Outcomes of Re-Switching Anti–Vascular Endothelial Growth Factor after Brolucizumab-Associated Inflammation in Age-Related Macular Degeneration

Article information

Abstract

Material and Methods

Results

A total of 16 eyes from 16 patients were included in the analysis. The mean age was 75.4 ± 8.1 years, and there were 10 male and 6 female patients (Table 1). Prior to initiating brolucizumab treatment, 13 patients were treated with aflibercept, 2 patients with bevacizumab, and 1 patient with ranibizumab. On average, patients had undergone 27.9 ± 13.1 injections. All patients had shown an insufficient response to previous anti-VEGF injections, prompting a switch to brolucizumab. Of these, 13 patients switched to achieve anatomical improvement, while 3 patients switched to increase treatment interval. A total of 10 patients received only one injection of brolucizumab, while 3 patients received two injections, and the remaining 3 patients underwent three, seven, and eight injections, respectively. Patients who received more than three brolucizumab injections maintained stable anatomical outcomes throughout the treatment period and were treated at intervals of 2 to 3 months. These patients were managed using a treat-and- extend regimen based on the presence of SRF or IRF. An average of 2.1 ± 2.2 brolucizumab injections were administered before the occurrence of IOI. In terms of IOI, seven patients developed anterior chamber inflammation, four patients had vitritis, three patients had both anterior chamber inflammation and vitritis, and two patients had anterior chamber inflammation, vitritis, and retinal vasculitis. Nine patients were managed exclusively with topical prednisolone, six received a combination of topical and oral prednisolone, and one patient did not receive any treatment for IOI. Complete remission of IOI was achieved in an average of 5.9 ± 5.2 weeks following onset, with the majority of patients achieving remission within 6 weeks. Following complete remission of brolucizumab-induced IOI, 12 patients switched back to aflibercept, 1 patient to bevacizumab, 1 patient to combination of ranibizumab and bevacizumab, and 1 patient to aflibercept and bevacizumab, all due to recurrent exudation, with an interval ranging from 0 to 10 weeks (2.6 ± 3.4 weeks). One patient did not receive anti-VEGF therapy, as no evidence of retinal fluid accumulation was observed. Among the 15 patients who switched to aflibercept, ranibizumab, or bevacizumab, 11 (73.3%) resumed the same anti-VEGF agent used prior to brolucizumab treatment, whereas 4 (26.7%) changed to a different agent at least once compared with their pre-brolucizumab regimen. Following the transition back to the previous anti-VEGF agents, no recurrence of IOI was observed in all cases.

Table 1

Baseline demographics and characteristics of patients (n = 16)

Number and timeline of anti-VEGF injections 6 months before and after the brolucizumab treatment is shown in Fig. 1. There was no significant difference between the number of anti-VEGF injections between the 6 months before and after the brolucizumab treatment period (2.8 ± 0.7 vs. 2.5 ± 0.9; p = 0.096, Wilcoxon matched-pairs signed rank test).

Fig. 1

Time and type of anti–vascular endothelial growth factor injections before and after brolucizumab treatment. Number after “Br” indicates the number of total brolucizumab injections before intraocular inflammation. MNV = macular neovascularization; PCV = polypoidal choroidal vasculopathy; Af = aflibercept; Rn = ranibizumab; Bv = bevacizumab; Br = brolucizumab.

Four weeks after the anti-VEGF injection just before the initiation of brolucizumab, two patients (12.5%) achieved complete resolution of SRF and IRF. These two patients switched to brolucizumab to extend the interval between injections. Additionally, 13 patients (81.3%) presented with SRF, and 3 patients (18.8%) had persistent IRF. During the brolucizumab treatment, 15 patients (93.8%) achieved complete resolution of SRF and IRF, while 1 patient (6.3%) had persistent SRF, and no patients exhibited IRF. After switching back to other anti-VEGF agents, six patients (37.5%) retained dry macula throughout the entire 6-month follow-up period while nine patients (56.3%) presented with SRF and three patients (18.8%) exhibited IRF (Figs. 2A–2C, 3A–3D). The proportion of patients achieving dry up was 46.7% after the first other anti-VEGF injection following the brolucizumab treatment, 53.3% after the second injection, and 33.3% after the third injection (Supplementary Fig. 1). Based on the proportion of patients achieving dry macula, significant differences were observed between T1 and T2 (12.5% vs. 93.8%; p < 0.01, chi-square test) and between T2 and T3 (93.8% vs. 43.8%; p < 0.05, chi-square test).

Fig. 2

Percentage of patients with complete dry up macula. (A) Four weeks after the last anti–vascular endothelial growth factor (anti-VEGF) injection before initiation of brolucizumab (T1). (B) Four weeks after the last brolucizumab injection (T2). (C) Four weeks after the last anti-VEGF injection during the 6-month follow-up period after the last brolucizumab (T3). SRF = subretinal fluid; IRF = intraretinal fluid.

Fig. 3

Optical coherence tomography images of the representative case (case 1). (A) First visit in treatment-naive state. (B) Four weeks after the last aflibercept injection before switch to brolucizumab. (C) Four weeks after one brolucizumab injection. (D) Four weeks after receiving two times of re-switched aflibercept.

In comparison of anatomical outcome at T3 between patients who received a single brolucizumab injection and those who received multiple injections, there was no significant difference in the proportion of dry macula (40.0% vs. 50.0%; p > 0.999, chi-square test). To further explore treatment response by nAMD subtype, anatomical outcomes at T3 were analyzed: all four patients with type 1 macular neovascularization exhibited fluid accumulation, whereas among two patients with type 2 macular neovascularization, one developed fluid accumulation and the other maintained dry macula. Within the polypoidal choroidal vasculopathy group, six patients retained dry macula at T3, whereas the other four showed evidence of fluid accumulation.

To assess the long-term anatomical outcomes of switched anti-VEGF agents, the proportion of patients with dry macula was evaluated at 12 months after the last brolucizumab injection. Six patients achieved complete fluid resolution, whereas 10 exhibited fluid accumulation. No significant difference in the proportion of patients achieving dry macula was observed between T3 and the 12-month time point after the last brolucizumab injection (43.8% vs. 37.5%; p > 0.999, chi-square test).

BCVA was 0.40 ± 0.26 logMAR at T1, 0.42 ± 0.24 log-MAR at T2, and 0.48 ± 0.44 at T3. BCVA at T2 and T3 showed no significant change compared to T1 (p = 0.623 and p = 0.176, respectively, Wilcoxon matched-pairs signed rank test) (Fig. 4A). CMT was 345.8 ± 109.4 μm at T1, 301.0 ± 94.5 μm at T2, and 326.8 ± 127.9 μm at T3. In comparison of T2 and T3 from T1, both showed no significant difference (p = 0.131 and p = 0.294, respectively, paired t-test) (Fig. 4B). CCT was 145.8 ± 59.8 μm at T1, 146.9 ± 64.6 μm at T2, and 154.6 ± 66.3 μm at T3. As with analysis in CMT, no significant change was found in comparison of T2 and T3 from T1 (p = 0.696 and p = 0.604, respectively, Wilcoxon matched-pairs signed rank test) (Fig. 4C).

Fig. 4

Treatment outcomes. (A) Best-corrected visual acuity (BCVA) changes. (B) Central macular thickness (CMT) changes. (C) Central choroidal thickness (CCT) changes. Error bars indicate 95% confidence interval. logMAR = logarithm of minimal angle of resolution; T1 = 4 weeks after the last anti–vascular endothelial growth factor (anti-VEGF) injection before switching to brolucizumab; T2 = 4 weeks after the last brolucizumab injection; T3 = 4 weeks after the last anti-VEGF injection during the 6-month follow-up period after the last brolucizumab.

BCVA, CMT, and CCT demonstrated no significant difference at the 12-month time point after the last brolucizumab injection compared with T1 (BCVA: 0.55 ± 0.50 logMAR vs. 0.40 ± 0.26 logMAR, p = 0.136, Wilcoxon matched-pairs signed rank test; CMT: 365.2 ± 225.4 μm vs. 345.8 ± 109.4 μm, p = 0.614, paired t-test; CCT: 157.3 ± 93.5 μm vs. 145.8 ± 59.8 μm, p = 0.528, Wilcoxon matched-pairs signed rank test).

In evaluating pretreatment and posttreatment outcomes with respect to the intended treatment goals, one patient who received brolucizumab to extend the treatment interval demonstrated an increase in the injection interval from 8 weeks during the pre-brolucizumab period to 12 weeks during the brolucizumab treatment period. For the remaining two patients in whom brolucizumab was administered with the aim of extending the treatment interval, assessment of treatment goals was not possible, as only a single brolucizumab injection was administered due to the development of IOI. Among the 13 patients who received brolucizumab for anatomical improvement, the proportion with a dry macula increased from 0% at T1 to 92.3% at T2.

Discussion

In this study, patients who showed suboptimal response to previous anti-VEGF agents achieved superior anatomical outcomes after switching to brolucizumab. However, following its discontinuation due to IOI and the resumption of previous anti-VEGF agents, the anatomical outcome gradually worsened.

The occurrence of IOI following brolucizumab treatment has been a significant concern in the management of nAMD. A post hoc analysis of HAWK and HARRIER trials revealed that among 1,088 brolucizumab-treated eyes, 49 eyes (4.5%) experienced at least one IOI-related adverse events [11]. The 96-week outcome of HAWK and HARRIER trials showed that 4.7% patients experienced IOI in the HAWK group, compared to 0.6% in comparative aflibercept group [13]. IOI manifested in various entities including iritis, uveitis, chorioretinitis, keratic precipitates, retinal vasculitis, and vitritis, with iritis and uveitis being the most frequently reported.

Real-world data have confirmed and expanded upon these findings. Park et al. [14] reported a cumulative incidence of clinically significant IOI of 3.47% within 180 days of initiating brolucizumab treatment in a nationwide cohort study in the Korean population, significantly higher than the incidence observed with ranibizumab (0.36%) or aflibercept (0.49%). Kim et al. [15] revealed overall IOI incidence rate of 41 eyes out of 294 eyes (13.9%), with most cases being anterior uveitis (26 eyes, 8.8%), followed by retinal vasculitis (7 eyes, 2.4%) and occlusive retinal vasculitis (1 eye, 0.3%). The majority of cases were successfully managed with topical or subconjunctival corticosteroid treatment, and no subsequent complications were observed [16].

The management of nAMD patients who develop IOI after brolucizumab treatment presents a significant challenge, mainly on the consequent treatment option. Inoda et al. [4] returned to aflibercept after discontinuing brolucizumab while Bodaghi et al. [12] and Witkin et al. [17] retreated patients with another anti-VEGF agent after discontinuation. However, the authors did not provide data on outcomes following the discontinuation of brolucizumab. Lin et al. [18] proposed faricimab as a treatment option after successful treatment in eyes with IOI due to brolucizumab. Dugel et al. [13] continued brolucizumab treatment in majority of patients with IOI with only 0.95% of vision loss of more than five ETDRS (Early Treatment Diabetic Retinopathy Study) letters. Kim et al. [19] continued brolucizumab treatment in 18 eyes out of 38 eyes with IOI, with 5 cases of recurrent IOI but no severe vision loss. Dugel et al. [13] and Kim et al. [19] rechallenged brolucizumab in patients with mild IOI (predominantly iritis or uveitis) successfully managed with topical or oral prednisolone, while Chakraborty et al. [20] similarly observed that most brolucizumab-induced IOI cases presented with mild anterior chamber inflammation, with only a single instance of branch retinal occlusion. Despite concerns about IOI and difficulty of finding alternative treatment methods, brolucizumab demonstrates remarkable efficacy in achieving fluid dry up in refractory nAMD patients with poor response to previous anti-VEGF a gents. Ota et al. [21] demonstrated that full absorption of fluid in 12 eyes (52.2%) and reduction in 8 eyes (34.8%) occurred among 26 eyes in patients who switched from aflibercept due to insufficient response. Yeom et al. [22] also revealed that fluid was reduced in over 76 eyes out of 81 eyes (93.8%) in the first month after switching to brolucizumab and 31 eyes (38.3%) achieved dry macula after a single injection. In our study, we have shown comparative efficacy of brolucizumab in terms of fluid reduction compared to above studies, with 93.8% of patients achieving dry macula during brolucizumab treatment. In addition, returning to previous anti-VEGF agents showed gradual increase in percentage of patients with SRF and IRF (Supplementary Fig. 1). While IOI still remains as a major concern for brolucizumab, our results suggest brolucizumab remains a valuable option, particularly for patients refractory to previous anti-VEGF treatments [23,24]. Nevertheless, rechallenging brolucizumab should be reserved for patients with mild IOI manifestations, such as iritis or anterior uveitis, as re-treatment in those who have experienced severe IOI may pose a risk of vision-threatening complications.

Our study has several limitations. First, as this study was conducted at a single tertiary center and incidence of IOI is relatively low, the sample size was relatively small, making it difficult to generalize findings from this study. Also, after discontinuation of brolucizumab treatment, the frequency or dosage of other anti-VEGF agents were not changed compared to previous regimens, potentially indicating a risk of undertreatment. However, this may be partly attributed to the regulatory constraints on aflibercept administration prior to 2022 in Korea, which mandated a minimum interval of 8 weeks between injections. Additionally, the unavailability of double-dose aflibercept (8 mg) until its market introduction in 2024 limited treatment options. Further studies with a larger number of patients from multicenter may need to be conducted.

In conclusion, switched brolucizumab demonstrated superior anatomical outcomes compared to previous anti-VEGF agents. However, after discontinuation of brolucizumab due to IOI, the use of previous anti-VEGF agents resulted in suboptimal efficacy. In light of these findings, next generation anti-VEGF agents such as faricimab or double-dose aflibercept should be evaluated in larger cohorts to assess their potential as safe and effective alternatives to brolucizumab.

Supplementary Materials

Article information Continued