Outcome of Switching Anti-VEGF Agents at Fixed Intervals to Treat Eyes With Macular Edema Secondary to Branch Retinal Vein Occlusion

Objective To determine whether switching intravitreal injections of ranibizumab or a�ibercept to the other will affect the interval of subsequent injections in patients with macular edema secondary to branch retinal vein occlusion (BRVO). Results The medical charts of 16 eyes of 16 patients were reviewed. After injections of ranibizumab or a�ibercept for >3 times/year, the type of anti-vascular endothelial growth factor agent was switched. A comparison was made of the average injection interval for the one year before switching and the interval of the �rst two injections post-switching. In the group that switched from ranibizumab to a�ibercept (R to A group, n=8), the interval before was 85.4±6.6 days which was signi�cantly increased to 125.9±23.2 days post-switching (P=0.0013). In the A to R group (n=8), a signi�cant shortening from 103.9±11.8 days before to 85.9±12.6 days was observed (P=0.007). In the R to A group, the central retinal thickness at a recurrence was 522.1±79.6 µm which decreased signi�cantly to 419.3±64.5 µm post-switching (P=0.005). The results suggest that the intervals can be different between ranibizumab and a�ibercept in patients with macular edema secondary to BRVO who required repeat injections.


Introduction
The major cause of the decrease of visual function in eyes with a branch retinal vein occlusion (BRVO) is macular edema.The retinal ischemia caused by the blockage induces an increase in the levels of intraocular vascular endothelial growth factor (VEGF) which then enhances the vascular permeability of the blood vessels.This then results in a leakage and macular edema. [1]ti-VEGF agents have been shown to be an effective treatment for the BRVO-induced macular edema. [2,3] A present, two anti-VEGF agents, ranibizumab and a ibercept, are widely used to treat patients with a BRVO.The use of a ibercept for central retinal vein occlusion (CRVO) with macular edema requires fewer injections than ranibizumab suggesting that the duration of its effectiveness may differ from that of ranibizumab. [4]However, a direct comparison of the duration of effectiveness has not been determined for these two agents for BRVO.Relevant to our study, two earlier studies found that the interval of injections was prolonged when ranibizumab was switched to a ibercept in patients with macular edema secondary to BRVO. [5,6] owever, the prolongation of injection interval after switching agents may be in uenced by the time-dependent attenuation of the natural progression of the disease process. [7,8] or this reason, it is important to examine cases in which patients have been switched from a ibercept to ranibizumab.
Thus, the purpose of this study was to compare the interval of anti-VEGF injections after ranibizumab was switched to a ibercept or when a ibercept was switched to ranibizumab in patients with macular edema secondary to BRVO.To accomplish this, we reviewed the medical records of patients who had undergone a switching of ranibizumab to a ibercept (R to A) or from a ibercept to ranibizumab (A to R) on the interval of the injections, and the degree of macular edema.

Methods
Sixteen eyes of 16 patients with macular edema associated with a BRVO were studied.At least 3 intravitreal injections of an anti-VEGF agent, ranibizumab (Lucentis®; Novartis Pharma, Basel, Switzerland) or a ibercept (Eylea®; Bayer Pharma AG, Berlin, Germany) were given for one year, the preswitch period, between July 2015 and December 2018.The patients were then switched to the other type of anti-VEGF agent.The reason for switching agents was because the patient who had an aversion to get repeated same injections accepted our suggestion to change to another type of agent.
The general protocol of our hospital encourages patients to visit their doctor every month at which time optical coherence tomography (OCT; 3D OCT-2000, Topcon) and visual acuity measurements were performed.The OCT images were used to measure the central retinal thickness (CRT), and if the CRT was ≥ 350 µm, a recurrence of the macular edema was considered to have occurred.Then, an anti-VEGF intravitreal injection was given with the pro re nata (PRN) treatment protocol.Patients who had undergone other ocular treatments such as topical steroids, intraocular surgery, and scatter and grid laser photocoagulation at the pre-switch period were excluded.
Retrospective investigations of the medical records of the patients were performed.A comparison between an average injection interval for the pre-switch period and the interval of the rst two injections after switching the anti-VEGF agents was considered the primary endpoint.Comparisons of the CRT before and after switching agents were considered at the time of a recurrence and at the time one month after the injection.

Statistical analyses
The signi cance of the differences in the pre-injection demographic data of the R to A and the A to R groups was determined by Mann-Whitney U tests or Fisher's exact tests.The signi cance of the changes in the injection intervals and CRT before and after switching agents was made using paired t-tests.A Pvalue < 0.05 was taken to be statistically signi cant.

Results
The patients were divided into two groups: the R to A group (4 men and 4 women) and the A to R group (5 men and 3 women).The demographics of the patients in the two groups are summarized in Table 1.There were no signi cant differences in the mean age, male-female ratio, best-corrected visual acuity, CRT, estimated duration from onset, frequency of anti-VEGF injections in the pre-switch period, and previous photocoagulations between the two groups.In the R to A group. the mean injection interval before switching was 85.4 ± 6.6 days which was signi cantly increased to 125.9 ± 23.2 days after the switch (P = 0.0013; Fig. 1a).In the A to R group, there was a signi cant shortening of injection interval from 103.9 ± 11.8 days before to 85.9 ± 12.6 days after switching agents (P = 0.007; Fig. 1b).In the R to A group, the percentage increase in the injection interval before and after switching agents was ± 10% in 1 eye (12.5%), 10-30% in 1 eye (12.5%), and ≥ 30% in 6 eyes (75%).In the A to R group, the percentage shortening of the intervals before and after switching agents was ± 10% in 2 eyes (25%), 10-30% in 5 eyes (62.5%), and ≥ 30% shortening in 1 eye (12.5%;Fig. 1c).
The mean CRTs at one month after the anti-VEGF injections was 228.7 ± 31.8 µm before and 230.4 ± 45.2 µm after the switch in the R to A group (P > 0.05; Fig. 2a) and 263.4 ± 28.0 µm before and 274.5 ± 30.4 µm after the switch in the A to R group (P > 0.05; Fig. 2c).The mean CRTs at a recurrence in the R to A group, was 522.1 ± 79.6 µm before and 419.3 ± 64.5 µm after the switch, and there was a signi cant difference in the degree of macular edema (P = 0.005; Fig. 2b).In contrast, the CRTs was 433.8 ± 79.6 µm and 472.5 ± 75.3 µm at a recurrence in the A to R group, and these ndings showed that the differences in the degree of macular edema was not signi cantly changed (P > 0.05; Fig. 2d).

Discussion
We observed a prolongation of the injection interval after switching to a ibercept and shortening of injection interval after switching to ranibizumab.These ndings suggest that the prolongation and shortening of the injection interval were related to a difference in the duration of effectiveness of the two agents.
In the R to A group, average injection interval before the switch was 85 days, and it increased signi cantly to 126 days after the switch.The increase in the injection intervals was 41 days which is 1.48 times longer than the average injection interval during the pre-switch period.In a prospective study, Casselholm et al [4] compared the injection intervals for ranibizumab and a ibercept in eyes with a CRVO performed with the treat-and-extend protocol.They reported that the average injection interval for ranibizumab was 6.6 week and that for a ibercept was 10.0 weeks.These ndings indicated that the effects of a ibercept lasted 1.5X longer than ranibizumab.Thus, their ndings are comparable to our percentage prolongation of the injection interval in the R to A group.
There have been many studies on the effectiveness of switching from ranibizumab to a ibercept in patients with macular edema secondary to a BRVO.The patients studied can be classi ed into two types; one type is the patient who was not responding to ranibizumab and were then switched to a ibercept, [5,9,10] and the other type was the patient who had a difference in the duration of the effectiveness between two agents as indicated by the lengthening of the interval after switching to a ibercept. [6]The di culty in comparing these results is that there is a spontaneous healing in more than 30% of cases if macular edema secondary to BRVO develops within one year after the onset of BRVO. [11]The RELATE study examined patients who had received ranibizumab treatment with a loading phase of 6 PRN injections for 3 years, and they found that the number of injections required for successful treatment decreased with time; 9.7 times in the rst year, 6.9 times in the second year, and 4.0 times in the third year. [12]Thus, we cannot eliminate the possibility that the prolongation of injection interval after switching to a ibercept was related to a natural attenuation of the disease progression.Therefore, after switching to ranibizumab, a prolongation of injection interval could also be expected.However, to our knowledge, there have been no studies performed to evaluate the change in the injection interval after switching therapy from a ibercept to ranibizumab in patients with BRVO.In our cases, in the A to R group, average injection interval before the switch was 104 days and after the switch to ranibizumab was 86 days.Therefore, the difference between the two injection intervals is 18 days, that is, 1.21 times shorter than the average injection interval after switching to ranibizumab.The results showed no prolongation of injection interval due to a time-dependent attenuation of the natural progression of the disease process.
The reason for the longer effectiveness of a ibercept is because it has different biochemical characteristics from ranibizumab.For example, a ibercept can bind not only to VEGF-A but also to other VEGF family members, such as VEGF-B and placental growth factor (PlGF).In addition, a ibercept has a strong a nity for VEGF-A. [13]The intraocular PlGF levels are elevated in patients with macular edema secondary to BRVO suggesting that it may be involved in the recurrence of macular edema. [14]In animal experiments, Stewart et al. reported that a ibercept had longer binding activity than ranibizumab. [15]In addition, Niwa et al. compared the intraocular kinetics of ranibizumab and a ibercept after intravitreal injections in monkeys. [16]They reported that the half-life of both agents in the aqueous after intravitreal injections was not signi cantly different, but a ibercept suppressed the VEGF level for a longer time.From these results, we can conclude that a ibercept has a longer period of effectiveness.This was true even after the agents decreased below the detection level which suggested that a ibercept has a better potential of preventing the recurrence of macular edema.Saishin et al. examined the effects of intravitreal injections of ranibizumab and a ibercept bimonthly on CRVO patients, and they reported that the CRT measured at the time of recurrence was thicker in the ranibizumab than the a ibercept group. [17]heir ndings are consistent with the results of our study in showing the difference in the CRTs at the time of recurrence in the R to A group.
In conclusion, the injection interval can be different between ranibizumab and a ibercept in patients with macular edema secondary to BRVO.Because the difference of injection interval affects patient's medical cost, economic, or psychological burden, it is necessary to focus on the selection of agents for BRVO treatment as well as relieving the effects of macular edema by the anti-VEGF agents.

Limitations
Our study was limited due to its retrospective nature and small sample size.Another limitation is the different assessment period among pre-switching and after switching agents.On one hand, a one-year assessment period was set before switching agents, but the assessment period after switching agents was not the same as the one before switching agents.This is because patients have a tendency to ask physicians to switch back to the previous one when a shortening of the injection interval occurs due to switching agents.This indicates that the prolongation or shortening of the injection interval becomes a su cient reason for patients to select the agent.

Figures
Figure 1 Injection intervals before and after switching anti-vascular endothelial growth factors (anti-VEGFs).a: Group that switched from ranibizumab to a ibercept (R to A). P-value, paired t test.b: Group that switched a ibercept to ranibizumab (A to R). P-value, paired t test.c: Percentage change of injection interval before and after switching anti-VEGF agents.
Central retinal thickness (CRT) before and after switching anti-VEGF agents.a: In the R to A group, the CRT before and after switching agents at one month after the anti-VEGF injection.b: In the R to A group, the CRT before and after switching agents at a recurrence.c: In the A to R group, the CRT before and after switching agents at one month after the anti-VEGF injection.d: In the A to R group, the CRT before and after switching agents at a recurrence.P-value, paired t test.
Values are the means ± standard deviations.n, number of eyes; BCVA, best-corrected visual acuity; log-MAR, logarithm of the minimum angle of resolution; CRT, central retinal thickness.a Mann-Whitney U test, b Fisher's exact test.

Table 1 .
Demographics of patients