Safety Outcomes of Brolucizumab in Neovascular AMD

Treatment options for wet age-related macular degeneration (AMD) include off-label use of bevacizumab or use of an FDA-approved agent, including ranibizumab, aflibercept, brolucizumab or the port delivery system. Brolucizumab, a humanized monoclonal single-chain antibody fragment that binds and inhibits vascular endothelial growth factor A (VEGF-A), was initially approved in 2019 for the treatment of neovascular AMD following two phase 3 clinical trials, HAWK and HARRIER. These clinical trials found a 4% rate of intraocular inflammation following intravitreal brolucizumab compared to 1% for aflibercept.

In the months following FDA approval, the American Society of Retina Specialists (ASRS) Research and Safety in Therapeutics (ReST) Committee started to receive reports of cases of occlusive retinal vasculitis following brolucizumab administration. In a series of 26 eyes of 25 patients with retinal vasculitis following brolucizumab submitted to the ReST committee, 91% had arterial involvement, 79% had venous involvement, 48% had involvement of the choroidal vessels, and 83% had occlusive disease. Novartis initiated an internal review of reports of intraocular inflammation following brolucizumab and established an external safety review committee. The safety review committee did a post-hoc analysis of the HAWK and HARRIER trials and reported signs of retinal vasculitis in 3.3% (36 of 1,088) of treated eyes, and concomitant signs of retinal vasculitis in 2.1% (23 of 1,088) of treated eyes.

Real-world data on incidence of intraocular inflammation and retinal vasculitis is more limited. In this study, the authors examined a large cohort of eyes receiving brolucizumab using data from the Intelligent Research in Sight (IRIS) Registry and Komodo Healthcare Map. These large patient databases were used to identify eyes that were treated with at least one intravitreal brolucizumab injection (index date). Data was available for up to 3 years prior to the index date and up to 6 months following the index date. Eyes were assessed for adverse events using ICD 10 codes for intraocular inflammation (IOI), retinal vascular occlusion, and retinal vasculitis. Measures were taken to ensure the ICD 10 diagnoses identified after brolucizumab injection were a new event and not carry-over diagnoses.

The study included 10,654 eyes from the IRIS registry and 11,161 eyes from Komodo and found that over a median follow-up of 95-97 days, the rate of intraocular inflammation was 2.4% following intravitreal brolucizumab. Additionally, they found the rate of retinal vasculitis to be 0.6%. Risk factors associated with the development of inflammation (including occlusive retinal vasculitis) were female gender, prior history of retinal vascular occlusion, and prior history of intraocular inflammation. The authors noted that those with a history of IOI or retinal vascular occlusion in the 12 months prior had an increased risk (OR 4.69 for IRIS and 5.98 for Komodo) for IOI and/or retinal vascular occlusion in the 6 months following brolucizumab initiation. Female gender was associated with a higher incidence of IOI (2.9-3.0% vs 1.3-1.4%) though the proportion of female patients in the study (57%) was higher than males. Age, history of prior anti-VEGF therapy, and follow-up duration did not appear to impact risk of IOI and/or retinal vascular occlusion. Interestingly, prior IOI or occlusion were associated with a much higher risk of retinal vasculitis (OR 12.53 for IRIS and 12.3 for Komodo). Overall, the study observed that men without a history of prior IOI or occlusion were at the lowest risk for inflammation following brolucizumab (~1.2%), while women with prior IOI or occlusion were at the highest risk (~11-14%).

The authors acknowledge certain limitations inherent to a retrospective study of this design, including an inability to access patient records to confirm coded adverse events and expand on the severity of such events. Additionally, with a median follow-up of only 3 months with ~80% in the IRIS database and ~70% in Komodo having only received 2 or fewer injections, the true prevalence of IOI may be higher. Nevertheless, for those interested in continuing brolucizumab use in their practice, the study provides a framework for risk stratification to help identify patients for whom brolucizumab must be avoided.