Purpose: To describe the association between the use of various types of topical ocular medications and acquired lacrimal canalicular obstruction in 14 patients.
Methods: The records of all patients in the author's practice with either lacrimal canalicular or punctal occlusion associated with the use of topical ocular medication were reviewed.
Results: Fourteen cases were identified. The obstructions occurred at any point from the punctum to the common canaliculus, but most commonly occurred 2-5 mm from the lacrimal punctum. There was an association with various degrees of clinically apparent subconjunctival scarring maximal at the inner canthus, rarely to a severe degree, with symblepharon, medial canthal keratinization and cicatricial medial entropion. In some cases, no subconjunctival scarring could be clinically detected. Topical medications used were often multiple and included prednisolone acetate/phenylephrine hydrochloride (n = 5), timolol maleate (n = 5), pilocarpine (n = 3), dipivefrine hydrochloride or adrenaline (n = 3), chloramphenicol (n = 3), tobramycin (n = 3), indomethacin (n = 2), ecothiopate iodide (n = 1), betaxolol (n = 1), dexamethasone (n = 1), tropicamide (n = 1) and the long-term use of naphazoline and various artificial tear preparations (n = 1). The duration of exposure ranged from 3 weeks to 20 years, with seven patients having used drops for 3-6 weeks. Seven patients had surgical repair, three by dacryocystorhinostomy (DCR) and glass by-pass tube (all successful), three by canalicular repairs (one failed) and one by DCR and canalicular repair that restenosed at the puncta, who then had successful punctoplasty and silicone intubation.
Conclusions: Lacrimal canalicular obstruction may occur after relatively short-term exposure to topical ocular medications or as part of a more widespread cicatricial reaction in patients on long-term medication. While a direct causal relationship cannot be confirmed, there appears to be a strong association and the site of the obstructions makes other causes unlikely.