The VEGF family includes five protein members: VEGF-A, VEGF-B, VEGF-C, VEGF-D and placental growth factor (PlGF) (11,12). Among them, the main role of VEGF-A is to induce endothelial cell activation and increase vascular permeability by binding to its corresponding receptors (VEGFR1 and VEGFR2) (13). VEGF-A has several isoforms, such as VEGF111, VEGF121, VEGF145, VEGF165, VEGF189 and VEGF206. For example, VEGF165 was the first isoform to be identified and is considered to be the classic pro-angiogenic VEGF-A subtype (14). VEGF-B has potential for the treatment of coronary artery disease and heart failure. Different from other proangiogenic factors of the VEGF family, in the heart, VEGF-B has the highest expression in cardiomyocytes and is involved in cardiac remodeling after myocardial infarction. Thus, VEGF-B regulates myocardial contraction and metabolism, and VEGF-B exerts cardioprotective effects, protecting cardiomyocytes from ischemia through physiological hypertrophy, Specifically, VEGF-B may activate the downstream akt/mtorc1 pathway to mediate physiological hypertrophy (15). VEGF-C has structural homology with VEGF-D and serves a major role in lymphangiogenesis (16). PLGF can bind to VEGFR1, thereby preventing VEGF from binding to VEGFR2. It is also a factor that promotes abnormal angiogenesis in the retina and subretina (11).

VEGFR2 is a tyrosine kinase receptor that consists mainly of an extracellular, a transmembrane and an intracellular structural domain (17). The extracellular domain has seven immunoglobulin homologous structural domain repeats (18). The intracellular domain consists of a kinase structural domain including several tyrosine residues, and VEGF binding to VEGFR2 leads to the phosphorylation of the tyrosine residues and activation of the kinase domain. The key phosphorylation sites include Y1054 and Y1059, whereas the coreceptor neuropilin-1 (NRP-1) is involved in developmental angiogenesis by binding to VEGFR2 (19-21). In addition, phosphorylation of VEGFR2 can activate signaling pathways, such as those promoting endocytosis of the key EC adhesion molecule vascular endothelial (VE)-cadherin, leading to increased vascular permeability (22). The phosphorylation site Y949 of VEGFR2 is targeted to limit vascular permeability in retinopathy through downstream signaling. When the VEGFR2 Y949 signaling pathway is impaired, phosphorylation of the vascular endothelial VE-cadherin Y685 site is reduced. These results indicate that targeting VEGFR2-regulated VE-cadherin phosphorylation may inhibit edema (23). VEGFR1 and VEGFR3 also contain the same three domains as VEGFR2, but their biological functions are different. VEGFR1 can act as a decoy receptor to limit VEGFA/VEGFR2 activity in the physiological environment, while VEGFR3 mainly binds VEGF-C/VEGF-D to regulate the formation of lymphatic vessels (4). A graphical representation of VEGF and its receptors is illustrated in Fig. 1.

VEGF and its receptor VEGFR play a role not only in normal physiological blood vessel growth but also in pathological angiogenesis, tumor growth and metastasis (24,25). VEGF usually regulates angiogenesis by binding to VEGFR1 or VEGFR2. Notably, VEGF mainly binds to VEGFR2, which has more potent proangiogenic activity and higher tyrosine kinase activity than VEGFR1(26), whereas VEGFR1 is more commonly involved in inflammatory pathways (27).

VEGF binds to VEGFR2 and activates downstream signaling, including the PI3K, Akt, Ras and MAPK pathways, to promote cell proliferation, migration and differentiation (28). Among them, proliferation and migration occur in ECs, thereby regulating vascular patterns by controlling δ-notch signaling (29). VEGF also protects neurons from ischemia via VEGFR2(30), and VEGFR2 deficiency results in abnormal neuronal angiogenesis (31).

Retinal neovascularization occurs in a group of ischemic retinopathies resulting from damage to the retinal vasculature, among which DR and retinal vein occlusion (RVO) are the most common. Subretinal or choroidal neovascularization (CNV) occurs in diseases involving the outer retina or Bruch's membrane, and nvAMD is the most common type of such neovascularization (32).

Overview of DR. DR is a tissue-specific neurovascular complication caused by type 1 and type 2 diabetes mellitus. It is estimated that 440 million individuals will have diabetes mellitus by 2030, and DR affects ~29% of diabetic patients (33). The two main categories are the PDR and non-PDR types. The pathophysiological changes in DR include several aspects of neurodegeneration, inflammation and oxidative stress. Current treatments for DR, including anti-VEGF therapy, steroids, laser photocoagulation and vitrectomy, have limitations and side effects such as visual acuity and visual field damage, and new therapeutic strategies need to be explored (34). It has been shown that only laser photocoagulation and anti-VEGF injections are effective treatments in cases of severe retinopathy and that traditional Chinese medicine may be promising in reducing VEGF levels, inflammation, oxidative stress, apoptosis and angiogenesis in DR (35). For example, control of late glycosylation by intraocular injection of genipin may be a strategy to prevent retinopathy (36).

Overview of RVO. RVO is one of the most common retinal vascular diseases (37). Obstruction of the main retinal veins is known as central RVO (CRVO), and obstruction of the smaller veins is known as branch RVO (BRVO). The prevalence rates of CRVO and BRVO are 0.1-0.4 and 0.6-1.2%, respectively, worldwide (38). The pathogenesis of RVO is multifactorial and involves complex interactions between multiple vascular and inflammatory mediators. VEGF is a potent mediator of vascular permeability and inflammation and plays a central role in the pathogenesis of RVO; additionally, several cytokines, including IL-6, IL-8 and C-C motif chemokine 2 (CCL2) (39,40), have been reported to be involved.

nvAMD. nvAMD is an important cause of vision loss in elderly individuals, the worldwide prevalence of the disease is 8.7% (41) and it mainly affects the deeper retinal layers of the macula and the surrounding choroidal system. nvAMD pathogenesis is closely related to age, oxidative stress and lipid metabolism. The molecular pathways of macular neovascularization include angiogenesis and arteriogenesis. For example, platelet-derived growth factor (PDGF), angiopoietin (Ang)-1 and Ang-2 play crucial roles in regulating angiogenesis and influencing vascular growth, maturation and stabilization, and VEGF is an important factor in the development of CNV and retinal leakage (42,43).

Key pathological processes in DR include impaired capillary perfusion and subsequent tissue ischemia; these processes can lead to retinal microinfarcts and collateral vessel formation, in which VEGF-A promotes neovascularization by activating VEGFR2 (44,45). Unlike physiological processes where vascular growth is balanced by anti-angiogenic factors, in the pathological process of DR, pathological vascular structural disturbances caused by pericyte loss are associated with high VEGF expression (46). VEGFR2 is highly expressed in diabetic microvessels; therefore, it may be used as a biomarker for the early diagnosis of DR. For example, a recent study revealed that the application of VEGFR2 nanoprobes for in vivo detection may have translational potential in the early diagnosis of DR (47).

nvAMD. nvAMD is characterized by CNV, in which the VEGF/VEGFR2 pathway is involved. Choroidal ECs secrete CCL2, which attracts macrophages to CNV lesions and promotes macrophage M1 polarization; additionally, M1-type macrophages/microglia secrete IL-6 and TNF-α, which synergize with VEGF to promote pathological angiogenesis and contribute to CNV development (48,49). Although intravitreal injection of anti-VEGF drugs has become the first-line treatment for nvAMD, this treatment has a number of drawbacks, including the need for repeated injections, poor or no response in some patients and complications such as retinal fibrosis (41) Therefore, in addition to the traditional VEGF/VEGFR2 pathway-targeted therapy, VEGF-C and VEGF-D signaling also plays key roles. Hypoxia-induced retinal VEGF-C expression induces pathological retinal neovascularization to a similar extent as VEGF-A, and VEGF-D has been shown to promote angiogenesis in corneal angiogenesis models (50).

Neovascularization begins with stimulation by hypoxia-inducible factor (HIF). Under normoxia, HIF-1α is unstable and rapidly degraded, whereas under hypoxic conditions, HIF-1α dimerizes with HIF-1β, which translocates to the nucleus and activates transcription of the target gene VEGF, leading to the production of VEGF by stromal cells; VEGF then diffuses into pre-existing blood vessels before binding to VEGFR2(51). The binding of VEGF to VEGFR2 results in disruption of the basement membrane of ECs, which are transformed into tip cells with high migratory potential. When tip cells form, the activation of VEGFR2 receptors follows VEGF gradients or other proangiogenic stimuli to direct neovascularization (52,53).

The ERK/MAPK mechanistic signaling pathway involves a variety of kinases, and ERK1/2 are members of the classical MAPK cascade (54). This pathway is associated with a variety of diseases, including several types of cancer and cardiovascular disease (55,56). Moreover, the ERK/MAPK pathway is among the main targets of VEGFR2 activation. The binding of VEGF-A to VEGFR2 phosphorylates phospholipase C-γ (PLC-γ) and activates the ERK/MAPK pathway; subsequently, adipose mesenchymal stem cells differentiate into ECs to promote neointima formation (13,57).

Fucoxanthin may play an important role in the prevention of neovascularization. Fucoxanthin interacts with VEGF, thereby impairing the ability of VEGF to activate VEGFR2 and its associated downstream signaling, namely the phosphorylation of MEK and ERK; thus, fucoxanthin acts as an anti-angiogenic agent (58). For specific fundus neovascularization diseases, such as nvAMD, the search for new therapeutic targets is particularly important, since anti-VEGF treatment still results in poor end-stage visual acuity. Zhou et al (59) reported that blockade of the VEGF signaling pathway in human retinal pigment epithelial (RPE) cells increased IL-8 secretion through the MEK/ERK1/2 axis, whereas overactivation of the VEGF pathway decreased IL-8 production. IL-8 expression is upregulated in hRPE cells after VEGF signaling inhibition, which is associated with nvAMD. Therefore, IL-8 could serve as an alternative therapeutic target for nvAMD (59). A recent study revealed that elevated levels of YKL-40 and VEGF and activation of the ERK pathway were observed in the neural retina and RPE/choroidal tissues of laser-induced CNV mice. Moreover, after intravitreal injection of the anti-YKL-40 antibody, the levels of YKL-40 and phosphorylated proteins in the ERK pathway decreased; these results indicated that the anti-YKL-40 antibody inhibited the activation of the ERK pathway. These results have indicated that YKL-40 could be a new target for the diagnosis and treatment of CNV (60). The association between the ERK/MAPK pathway and VEGFR2 activation indicates an unprecedented role for this receptor in response to stimulation by various mechanical factors. Specifically, ERK activation is related to the phosphorylation of VEGFR-2 at Y1175 and Y1214(61). Although the MAPK pathway is prevalent in numerous diseases, the specificity of VEGFR2 as a mechanoreceptor could provide unique targets and concepts for the angiogenic process.

c-Src pathway. c-Src is a cytoplasmic tyrosine kinase associated with cell or endosomal membranes, which also plays a crucial role in regulating VEGF signaling. c-Src is similarly associated with a variety of diseases, including cancer and cognitive disorders (62,63). c-Src directly interacts with VEGFR2 in response to VEGF stimulation, thereby acting as a mechanotransduction protein downstream of VEGFR-2 signaling (64). c-Src serves a role in several cellular processes, including adhesion, motility, proliferation and differentiation (65). In addition, c-Src activity is regulated, at least in part, by mechanical factors such as shear stress and matrix adhesion. c-Src lies upstream of other effectors, such as the ERK pathway, and has been shown to phosphorylate tyrosine residues on VEGFR2, which further activates both VEGFR2 and c-Src by creating a positive feedback loop (66). A previous study suggested that in the Akimba model of diabetic retinopathy, the inhibition of c-Src family kinases with highly specific inhibitors could be an attractive therapeutic intervention for retinal vasculopathy (67). Another study further demonstrated the potential relationship between VEGFR2 and c-Src. This study indicated that PLC-γ is induced downstream of VEGFR2 phosphorylation at Y1173 (pY1173). The Y1173/PLC-γ/endothelial nitric oxide synthase (eNOS)/c-Src pathway was examined in both the healthy and tumor vasculature of VEGFR2Y1173F/+ mice; this pathway was associated with reduced PLC-γ and eNOS activation and suppressed vascular leakage. The PLC-γ pathway downstream of VEGFR2 py1173 can couple VEGFR2 to c-Src. High PLC-γ expression is associated with angiogenic activity and poor prognosis (68). Overall, c-Src acts as a mechanotransduction factor to regulate VEGFR2 activation, which subsequently affects angiogenesis and could be a target for future studies.

PI3K/Akt pathway. The activation of the PI3K/Akt pathway stimulates EC proliferation, mainly by recruiting Akt to the cell membrane and inhibiting apoptosis; however, the activation of PLC-γ leads to the modulation of the intracellular calcium concentration or the production of eNOS through the activation of nuclear factor of activated T cells. This ultimately increases vascular permeability and affects EC proliferation and survival through the aforementioned processes (69,70). The molecular chaperone heat shock protein 90 (HSP90) is a promising molecular target, and HSP90 inhibitors have been indicated to induce anti-angiogenic effects by affecting the PI3K/Akt/eNOS signaling pathway in ECs and downregulating VEGFR2 expression (71). In another study, a peptide (VGB3) mimicked the interaction between VEGF-B and VEGFR1, and binding to VEGFR2 abolished the PI3K/AKT/mTOR pathway and inhibited the proliferation and tube formation of human umbilical vein ECs (HUVECs) (72). Thus, inhibition of the PI3K/Akt axis by inhibitors of VEGFR2 could reduce angiogenesis, and these inhibitors are expected to be applied to further models of neovascular disease in the future.

Rho GTPase family pathway. The Rho family of small GTPases are molecular switches capable of converting extrinsic stimuli into cytoskeletal rearrangements. In vascular ECs, Cdc42, Rac1 and RhoA control cell migration and cell-cell junction formation downstream of angiogenic and inflammatory cytokines. For example, in cell migration, Cdc42 leads to filopodia formation by promoting actin linear extension. Rac1 promotes lamellipodia formation through the WAVE2 complex and arp2/3. RhoA regulates vascularization and permeability by promoting actinomyosin contraction and serves an important role in EC migration and vascular stabilization (73,74).

During angiogenesis, VEGF attracts ECs, whereas semaphorin 3E (Sema3E) repels them. The small GTPase RhoJ plays a multifaceted role in this EC-directed migration. In the GTP-bound state, RhoJ interacts with the cytoplasmic structural domain of plexin D1. RhoJ released from plexin D1 induces cell contraction upon stimulation with Sema3E; this further mediates the Sema3E-induced binding of plexin D1 to VEGFR2, transphosphorylation of VEGFR2 at Y1214 and activation of p38, leading to reverse EC migration. Upon stimulation with VEGF-A, RhoJ promotes the formation of an all-receptor complex consisting of VEGFR2, plexin D1 and NRP-1, which prevents the degradation of internalized VEGFR2 and maintains intracellular signaling. Upon conversion to the GDP-bound state, RhoJ is diverted from plexin D1 to VEGFR2, which terminates VEGFR2 signaling. As confirmed in an oxygen induced retinopathy mouse model, RhoJ deficiency in the ECs of ischemic retina effectively inhibits abnormal angiogenesis (75).

In addition, an increase in RhoA activity can potentially cause anti-angiogenic effects, as shown by Hauke et al (76). Hauke et al (76) reported that, compared with dominant-negative RhoA, active and wild-type RhoA markedly inhibited EC proliferation, migration, tube formation and vascular sprouting in vitro. In addition, active RhoA reduced HUVEC-associated angiogenesis in vivo. These results indicated that RhoA itself may have an anti-angiogenic effect rather than being dependent on the RhoA/Rho-associated coiled-coil kinase axis for its action. In another study, Katari et al (77) focused on the Rho/Yes-associated protein (YAP)/VEGFR2 mechanotransductional pathway regulating the transient receptor potential cation channel subfamily V member 4 (TRPV4) activity in the tumor microenvironment, and identified endothelial TRPV4 as a new alternative therapeutic target for tumor angiogenesis; however, its role in fundus neovascularization disease is unknown, and more studies are needed to confirm its role in vascular normalization.

Based on the above studies, Rho GTPases could be considered potential targets for the treatment of abnormal angiogenesis and hyperpermeability in retinal vascular diseases.

STAT pathway. STAT family members are activated downstream of VEGFR2 and are particularly important for EC survival and proliferation. Ramshekar et al (78) explored retinal endothelial STAT3 as a downstream effector of VEGF-triggered signaling and examined its ability to promote the development of vascularization in the vitreous body and delay the expansion of intraretinal vascularization. The study showed that subretinal delivery of lentiviral vectors driven by a Cdh5 promoter-expressing short hairpin RNA targeting STAT3 reduced vitreous vascularization but did not prolong intraretinal vascularization; these results indicated that VEGF-triggered activation of STAT3 in the retinal endothelium was needed for intravitreal vasculature formation, but not for intraretinal vascularization. In addition, Yu et al (79) identified apatinib as a potential target for alleviating neovascularization and fibrosis in nvAMD; its mechanism of action was related to the inhibition of VEGFR2 activation, which prevented angiogenesis and fibrosis through the downregulation of STAT3 phosphorylation. In summary, the STAT pathway plays a key role in inhibiting neovascularization, and the discovery of new clinical inhibitors can aid in the treatment of retinal and choroidal vascular diseases. The downstream mediators of the VEGFR2 signaling pathway are illustrated in Fig. 2.

Anti-VEGF therapy plays a crucial role in blocking pathological angiogenesis. However, a large body of clinical data indicates that VEGF-targeted therapies may be limited by decreased visual acuity after repeated administration, with a loss of efficacy after the initial response (80). Patients with disease recurrence may develop resistance to anti-VEGF therapy through unknown mechanisms. Therefore, a single anti-VEGF therapy does not adequately prevent intraocular neovascularization, and other new molecular pathways need to be targeted to further improve or effectively complement the anti-VEGF effect. The following is an overview of the main current therapeutic agents.

Bevacizumab. Bevacizumab (Avastin^®) is a humanized anti-VEGF antibody developed by Genentech, Inc., which inhibits VEGF signaling by binding to VEGF-A, VEGF-C and VEGF-D and preventing their interaction with VEGFR2(81). Bevacizumab has been reported to be effective in treating nvAMD when it is administered intravitreally. Bevacizumab is also less costly than other drugs, such as ranibizumab, pegaptanib and aflibercept, and is effective in the long-term treatment of nvAMD (82,83). However, patients whose eyes were treated with aflibercept were almost three times more likely to stop treatment after 1 year than patients whose eyes were treated with bevacizumab (43 vs. 15%). These results reveal the superiority of aflibercept over bevacizumab and have important clinical implications for treatment selection in patients with nvAMD (84). Therefore, more prospective studies focusing on the comparative impact and efficacy of anti-VEGF therapeutics on neovascularization diseases are needed.

Despite some of the clinical benefits of anti-VEGF therapy, monthly injections may lead to numerous ocular complications, including hemorrhage, infection and endophthalmitis. Reddy et al (85) determined the efficacy of mesenchymal stem cell-derived small extracellular vesicles (MSC-sEVs) loaded with the anti-VEGF drug bevacizumab by measuring the efficacy of this drug in a rat model, and the results indicated that bevacizumab-loaded MSC-sEVs reduced the frequency of intravitreal injections needed to treat DR. The reduction in the frequency of injections may reduce ocular complications and improve patient compliance, providing the possibility of improved patient treatment.

Ranibizumab. Ranibizumab (sold under the brand name Lucentis^® and developed by Genentech, Inc./Novartis AG) is a recombinant humanized monoclonal fragment antigen-binding (Fab) region that targets VEGF. Like bevacizumab, ranibizumab prevents the activation of VEGFR1 and VEGFR2 by binding with high affinity to all VEGF isoforms, thereby inhibiting angiogenesis (86).

For different diseases, ranibizumab and aflibercept have different effects. Because the differences between the effects of aflibercept and ranibizumab on the choroid of patients with BRVO-macular edema (ME) are not known, Kishishita et al (87) included 36 patients with BRVO-ME who were treated with intravitreal injections of aflibercept or ranibizumab and observed changes in the choroidal thickness of the subcentral concave area for a follow-up period of 12 months or longer. The results revealed that the effects of ranibizumab and aflibercept on choroidal thickness in patients with BRVO-ME were the same. However, in another study, ranibizumab provided a complementary solution to the poor outcome of aflibercept treatment. In patients with nvAMD who were >50 years old and exhibited an inadequate response to aflibercept, treatment with ranibizumab for 6 months improved visual acuity (88). To investigate the advantages and disadvantages of the injection modalities reported in clinical trials, a 2-year study by Debourdeau et al (89) examined 3,313 eyes with nvAMD; 1,243 eyes were categorized as ‘pro re nata’ (PRN) and 2,070 eyes were initiated on the ‘treat-and-extend’ (T&E) program. The PRN protocol applies to regular monthly visits after the lesion has become quiescent, while the T&E protocol adds treatment intervals after stabilization of the lesion to maintain therapeutic stability. After treatment with ranibizumab and aflibercept, eyes treated with the T&E program had better visual acuity (VA) outcomes compared with eyes treated with PRN.

Owing to the high cost of current anti-VEGF treatments, patients with nvAMD require frequent intravitreal injections for optimal visualization, which may lead to undertreatment for patients with a high financial burden. As a result, the ranibizumab port delivery system (PDS) was developed, and Eichenbaum et al (90) proposed surgical implantation of this refillable device in the vitreous cavity, which would enable the sustained release of ranibizumab. Although ranibizumab is a widely used anti-VEGF drug, the risk of adverse events and safety as well as the efficacy of PDS need to be further investigated. A study by Lowater et al (91) revealed that patients receiving PDS had a high incidence of adverse events, including 25% experiencing vitreous hemorrhage and 20% experiencing hyphemia. Future studies include further improving the therapeutic safety of PDS. An ongoing clinical trial (92) is further exploring the potential of PDS with ranibizumab. Carlà et al (92) reported that the level of clinical efficacy (the incidence of adverse reactions was determined according to the examination of the eyes of the patients) of ranibizumab given continuously with PDS was comparable to the efficacy of the IVI treatment in nvAMD. However, a high incidence of adverse effects remains; vitreous hemorrhage occurred in 68% of patients and endophthalmitis occurred in 18% of patients. Therefore, future studies are needed to better define the long-term efficacy of PDS and improve patients' vision.

Aflibercept. Aflibercept was developed by Regeneron Pharmaceuticals, Inc (93). It is a recombinant fusion protein that fuses the extracellular immunoglobulin-like (Ig) structural domain 2 of VEGFR1 and the extracellular Ig structural domain 3 of VEGFR2 to the fragment crystallizable (Fc) portion of human IgG1; this process enables stronger binding of aflibercept to VEGF-A and VEGF-B than the previously used ranibizumab or bevacizumab (94). Aflibercept was approved for the treatment of AMD in 2011 and has subsequently been used to treat certain DRs (95).

For the treatment of AMD, Cao et al (84) examined 122 eyes of 106 patients with nvAMD who received injections of aflibercept for 3 consecutive months (n=70) or bevacizumab (n=52), followed by a treatment-extension interval regimen. The results revealed that patients whose eyes were treated with aflibercept were three times more likely to discontinue treatment after 1 year compared with those whose eyes were treated with bevacizumab (43 vs. 15%). These observations revealed that aflibercept was superior to bevacizumab and had important clinical implications for patient treatment selection. In another study, Kucukevcilioglu et al (96) focused on a multicenter comparison of the 24-month efficacy of ranibizumab, aflibercept and the ranibizumab-aflibercept switch. Their results indicated that visual outcomes, including VA and central macular thickness, were similar in non-switchers (aflibercept and ranibizumab groups) and switchers (from ranibizumab to aflibercept) after 2 years of follow-up; moreover, patients that received aflibercept required fewer injections, office visits or additional treatments and the treatment was beneficial for patients whose families were not financially stable. To investigate the effects of different anti-VEGF drugs on AMD, Kanadani et al (97) utilized the T&E protocol in an observational study of 131 patients with exudative nvAMD and compared four anti-VEGF drugs: Ranibizumab, aflibercept, bevacizumab and aflibercept. The study revealed that intravitreal aflibercept administration resulted in better visual and anatomical improvements in the T&E protocol, with notably fewer injections, compared with the other drugs tested. However, large multicenter randomized clinical trials with longer follow-up periods are needed to assess whether this therapeutic route is effective and whether it provides better results in the treatment of retinal vascular diseases.

Faricimab. The Ang/Tie pathway plays an important role in maintaining vascular stability and synergizing with VEGF. However, under pathological conditions, Ang-2 and VEGF-A synergistically can cause vascular leakage and neoangiogenesis. In addition, Ang-2 can increase vascular permeability by activating FAK phosphorylation via β1 integrins, inducing cytoskeletal remodeling and pericyte loss (98). Faricimab is a bispecific anti-VEGF/Ang-2 antibody and exhibits improved vascular stability and reduced retinal inflammation compared with single anti-VEGF therapy (99). To evaluate the safety and efficacy of faricimab, Khanani et al (100) conducted a phase III clinical trial for the treatment of nvAMD, and concluded that a fixed dosing regimen every 16 weeks was required in the first year based on the patients' disease activity, while this was incorporated into the second year in a personalized treatment interval (PTI) format. The PTI approach aims to reduce patient burden by customizing treatment intervals to meet the needs of individual patients. Moreover, Heier et al (101) demonstrated that extended treatment intervals with faricimab (every 16 weeks) reduced the patient burden of treatment compared with aflibercept, which inhibited the VEGF pathway alone and was given every 8 weeks, and provided visual benefit for patients with nvAMD and diabetic macular edema. To assess ocular anatomical and functional outcomes in patients with nvAMD treated with vitreous faricimab, Pandit et al (102) examined the central fovea thickness (CFT), maximal fiber-vascular pigment epithelial detachment (fvPED) and Snellen visual acuity and compared them with those of patients who were previously switched to faricimab after other anti-VEGF treatments. The proportion of hydrops in the retina before conversion was 36.7%, which decreased to 24.8% after conversion. The proportion of eyes with subretinal fluid was 53.2% before conversion and decreased to 26.6% after conversion. The results revealed that vitreous faricimab improved anatomical outcomes compared with those of patients with previously treated nvAMD while maintaining visual acuity in the short term. Moreover, Szigiato et al (103) similarly observed the short-term efficacy of faricimab and evaluated the CFT, fvPED, subretinal fluid and intraretinal fluid of patients; their results indicated a notable reduction in CFT and fvPED and stabilization of visual acuity, as well as a better visual outcome, occurred in patients that switched to faricimab after previous anti-VEGF therapy. However, long-term clinical studies are needed to confirm the long-term efficacy of faricimab. A comparison of different anti-VEGF drugs is presented in Table I (104-106).

In gene editing technology, the creation of dominant-negative VEGFR2 using the multifunctional prime editing system could block aberrant retinal angiogenesis in a mouse model of oxygen-induced retinopathy (107). Additionally, advances have been made in the development of molecular inhibitors that target VEGFR2. CLK inhibitors (MU1210 and T3-CLK) have been shown to reduce the mRNA and protein expression, as well as the downstream signaling of VEGFR2. This was partly due to the reduced activity of the WNT/β-catenin pathway, since the activation of this pathway induced VEGFR2 expression, whereas the knockdown of β-catenin blocked VEGFR2 expression. Notably, no alternative splicing of VEGFR2 was detected. Therefore, C81 may be a promising compound for the treatment of diseases that depend on angiogenesis and inflammation, as they impair both processes (108). Furthermore, Tang et al (109) reported that acrizanib, which is a small-molecule inhibitor of VEGFR2, acts by differentially inhibiting multiple phosphorylation sites of VEGFR2 in ECs, namely Y951, Y996, Y1059, Y1175 and Y1214. Among them, the highest (Y1214) and lowest (Y996) inhibition differed by ~2.5-fold. This has an important impact on physiological angiogenesis and pathological neovascularization. In summary, inhibitors developed on the basis of VEGFR2 may attenuate angiogenesis and provide better treatment outcomes.

Inhibiting neoangiogenesis by interfering with the VEGF and Ang-2/Tie2-related pathways is a novel concept. Using in vitro and in vivo experiments, Lei et al (110) demonstrated that 5α-hydroxycostic acid (isolated from the natural plant Viburnum vulgare) had a therapeutic effect on neovascularization in a rat CNV model by inhibiting cell proliferation and angiogenesis. These results indicated that 5α-hydroxycostic acid may inhibit neovascularization by interfering with the Ang-2/Tie2-related pathway and may be a candidate for the treatment of CNV. In addition, Liu et al (111) suggested that intravitreal injection of borsub, which is a reversible proteasome inhibitor, could ameliorate CNV by antagonizing the VEGF-A/VEGFR2 and PDGF-D/PDGF receptor-β pathways. These results indicate that the development of multitarget therapies may hold potential in addressing the shortcomings of inadequate single-target therapies.

Therapeutic approaches involving gene editing are also being developed. Zeng et al (112) depleted VEGFA using a novel CRISPR/Cas9 system, which inhibited the proliferation, migration and tube formation of HUVECs in vitro and ultimately inhibited corneal neovascularization in vivo. In addition, Huang et al (113) performed recombinant adeno-associated virus (AAV) serotype 1-mediated CRISPR/Cas9 editing of the genomic VEGFR2 locus, which suppressed angiogenesis in mouse models of oxygen-induced retinopathy and laser-induced CNV. These results indicate that gene editing may be used in the future for the treatment of fundus diseases.

The drugs bevacizumab, ranibizumab and aflibercept are expensive and require frequent intraocular injections; therefore, they can be a heavy burden for economically disadvantaged families (114). Challenges also remain in the selection of specific anti-VEGF drugs. From a practical point of view, ophthalmologists usually choose a specific anti-VEGF drug because it improves vision and reduces the cost or number of injections. However, numerous challenges remain in the treatment of patients with nvAMD because certain anti-VEGF drugs are limited to disease-specific reimbursement (115).

Poor visual outcomes. While it has been demonstrated that continued injections are needed to achieve better vision, the financial situation of the patient's family and compliance are key factors, and financial difficulties or a lack of compliance may lead to inadequate treatment and increased medical risk. Both can lead to suboptimal vision outcomes (116). Subretinal fibrosis is an end-stage sequela of nvAMD and can lead to permanent central vision loss in patients with nvAMD. Even when adequate anti-VEGF therapy is given, ~1/3 of patients still develop irreversible visual impairment due to subretinal fibrosis (117). Moreover, geographic atrophy secondary to AMD is a progressive, irreversible loss of vision involving photoreceptors, RPE cells and choroidal capillaries, and its progression may be more rapid in patients receiving anti-VEGF therapy (118). In addition, normal retinal structures may be affected by certain treatments. For example, Zhuang et al (119) further explored the short-term changes in the normal retinal vasculature after anti-VEGF treatment in patients with nvAMD; their results indicated that anti-VEGF treatment transiently affected the relatively normal retinal vasculature, which could lead to edema of the nerve fibers on the nasal side of the optic disc. Therefore, anti-VEGF therapy may affect both healthy and diseased tissues of the fundus, thereby affecting the regression of patients' vision.

Mechanisms related to drug resistance: Changes in VEGF isoform expression and epigenetic regulation. The VEGF isoform VEGF165 is either in a diffusible state (one half) or bound to heparan sulfate proteoglycan on the cell surface (the other half), whereas VEGF121 lacks a heparin-binding domain and diffuses freely in tissues (120). Serine/arginine protein kinase 2 binds and activates its downstream target serine/arginine splicing factor 1 (SRSF1), which increases splicing and expression of VEGFA165(121). Another study showed that 30 kPa extracellular matrix rigidity inhibited the expression and nuclear accumulation of YAP to regulate the expression of SRSF1 via runt-related transcription factor 2, which subsequently inhibited the expression and secretion of VEGF165 in tumor cells (122).

In epigenetic regulation, the DNA methylation-regulated VEGFR signaling pathway may be involved in the development of diabetic cardiovascular disease. VEGF-B hypomethylation may be a potential biomarker for early intervention in patients with this disease. The expression of DNA methyltransferases, such as DNA methyltransferase (DNMT)1, DNMT3a and DNMT3b, may serve as potential biomarkers of the anti-VEGF diabetic ME response (123).

Since anti-VEGF therapy is limited by the low response rates of single drugs, drugs targeting multiple neoangiogenic pathways are required. The application of existing anti-VEGF drugs should be optimized and prognostic biomarkers should be evaluated to identify those that predict effective responses. Biomarkers need to be prospectively validated in independent randomized trials; however, novel drug discovery and the development of multitarget therapeutics are also important. In this context, the bis-antibody-based drug RO-101 is an ideal candidate for the treatment of retinal diseases. In preclinical models, RO-101 has shown similar or greater effects than current anti-VEGF treatments, including neointimal growth abrogation with comparable or longer half-lives. In addition, RO-101 has shown strong binding affinity for VEGF-A and Ang-2, and is biocompatible with retinal tissue in animal studies, indicating potential compatibility for use in humans (124). Therefore, further studies are expected to validate its effectiveness and value in the clinic. In addition, a new technology has been developed in which carbon nanodonut (CND) with a donut-like structure is synthesized using sodium alginate (SA) and 1,8-diaminooctane (DAO) as raw materials. In human umbilical vein endothelial cells, CND can reduce the levels of reactive oxygen species and proinflammatory cytokines (such as IL-6 and IL-1β) five-fold compared with bevacizumab. In addition, CND has a strong affinity for VEGFA165, with a dissociation constant of 2.2x10^-14 M; this value is >1,600 times stronger than that of the commercial drug bevacizumab (Avastin^®). Therefore, SA/DAO-CNDs have the potential to be used as drugs for the treatment of various angiogenesis-related ocular diseases (125). Patient compliance and financial pressure are equally challenging. The high cost of anti-VEGF therapy and the need for multiple injections have led to decreased compliance and financial stress for several patients. However, investigating sustained-release materials that can prolong the duration of drug action, minimize the number of intravitreal injections, and improve patient compliance is a viable solution to this problem and can lead to significant therapeutic benefits. Two extended-release materials are currently undergoing clinical trials (NCT03677934 and NCT03953079), including PDS and GB-102. PDS mainly targets VEGF-A, while Gb-102 is a tyrosine kinase inhibitor that targets VEGF-A and PDGF. Their role is to reduce the burden of frequent intravitreal injections on patients and doctors. Although basic research indicates that extended-release drugs have strong therapeutic applications, the preparation of drug carriers with good compatibility with drugs, long release times and prolonged duration of action is still a major challenge at present (126,127).

Gene therapy applied directly to the eye has also shown potential for long-lasting drug applications. In one study, an AAV2 vector was used to fuse a truncated form of soluble VEGFR2 to the FC portion of human IgG1. AAV2-mediated gene delivery of sVEGFR-2-Fc was effective in inhibiting laser-induced CNV in mice and was considered a promising tool for developing therapeutic tools for retina-related diseases and preventing neovascularization (128). In addition, genes encoding anti-VEGF Fab proteins have shown potential in the treatment of retinal diseases. RGX-314 uses a novel recombinant AAV8 vector to induce the production of anti-VEGF Fab proteins. Various phase II and III studies are currently underway to evaluate the efficacy of RGX-314 in patients with nvAMD (129,130). In a study, RGX-314 (currently known as ABBV-RGX-314) expression consistently inhibited VEGF-A and was safe and effective in treating patients with nvAMD (131). Another type of gene therapy involves ixoberogene soroparvovec (ixo-vec), formerly known as ADVM-022, which is administered by intravitreal injection. This technology utilizes AAV.7m8 as a vector to generate an aflibercept-carrying transgene for the treatment of nvAMD, and the phase II LUNA trial assessing the safety and efficacy of ixo-vec is currently in progress (132). Finally, 4D-150, which contains transgenes that target aflibercept and VEGF-C, is currently planned to enter a phase III clinical trial for nvAMD (133). However, gene therapy targets, routes of administration and potential safety issues require further validation.

Understanding the structural biology of VEGFR2 has markedly advanced drug design. As previously described, the complete VEGFR2 receptor comprises an extracellular region, a transmembrane domain and an intracellular kinase domain. Insights from ligand-binding domain and coreceptor complex structures, such as the VEGF-VEGFR2 complex and VEGFR1/VEGFR2 extracellular fusion region, have guided the development of therapeutic agents such as ranibizumab and aflibercept, which are now widely used to treat AMD. Furthermore, by targeting the catalytic structural domain, ‘DFG-out’ (type II) VEGFR2 inhibitors effectively suppress VEGFR2 phosphorylation, reducing neoangiogenesis and overcoming drug resistance (134). Future research directions include capturing the transient states of VEGFR2 structural activation to design conformation-specific inhibitors, integrating genomic, proteomic and structural data for precision-targeted therapies and developing nanomedicines capable of penetrating the BRB based on structural insights. Thus, structural biology serves not only as a fundamental tool for elucidating VEGFR2 function but also as a cornerstone for overcoming current therapeutic limitations and advancing next-generation ophthalmic treatment.

Single-cell RNA sequencing (scRNA-seq) technology offers novel insights into the cellular heterogeneity and molecular pathways of retinal diseases (135). Gene expression profiles vary across distinct retinal cell types and disease microenvironments. In AMD, which predominantly affects the macula, pathological alterations are most pronounced in macular neuroglial cells, microglia and astrocytes during disease progression. Single-cell sequencing may reveal gene expression changes in different cell types (136). Transcriptomic analysis of retinal samples from patients with DR at various stages revealed dysregulated mechanisms, including the Hippo signaling pathway and gap junction formation (137). scRNA-seq enables the identification of disease-specific cellular subpopulations and facilitates the prognosis of anti-VEGF therapeutic efficacy through cell type-specific gene expression signatures, thereby guiding personalized treatment strategies. Future directions include integrating single-cell assay for transposase-accessible chromatin with high-throughput-seq (to delineate epigenetic regulation) with the phosphoproteomic profiling of VEGFR2 activation to dissect drug resistance mechanisms, as well as establishing comprehensive ophthalmic single-cell databases to inform individualized clinical decision-making.