Reinjection in Patients with Intraocular Inflammation Development after Intravitreal Brolucizumab Injection
Article information
Abstract
Purpose
To investigate the outcomes of brolucizumab reinjection after intraocular inflammation (IOI) development.
Methods
This retrospective study analyzed patients with brolucizumab injections from April 2021 to January 2024. Patients who developed IOI after brolucizumab were included and categorized into subgroups depending on reinjection, discontinuation, and further IOI development.
Results
A total of 472 eyes of 432 patients received brolucizumab injections. Thirty-eight cases developed IOI at least once, and 25 continued brolucizumab. Sixteen cases had no more IOI events, and nine experienced a second or more IOI events. Among the nine cases, three maintained brolucizumab injections despite IOI recurrence. The incidence of IOI was 8.1% based on the number of eyes (38 of 472 eyes) and 2.0% based on the number of brolucizumab injections (50 of 2,468 injections). The incidence of occlusive retinal vasculitis was 0.2% (1 of 472 eyes). The recurrence rate was 23.7% (9 of 38 eyes). The average number of injections between the first brolucizumab injection and the injection date on which IOI first developed was 2.15 times in the no-reinjection group, 3.44 times in the no-IOI-recurrence group, and 2.0 times in the second-IOI-episode group. Time to IOI occurrence in cases with first IOI episode was 18.60 ± 16.73 days, with 15 cases developing IOI within 1 week.
Conclusions
This study elucidates the real-world incidence of brolucizumab associated IOIs, with a description of information related to reinjections after the IOI episodes. A comprehensive understanding of brolucizumab reinjection is essential for its optimal utilization.
Age-related eye diseases that lead to low vision and blindness are a growing challenge for public health professionals [1]. Among them, age-related macular degeneration (AMD) is a degenerative disease associated with aging of the macula [2]. Currently, AMD affects more than 50% of people over 80 years. It is the leading cause of blindness worldwide, and anti–vascular endothelial growth factor (anti-VEGF) therapy has become the standard treatment for neovascular (wet) AMD (nAMD) [3–5]. Five molecules are currently approved and marketed for the symptomatic treatment of nAMD: bevacizumab, ranibizumab, aflibercept, brolucizumab, and faricimab [3,6].
Brolucizumab (Beovu, Novartis) is a human VEGF inhibitor indicated for the treatment of nAMD, diabetic macular edema, and macular edema secondary to retinal vein occlusion [7]. It was approved by the US Food and Drug Administration for intravitreal use in 2019 [7]. T his is a relatively low molecular weight (26 kDa) humanized single-chain antibody fragment, which binds to the three major isoforms of VEGF-A (VEGF110, VEGF121, and VEGF165), thereby preventing interaction with receptors VEGFR-1 and VEGFR-2 [3,7,8]. The HAWK and HARRIER phase 3 clinical trials showed brolucizumab to be noninferior to aflibercept in terms of visual acuity gains and showed greater fluid resolution compared with aflibercept. The q12w potential of brolucizumab 6 mg treatment was >50% through week 48 and was maintained to week 96 [9,10]. Longer durability and better outcomes are the two major challenges in the selection of anti-VEGF agents [4]. Therefore, brolucizumab may function as a durable agent, allowing high molar doses and, thus, fewer injections for patients [3]. As a result, it is now widely used in the treatment of patients with nAMD. However, intraocular inflammation (IOI) is a well-known side effect of the drug with symptoms that can range from mild inflammation to vision-threatening conditions, such as retinal vasculitis and vascular occlusion [11–13]. Thus, it is common to discontinue the drug or switch to another agent for safety reasons when inflammation occurs after brolucizumab injection.
Few studies have addressed the reinjection of brolucizumab after IOI. Witkin et al. [14] mentioned second brolucizumab injections in 3 of 34 eyes with IOI related to brolucizumab. Two received a subtenon triamcinolone injection at the same time as the brolucizumab injection, and no recurrence of inflammation occurred in all three eyes. Singer et al. [12] found that 36 of 49 eyes continued brolucizumab treatment after the first IOI-related adverse effect in a post hoc analysis of the HAWK and HARRIER trials. Of these, 24 completed the trial with brolucizumab treatment only, and the remaining 12 eyes discontinued brolucizumab injection later for various reasons. However, further information is required to comprehensively understand the features of brolucizumab reinjection.
This study aimed to investigate the outcomes of patients who had experienced IOI following brolucizumab injection, with a particular focus on those who had received brolucizumab reinjections despite IOI development.
Materials and Methods
Ethics statement
This study was approved by the Institutional Review Board of Nune Eye Hospital (No. 2024-08-001), with a waiver of informed consent due to the retrospective design. All study protocols adhered to the principles of the Declaration of Helsinki.
Study design
This study included all patients who received brolucizumab between April 2021 and January 2024, regardless of age or diagnosis. The incidence of IOI was investigated, and those who developed their first IOI were divided into two groups based on whether they continued brolucizumab injections or not (no-reinjection group and reinjection group). In the reinjection group, the patients were further divided into two groups according to a second IOI event (no-IOI-recurrence group and second-IOI-episode group). In this group, those who continued brolucizumab injections and those who finally discontinued brolucizumab were identified. Patients with a previous history of uveitis were excluded, as well as those with adverse events other than inflammation (IOI).
Data, including the number and date of brolucizumab injection on which IOI first developed, and symptoms and treatments that were administered at the time, were collected. Injection information prior to brolucizumab injection was collected, such as the kind of medication used and the total number of injections. This information was obtained except for treatment-naive patients. If brolucizumab was discontinued for any reason and switched to another agent, the type of medication was also identified.
Visual acuity was measured on the date of the injection, and 1 and 3 months after the injection date that triggered the IOI episode. Finger count was converted to 0.014 [15] and subsequently converted to logarithm of the minimum angle of resolution (logMAR) visual acuity. Cases with insufficient data during follow-up were excluded from the analysis.
Age, sex, and the injection site were also recorded. Age was defined as the age at which the first IOI occurred. The presence or absence of specific diagnoses, such as AMD, central serous chorioretinopathy (CSC), and diabetic retinopathy (DMR), was identified.
Statistical analysis
Data were analyzed using descriptive statistics in IBM SPSS ver. 21 (IBM Corp). No tests of statistical significance were performed. The data were transformed into a graph using Microsoft Excel 2013 (Microsoft Corp).
Results
A total of 472 eyes in 432 patients were included. Thirty-eight cases developed IOIs, while 434 cases did not. Among the 38 cases with IOI, 13 discontinued brolucizumab. Eleven cases discontinued treatment due to IOI, and two due to other factors, including one case with visual acuity limitations and one case lost to follow-up. In 25 cases, brolucizumab was reinjected without discontinuation after the first IOI and a second IOI occurred in 9 of these cases. The remaining 16 cases did not experience additional IOIs. Among these, 14 continued brolucizumab, and 2 discontinued treatment due to decreased drug efficacy. Of the nine cases with a second IOI, three maintained brolucizumab injections and six discontinued treatment, five due to IOI and one due to decreased drug efficacy (Fig. 1). Eight of the nine cases had two IOIs, and one had four IOIs. The case with four IOIs was the one in which brolucizumab was discontinued due to decreased drug efficacy, as previously mentioned. This patient received a total of 13 injections of brolucizumab until discontinuation. For those with brolucizumab discontinuation, patients switched to other agents or decided to stop intravitreal injections. In the no-reinjection group, three switched to bevacizumab, seven to aflibercept, and two discontinued anti-VEGF injection. Two cases in the no-IOI-recurrence group switched to faricimab. In the second-IOI-episode group, two cases switched to bevacizumab, two to aflibercept, and one to faricimab.
Distribution of patients with brolucizumab injections. Cases with intraocular inflammations (IOIs) were further grouped according to brolucizumab reinjection.
The incidence of inflammation after brolucizumab injection in our institution was 8.1% based on the number of eyes (38 of 472 eyes) and 2.0% based on the number of brolucizumab injections (50 of 2,468 injections). The incidence of occlusive retinal vasculitis was 0.2% (1 of 472 eyes). The recurrence rate after reinjection was 23.7% (9 of 38 eyes).
The mean age of the 38 cases with at least one IOI was 70.16 years, and 71.52 years in the reinjection group. Cases with at least one IOI episode were diagnosed as AMD in 37 cases, CSC in one case, and DMR in one case. In the re-injection group, there were 25 cases of AMD and one case of DMR. The majority of patients received brolucizumab for AMD, and DMR comorbidity was seen in a patient with AMD in one case (Table 1). The incidence of IOIs analyzed by diagnosis is as follows. For those who received brolucizumab for the treatment of AMD, 8.1% (37 of 455 eyes) developed IOI. The incidence was 20% (one of five eyes) in CSC, and 0% (zero of three eyes) in diabetic macular edema. However, the incidence increased to 20% (one of five eyes) when AMD patients with DMR were treated with brolucizumab.
Demographic data of patients with at least one IOI episode of brolucizumab injection (n = 38)
The most common symptoms in patients with IOIs were anterior chamber cells and vitreous cells with opacity observed by clinicians. Cases with only patient symptoms, such as increased floaters or optical coherence tomography findings like hyperreflective dots, were also present. One case revealed segmental branch retinal artery occlusion (BRAO) in the first IOI episode. Fluorescein angiography (FAG) was performed in 13 cases with the first IOI episode. One case developed retinal vascular sheathing, but FAG was not performed in this case. In the second and third IOI episodes, no FAGs were performed. The only case with four IOIs had a FAG in the fourth IOI episode, and vascular leakage was detected. However, this patient discontinued brolucizumab after the fourth IOI due to the lack of drug efficacy rather than IOI (Fig. 2A). In the first IOI episode, observation was done in nine cases (23.7%) without steroid treatment. Ten (26.3%) were treated with subtenon triamcinolone injection, and seven (18.4%) had intravitreal triamcinolone injections (IVTAs). Treatments for the second IOI episode included observation in four cases (44.4%), subtenon in one case (11.1%), and IVTA in two cases (22.2%). The third and fourth IOI cases occurred in the same eye, with subtenon triamcinolone injection in the third IOI and IVTA in the fourth IOI (Fig. 2B).
(A) Symptoms and (B) treatments during intraocular inflammation (IOI) events. FAG = fluorescein angiography; IIOP = increased intraocular pressure; OCT = optical coherence tomography; ITVA = intravitreal triamcinolone injection; IOP = intraocular pressure.
The interval between the first brolucizumab injection and the injection date with the first IOI episode was shorter in the no-reinjection group (111.15 days) compared to the reinjection group (172.44 days). When comparing the subgroups of the reinjection group, this interval was shorter in patients with two or more IOI episodes (second-IOI-episode group, 91.89 days) compared to patients without additional IOIs (no-IOI-recurrence group, 217.75 days). In the second-IOI-episode group, the second IOI tended to occur about 304.56 days after the first IOI development (Fig. 3A). In terms of injections, the number of injections between the first brolucizumab injection and the injection date with the first IOI episode was lower in the no-reinjection group compared to the reinjection group (2.15 vs. 2.92). Reinjection group subgroups showed 3.44 injections in the no-IOI-recurrence group, and the second-IOI-episode group showed lesser injections of 2 until the IOI development. In the second-IOI-episode group, second IOIs tended to develop after 2.44 additional brolucizumab injections from the first IOI (Fig. 3B). In cases with the first IOI episode, IOI occurred on an average of 18.60 ± 16.73 days after the injection that triggered the IOI episode. Of these, 15 cases developed IOI within 1 week. For the second IOI episode, the average was 11.13 ± 12.74 days, with five cases developing IOI within 1 week (Fig. 3C, 3D).
Detailed information of each intraocular inflammation (IOI) episode. (A) Time interval between injections. (B) Number of average brolucizumab injections before an IOI episode. Time from injection date to IOI occurrence: (C) IOI first episode and (D) IOI second episode.
The change in the best-corrected visual acuity (BCVA; in logMAR) 1 and 3 months from the first IOI injection date was measured in each group. The no-reinjection group showed the greatest initial decline among all groups, but these changes recovered after drug discontinuation. The reinjection group had a small initial drop and the levels remained constant. In subgroups, the no-IOI-recurrence group had the least amount of visual loss. The second-IOI-episode group showed a gradual decline in BCVA over time. The BCVA on the injection day was worse in the second IOI episode than the first one, suggesting that visual acuity does not recover between the first and second episodes and gradually deteriorates with repeated inflammation (Fig. 4).
Best-corrected visual acuity changes (ΔBCVA) in logarithm of the minimum angle of resolution (logMAR) 1 and 3 months from the injection date that triggered the first intraocular inflammation (IOI) episode.
Discussion
The incidence of inflammation after brolucizumab injection in our institution was 8.1% (38 of 472 eyes) and 23.7% (9 of 38 eyes) experienced recurrence after reinjection. The incidence was previously reported to be 4.6% (50 of 1,088 eyes) in the HAWK and HARRIER trials [11] and 10.5% (53 of 505 eyes) in the OCTOPUS and SWIFT trials [13]. Severe effects, such as vascular occlusion, were uncommon, with one case of BRAO in our study. This was a unique case where the patient developed BRAO just after the first brolucizumab injection but continued brolucizumab without switching to another agent. The patient started brolucizumab due to the insufficient effects of a previous injection. Although a retinal vein occlusion event occurred, the lesion was small and segmental and was localized at the extrafovea (superotemporal to the disc) without affecting the patient’s symptoms or the AMD lesion itself. The patient did not receive additional steroid treatment, and after discussion, the physician decided to reinject brolucizumab. For almost 3 more years, no further IOI events occurred, and the patient is currently maintaining treatment and extension. Our study showed the real-world incidence of IOI after brolucizumab injection; however, the decision can be made more sensitively in the clinical field, and suspicious cases might also be considered and treated as mild IOIs.
FAGs were performed less frequently in our study. Among the 38 cases with the first IOI episode, 13 cases underwent FAG. The low rate of FAG can be explained by two factors. First, the symptoms of IOIs were relatively mild. Second, as experience with IOIs has increased and more data has accumulated, clinicians have been able to diagnose IOIs more quickly without further evaluation, leading to faster treatment.
In this study, five types of treatment options were considered for brolucizumab-induced IOIs. The first option was observation without using steroids. In some cases, inflammation induced an increase in intraocular pressure, and when the patient was treated only with intraocular pressure–reducing agents, such as glaucoma eye drops, then they were categorized into the observation group. Second was the use of steroid eye drops only, and we classified these two as relatively mild treatments. More aggressive treatments are classified as the usage of oral steroids, subtenon triamcinolone injection, and intravitreal triamcinolone injection. According to this classification, the no-reinjection group was treated more aggressively (61.5%) than the reinjection group (44.0%), indicating that brolucizumab was discontinued for safety in cases of severe inflammation. In contrast, reinjection was often associated with relatively mild inflammation. Of the patients who experienced a second IOI, 44.4% underwent observation without using any steroid agents. These findings suggest that patients who respond effectively to brolucizumab could be monitored and reinjected with caution, even if mild inflammation occurs. In 2021, Witkin et al. [14] reported a study on brolucizumab-induced IOIs and reinjection. In this study, among the 34 patients who experienced IOI following brolucizumab administration, three underwent brolucizumab reinjection, with subtenon injection performed concomitantly in two cases. No further IOI recurrence was observed in any of the three cases. This suggests that the concomitant use of brolucizumab and steroids may be a good option for patients with nonsevere IOIs. In the case of patients for whom the benefit of the injection outweighs the adverse effects, reinjection may be considered. Therefore, it is important for the physician to predict the likelihood of IOI recurrence after reinjection for decision-making. Based on our results, the time and the number of injections from the first brolucizumab injection to the injection day of the first IOI event was shorter in those who had a second IOI in the two reinjection subgroups (no-IOI-recurrence group and second-IOI-episode group). This suggests that if the first IOI occurs relatively early, the probability of IOI recurrence after reinjection may be higher.
In many cases, IOIs developed within 1 week, with an average of approximately 2 weeks from the time of brolucizumab injection. The upper bounds of the 95% confidence interval was 35.33 days in cases with the first IOI episode, and 23.87 days in cases with the second IOI episode. Based on these findings, we suggest that the appropriate time for follow-up is twice, i.e. 2 weeks, and 1 month after the injection. If available, three follow-up visits, including an additional visit at 1 week, would also be beneficial.
When comparing visual acuity in the no-reinjection group and the reinjection group, both groups show an overall decrease immediately after developing IOI. It is likely that in the acute phase, visual acuity decreases due to inflammation. The no-reinjection group, which discontinued brolucizumab without a subsequent reinjection, seemed to recover visual acuity due to the absence of further inflammation. In contrast, the second-IOI-episode group, which had a second IOI, tended to show a gradual decline in visual acuity over time, which is likely due to the accumulation of inflammation. However, the absolute visual acuity values in the group that did not undergo further injections were consistently lower than those in the reinjection group, making it challenging to conclude that this group had an advantage in terms of visual acuity.
The no-reinjection group demonstrated a greater degree of visual impairment immediately after inflammation compared to the other groups. This may have contributed to the decision to discontinue the use of brolucizumab. This group had the lowest visual acuity prior to injection among all groups, suggesting that the threat of visual decline was significant, even with relatively mild inflammation. In contrast, the group that opted for reinjections despite experiencing a first IOI (reinjection group) had less visual loss in the acute phase. This suggests that the inflammation was not as severe as in the group that decided to discontinue the drug after the first IOI (no-reinjection group). Of these, the group with recurrent IOIs (second-IOI-episode group) had relatively worse visual acuity after experiencing inflammation than the group without recurrent IOIs (no-IOI-recurrence group). Nevertheless, the decision to continue the injections appears to have been influenced by the fact that the visual acuity in this group was relatively high, suggesting that the extent of the lesions was not severe enough to immediately stop the injections. This suggests that the risk of inflammation could be tolerated to maintain the injections if there are no serious side effects. Despite the inflammation issues, reinjection of brolucizumab can be considered cautiously by weighing the pros and cons.
This study had some limitations. One is the relatively small number of total cases: only 38 patients with IOI were enrolled. However, more cases will accumulate as the duration and rates of brolucizumab usage increase. As this is a retrospective study analyzing clinical data, further research with a randomized, controlled, prospective study is needed.
In summary, a comprehensive understanding of brolucizumab reinjection is essential for its optimal utilization. We should understand that reinjection does not necessarily mean a recurrence of an inflammation. A good understanding of the medication is the most efficient method to achieve positive results.
Notes
Conflicts of Interest
None.
Acknowledgements
None.
Funding
None.