Introduction

Branch retinal vein occlusion (BRVO) is a common vascular retinal disorder [1]. Loss of vision due to BRVO is usually caused by macular edema (ME) [2]. Various treatments have been used, such as anti-vascular endothelial growth factor (VEGF) therapy [3–5], intravitreal administration of steroids [6, 7], laser treatment [8], and surgical procedures [9, 10]. Recently, anti-VEGF therapy has become widely used for treating ME associated with BRVO [3]. However, treatment success is often temporary or ineffective despite multiple intravitreal injections. To our knowledge, only a few studies deal with ME recurrence associated with BRVO after anti-VEGF therapy [11, 12], and the predictive factors are still unclear. The purpose of this study was to investigate predictive factors of IVB therapy for recurrent ME associated with BRVO.

Methods

All procedures conformed to the tenets of the Declaration of Helsinki, and the study design was approved by the Institutional Review Board of Kobe City Medical Center General Hospital. The review board waived the need for written informed consent because the study design consisted of a retrospective chart review. We retrospectively examined consecutive nonrandomized patients who underwent 1.25 mg of intravitreal injection of bevacizumab (IVB) from 2009 to 2013 as a primary treatment for ME associated with BRVO and who were followed monthly for at least 5 months. Eyes that had undergone pars plana vitrectomy prior to IVB for BRVO were excluded. Eyes with previous scatter photocoagulation (PC) were included, but those with previous grid-pattern PC were excluded. Best-corrected visual acuity (BCVA) was measured using the Landolt ring and converted to the logarithm of the minimal angle of resolution (logMAR). Fluorescein angiography was performed, and eyes with retinal nonperfusion over five disc areas were diagnosed as having ischemic BRVO. Optical coherence tomography (OCT) was performed to measure central retinal thickness (CRT) using 3D OCT-1000 (Topcon, Tokyo, Japan) or Spectralis Heidelberg retina angiograph plus OCT (Heidelberg Engineering, Dossenheim, Germany); 3D OCT-1000 images were obtained using TrueMap™ (Version 2.12, Topcon, Japan) software. Spectralis OCT images were obtained using Spectralis Family Acquisition Module (Version 4.0.2.0, Heidelberg, Germany) and Heidelberg Eye Explorer (Version 1.6.1.0, Heidelberg, Germany) software. Horizontal and vertical scans of the macula were recorded for diseased eyes. Measurements were performed under pupillary dilation. The eye-tracking system of the device was used to assure that scans were performed in the correct position.

Based on postoperative CRT, patients were classified into two groups: an effective group if CRT decreased to ≤250 µm within postoperative 3 months, and an ineffective group if CRT remained >250 µm throughout the first 3 months (Fig. 1). The effective group was then divided into two subgroups: a recurrent group if ME was once resolved to ≤250 µm but recurred again, and a nonrecurrent group if ME resolution was maintained throughout the follow-up period without additional injections (Fig. 1). We examined any differences between groups in age, gender, estimated elapsed time from disease onset to IVB, baseline BCVA, baseline CRT, and follow-up period after IVB therapy.

Fig. 1
figure 1

Optical coherence tomography of representative cases in each group. Effective (nonrecurrent) (a–c), effective (recurrent) (d–f), and ineffective (g–i) groups at baseline (a, d, g), 0–3 months after IVB (b, e, h), and 3–6 months after IVB (c, f, i). IVB intravitreal bevacizumab injection

Full size image

SPSS statistics 22 for Windows (SPSS Inc., Chicago, IL, USA) was used for the statistical analysis. Welch’s t test was used to compare continuous variables between the groups. Fisher’s exact test was applied for categorical variables. The level of significance was set at P < 0.05 for all statistical tests. Continuous data are expressed as mean ± standard deviation (SD).

Results

In total, 52 eyes of 52 Japanese patients were evaluated. There were 27 men and 25 women, and average patient age was 68.3 ± 11.4 years. The major branch of the retinal vein was occluded in 28 eyes and the macula branch in 24. Fluorescein angiography showed that 16 eyes were ischemic and 31 were nonischemic; angiography was not performed in five eyes. Five eyes had undergone scatter PC prior to initial IVB. The effective group consisted of 37 eyes (71 %) and the ineffective group of 15 (29 %). Within the effective group, the nonrecurrent subgroup consisted of 11 eyes (30 %) and the recurrent subgroup of 26 (70 %). The first recurrence of ME in the recurrent group occurred within 5 months after primary IVB in all cases. There was no significant difference between effective and ineffective groups in age, gender, occlusion site, retinal ischemia, history of scatter PC prior to IVB, estimated elapsed time from disease onset to IVB, baseline BCVA, baseline CRT, and follow-up period (Table 1). However, elapsed time for the nonrecurrent group was significantly shorter than for the recurrent group (15.7 ± 8.9 vs. 29.7 ± 29.5 weeks, respectively; P = 0.036); there was no significant difference in other factors (Table 2). A scatter diagram of the elapsed time in each group is shown in Fig. 2. There were no vision-threatening complications in this study.

Table 1 Characteristics of effective and ineffective groups
Full size table
Table 2 Characteristics of nonrecurrent and recurrent groups
Full size table
Fig. 2
figure 2

Scatter diagram showing elapsed time of each patient. a Distribution of elapsed times overall. Error bars standard deviation. b Enlargement of the region of short elapsed time (≤50 weeks)

Full size image

Discussion

Several studies suggest macular ischemia [13], baseline BCVA [4], age [4], duration from onset to primary IVB [4, 14], baseline CRT [4], and vitreomacular adhesion [15] as predictive factors for BRVO prognosis. These varying data may have resulted from studies with different designs, such as the number of eyes (38–205), IVB dose (1.25–2.0 mg), follow-up interval (4–12 weeks), final follow-up time (3–12 months), definition of recurrence (CRT >250 μm or >30 % after primary IVB), patient history/risk factors of BRVO (hypertension, hyperlipidemia, diabetes mellitus, thrombophilia, hypercoagulation, systemic and inflammatory diseases, medications, and ocular conditions) [4, 16–18], and exclusion criteria. In the study we report here, we found no significant differences in age, gender, estimated elapsed time from disease onset to IVB, preoperative visual acuity, and preoperative CRT between effective and ineffective groups. However, within the effective group, elapsed time in the nonrecurrent group was significantly shorter than in the recurrent group, suggesting that duration from symptom onset to initial IVB could affect the ME recurrence rate but not the efficacy rate after single IVB injection. To our knowledge, there are only a few studies dealing with ME recurrence associated with BRVO after anti-VEGF therapy. Yasuda et al. report that in 62 IVB-effective eyes with BRVO, 21 (34 %) showed no ME recurrence [12]. This rate was consistent with our results showing that within the effective group, 11 of 37 eyes (30 %) showed no recurrence. Those authors also report that all seven eyes with “rebound” ME had undergone initial IVB treatment within 8 weeks from symptom onset, suggesting that the rebound was more likely to occur when IVB therapy was initiated at a relatively early stage of ME, before edema had reached the maximum degree [12]. They defined ME rebound when the ratio of foveal thickness and recurrence/foveal thickness at baseline became ≥110 % after an initial decrease of foveal thickness [12]. This definition of rebound is quite different from our definition of recurrence; therefore, there is no discrepancy between their report and our finding that recurrence also occurred when the initial IVB was undergone after 8 weeks from symptom onset (Fig. 2). Hanada et al. report that, in a total of 95 IVB treatment sessions conducted on 37 eyes of 37 patients with BRVO, 25 injections were given, with a follow-up period of ≥6 months without additional treatment [11]. In their study, the probability of retreatment with IVB was ~70 % after each individual injection, regardless of the number of injections. However, predictive factors for retreatment were not described [11]. Our study showed that early IVB treatment may suppress ME recurrence associated with BRVO. Although it is possible that cases with spontaneous ME resolution may have been masked by early IVB treatment, it is unlikely to be the sole reason for the shorter elapsed time in the nonrecurrent group, because the elapsed times between the effective and ineffective groups were not significantly different. The mechanism of how early IVB treatment suppresses ME recurrence is unclear; however, a previous study reports that anti-VEGF therapy for retinal vein occlusion reduced the occurrence of retinal nonperfusion and could eliminate a positive feedback loop in which retinal ischemia results in high levels of VEGF that promote worsening of retinal ischemia [19]. It is suggested that early anti-VEGF therapy can decrease the level of VEGF in the acute phase of the disease and interrupt the positive feedback loop of VEGF, suppressing ME recurrence. It is also suggested that PC downregulates VEGF expression in the ischemic retina [20, 21]. In eyes previously treated with PC, VEGF levels might not be high but continue to be expressed over a considerable period; or VEGF might not play a major role in prolonged ME. It is possible that ME is more likely to recur after single IVB in these eyes. Indeed, four of 26 eyes in the recurrent group but none of 11 eyes in the nonrecurrent group had undergone scatter PC prior to IVB (Table 2). However, the number of eyes previously treated with PC was too small to evaluate its role in ME recurrence.

An intravitreal anti-VEGF injection has a low probability of vision-threatening complications. It is reported that rhegmatogenous retinal detachment (RD) was found in 0.013–0.02 % and endophthalmitis in 0.02–0.095 % of eyes after anti-VEGF injections [22–28]. Although the complication rate after a single injection was very low, repeated injections could lead to an increase in the risk of complications, such as RD and endophthalmitis.

The main limitations of this study were the retrospective design, the small number of patients, and short follow-up period. Also, the nonperfusion area must be evaluated, because it is related with VEGF and the prognosis of visual acuity [29]. To investigate predictive factors for BRVO prognosis more thoroughly in the future, a randomized prospective study, including an anatomical and physiological analysis, is recommended.

In conclusion, early IVB treatment against BRVO may suppress ME recurrence and reduce the number of IVB injections, minimizing both the risk of vision-threatening complications and the economic burden to the patient.