To examine the possible link of acute-onset glaucoma with topiramate.
A case-control study was conducted among a cohort of subjects who had visited an ophthalmologist in the Province of British Columbia, Canada from 2000 to 2007. Cases were identified as those newly diagnosed with glaucoma (ICD-9 360). For each case, 5 controls were selected and matched to the cases by age and calendar time using density-based sampling. Crude and adjusted rate ratios (RRs) for current and past use of topiramate were computed. As a sensitivity analysis, the risk of glaucoma with a positive control drug (an oral steroid) and a negative control drug (inhaled albuterol) was also assessed.
From the initial cohort of 989 591 subjects, 178 264 cases of glaucoma and 891 320 controls were identified. There was a slight increase in the risk of glaucoma among current users of topiramate (RR = 1.23 [95% confidence interval (CI), 1.09-1.40]). This risk was further elevated among new users of the drug (RR = 1.54 [95% CI, 1.09-2.17]). No increase in the risk of glaucoma requiring drug therapy was observed among current topiramate users (RR = 1.09 [95% CI, 0.80-1.61]).
We found an increase in the risk of glaucoma with first-time users of topiramate. Future studies are needed to confirm these findings.
Topiramate is an anticonvulsant mainly used in the treatment of epilepsy as well as prophylactic therapy for migraine headaches. In 2009, the annual sales of topiramate in the United States exceeded $2 billion. One rare, but potentially serious adverse event that is linked to topiramate use is acute angle-closure glaucoma, which may lead to blindness. In many of the cases glaucoma was reported within days of using topiramate and symptoms subsided after stopping topiramate. Currently, data on the incidence of topiramate-induced glaucoma are lacking. Given the popularity of this drug in different clinical settings, especially in younger adults, information on this adverse event will be vital to practicing clinicians.
We used a case-control study design to examine the possible association between topiramate and glaucoma. The British Columbia Linked Health Database (BCLHD), a provincially linkable database, was used as the main source of the data for this study. The data captures physician visits (including in-patient procedures), hospital admissions, demographics, and prescription drug use for approximately 4.5 million residents for the Province of British Columbia, Canada. The prescription drug database captures information on drug dose, day supply, and quantity. The database has been used extensively in health services and pharmacoepidemiologic research. The cohort included all those who visited an ophthalmologist’s office from 2000 to 2007.
Case and Control Selection
Cases were defined as those with a first International Classification of Disease, 9th edition (ICD-9) code (365) recorded by an ophthalmologist. We included all subtypes of 365 as the accuracy of ICD-9 codes for different types of glaucoma in the BCHLD has not been established. The first date of diagnosis was referred to as the index date. Prior cases of glaucoma were excluded by going back 1 year prior to the index date. In a separate analysis, we also defined glaucoma as having received a glaucoma drug within 90 days of glaucoma diagnosis. This approach has been used in previous studies of drug-induced glaucoma. For each case, 5 controls were selected from the cohort using a density-based sampling method. This method of control selection allows for the close approximation of the rate ratio (RR) to the odds ratio. Controls were selected if they had satisfied the following criteria: 1) each control had to have been followed at least as long as the time of the case (index date) and thus was at risk of developing glaucoma; and 2) controls were matched to the cases within 1 year of birth date (±1 year). In order to control for prescribing trends for topiramate, cases and controls were also matched by cohort entry (±30 days).
Descriptive statistics were used to examine data demographics. We constructed a conditional logistic regression model to estimate rate ratios and adjusted for the following covariates: sex, coronary artery disease, hypothyroidism, hypertension, diabetes, and statin, corticosteroid, and thiazide use. Current users of topiramate were classified as those who had an overlapping prescription on the index date. Past users were defined as those who did not have an overlapping topiramate prescription. Since topiramate-induced glaucoma is believed to be an acute event with possibly new users being at a higher risk, we assessed the risk of glaucoma with first-time users of topiramate. In order to examine the validity of the results, we further assessed the risk of glaucoma with 2 unrelated drugs. Inhaled albuterol was used as a negative control as this drug is not expected to increase the risk of glaucoma. Oral corticosteroids were used as positive controls as they have shown to be associated with an increased risk of glaucoma. All rate ratios were compared to non-users of a study drug. We also performed two separate analyses, one defining glaucoma as a registered ICD-9 365 code, and the other a drug therapy for glaucoma within 90 days of receiving an ICD-9 code.
From the initial cohort of 989 591 subjects, we identified 178 264 cases of glaucoma and 891 320 corresponding controls ( Table 1 ). Cases were more likely to have hypertension and use thiazide diuretics. The incidence of glaucoma in the entire cohort was 40 per 1000 users. There was a slight increase in the risk of glaucoma among current users of topiramate (RR = 1.23 [95% confidence interval (CI), 1.09−1.40]). This risk was further elevated among new users of the drug (RR = 1.54 [95% CI, 1.09−2.17]) ( Table 2 ). The risk of glaucoma among younger subjects (<40 years) was doubled, although because of the small number of cases this value did not reach statistical significance (RR = 2.01 [95% CI, 0.83−4.84]). The median onset from the start of a topiramate prescription to the diagnosis of glaucoma was 28 days. As expected, current albuterol use was not associated with an elevated risk of glaucoma, whereas oral steroid users were at a higher risk of developing glaucoma ( Table 2 ). No increase in the risk of glaucoma requiring drug therapy was observed among current topiramate uses (RR = 1.09 [95% CI, 0.80−1.61]).
|Number of subjects||178 264||891 320|
|Age at index date (y; mean ± SD)||65.0 ± 15.6||65.0 ± 15.6|
|Male sex (%)||42.1||42.9|
|Time to diagnosis of glaucoma (y; mean ± SD)||1.4 ± 1.8||1.4 ± 1.8|
|Comorbidity 1 year prior to index date (%)|
|Coronary artery disease||8.2||8.1|
|Cases (178 264)||Controls (891 320)||Crude RRs||Adjusted RRs (95% CIs) a|
|Non-use||177 400||887 514||1.00 b||1.00|
|Current users (N)||304||1229||1.24||1.23 (1.09−1.40)|
|First-time users (N)||43||138||1.56||1.54 (1.09−2.17)|
|Prevalent current (N)||261||1091||1.20||1.19 (1.04−1.37)|
|Past use (N)||560||2577||1.09||1.08 (0.99−1.190)|
|Current users of albuterol (N)||10 681||53 731||0.99||0.99 (0.97−1.01)|
|Oral steroid use|
|Cumulative use (>800 days)||1404||4396||1.61||1.60 (1.51−1.70)|