Understanding Retinal Vasculitis Associated with Brolucizumab: Complex Pathophysiology or Occam’s Razor?

KEYWORDS:

• Brolucizumab
• retinal vasculitis
• occlusive vasculitis
• intraocular inflammation

Recent reports of retinal vasculitis (some cases were also occlusive) after brolucizumab injection has ignited a poorly understood immunogenicity alert in the form of intraocular inflammation and retinal vasculitis. It is a complex phenomenon that has not been reported previously with other anti-VEGFs. Herein, the authors discuss the various pathogenic mechanisms and the potential role of anti-drug antibodies (ADA).

Prior to its launch, brolucizumab (Beovu™) had intrigued the ophthalmic community with the promise of superior drying and 8–12 weeks dosing in patients of neovascular age related macular degeneration (nAMD).^1–3 The enthusiasm though has been dampened by the incidences of occlusive retinal vasculitis, an adverse event that has not been previously reported with the use of other anti-VEGF agents injected intraocularly for the same indication.^4–8

The Research and Safety in Therapeutics (ReST) Committee of American Society of Retina Specialists (ASRS) reported occlusive vasculitis in 26 eyes of 25 patients, out of 37000 patients who were administered over 70000 injections of brolucizumab until April 1, 2020.⁷ The updated US product label reports a 1% risk of retinal artery occlusion and a 4% rate of intraocular inflammation (IOI).⁹ While the additional Safety Review Committee for HAWK and HARRIER phase 3 trials reported the incidence of IOI at 4.6%, 3.6% of the cases had retinal vasculitis associated with IOI, out of which an overall 2.1% incidence of retinal vascular occlusion was noted.¹⁰ The post-marketing surveillance reports occurrence of retinal vasculitis plus vascular occlusion at 3.73 events/10000 injections as of 8^th May 2020.¹¹

Due to the rarity of the events, a thorough understanding is difficult to achieve, and thus hypotheses based on anecdotal data have been reported by some authors. With the varied presentation, no single “fit-all” dogma can be derived, and thus complex pathophysiology can be assumed responsible. In this article, the authors aim to assess the implicated mechanisms to achieve a better working understanding of the pathophysiology and help to determine if the process is indeed complex, or if there is an Occam’s razor to explain it?

According to published reports,^4–8,10 few clear observations can help us to understand this complex phenomenon. First being that IOI preceded vasculitis by a few weeks, second, predominantly arterial involvement, third, all the patients had previous anti-VEGFs injections in the past (no treatment naïve patients), fourth all the affected patients were female, and lastly, vasculitis was found to be noninfectious and not related to post-procedure endophthalmitis. The published report of the independent SRC reported 50 treatment naïve patients from HAWK and HARRIER trials developing IOI, with 7 eyes developing IOI+retinal vascuilitis+retinal occlusion 25 days (3–49) after the last injection. The report doesn’t specify the gender distribution of the cases.¹⁰ Furthermore, to understand this phenomenon, it is important to highlight the key differences between brolucizumab and other anti-VEGFs being used in the past. The first being that brolucizumab is the smallest molecule to date with no Fc portion and, second, the preexisting anti-drug antibodies (ADA) were found to be highest with brolucizumab compared to other anti-VEGFs.¹²

Extensive uveitis work-up was negative in the cases published. An expert panel disproved the need for a systemic workup to find a probable source of an embolic occlusion.¹³ Furthermore, the involvement of only the injected eye (no bilateral event) and mean age-group makes it less likely to be due to immunogenicity secondary to systemic autoimmune mechanism. Interestingly, all 15 published cases were females. The ASRS communication revealed that 88% of the cases reported to the ReST committee were females. We agree that female preponderance is seen in cases of systemic autoimmune diseases although this limited data set cannot conclusively implicate any causative association. In the brolucizumab data, 8 out of the 15 case reports revealed an underlying presence of DM, arthritis or multiple sclerosis, all of which have some auto-immune pathophysiology. Three cases were reported in the patients having breast cancer. Rosenbaum et al., have shown that patients with retinal vasculitis rarely suffer from systemic vasculitis and it is more commonly associated with local uveitis.¹⁴ The systemic immune reaction may thus not be a causative factor but can be a predisposing factor in the reactions related to vasculitis post-brolucizumab. The recently published report from Iyer et al gives us the first histopathological evidence about the nature of inflammation in the vitreous of one such case.⁸ It concludes the presence of inflammatory cells to be in favor of a type IV hypersensitivity reaction. The report though didn’t analyze the histopathological sample from the vessels involved to provide conclusive evidence.

It is possible that small size may allow higher molar concentration and incite a strong local immune reaction leading to inflammation. We have highlighted the role of type III hypersensitivity reaction (HSR) in the past.¹⁵ The majority of vasculitic diseases involve the deposition of antigen-antibody complex, which is Type III HSR. These deposits have been shown in the capillary bed and vessel walls and can lead to occlusive vasculitis. Arthus reaction, a subtype of Type III HSR has been reported in patients on systemic monoclonal antibody (mAb), including systemic anti-VEGF therapies. Such reactions are assumed to be due to high antigen load, which leads to a subsequent increase of antibodies.^16,17 Arthus reaction is more frequent in patients with auto-immune conditions. The higher molar concentration of brolucizumab (11 and 22 times greater than aflibercept and ranibizumab, respectively), if antigenic, may produce a higher rate of antibody formation. Type III HSR is due to the formation of biologic/ADA immune complexes in the circulation. When these complexes are in the correct stoichiometric ratio, they are deposited in tissues and cause inflammation and tissue damage. The requirement of the correct ratio to have tissue deposition might explain why vasculitis is seen in some individual rather than in clusters as each individual having differing amounts of ADAs leading to different stoichiometric ratios.¹⁸

One of the most distinguishing features of brolucizumab is the presence of pre-treatment anti-drug antibody (ADA), found to be 35–52% in treatment naïve patients.¹² This is in contrast to other agents such as ranibizumab and aflibercept who had pre-treatment ADA below 5%.^19,20 The higher incidence of pre-treatment ADA is associated with smaller biologics.²¹ We have highlighted the role of ADA in cases of immunogenicity due to biologics in the past.¹⁵ A higher load of antibodies is required in type III HSR to initiate aggregation.²² The analysis of ADA in phase III trials have revealed that the incidence of sterile IOI has been associated with the occurrence of treatment-emergent ADA in both ranibizumab and brolucizumab.^12,19 The relation between inflammation and ADA against ranibizumab interestingly is seen only in patients of nAMD and not with diabetic macular edema (DME) or retinal vein occlusion (RVO).¹⁹ ADA if bound to a soluble antigen (probably the drug molecule in this scenario) may form a deposit along the vessel and cause occlusive vasculitis.

To summarize, while the anecdotal evidences limit us from drawing any definitive conclusions, the occurrence of IOI before the onset of occlusive vasculitis, the timing injection to event, the significantly higher preexisting ADA against brolucizumab, the incidence of treatment-emergent ADA with severe IOI in phase 3 trials and the deposits which appear phenotypically similar to immune complex aggregates (Type III) all point toward the significant role of ADA in these reactions compared to any other causative factor. ADA assay until now has not been assessed in routine practice, and detailed analysis was also not widely employed during the clinical trials. Given the emergent adverse events of IOI and vasculitis with brolucizumab, the further role of ADA should be evaluated in future clinical trials. We strongly believe the need of assessing the origin and potential treatment of the problem as well as the search for biomarkers in identifying susceptible individuals.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.