Anti-glaucoma agents-induced pseudodendritic keratitis presumed to be herpetic simplex keratitis: a clinical case series

Abstract

Anti-glaucoma agents-induced corneal toxicity may be misdiagnosed as herpetic simplex keratitis (HSK). In our study, nineteen glaucoma patients were presumed to have HSK before referral. Corneal lesions were classified into (I) linear pseudodendritic lesions formed by elevated opacified cells, (II) linear pseudodendritic lesions formed by grouped superficial punctate keratitis (SPK), (III) satellite full-thickness epithelial defects, (IV) satellite lesions formed by elevated opacified cells, and (V) geographic lesions formed by grouped SPK. We observed thirty-one events, with 15 in the lower and 16 in the central corneas. There were 21 (67.7%) type II, five (16.1%) type V, two (6.5%) of each for types III and IV, and one (3.2%) type I events. Among linear lesions (types I and II), 17 (77.3%) had horizontal and 5 (22.7%) had curvilinear orientations. Exposure duration to the last-added anti-glaucoma agent was three days to 14.5 years. About half of the events (16/31, 51.6%) used prostaglandin analogues, and 30/31 (96.8%) applied benzalkonium chloride (BAK)-containing agents. All lesions resolved within two months after decreasing offending medications or enhancing protection of ocular surface. In conclusion, anti-glaucoma agents-induced pseudodendritic keratitis presents majorly in central-lower cornea as horizontally linear lesions, and BAK-containing agents are observed in the most events.

Figure 1

Examples of type I–V anti-glaucoma agent-induced pseudodendritic lesions, in comparison with typical superficial punctate keratitis (SPK). (a) Type I. Linear pseudodendritic lesions with elevated opacified cells. (b) Type II. Linear pseudodendritic lesions formed by grouped superficial punctate keratitis. (c) Type III. Satellite full thickness epithelial defects. (d) Type IV. Satellite lesions formed by elevated opacified cells. (e) Type V. Geographic lesions formed by grouped superficial punctate keratitis. (f) Typical superficial punctate keratitis in patients with dye eye diseases (DED) without pseudodendritic presentation.

Figure 2

Examples of the orientation and location of anti-glaucoma agent-induced pseudodendritic lesions. The orientation was only measured in linear lesions (type I and II). The white dashed lines illustrate the contour of corneal lesions. The black solid lines divided the corneas as three equal parts. (a) Horizontal orientation. (b) Curvilinear orientation. (c) Lower-located pseudodendritic lesions, with more than 2/3 of the lesion involved the lower cornea. (d) Central-located pseudodendritic lesions, with more than 2/3 of the lesions involving the central cornea.

Figure 3

A 75-year-old female (case 10, second episode) with Sjogren syndrome who used latanoprost 0.005% (Xalatan, Pfizer, New York, NY, US, with 0.02% BAK) for one month and dorzolamide 2%/timolol 0.5% combination (Cosopt; Merck & Co., Inc., Whitehouse Station, NJ, US, with 0.0075% BAK) for three days. She presented with type II pseudodendritic lesion from by grouped superficial punctate keratatitis in the left eye with the presumed diagnosis of HSK by the referral ophthalmologist (a, b). Dorzolamide 2%/timolol 0.5% combination was discontinued, followed by adding preservative free bimatoprost 0.03% (Lumigan, Allergan, Madison, NJ, US) and lubricants. The pseudodendritic lesion resolved within two weeks (c, d).

Figure 4

A 69-year-old male (case 7) with dry eye disease used latanoprost 0.005% (Xalatan, Pfizer, New York, NY, US, with 0.02% BAK) for one year. He was referred to our clinic with the presumed diagnosis of HSK, and centrally located type II linear pseudodendritic lesions formed by grouped superficial punctate keratitis was found in the left eye (a, b). The viral culture was negative. The topical anti-glaucoma agents were discontinued followed by application of therapeutic soft contact lenses. The main lesion resolved within two weeks (c, d).