Reduction/oxidation (redox) imbalances, which are caused by imbalances in pro-oxidant/antioxidant reactions increase oxidative stress (OS), resulting in retinopathies; ROS are one of the most important oxidants (15). RPE cells secrete a variety of cytokines, including VEGF (an angiogenic stimulator) and PEDF (an angiogenic inhibitor) (14,16). Physiologically, in the retina, there is a balance between PEDF and VEGF, which is essential for retinal health. VEGF-induced angiogenesis is normally inhibited by PEDF (17). An increased VEGF/PEDF ratio can lead to retinopathies (18). In a model of ischemia-induced retinal neovascularization, the retina shows a 5-fold increase in VEGF and a 2-fold decrease in PEDF, compared with the age-matched controls (16). Advanced glycation end product (AGE), a senescent macro-protein derivative, formation of which is accelerated in patients with diabetes, elicits OS and induces vascular inflammation and is therefore involved in retinopathies (19,20). PEDF inhibits AGE-induced retinal vascular hyperpermeability and angiogenesis by blocking the generation of ROS and subsequent ROS-mediated upregulation of VEGF (21). These findings suggest that an imbalance between angiogenic stimulators and inhibitors contributes to retinopathies.

Inflammation is a critical factor in the pathogenesis of certain retinopathies (14). Endoplasmic reticulum (ER) stress is an inducer of the inflammatory response, which can cause AMD via the interaction of C-reactive protein (CRP) and serum amyloid P protein (14). CRP is an inflammatory biomarker and risk factor for AMD (22). VEGF contributes to the inflammatory process following laser-induced retinopathy (23). In human RPE cells, CRP can induce VEGF and interleukin-8 (IL-8) expression (24). The increased level of inflammation occurs in proliferative DR eyes with larger non-perfusion areas and the expression of IL-8 positively correlates with the non-perfusion area (25). Inflammatory mediators [such as VEGF, hypoxia-inducible factor (HIF)-1α, IL-8 and others] are markedly increased (4-10-fold) compared with the control retina in laser-induced retinopathy (23). Additionally, there is upregulation of VEGF and IFN-γ-inducible protein-10 (IP-10), characteristic inflammatory cytokines present in the aqueous humor of patients with AMD (26,27). These indicate that the VEGF pathway is involved in retinal inflammation and damage and plays a role in the inflammation-induced pathogenesis of retinopathies.

DR and RVO are the two most common ischemic retinopathies and upregulation of VEGF further aggravates retinal ischemia and hypoxia, commonly leading to loss of vision (28). Noninvasive perfusion measurement data have suggested that the decreased choroidal circulation and subsequent hypoxia/ischemia can aggravate the pathogenesis of AMD (29-31). In human RPE cells, ER stress upregulates the expression of VEGF via the actions of activating transcription factor (ATF) (32,33). Furthermore, the excessive proteasomal inhibition generates ER stress and induces the expression of VEGF, HIF-1α and angiopoietin-2 to support neovascularization in AMD (34). Hypoxia is also an efficient inducer of VEGF expression (35). VEGF, as an HIF-1α-regulated angiogenic mediator, and HIF-1α are upregulated in the retinal tissue of DR (36).

VEGF upregulation in retinopathies is also associated with other factors that collectively affect the pathological process, leading to increased VEGF levels. The missense mutation in R345W of fibulin-3 protein causes an accumulation of a misfolded protein within the ER and the retention of the protein in the ER stimulates ER stress, resulting in upregulation of VEGF in human RPE cells (37). Alternatively, increased levels of homocysteine (an ER stressor) are involved in wet AMD (w-AMD) (38). Homocysteine is a toxic metabolite that induces ER stress in the liver in patients who abuse alcohol (39). Notably, excessive alcohol consumption is also considered to be a risk factor for AMD (40). Homocysteine can activate the protein kinase R-like ER kinase (PERK)/ATF4 pathway, which is an important component of ER stress, inducing upregulation of VEGF expression in human RPE cells (32). OS induces the upregulation of ATF4 via arsenite (an oxidative stressor) in human RPE cells (14). Upregulation of ATF4 is sufficient to activate the VEGF promoter (32). These findings suggest that the homocysteine/arsenite induces the expression of VEGF via an ATF4-dependent mechanism.

The commonality among w-AMD, DME, DR and RVO is the vascular permeability and pathological neovascularization induced by VEGF. Intravitreal injections of anti-VEGFs can resolve macular edema, subretinal fluid and neovascularization (12). Although anti-VEGFs were initially developed as a cancer treatment, bevacizumab was first approved by U.S. Food and Drug Administration (FDA) in 2004 for colon cancer treatment (41); however, they are now widely used in various retinopathies, including RVO, DR, AMD and DME. Bevacizumab, a humanized monoclonal antibody with a molecular weight of 149 kDa, has been used in ophthalmology since 2005 as an off-label therapy (42). Pegaptanib, an anti-VEGF drug used in certain retinopathies (w-AMD), was approved by the FDA in 2004 (41). Ranibizumab, approved by the FDA in 2006 for w-AMD, contains the Fab fragment of a humanized IgG1 κ isotype murine monoclonal antibody and has a molecular weight of 48 kDa, almost three times smaller than bevacizumab (43). Aflibercept, with a molecular weight of 115 kDa, was approved by the FDA in 2011 (42).

Recently introduced anti-VEGF drugs for retinopathies include brolucizumab and faricimab. Brolucizumab was approved by the FDA in 2019. Brolucizumab is a single-chain humanized antibody fragment, designed to bind VEGF-A (43,44). Faricimab, a bispecific heterodimeric monoclonal human antibody, has a greater molecular weight than bevacizumab and was approved by the FDA for DME and w-AMD in 2022 (45). Due to their shorter presence on the market, they have not yet been extensively used and studied.

Intraocular inflammation is a primary ocular adverse effect associated with anti-VEGF therapies (46). The incidence of significant ocular inflammation following ranibizumab injection for AMD ranges from 1.4-2.9%, whereas the incidence following bevacizumab injection is comparatively lower, between 0.09-0.4% (47). An acute increase in IOP is relatively common immediately after an anti-VEGF injection, with most studies reporting spontaneous normalization of IOP within several hours (42,48,49). The incidence of rhegmatogenous retinal detachment post-anti-VEGF injections is low, ranging from 0-0.67% (47,50). Of note, there was no statistically significant difference in the occurrence of rhegmatogenous retinal detachment and retinal tear between eyes treated with anti-VEGF treatments and the control group (46).

Following bevacizumab injection, rare ocular side effects include anterior ischemic optic neuropathy in w-AMD and angioid streaks in pseudoxanthoma elasticum (51,52), central retinal artery occlusion (RAO) and RVO in central RVO, AMD and proliferative DR (53) and hemorrhagic macular infarction in central RVO (54), acute retinal ischemic changes in diabetic rubeosis (55) and sixth nerve palsy (56). Among 4,069 injections (ranibizumab or aflibercept) administered for the treatment of DME (464 cases), w-AMD (187 cases) or RVO (156 cases), only 18 cases (0.44%) exhibited transient central RAO, including four severe cases (ranibizumab: two cases, aflibercept: two cases) (57).

The literature search was conducted on January 1, 2025, using PubMed, Clinicalkey (Elsevier), SpringerLink, Cochrane Library and CNKI data-bases. The search used combinations of the following search terms: (anti-VEGF OR aflibercept OR bevacizumab OR ranibizumab OR brolucizumab OR faricimab) AND (papillary OR circumpapillary OR juxtapapillary OR intrapapillary OR optic disk OR optic nerve head OR optic papilla) AND (p-RNFL OR blood flow OR blood vessel OR capillary OR microvascular OR microcirculation). Filters applied included human studies and articles were initially screened based on the titles or abstracts. The inclusion criteria were: i) Original research articles and ii) participants included were aged over 18 years. Case reports were excluded.

Studies that assessed p-RNFL thickness using optical coherence tomography angiography (OCTA) and blood circulation parameters using laser speckle flowgraphy (LSFG) [measuring mean blur rate (MBR), relative flow volume (RFV) and blowout score], Canon laser blood flowmeter (CLBF; measuring vessel diameter), scanning laser Doppler flowmetry (SLDF; measuring blood flow) and OCTA [measuring vessel density (VD), vessel length density (VLD), capillary volume (CV), perfusion density (PD) and perfusion index (PI, or flux index)] were included. All studies included in the present review involved patients with RVO, DR, AMD or DME. Previous studies and clinical trials examining the effects of anti-VEGF treatments on p-RNFL and papillary/peripapillary blood circulation are summarized in Tables I and II, respectively. The studies included in the present review encompass an analysis of 3,077 affected eyes treated with anti-VEGF treatments for retinopathies, as detailed in Tables I and II. The underlying mechanisms for the effects of anti-VEGF treatments on p-RNFL and papillary/peripapillary blood circulation are illustrated in Figs. 1 and 2, respectively.

Martinez de-la-Casa et al (61) reported significant p-RNFL thinning in patients treated with ranibizumab over 12 months, compared with an untreated control group. Although IOP elevation was observed, no statistical correlations were found between changes in the outer macular ring and changes in p-RNFL (61). It appears unlikely that post-injection p-RNFL thinning is secondary to changes in reduced macular edema possibly extending to the peripapillary region. IOP fluctuations induced by anti-VEGF treatments could potentially damage the p-RNFL (62). Subsequent continuation of this study revealed the long-term effects (8-year follow-up) of ranibizumab on p-RNFL, with statistically significant thinning observed in both the treated and control groups (63). Of note, IOP remained unchanged throughout the study (63). Parlak et al (64) also observed significant thinning in the affected eyes treated with ranibizumab, with no significant change in IOP over 12 months. The p-RNFL thickness decreased with age at a rate of -0.21 µm/year (65). Advanced age in patients with w-AMD is associated with a significant decrease in p-RNFL thickness (66) and higher stages of w-AMD are associated with a thinner p-RNFL (67). These studies suggest that aging and the natural progression of w-AMD may contribute to p-RNFL thinning. Additionally, unexpected p-RNFL thinning in the untreated fellow eyes was reported by Valverde-Megías et al (63) and Parlak et al (64), indicating a potential systemic effect of anti-VEGF on untreated fellow eyes. Another study indicated a significant decrease in p-RNFL thickness in treated eyes over 12 months, with no significant difference in the fellow eyes (68). The authors attributed the thinning of p-RNFL thickness to the anatomical improvement of macular lesions, although a direct effect of ranibizumab on decreased p-RNFL could not be excluded.

Given that ranibizumab may affect fellow eyes of in patients with w-AMD via systemic circulation, certain studies included age-matched healthy subjects as controls. Long-term follow-up revealed no statistically significant differences in p-RNFL thickness (69,70) and IOP (70) between treated, fellow and healthy eyes. Therefore, p-RNFL thinning in fellow eyes may not be associated with a systemic effect of ranibizumab on untreated fellow eyes. The absence of p-RNFL thinning due to ranibizumab has also been reported by Zucchiatti et al (71) over 12 months, El-Ashry et al (72) over 1 month and Sobacı et al (73) over a mean of 14 months. Notably, Horsley et al (74) reported an increase of 5.88 µm in the p-RNFL thickness during a follow-up of 20.1±3.6 months despite no statistical difference (P=0.35) and there were no sustained IOP elevations for any of the patients with w-AMD. This suggests that ranibizumab may not affect p-RNFL in w-AMD.

The safety of aflibercept in relation to p-RNFL thickness has been evaluated in both long-term and short-term studies. Following aflibercept injections, a significant reduction in p-RNFL thickness was observed in treated eyes over a 12-month period, with no significant changes in intraocular pressure (IOP) (68). Additionally, no significant differences in p-RNFL thickness and IOP were noted in the fellow eyes (68), suggesting that aflibercept does not affect fellow eyes through systemic circulation. Within 2-5 min post-injection, aflibercept induces immediate p-RNFL thinning, particularly in the nasal region and increases IOP, with a significant negative correlation between increased IOP and p-RNFL thinning (75). The IOP spike immediately following anti-VEGF injections may compromise the integrity of the p-RNFL (71), with the nasal p-RNFL being particularly susceptible to IOP fluctuations. Furthermore, the nasal sector p-RNFL is less affected by macular edema in w-AMD prior to anti-VEGF injection (75), potentially making it a more accurate indicator of damage caused by anti-VEGF injections leading to increased IOP, thereby justifying the significant thinning of p-RNFL in the nasal sector. In contrast to the aforementioned studies, Gunay et al (6) reported no significant changes in p-RNFL thickness in patients with w-AMD with macular edema over 12 months and no significant change in IOP was observed. Thus, the p-RNFL thinning attributed to the reduction of macular edema and increased IOP following aflibercept injections contradicts the findings of Gunay et al (6).

Regarding the effects of bevacizumab monotherapy on p-RNFL, Entezari et al (76) found a significant reduction in p-RNFL thickness in patients with w-AMD at 12 weeks, which returned values that were comparable to baseline values at 24 weeks. Furthermore, IOP elevation returned to baseline values within 30-60 min in most patients receiving two injections of bevacizumab (76), indicating that fewer bevacizumab injections and short-term IOP increases were probably insufficient to cause p-RNFL damage. However, another prospective study involving 135 patients with w-AMD found no statistical differences in p-RNFL thickness after 12.4±2.4 bevacizumab injections over 24 months (77). Similarly, Sobacı et al (73) observed no changes in p-RNFL after 5.1±1.3 bevacizumab injections over a mean period of 14 months, although IOP was statistically different compared with baseline values. These studies suggest that a greater number of bevacizumab injections do not result in significant p-RNFL damage.

AMD and DME are among the most common causes of blindness in adults. Several macular diseases are prevalent in the elderly and thus, conditions such as AMD and DME can occur concurrently (78). The standard treatment for DME and w-AMD involves anti-VEGF injections (79). Due to the chronic nature of w-AMD and DME, standard treatment regimens typically recommend multiple injections during the first year of treatment (80). Additionally, the duration of effectiveness of anti-VEGF agents is limited (81), necessitating repeated injections to maintain therapeutic effects. A transient and acute volume-related IOP increase is commonly observed following an anti-VEGF injection (82). Although the incidence of delayed and sustained IOP is low after a single or multiple anti-VEGF injections (83), an acute increase in IOP may cause loss of p-RNFL (81). Prophylactic anterior chamber paracentesis (ACP) combined with anti-VEGF treatments offers an effective alternative for preventing the acute rise in IOP (84). Thus, it is important to determine whether an anti-VEGF injection and the transient elevation in IOP adversely affect p-RNFL thickness in patients with w-AMD and DME.

In two short-term studies conducted by Khodabande et al (80) and Soheilian et al (81), patients with w-AMD or DME were randomly assigned to one of two groups: A group that was not treated with prophylactic ACP and the other which was treated with prophylactic ACP. The two studies consistently found that IOP was markedly higher in patients without ACP compared with those with prophylactic ACP following bevacizumab injections and no significant p-RNFL thinning was observed in patients with prophylactic ACP. However, the studies diverged in their conclusions regarding p-RNFL thickness changes in patients without ACP. Soheilian et al (81) reported significant p-RNFL thinning in patients without ACP, whereas Khodabande et al (80) found no significant thinning in this group. This discrepancy may be attributed to the older age and thinner baseline p-RNFL in the study by Soheilian et al (81) compared with those in the study by Khodabande et al (80) (66.4±4.9 vs. 62.9±10 years; 85.3±5.6 vs. 110±19 µm, respectively), potentially rendering them more susceptible to IOP fluctuations. Overall, ACP appears to mitigate acute IOP elevation and offers a certain degree of protection to the p-RNFL.

Regarding the effects of a combination of anti-VEGF treatments on p-RNFL, no statistically significant differences in p-RNFL thickness were observed among patients with w-AMD treated with ranibizumab alone (over 20.1±3.6 months; 13.4±3.6 injections), a combination of ranibizumab and bevacizumab (over 27.1±4.2 months; 17.7±3.7 injections), or a combination of ranibizumab, bevacizumab and pegaptanib (over 27.0±9.7 months; 16.0±5.5 injections) (74). Recent investigations have explored the correlation between the number of anti-VEGF injections and p-RNFL thickness. No significant effect on p-RNFL was observed in patients with w-AMD following a combination of anti-VEGF treatments (5-50 injections) (66,85). However, a study by Wang et al (86) with an extended follow-up period (46.8 ±42.0 months) suggested a dose-response relationship between the number of injections (a combination of bevacizumab, ranibizumab and aflibercept; 3-138 injections) and p-RNFL thinning, which became more pronounced after ~30 injections and 50 months of treatment. These findings indicate that prolonged treatment duration and a higher number of injections may be required to detect damage to the p-RNFL caused by active w-AMD.

Progressive optic neuropathy, characterized by a deficit in p-RNFL and corresponding thinning of the neuroretinal rim tissue, can lead to progressive visual field defects (87,88). Thinning of both p-RNFL and Bruch's membrane opening minimum rim width (BMO-MRW) occurs in early progressive optic neuropathy (89). Moreover, a decrease in the minimum rim area precedes reductions in p-RNFL and visual field loss (90). Theoretically, BMO-MRW serves as a sensitive marker for evaluating the effects of anti-VEGF injections on the optic nerve head. BMO-MRW markedly decreases with an increasing number of anti-VEGF injections, while no significant effect on p-RNFL is observed (66). It appears that BMO-MRW is affected earlier than p-RNFL following long-term anti-VEGF injections, suggesting that BMO-MRW may be a more sensitive marker than p-RNFL. Given the relative lack of literature on this topic, further studies with larger sample sizes are required to draw more concrete conclusions. BMO-MRW declines with age at a rate of -1.34 µm/year and the association with age is stronger with BMO-MRW than with p-RNFL thickness (65). Therefore, the decrease in BMO-MRW after anti-VEGF injections may not necessarily be drug-induced, but is instead likely age-induced.

A significant yet transient elevation in IOP immediately following anti-VEGF administration is relatively common. Typically, IOP returns to baseline levels within 30 min to several hours post-injection (42,48,49,91). The prevalence of sustained IOP increase ranges from 3.4-11% (92). ACP offers a less painful and effective alternative to mitigate the acute rise in IOP following anti-VEGF administration (84). Theoretically, regular administration of ACP may reduce IOP spikes post-anti-VEGF administration, thereby preventing repeated stress on the p-RNFL. In patients with w-AMD who do not receive ACP, the loss in p-RNFL thickness (-2.16±3.60 µm) is markedly greater than in those who receive ACP regularly (93). The p-RNFL thickness declines with age at a rate of -0.21 µm/year (65). Patients with w-AMD who do not receive ACP experience a loss of -0.86 µm/year in p-RNFL thickness after an average of 2.5 years (93). This indicates a greater loss of p-RNFL thickness in long-term anti-VEGF injections for w-AMD than the age-related p-RNFL loss. Furthermore, there is a negative correlation between temporal p-RNFL and IOP following a combination of anti-VEGF injections (94). This suggests that higher IOP measurements are associated with a greater decrease in p-RNFL thickness, indicating that p-RNFL thickness loss is due to IOP spikes rather than age. In patients with w-AMD or DME treated with more than six ranibizumab or aflibercept injections, a significant acute and transient IOP increase is observed 5 min post-injection and significant p-RNFL thinning is observed by the third month (92). Additionally, similar immediate post-injection changes, such as IOP increases, are observed after 1 year (92). This suggests that repeated anti-VEGF injections could lead to irreversible changes in the optic nerve head structures. It appears that IOP elevation may be associated with p-RNFL thinning. The p-RNFL thickness may also be influenced by other factors, such as macular edema, in patients with w-AMD or DME treated with a combination of anti-VEGF treatments. The p-RNFL thickness in these retinopathies affecting the macula is greater than in age-matched control eyes (59). Although changes in p-RNFL thickness are markedly associated with changes in central macular thickness in patients with DME, no such association exists in patients with AMD (60). Whether changes in macular edema in patients with w-AMD or DME affect p-RNFL thickness following a combination of anti-VEGF injections and to what extent, remains unclear.

Stereotactic radiotherapy (SRT) is a novel adjuvant approach for treating w-AMD (95). The mean number of anti-VEGF injections decreases by nearly 50% during the 12 months following SRT compared with the preceding year, while visual acuity improves (96). A single dose of SRT markedly reduces the need for anti-VEGF injections over 2 years without compromising visual acuity (97). Morphological changes only affect the outer retinal layers following SRT (96). It is necessary to investigate whether SRT, in conjunction with decreased anti-VEGF injections, affects p-RNFL thickness. It has been demonstrated that SRT in conjunction with ranibizumab and aflibercept does not lead to significant changes in p-RNFL thickness over 12 months in patients with w-AMD (94). Radiation of SRT can induce microvascular changes; however, only 1% of affected eyes exhibit a worsening of vision (97). Therefore, SRT may be a potential first step in preventing a worsening of vision as a result of a decrease in anti-VEGF injections. A combination of anti-VEGFs and SRT is thus a novel direction for the treatment of w-AMD.

Radial peripapillary capillaries (RPC) constitute the vascular network located within the RNFL (98). The relationship between RPC density and p-RNFL thickness has been extensively investigated in both healthy individuals (99) and those with pathological peripapillary retinas (100). A study suggested that there was a positive correlation between RPC-VD and p-RNFL thickness (101). However, a recent study by Zhuang et al (102) reported significant negative associations between transient reductions in RPC-VD and increases in p-RNFL thickness, primarily in the nasal regions, 1 week after treatment with a combination of conbercept, aflibercept and ranibizumab. These increases in p-RNFL thickness may be attributed to nerve fiber layer edema. Given that both RPC-VD and p-RNFL thickness returned to baseline values within 1-3 months post-injection, it appears that nerve fiber layer edema contributes to these negative correlations.

Brolucizumab has recently been approved for the treatment of w-AMD (103). It offers the potential for greater tissue penetration compared with earlier anti-VEGF agents and can deliver higher molar doses than larger molecules (104,105). Patients with refractory w-AMD who exhibit poor responses to other anti-VEGF treatments are often transitioned to brolucizumab. No significant changes in p-RNFL thickness have been observed within 3 months following a single brolucizumab injection, which is subsequently followed by a combination of aflibercept, bevacizumab and ranibizumab (3-37 injections) (62). However, a significant decrease in temporal p-RNFL thickness is noted 1 month after the brolucizumab injection, with this reduction dissipating after 3 months as macular edema increases (62). Therefore, the decrease in temporal p-RNFL thickness following brolucizumab injection in patients with w-AMD may be associated with improvements in macular edema. A single brolucizumab injection appears to be safe concerning p-RNFL damage in patients with intractable w-AMD.

In examining the long-term effects (≥12 months), a significant thinning of the p-RNFL was observed in patients with w-AMD from 12 months to 8 years post-injection (61,63,68,86,92-94) and in patients with DR after 24 months (119). Regarding short-term effects (<12 months), significant p-RNFL thinning was noted in patients with w-AMD from 2 min to 4 months post-injection (60,75,76,81,108,109,168), in patients with DR from 1-6 months post-injection (110,112) and in RVO patients from 3-9 months post-injection (127,154-156).

Significant p-RNFL thickening was observed in patients with w-AMD from 3-47.9 months post-injection (58,102,111). A slight increase in p-RNFL thickness post-injection was found in patients with w-AMD (74) or non-proliferative DR concomitant with DME (114), although this increase did not reach statistical significance.

No changes in p-RNFL thickness have been reported in several studies, indicating no significant alterations in p-RNFL thickness in patients with w-AMD during a follow-up period of 1-96 months post-injection (6,69,70-73,77,80). Of note, repeated anti-VEGF injections appeared to have no detrimental effects on p-RNFL.

Contradictory results have been described in the effects of anti-VEGF on p-RNFL in retinopathies, such as p-RNFL thinning, p-RNFL thickening and no change. The present study systematically categorize factors contributing to discrepancies as IOP, edema, normal course of retinopathies, aging and co-morbidities. For p-RNFL thinning, studies are divided into long-term and short-term based on study duration.

For long-term p-RNFL thinning, data are mixed regarding the relationship between post-injection IOP increase and p-RNFL thinning. Certain studies suggest that p-RNFL thickness loss could be attributed to IOP spikes rather than age, with repeated anti-VEGF injections potentially leading to changes in optic nerve head structures (92-94). However, other findings indicate no changes in IOP accompanying p-RNFL thinning (63,68). Thus, the involvement of post-injection IOP elevation in long-term p-RNFL thinning remains contested, necessitating further investigation. For short-term p-RNFL thinning, the relationship between elevated IOP and p-RNFL thinning is also debated. A significant negative correlation between increased IOP and p-RNFL thinning, with IOP spikes immediately after anti-VEGF injection potentially affecting p-RNFL integrity, has been observed (71,75). Conversely, several studies have reported that short-term increased IOP may not be severe enough to cause p-RNFL damage (76,108). In cases where no changes in p-RNFL thickness were observed, sustained IOP was not detected (6,71-73,77).

The long-term thinning of the p-RNFL is probably attributable to the anatomical improvement of macular lesions. By contrast, the short-term thinning of p-RNFL may be explained by the regression of retinal/macular edema following anti-VEGF injections (119,156). Another potential explanation is that in patients newly diagnosed with w-AMD, baseline p-RNFL values may have been elevated due to edema spreading from the macula, rather than actual thinning of the p-RNFL (111). Regarding the thickening of p-RNFL, there are two perspectives on the effect of edema. One view suggests that anti-VEGF injections may transiently affect relatively normal retinal vessels, leading to decreased blood flow and subsequent nerve fiber edema (102). Alternatively, the thickened p-RNFL may result from fluid spreading from the thickened central retina in patients with w-AMD (111). In studies reporting no changes in p-RNFL thickness, there is no mention of changes in patient edema (6,71-73,77).

The normal progression of retinopathies may also contribute to the long-term thinning of p-RNFL, as observed in diabetic retinas (119), suggesting that the thinning could be a natural course rather than an effect of anti-VEGF treatment.

Aging is another factor, with p-RNFL thickness declining by 0.21 µm per year (65). Advanced age in patients with w-AMD is markedly associated with reduced p-RNFL thickness (66) and higher stages of w-AMD are associated with a thinner p-RNFL (67). Additionally, in patients with unilateral w-AMD, the fellow eyes exhibited thinner p-RNFL over a 12-month study period (42), indicating that aging may be linked to thinning.

Co-morbidities also play a role in the long-term thinning of p-RNFL. Hypertensive retinopathy, for example, involves high blood pressure damaging both retinal microcirculation and RNFL (169). Diabetes mellitus is another example, with a thinner p-RNFL observed in patients without diabetic retinopathy (170). Abnormalities in blood vessels by diabetes mellitus and hypertension (systemic diseases) give rise to RVO and glaucomatous optic neuropathy (125). Glaucoma is more prevalent in patients with RVO compared with healthy individuals (126). In patients with glaucoma, RNFL declines by 0.76±0.85 µm per year during 3.7±1.5 years (171). Visual field damage in glaucoma, corresponding with RNFL thinning, is observed in the fellow eyes of certain patients with unilateral RVO (172). These findings suggest that RVO and glaucoma may share systemic risk factors, reflecting a common pathogenic mechanism (172). Consequently, the long-term thinning of p-RNFL in patients with retinopathy may be attributed to coexisting glaucoma.

No changes in the p-RNFL thickness have been observed in the fellow eyes, accompanying the post-injection thinning in the affected eyes after 1 year, indicating no effects via the systemic route (68). However, significant thinning was detected in both the fellow and affected eyes in an 8-year follow-up (63). The effect should have been cumulative over aging, although the systemic route was not excluded.

Studies have reported that there are no associations between the number of injections (from 3-50 injections) and the p-RNFL thickness (64,66,68-70,85). However, other studies have suggested that there were associations and the injection number ranged from 11.27±9.31 (58), 19.29±6.92 (94) and ≥30 (86). Additionally, no associations have been reported between the types of anti-VEGF treatments and the p-RNFL thickness (132).

An increase in VD was observed in patients with branch RVO (154-156), central RVO (157) and DR (147) following anti-VEGF treatment. There are three potential explanations for the observed improvement in blood circulation: i) The anti-VEGF treatment reduces inflammatory features in macular edema (156,173-175); ii) the treatment enhances the compression of blood vessels and re-perfuses occluded capillaries (154); or iii) there may be an overestimation of blood circulation parameter values. Prior to anti-VEGF injection, these values might be underestimated due to macular edema, which causes a masking effect or retinal vessel displacement (157) and the movement rate of red blood cells may fall below the detection limit of OCTA due to increased blood viscosity and a decreased blood flow velocity in patients with retinopathies (154). Consequently, post-injection values may be overestimated, suggesting that a statistically significant increase after anti-VEGF injections may not reflect a genuine statistical gain.

Inhibited blood circulation has been observed in the affected eyes. Post-injection decreases in MBR were observed in patients with w-AMD within 30 min-3 months (102,132-134). A slight reduction in MBR post-injection was also noted in patients with central RVO (159) and branch RVO (160). Several studies have reported a decrease in post-injection MBR in patients with DME within 1-4 weeks (12,161,162). Additionally, post-injection decreases in VD (102), flux index (85), arterial diameter (131), perfusion density (140), vascular caliber (163) and CV (142) have been documented.

Potential reasons for the inhibited blood circulation due to anti-VEGF treatments are: i) Vasoconstriction, a possible pharmacological effect of anti-VEGF treatments accompanied by regression of macular edema and improvement in visual acuity (12,161,162) and retinal sensitivity (159); ii) increased IOP; certain studies have suggested a significant post-injection increase in IOP, accompanied by a significant reduction in blood circulation parameters (140). However, other findings indicate no correlation between post-injection changes in IOP and blood circulation parameters (85,102,132). Thus, the involvement of acute increased IOP after anti-VEGF injections in the reduction of peripapillary perfusion remains contested; iii) upregulation of pro-inflammatory cytokines in patients with w-AMD due to anti-VEGF treatments. In certain patients, IP-10 and IL-6 levels increased following anti-VEGF injections (26). The up-regulation of adhesion molecule levels during inflammation caused leukocyte arrest, potentially leading to capillary lumen obstruction (156); iv) measurement error due to obscuring the border between tissue and vessel on LSFG images, caused by blood stasis or capillary vessel dilation (159).

Several short-term studies have reported no significant alterations in blood circulation parameters following an anti-VEGF injection (7,114). While these findings suggest an absence of short-term effects, further research employing prospective designs and extended follow-up periods are necessary to validate these results.

The potential mechanisms underlying the effects of anti-VEGF on papillary and peripapillary microcirculation in retinopathies warrant discussion. VEGF-induced vasodilation may enhance blood flow to neuronal tissues, potentially contributing to neuroprotection in the retina, thereby indicating the neuroprotective role of VEGF (176). Ischemia is one of the important causes of retinopathy. The pathological mechanisms of retinal ischemia-reperfusion injury are complex, involving oxidative stress (OS), inflammation and neuronal apoptosis (177). OS-induced inflammatory cytokines, such as intercellular adhesion molecule-1, are implicated in VEGF transcription, with increased VEGF expression further exacerbating inflammation (178).

The results suggest that anti-VEGF injections mitigate certain types of damage induced by ischemia-reperfusion injury, such as increased VD in patients with branch RVO post-anti-VEGF injections (154-156), potentially due to reduced expression of inflammatory factors (156), decreased retinal cell apoptosis with significant downregulation of cleaved caspase-3 (177) and inhibition of oxidative stress via activation of the Akt/Nrf2 pathway by the anti-VEGF treatment and N-Acetylserotonin (177). However, potential reasons for reduced blood circulation by anti-VEGF treatments include vasoconstriction (161), upregulation of IL-6, a pro-inflammatory cytokine (26) and increased IOP (140). Although vasoconstriction may reduce edema (162), decreased volumetric blood flow to neuronal tissues could adversely affect the optic nerve. IL-6 induces phosphorylation (S47) of Sirtuin 1 through activation of PI3K/AKT/mTOR signaling, thereby inhibiting Sirtuin 1 activity and increasing acetylation of E2F transcription factor 1, consequently elevating apoptosis in RPE cells under OS (27). Additionally, elevated adhesion molecule levels during inflammation can cause leukocyte arrest, potentially obstructing the capillary lumen (156) and resulting in decreased blood flow to neuronal tissue. The volume of anti-VEGF injections could induce ischemia mediated by elevated IOP, possibly leading to decreased blood flow to neuronal tissue (8). However, the increase in IOP due to the volume of anti-VEGF injections typically returns to baseline over time. Therefore, further investigation is required to determine whether anti-VEGF affects the optic nerve due to decreased volumetric blood flow to neuronal tissues caused by increased IOP and if so, whether retinal neuronal cell damage is related to recurrent vascular defects, or whether retinal blood vessels become more susceptible to damage from repeated elevated IOP in patients with retinopathy undergoing multiple anti-VEGF injections remains to be determined, particularly in those with pre-existing vasculopathy or glaucoma.

In patients with retinopathies post-injection, no statistically significant changes in the blood circulation parameters were observed in the fellow eye, accompanying the changes in the affected eyes (12,133,134,154,157). It is possible that anti-VEGF treatments do not enter the systemic circulation in a concentration at a high enough level to affect the circulation of the fellow eyes.

No correlations have been observed between blood circulation parameters and the different types of anti-VEGF treatments (132,162) or the number of anti-VEGF injections (85).