Brolucizumab IVI tied to occlusion of retinal arteries, perivenular abnormalities

Intravitreal injection (IVI) of brolucizumab for the treatment of neovascular age-related macular degeneration results in variable occlusion of large or small retinal arteries, or both, and perivenular abnormalities, which characterize retinal vasculitis and intraocular inflammation (IOI), a study has found.

“It may span from peripheral vasculitis to occlusion of large retinal arteries around the optic nerve or macula with severe vision loss,” the authors said. “A high index of suspicion is required because vitreous cells may obscure visualization of retinal details.”

Fifteen eyes from 12 patients identified from 10 US centres were included in this retrospective case series. Patient demographics, ophthalmologic examination results, and retinal imaging findings were reviewed to assess the features and outcomes of eyes with retinal vasculitis and IOI after IVI of brolucizumab 6 mg/0.05 ml.

Prior to brolucizumab switch, the number of previous antivascular endothelial growth factor (anti-VEGF) IVIs ranged between 2 and 80 in the affected eye. Retinal vasculitis and IOI were diagnosed at a mean of 30 days following brolucizumab IVI.

Before brolucizumab IVI, the mean visual acuity (VA) was 0.426 logarithm of the minimum angle of resolution (logMAR; Snellen equivalent, 20/53). VA at diagnosis of retinal vasculitis was 0.981 logMAR (Snellen equivalent, 20/191; range, 20/25–20/1,600; p=0.008).

Affected eyes exhibited IOI with variable combinations of focal or elongated segmental sheathing and discontinuity of small and large retinal arteries, sclerotic arteries, regions of vascular nonperfusion, cotton-wool spots, Kyrieleis plaques, irregular venous caliber with dilated and sclerotic segments, perivenular haemorrhages, and foci of phlebitis.

Delayed retinal arterial filling, retinal vascular nonperfusion, and variable dye leakage from affected vessels and the optic nerve were shown in fluorescein angiography. Two patients displayed ambiguous systemic evaluation for embolic causes, and three showed negative laboratory assessment for uveitis.

Various combinations of corticosteroids (eg, systemic, intravitreal, and topical) were administered. Two eyes underwent vitrectomy with no improvement in vision. Mean VA was 0.833 logMAR (Snellen equivalent, 20/136), which decreased compared to baseline, after a mean follow-up of 25 days (p=0.033).