Drug-induced Bilateral Secondary Angle-closure Glaucoma

A Literature Synthesis

Rory M. Murphy, MD; Belal Bakir, MD; Colm O'Brien, MD; Janey L. Wiggs, MD, PhD; Louis R. Pasquale, MD


J Glaucoma. 2016;25(2):e99-e105. 

In This Article


Classification According to Likelihood of Causing Drug-induced Bilateral 2° ACG

We evaluated 56 publications containing 60 reports of 29 distinct drugs or drug groups in this review. No case reports were found to have a definite Naranjo score of ≥9. This was mostly due to the absence of case reports with documented placebo testing, drug rechallenging, prior mention of similar reaction to the same or closely related drugs, or testing to determine whether the drug was present in toxic levels in blood or other body fluid. The following drug/drug groups were found to have probable Naranjo scores (5 to 8) and likely bilateral 2° ACG scores (≥4) in at least 1 case report (see the Materials and methods section for reference numbers): acetazolamide, anorexiant mix (see the Materials and methods section for ingredients), bupropion, cabergoline, ecstasy, escitalopram, flucloxacillin, hydrochlorothiazide, hydrochlorothiazide/triamterene, indapamide, mefanamic acid, methazolamide, oseltamivir, sulfasalazine, topiramate, and venlafaxine (Table 2). Interestingly, 2 drugs are oral carbonic anhydrase inhibitors (acetazolamide and methazolamide) used to treat glaucoma. The other drugs in this category were in a variety of drug classes used for diverse indications. Interestingly, several drugs listed in Table 2 were involved in serotonin, dopamine, and norepinephrine metabolism.

Five drugs were found to have probable Naranjo scores (5 to 8) and low 2° ACG scores <4 (see the Materials and methods section for reference numbers): chlorthalidone, flavoxate, metronidazole, paroxetine, and sulfamethoxazole/trimethoprim (Table 3). Chlorthalidone, metronidazole, and sulfamethoxazole/trimethoprim were reported to cause bilateral acute myopia, but not glaucoma. Chlorthalidone use was not associated with shallowing of the anterior chamber or elevated IOP despite documented uveal effusion on B-scan ultrasonography; in 2 cases the authors noted retinal striae and attributed induced myopia to uveal spasm as it partially resolved with cycloplegia.[35] In the metronidazole case, acute myopia was accompanied by shallowing of the anterior chamber but IOP did not increase substantially and B-scan ultrasound or ultrasound biomicroscopy assessing for uveal effusion was not performed; interestingly, rechallenge did cause a repeat in myopic shift. The circumstances surrounding the sulfamethoxazole/trimethoprim case were similar to the metronidazole case but the patient was not rechallenged. Flavoxate and Paroxetine cases exhibited features consistent with bilateral 1° ACG, typically in patients predisposed to this entity.

Basic Detox Nutrient[30] and Acetylsalicylic acid[13] received a Naranjo score in the possible range (a score of 3 for both) but good to excellent bilateral 2° ACG scores (scores of 6 and 5, respectively). Basic Detox Nutrient is marketed as a multivitamin that also contains several putative detoxifying agents including methyl-sulfonyl-methane, the ingredient the authors attributed to bilateral 2° ACG. Similarly in the acetylsalicylic acid case, the patient also consumed alcohol. These case reports are penalized in the Naranjo scoring system because it was unclear which drug entity was associated with bilateral 2° ACG.

Clinical Data

Clinical data are shown in supplemental table (Supplemental Digital Content 1, http://links.lww.com/IJG/A61).

Patient Demographics

There was a large range in patient demographic characteristics included in this review. Age was not a discriminatory factor in the development of drug-induced bilateral 2° ACG, with a span from as low as 18 (topiramate) to 76 years of age (acetazolamide). In fact a 5-year-old child developed topiramate-induced bilateral 2° ACG in the review by Abtahi et al.[10] Female sex was preponderant, making up 70.2% of the probable Naranjo and likely drug-induced, bilateral 2° ACG case reports.

Drug Dosage and Duration

Of the drugs with probable Naranjo scores (5 to 8) and likely bilateral 2° ACG scores (≥4), duration of drug usage until onset of symptoms ranged from as little as 3 hours with acetazolamide use[27] to 262 days with topiramate use.[63] Drug dosage varied from 250 to 750 mg in acetazolamide-induced 2° ACG and from 25 to 100 mg in topiramate-induced bilateral 2° ACG. Aref et al[44] reported a 70-year-old male who developed bilateral 2° ACG after consuming 50 to 100 mg of methazolamide over approximately a 12-hour period. Interestingly, Kwon et al[45] described a 51-year-old male who also developed bilateral 2° ACG after using 100 mg methazolamide for much longer (10 d); yet markers of episode severity such as the IOP level, magnitude of myopic shift, and recovery time were remarkably similar in the 2 cases indicating that methazolamide dose or duration can be quite variable when it is associated with bilateral 2° ACG.

Refractive Error and IOP

Myopic shifts of up to 10 D was reported in a case of bupropion-induced bilateral 2° ACG. Myopic shift was noted to occur in reports on acetazolamide, anorexiant mix, aspirin, bupropion, cabergoline, chlorthalidone, disothiazide, ecstasy, escitalopram, hydrochlorothiazide, hydrochlorothiazide/triamterene, indapamide, mefanamic acid, methazolamide, oseltamivir, promethazine, sulfamethoxazole/trimethroprim, topiramate, topiramate/sulfamethoxazole/trimethoprim, and venlafaxine. IOP was noted to be variable in these cases as well, ranging from 6 to 70 mm Hg in topiramate cases and from 22 to 78 mm Hg in acetazolamide cases.

Conjunctival Chemosis

Conjunctival chemosis is a phenomenon commonly reported in topiramate-induced bilateral 2° ACG.[59–65] Interestingly, chemosis was reported for other drugs implicated in bilateral 2° ACG, including acetazolamide,[24,27,28] anorexiant mix,[29] hydrochlorothiazide,[42] indapamide,[18] mefenamic acid,[43] and methazolamide.[44] Interestingly, there was no mention of conjunctival chemosis for paroxetine and flavoxate, which consistently had poor bilateral 2° ACG scores and likely represented cases of bilateral 1° ACG (see supplemental table for scores, Supplemental Digital Content 1, http://links.lww.com/IJG/A61).

Structural Analysis

Root chemical analysis showed that all drugs with probable Naranjo scale and likely bilateral 2° ACG score (≥4) had nitrogen and oxygen molecules and either cyclic pentagonal or hexagonal structure. Flucloxacillin was the only drug that had quadrangle, pentagonal, and hexagonal cyclic structures. Of these drugs, only 8 (acetazolamide, hydrochlorothiazide, indapamide, methazolamide, sulfasalazine, topiramate, methyl-sulfonyl-methane, and sulfamethoxazole/trimethoprim) had sulfonamide moieties.