The present study demonstrated that the vitreous level of VEGF was significantly higher in BRVO patients who showed less improvement of their best-corrected visual acuity after PPV (Figure 1). As a result, the VEGF level had a significant influence on the visual outcome according to multivariate analysis. These findings suggest that measurement of VEGF in the vitreous fluid may be useful for predicting the visual prognosis after PPV. In contrast to our findings, Shimura et al. have reported that the visual prognosis is not correlated with the vitreous level of VEGF. However, they did not perform retinal photocoagulation, while this was done in the present study. Retinal photocoagulation has been shown to induce the expression of various cytokines, such as interleukin (IL)-1. β, IL-6, IL-8, and VEGF, [13, 14] suggesting that the effect of photocoagulation on the production of VEGF may explain the difference between their findings and ours.
The Branch Vein Occlusion Study was a multicenter randomized clinical trial that established guidelines for use of retinal photocoagulation in the treatment of macular edema. The effectiveness of argon laser photocoagulation was demonstrated, but it was not recommended within 3 months of the onset of BRVO because spontaneous improvement can occur. However, some patients have poor visual acuity and persistent macular edema despite photocoagulation. Recently, PPV combined with posterior vitreous detachment has been reported to effectively reduce macular edema and improve visual acuity in BRVO patients[16, 17]. Therefore, we performed vitrectomy more than 3 months after the onset of BRVO in patients who had clinically detectable diffuse or cystoid macular edema, or who had persistent macular edema after photocoagulation. We also performed cataract surgery in 29 patients, because our subjects were relatively old with an average age of 63.1 years.
The macula is important for detailed vision, especially the fovea that consists entirely of cones. In humans, histological studies have shown that macular edema is associated with swelling of the Müller cells, especially in the outer plexiform layer of the neurosensory retina [19–21]. Accordingly, macular edema may affect retinal function and lead to visual impairment in BRVO patients.
The present study demonstrated that a higher VEGF level in the vitreous fluid at the time of surgery was significantly associated with more marked improvement of macular edema after PPV, which was in agreement with our findings in a previous study of PPV for BRVO. The vitreous level of VEGF was also significantly correlated with the improvement of BCVA and the decrease of foveal thickness. Intravitreal injection of VEGF has been reported to cause retinal edema, dilated and tortuous vessels, and capillary closure in adult primates,  while treatment with bevacizumab (a monoclonal antibody targeting VEGF) or ranibizumab (an Fab fragment that binds and neutralizes all isoforms of VEGF-A) improves macular edema in patients with BRVO [23–25]. Therefore, reduction of the intraocular VEGF level may be one of the mechanisms by which PPV improves macular edema in BRVO patients. In agreement with this hypothesis, we previously found that the vitreous level of VEGF was lower at the time of repeat vitrectomy in patients with macular edema due to retinal vein occlusion.
Interestingly, the improvement of visual acuity and the improvement of macular edema did not occur in parallel. Although the retinal thickness at the central fovea has been reported to influence visual acuity, [27, 28] we found that acuity sometimes showed little improvement even if macular edema resolved. This may have been because 1) improvement of visual acuity required longer than the 6-month follow-up period of this study or 2) improvement of vision failed to occur after edema resolved because of permanent photoreceptor cell damage due to macular ischemia in patients with high VEGF levels. Foveal bleeding could also influence the visual prognosis. Accordingly, a larger prospective and randomized study is needed to clarify the relation between visual acuity and macular edema in patients with BRVO.
In the present study, multivariate analysis showed that the vitreous VEGF level and retinal photocoagulation were significantly correlated with improvement of visual acuity and with the decrease of foveal thickness. Aiello et al. previously reported that the vitreous level of VEGF was reduced by retinal photocoagulation, and our findings would suggest that the decrease of VEGF due to retinal photocoagulation results in improvement of BCVA and a decrease of foveal thickness. However, the relations between retinal photocoagulation, the vitreous level of VEGF, improvement of visual acuity, and improvement of macular edema in patients with BRVO need to be investigated further.
The present study had several limitations. One major problem was the performance of laser photocoagulation prior to surgery. For the following reasons, we performed a combined analysis of the groups with (n = 26) and without (n = 28) laser photocoagulation. The first reason is that there was no significant difference of the VEGF level between the 26 eyes that received retinal photocoagulation and the 28 eyes without it (data not shown, p = 0.6755). Secondly, laser photocoagulation increases the expression of various cytokines, including VEGF, by cultured human retinal pigment epithelial (RPE) cells, but upregulation of VEGF occurs from 6 hours after photocoagulation and there is a return to the basal level by 72 hours. In contrast, Itaya et al. found the maximum VEGF level on day 3 in vivo, coinciding with the peak of macrophage infiltration. They suggested that the difference from in vitro data arose because infiltrating macrophages contributed more to upregulation of VEGF than RPE cells or because secretion of VEGF was induced by the interaction of macrophages with RPE cells. Furthermore, changes of retinal VEGF mRNA expression after laser photocoagulation were confined to RPE cells in miniature pigs, with reduced mRNA expression immediately after photocoagulation and a return to normal by 42 days. These results suggest that any increase of VEGF expression in the retina after photocoagulation is transient, and that the VEGF level decreases again relatively fast and then becomes stable. Because the average interval was 2.3 ± 1.5 months from laser photocoagulation to vitrectomy, photocoagulation is unlikely to have influenced VEGF levels. Another limitation of this study was the performance of cataract extraction at the time of vitrectomy. However, there were no significant differences of VEGF levels, improvement of macular edema, or improvement of visual acuity between the patients undergoing cataract surgery and those who did not. Multiple regression analysis demonstrated that cataract surgery was not correlated with the improvement of macular edema or the improvement of visual acuity. Therefore, cataract surgery did not influence our results. Further investigations will be needed to clarify the relation among the vitreous level of VEGF, improvement of macular edema, and improvement of visual acuity, as well as the possible effects of laser photocoagulation and cataract surgery.