This study examined the therapeutic effects of ranibizumab in patients with myopic CNV. Consecutive monthly intravitreal injections significantly improved BCVA, reduced CMT (measured by OCT), and lowered aqueous humor levels of VEGF-A, VEGF-D, and endoglin while increasing EGF levels. The results are similar to those from a study on the effects of ranibizumab in age-related macular degeneration (AMD) cases [23]. The cytokine concentrations may be quantitative indicators of ranibizumab’s effects in patients with myopic CNV. Ranibizumab significantly affected CNV secondary to pathologic myopia by improving BCVA values, reducing CMT, and modulating cytokines associated with angiogenesis. For different individuals, different treatment plans should be chosen.

The CNV group showed higher plasma concentrations of VEGF-A than the control group, whereas EGF and angiopoietin-2 concentrations were lower in the myopic CNV than in the control group. Moreover, compared to controls, the aqueous humor content of FGF-2 increased, whereas those of EGF and VEGF-A decreased in the eyes of patients with myopic CNV.

Plasma VEGF concentrations are higher in patients with active CNV secondary to AMD than in healthy individuals [24]. VEGF is the most important signaling protein in both vasculogenesis and angiogenesis, and the increase in plasma VEGF is a prerequisite for CNV generation. Therefore, it is not unexpected that a higher plasma concentration of VEGF-A was detected in the CNV than in the control group. However, there is a lack of consensus on whether the VEGF concentration in the aqueous humor is high in myopic CNV. Tong et al. [25] reported that VEGF concentrations in the aqueous humor grew significantly among patients with CNV secondary to AMD, polypoidal choroidal vasculopathy, and CNV secondary to pathologic myopia compared to healthy controls. In contrast, Sawada et al. [26] reported significantly lower VEGF concentrations in the aqueous humor of the myopic CNV group than those in the control group, which is consistent with our results. Vascular endothelial growth factor receptor (VEGFR)-2, which is related to vascular leakage, is mainly found in the retinal blood vessel cells of the leakage area, whereas VEGFR-3 is mainly located in deep retinal capillaries and is associated with VEGFR-2. VEGF-A induces VEGFR-2 and VEGFR-3 expression in the ischemic area, leading to retinal leakage, thereby promoting new blood vessel formation [6]. We hypothesize that VEGF-A may be localized in a small subfoveal area in combination with VEGFR-2 and VEGFR-3, resulting in decreased levels in the aqueous humor during CNV formation. Another possible explanation is that VEGF-A in the anterior chamber and vitreous cavity might be diluted due to the longer axial length and larger intraocular volume in patients with high myopia because of choosing cataract patients with normal axial length as the control group.

Aqueous humor EGF content is elevated in patients with AMD [11]. However, this was not the case in the patients with myopic CNV in this study. The EGF receptor resides primarily in the cell membrane, where it binds with a significant quantity of EGF. It transduces biological signals by invagination pinocytosis, leading to its internalization into the cytoplasm. Once internalized, lysosomes degrade EGF, and the EGF receptor is recycled back to the membrane. The expression of transforming growth factor-α (TGF-α) increases while CNV progresses [27]. TGF-α and EGF are 33–44% homologous in structure; therefore, TGF-α may competitively inhibit the expression of EGF, leading to a decrease in EGF levels in the plasma and aqueous humor of patients with myopic CNV.

Angiopoietin-2 exerts a key effect on vessel maturation, angiogenesis, and vessel regression. Its expression is predominantly observed in tumor tissue (especially in areas of tumor neovascularization) but not in normal tissue [28]. In myopic CNV, angiopoietin-2 and its receptor are abundantly expressed and localized in CNV lesions. Consequently, our comparison with the control group revealed lower concentrations of angiopoietin-2 in the plasma of patients with myopic CNV.

In a healthy retina, the mRNA of FGF-2 and its receptor are found in the ganglion cell layer and kernel, and transient retinal ischemia induces the synthesis of FGF-2 mRNA [29]. In vitro studies have found that the anti-VEGF antibody/anti-FGF-2 antibody complex can completely inactivate VEGF and FGF-2 in the retinal pigment epithelium containing CNV, better inhibiting the growth of vascular endothelial cells than simply adding the anti-VEGF antibody [30]. Unlike VEGF-A, EGF, and angiopoietin-2, FGF-2 is not expressed in retinal capillary cells and is limited to CNV lesions, which may result in higher aqueous humor levels of FGF-2 in patients with myopic CNV than in controls.

This research demonstrated reduced aqueous humor concentrations of VEGF-A, VEGF-D, and endoglin after consecutive intravitreal injections of ranibizumab, an anti-VEGF-A monoclonal antibody [8]. There is little doubt that the concentration of VEGF-A decreases after intravitreal injections [31]. As a result, CNV atrophy lowers the aqueous humor concentrations of the other two cytokines. VEGF-D works in the processes of angiogenesis and lymphangiogenesis. As there are no lymphatic vessels in the eyes, the VEGF-D expressed by the retinal pigment epithelium is related to ocular angiogenesis [32]. VEGF-D can also upregulate VEGF-A expression and promote angiogenesis [33]. Previous studies have shown that endoglin has strong immunogenicity in the endothelial cells of CNV membranes [13]. However, in this study, the aqueous humor contents of EGF increased after consecutive intravitreal injections of ranibizumab. CNV atrophy led to a decrease in TGF-α contents, weakening the inhibition of EGF. As a result, the aqueous humor content of EGF increased.

This study had some limitations. First, relatively few participants were recruited in the study. Although the outcomes were statistically significant for some detected cytokines, the molecular mechanisms involved should be explored further. Second, because of the restricted volume of the samples, not all cytokines that may be related to myopic CNV were detected. There are still a few molecules potentially associated with the disease that were not examined, such as TGF-β [34], matrix metalloproteinase [35], tumor necrosis factor [36], and pigment epithelium-derived factor [25]. Finally, we selected patients with cataract as the control group, so the refractive errors of the two groups were not equivalent. This was an oversight in the design of the experiment, which might affect the results.