Dear editor:

Recent publications have alerted clinicians to a syndrome of transilluminating iris depigmentation associated with the use of systemic fluoroquinolones and other antibiotics [1, 2]. This syndrome, termed bilateral acute iris transillumination by some authors [1], differs from the previously described bilateral acute depigmentation of the iris, which causes reversible atrophy of the iris stroma without iris transillumination [3]. In contrast, fluoroquinolone-associated uveitis preferentially targets the iris pigment epithelium, leading to irreversible iris transillumination [2].

We present a case of fluoroquinolone-associated uveitis with anterior segment optical coherence tomography (OCT) imaging to highlight some observations about this syndrome. We interpret pharmacokinetic data to help explain why oral, but not topical, moxifloxacin may cause fluoroquinolone-associated uveitis. We also discuss the possible impact of the presence of the native crystalline lens. Finally, we note that the anterior segment OCT findings suggest that both the iris stroma and iris pigment epithelium are affected in fluoroquinolone-associated uveitis.

Report of a case

A 70-year-old Caucasian man complained of blurry vision and photophobia in both eyes for 2 weeks. His past medical history was notable for lung cancer and an episode of pneumonia treated with oral moxifloxacin 3 days before the ocular symptoms began. His visual acuity was 20/30 OD and 20/40 OS with intraocular pressures (IOP) of 35 mmHg in both eyes. He had mild corneal edema and symmetric 3+ anterior chamber pigmented cells (Figure 1A,B). Gonioscopy revealed dense pigment obscuring all angle structures (Figure 1C,D). The posterior segments were normal. Serum laboratory evaluation and polymerase chain reaction of the right eye aqueous humor were negative for HSV, VZV, and CMV. He was treated with topical steroids and cycloplegic and IOP-lowering medications with little change in vision, IOP, or pupil size. Retroillumination images at presentation and 2 weeks later showed stable iris transillumination defects (Figure 1E,F,G,H), suggesting an acute, self-limited injury. Anterior segment OCT images of the iris demonstrated thinning, concavity, and adherence to the lens (Figure 2A,B). The patient died from respiratory failure after his 2-week follow-up visit.

Figure 1
figure 1

Clinical presentation of the anterior segment. Pigmented cells in the anterior chamber with deposition of a pigmented ‘wreath’ on the corneal endothelium adjacent to the limbus are seen in the right (A) and left (B) eyes. Gonioscopy shows pigment anterior to Schwalbe's line and extending on to the peripheral iris in the right (C) and left (D) eyes. Transillumination defects noted at presentation in the right (E) and left (F) eyes were stable 2 weeks later in the right (G) and left (H) eyes.

Full size image
Figure 2
figure 2

OCT of the anterior segment. The right (A) and left (B) eyes show significant stromal thinning with iris concavity and posterior synechiae.

Full size image

Comment

Previous reports have noted fluoroquinolone-associated uveitis only with systemic fluoroquinolones [1, 2]. Pharmacokinetic data may explain why topical administration has not been associated with this clinical presentation. After topical administration, there is a greater-than-tenfold higher concentration of moxifloxacin in aqueous (2.28 ± 1.23 μg/mL) than in vitreous (0.11 ± 0.05 μg/mL) [4], whereas oral administration produces similar aqueous (1.34 ± 0.66 μg/mL) and vitreous concentrations (1.58 ± 0.80 μg/mL) [5]. The steady serum and vitreous reservoirs of moxifloxacin during oral administration may maintain drug levels in the tissue at risk better than intermittent topical therapy.

The lens status of affected patients may also be relevant. To our knowledge, only phakic patients have been reported with fluoroquinolone-associated uveitis [1, 2]. Less drug diffuses posteriorly in phakic eyes [6], and posterior-to-anterior clearance may also be impaired in phakic eyes. Trapping of drug in the posterior chamber by synechiae between an intact lens and the iris, as demonstrated in our OCT (Figure 2), may result in higher drug concentrations adjacent to the posterior layers of the iris.

Why is intravitreal moxifloxacin injection apparently safe [7]? Experience gained from cidofovir-induced uveitis and hypotony indicates that toxicity to the nonpigmented epithelium of the ciliary body is uncommon after intravitreal therapy [8] compared to systemic administration [9], an effect that may be due to the more rapid clearance of intravitreally administered drug compared to recirculation from a serum depot. Additionally, in vitro toxicity studies have not examined the effect on the iris pigment epithelium [10], which may have different susceptibility to moxifloxacin toxicity. Finally, greater light exposure of the iris may have a bearing on the achieved toxicity.

Clinicians confronted with patients with acute onset of uveitis and pigment dispersion should be aware of the association between systemic fluoroquinolones and uveitis. Treatment with corticosteroids for prolonged pigment dispersion after the initial inflammatory phase is likely unnecessary and may contribute to glaucoma in steroid responders. Intraocular pressure would be predicted to improve as the load of pigment in the trabecular meshwork subsides.