuse. Rate ratios were adjusted for potential confounding variables. We identified 13 313 uveitis cases and 133 130 corresponding controls. Compared with nonusers of fluoroquinolones, current first-time users of moxifloxacin demonstrated the highest risk for uveitis (adjusted rate ratio, 2.98 [95% CI, 1.80-4.94]). Current first-time users of ciprofloxacin (adjusted rate ratio, 1.96 [95% CI, 1.56-2.47]) showed an increased risk for uveitis, while current first-time users of levofloxacin (adjusted rate ratio, 1.26 [95% CI, 0.90-1.77]) did not show a significant association with uveitis.
Conclusion and Relevance
Current use of moxifloxacin or ciprofloxacin appears to increase the risk for uveitis. A statistically significant association of current first-time use of levofloxacin with uveitis could not be identified. Levofloxacin may pose the least risk for uveitis compared with moxifloxacin and ciprofloxacin.
Oral fluoroquinolones (FQs) are one of the most commonly prescribed classes of antibiotics in the United States. In 2011, an estimated 23 million prescriptions for oral FQs were written in the United States.1 Oral moxifloxacin hydrochloride (Avelox) is a broad-spectrum, third-generation FQ with high oral bioavailability that is commonly prescribed for the outpatient treatment of pneumonia.
Recent reports have linked moxifloxacin to uveitis. In 2004, the first case report implicating oral moxifloxacin with bilateral uveitis was published.2 Subsequently, numerous case reports and series have shown an increase in the risk for uveitis with oral moxifloxacin.2- 8 A review of case reports from comprehensive registries including those of the World Health Organization and the US Food and Drug Administration found 40 cases of uveitis secondary to FQs, with 25 of these attributed to moxifloxacin.3 Information from case reports alone cannot demonstrate a causal link between oral moxifloxacin and uveitis, because discernment of the effect of the underlying infection from the effect of oral moxifloxacin is difficult. In contrast, causality may be best assessed by randomized clinical trials; however, these are often prohibitively difficult to perform for rare but significant adverse effects of medication. We therefore conducted a pharmacoepidemiologic case-control study to examine the possible association between oral moxifloxacin use and subsequent uveitis in a cohort of older men.
We used the LifeLink health claims database (IMS Health), a comprehensive database on hospitalizations, physician visits, and prescription drugs, as the main data source for this study. We had access to data on a cohort of 1 million men aged 40 to 85 years who were followed up from January 2001 through December 2011.
We conducted a time-matched, case-control study in our cohort. Ethics approval for the study was obtained from the University of British Columbia’s behavioral ethics board. Informed consent was not obtained because all the data were anonymized and we had no access to identifiable data. Cases were identified by probing for a first diagnosis of uveitis (code 364.00 from the International Classification of Diseases, Ninth Revision). The date of the first physician visit for uveitis was deemed the index date. Potential controls were generated using density-based sampling with the following criteria: (1) same cohort entry date as the case, (2) same follow-up duration as the case, and (3) similar age (±2 years) as the case. For each case, 10 controls were randomly selected from the eligible pool. This approach controls for time-window bias and time trends in antibiotic prescribing9 and allows for computation of the rate ratio (RR).10
We identified all oral FQ prescriptions within 1 year before the index date. We excluded all topical FQ prescriptions, because this route of administration has not been suggested to increase the risk for uveitis. Because the median onset of moxifloxacin-related uveitis has been shown to be approximately 2 weeks after the start of a prescription,3 we identified current moxifloxacin users as those who had received a moxifloxacin prescription within 15 days before the index date. First-time users were defined as those who had only received 1 prescription within 15 days of the index date. Repeated users were defined as those who had received 1 prescription within 15 days and at least 1 other prescription within 15 to 365 days before the index date. Recent (15-30 days) and distant (31-365 days) past use was also considered. Users of multiple different oral FQs were excluded. The risk for uveitis with 2 other FQs (ciprofloxacin hydrochloride and levofloxacin) was assessed to control for effects of an underlying infectious condition. The risk for uveitis with finasteride, a medication with no known associations with uveitis, to our knowledge, was also assessed. Nonusers of FQs were identified as the reference category.
We examined the demographics of cases and controls using descriptive statistics. We used conditional logistic regression analyses to compute crude and adjusted RRs. To avoid indiscriminant adjustment of covariates that may lead to overadjustment bias, we first identified potential confounding variables by drawing causal diagrams identifying true confounding variables (variables that appear to be associated with exposure and outcome) and excluding colliders (variables that can lead to selection bias).11 All analyses were performed with commercially available software (SAS, version 9.3; SAS Institute Inc).
We identified 13 313 cases and 133 130 corresponding controls. The mean (SD) age was 67.2 (13.8) years with follow-up of 2.8 (2.2) years for cases and controls. As expected, uveitis cases were more likely to have been diagnosed as having ocular trauma, diabetes mellitus, polymyalgia rheumatica, inflammatory bowel disease, and ankylosing spondylitis (Table 1). Compared with nonusers of FQs, current first-time users of moxifloxacin demonstrated the highest risk for uveitis (adjusted RR, 2.98 [95% CI, 1.80-4.94]) (Table 2). Current first-time users of ciprofloxacin also had an increased risk for uveitis (adjusted RR, 1.96 [95% CI, 1.56-2.47]), albeit lower than that of moxifloxacin. Levofloxacin use was not associated with uveitis (adjusted RR, 1.26 [95% CI, 0.90-1.77]). A comparison of the RR for current first-time users of moxifloxacin with those for current first-time users of ciprofloxacin and levofloxacin were generated. The RR was higher when oral moxifloxacin was compared with levofloxacin (2.38 [95% CI, 1.28-4.35]) than when oral moxifloxacin was compared with ciprofloxacin (1.52 [95% CI, 0.87-2.63]). Current repeated users of moxifloxacin and ciprofloxacin showed a lower RR for uveitis compared with current first-time users. As expected, no risk was observed among current repeat users of finasteride (adjusted RR, 1.08 [95% CI, 0.94-1.23]) (Table 3).
Oral moxifloxacin has been associated with uveitis in several case reports. We are able to quantify this association in the present study using a large-scale, case-control design. Our data support previous case reports suggesting that a higher number of uveitis cases develop with moxifloxacin compared with other FQs. These results also corroborate previous case reports suggesting that most of the cases of moxifloxacin-associated uveitis occur with the first dispensed prescription.3,5
The postulated mechanism behind FQ-induced uveitis is the direct toxic effects to the iris and ciliary body.12 All oral FQs exhibit high systemic absorption and vitreal penetration; however, moxifloxacin may have a higher tissue-binding affinity than levofloxacin. These pharmacodynamic characteristics may, in part, explain why the specific oral FQs are associated with different risks for uveitis.
The strengths of our study include a large sample size allowing appropriate quantification of this rare adverse event, the assessment of multiple oral FQs to control for potential confounding effects of the infectious condition (indication bias), and the use of a nonantibiotic control showing lack of association. The lack of a relationship between levofloxacin and uveitis increases the veracity of our conclusion regarding the association between moxifloxacin and uveitis.
The limitations of our study include the facts that our cohort was limited to older men and that our database only provides information on prescriptions rather than use. In addition, many cases of uveitis are currently idiopathic, and unknown variables may exist that should be adjusted for in future studies. Future studies, if able, would benefit from analyzing relationships between individual moxifloxacin dosing regimens and uveitis.
The results of the present study are consistent with those of case reports that have suggested an association between oral moxifloxacin and uveitis. Our results also suggest that ciprofloxacin may be associated with an increased albeit lower risk for uveitis. Clinicians should be aware of a possible risk for uveitis with oral moxifloxacin or ciprofloxacin when deciding to prescribe these medications.
Corresponding Author: Mahyar Etminan, PharmD, MSc, Pharmaceutical Outcomes Programme, Child & Family Research Institute, Department of Pediatrics, Faculty of Medicine, University of British Columbia, Room A4-195, 950 W 28th Ave, Vancouver, BC V5Z 4H4, Canada (firstname.lastname@example.org).
Submitted for Publication: March 25, 2014; final revision received July 29, 2014; accepted July 31, 2014.
Published Online: October 2, 2014. doi:10.1001/jamaophthalmol.2014.3598.
Author Contributions: Dr Etminan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Eadie, Etminan.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Etminan.
Administrative, technical, or material support: Eadie, Mikelberg.
Study supervision: Mikelberg.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.