The analysis of foveal microvascular anomalies in retinopathy of prematurity after anti-vascular endothelial growth factor therapy using optical coherence tomography angiography

Abstract

Background

To investigate the quantitative vascular and structural differences in the foveal region of the eyes in retinopathy of prematurity children with or without anti-vascular endothelial growth factor (VEGF) therapy and healthy children using optical coherence tomography angiography (OCTA).

Methods

This cross-sectional study analyzed 75 eyes from 44 subjects, categorized into four groups: ROP children treated with Conbercept or Ranibizumab, spontaneously regressed ROP, and healthy age-matched children. Using spectral-domain OCT and OCTA, we assessed parameters like central foveal thickness (CFT), foveal avascular zone (FAZ), superficial/deep capillary plexus (SCP/DCP), and choroidal vessel density (VD) at the fovea. Correlations between foveal microvasculature, preterm status and visual acuity were evaluated.

Results

Significant differences were found in FAZ area, CFT, and VD-SCP (parafoveal) among the groups. The FAZ area was smaller in ROP children (with/without treatment) than in healthy counterparts(p = 0.009). CFT was higher in the Ranibizumab and spontaneously regressed groups compared to healthy ones (p = 0.043, p = 0.037), while Conbercept-treated children showed no significant difference (p = 0.886). Foveal VD trends were higher in groups A, B, and C compared to group D. FAZ area correlated negatively with CFT, VD-SCP (foveal), and VD-DCP (foveal) (p < 0.001, p < 0.001, p = 0.001), and positively with choroidal VD (p = 0.012). CFT showed positive correlations with VD-SCP (foveal) and VD-DCP (foveal) (p = 0.003, p = 0.001).

Conclusion

ROP children exhibit a smaller FAZ area compared to healthy group, with no significant difference noted when comparing the use of different anti-VEGF agents. ROP children have a thicker CFT than healthy children, except for those treated with Conbercept. Furthermore, microvascular irregularities were correlated with central foveal thickness.

Introduction

Retinopathy of prematurity (ROP) is characterized by abnormal retinal vascular development in premature infants and remains a leading cause of preventable childhood blindness [1]. The reported incidence of ROP varies significantly, with notably higher rates in low- and middle-income countries compared to high-income countries [1, 2]. With advancements in reproductive technology and preterm care leading to increased survival rates among preterm infants, the incidence of ROP has risen substantially [3, 4].

While laser photocoagulation remains the gold standard for ROP treatment [3], the widespread application and significant successes in using anti-vascular endothelial growth factor (VEGF) drugs for intraocular neovascular diseases have led to an increasing number of ophthalmologists adopting and recommending intravitreal injections of these agents for ROP treatment [5–8]. The simplicity of the technique and lower risk of inducing myopia or peripheral visual field defects make it preferable to traditional laser photocoagulation therapy in clinical practice. Despite the growing body of research on the long-term effects of ROP on macular structure in recent years, comparative studies examining long-term retinal changes across different treatments are relatively scarce, particularly in relation to different types of anti-VEGF drugs, such as Ranibizumab versus Conbercept. Ranibizumab is a recombinant, humanised monoclonal antibody that binds to VEGF-A with a high affinity to making it inert, whilst Conbercept is a 143-kDa designed fusion protein that binds to all VEGF-A isoforms as well as the related VEGFR-1 ligands, VEGF-B and PlGF. Considering the difference in affinity for mentioned molecules, the latter binds to VEGF 50 times better than Ranibizumab.

Spectral-domain OCT (SD-OCT) technology has made remarkable strides in evaluating structural and quantitative retinal changes in various retinal diseases. Optical coherence tomography angiography (OCTA), an extension technique based on SD-OCT, utilizes the motion of red blood cells within vessel lumens as a contrast mechanism for visualizing microvascular beds in vivo. This technique forms the algorithmic basis for quantitative analysis. In contrast to traditional fundus fluorescein angiography (FFA), OCTA offers non-invasiveness, rapidity, high-resolution imaging, and three-dimensional capabilities. Furthermore, OCTA enables segmented observation of different vascular layers that may be difficult to discern and separate due to fluorescein sodium leakage and the two-dimensional nature of FFA [9–11]. This capability allows OCTA to elucidate morphological features of immature retinal microvasculature, aberrant shunt vessels, and neovascularization in various retinal and choroidal pathologies [11–13], thereby enhancing our understanding of pathogenesis. Its non-invasive nature has made OCTA one of the most widely used imaging techniques for both adults and children.

In this study, we compared quantitative vascular and structural parameters among three groups: children with ROP treated with anti-VEGF therapy (Conbercept or Ranibizumab), children with a history of spontaneously regressed ROP, and age-matched term-born healthy children. The insights gained from microvascular and structural changes will contribute to a deeper understanding of the pathological mechanisms underlying the disease and the visual development of affected children.

Methods

Patients

In this retrospective case-control study conducted at Peking University People’s Hospital (Beijing, China), all children diagnosed with ROP who underwent OCTA examinations from January 2019 to December 2021 were invited to participate. Full-term healthy children with matched ages were enrolled as normal controls. All the children in our study were of Chinese descent. The study was approved by the hospital’s ethics committee and adhered to the Declaration of Helsinki. Informed consent was obtained from all participants’ guardians. Inclusion criteria: (1) For the treatment received groups, individuals diagnosed with stage 3 ROP in zone 2, with or without plus disease, were enrolled. In contrast, for the group with spontaneously regressed ROP, individuals with a severity of the condition less than stage 3 in zone 2 were enrolled. All diagnoses were confirmed by two experienced ophthalmologists using a binocular indirect ophthalmoscope under anesthesia; (2) Did not receive any intraocular treatment or only received intravitreal anti-VEGF therapy (Ranibizumab or Conbercept) once; (3) Could endure OCTA examination with good cooperation. Exclusion criteria: (1) Received extra treatments during the follow-up period including additional retinal laser photocoagulation, vitrectomy and scleral buckling surgery, etc. (2) Axial length longer than 26 mm or refractive status (spherical equivalent) smaller than − 6.00 diopter of sphere (DS); (3) Combined with cataract, glaucoma, uveitis, or other retinal/choroidal diseases.

The staging, division, presence or absence of plus lesions, treatment indications and treatment procedures of ROP were confirmed by two experienced retinal specialists with reference to the International Classification of Retinopathy of Prematurity [14]. All the enrolled children were divided into four groups according to the treatment and disease process: Group A: children with the diagnosis of ROP, and received the intravitreal injection treatment of 0.25 mg (0.025 ml) Conbercept (Chengdu Kanghong Biotechnologies Co. Ltd.); Group B: children with ROP treated with 0.25 mg (0.025 ml) Ranibizumab (Lucentis; also Genentech, Inc.); Group C: children with a history of spontaneously regressed ROP without any ocular treatment; Group D: full-term and age matched healthy children.

All the participants had received the following eye examinations: best-corrected visual acuity (BCVA) assessment, cycloplegic refraction, slit lamp examination, and fundus examination. For analysis, the Snellen BCVA was transformed to logMAR format.

Cycloplegic refraction

After all the eyes of participants were dilated with compound tropicamide, a same optometrist would perform refraction using an automatic refractor (RM8900, Japan TOPCON company).

OCTA examination

The RTVue XR Avanti-OCTA instrument (Optovue, U.S.) system was employed for OCTA examinations in the present study. The device uses 840 nm super-luminescent diodes to scan a 3 mm × 3 mm macular region of retinal vessels at a frequency of 70,000 frames per second. The real-time tracking system could correct eye movements and reduce vascular artifacts. All subjects received repeated OCTA and OCT scans completed by the same proficient operator, with the highest-resolution images retained. Macular vascular densities (VDs) at the level of superficial capillary plexus (SCP) and deep capillary plexus (DCP) (including the whole-image, foveal, and parafoveal sections), as well as the size of the foveal avascular zone (FAZ) area (mm²) and central foveal thickness (CFT) were collected and analyzed. FAZ area, VD at different macular sections and levels of retina, and retinal thickness were automatically obtained from the built-in software. The FAZ area was defined as a clinically evident discrete avascular area in the central fovea without crossing capillaries throughout the entire retina. If blood vessels signals were present on the center of fovea, then the FAZ area was defined as zero. En face images of the SCP, defined as the layer from the internal limiting membrane (ILM) to the inner plexiform layer (IPL), and the DCP, defined as the layer from the outer border of the IPL to the outer plexiform layer (OPL), were visualized automatically by segmenting two separate slabs defined by arbitrary segmentation lines, which are created by the device software. With regard to VD, the “parafovea” section is defined as an annulus centered on the fovea with 1 and 3 mm inner and outer ring diameters. Choroid VD was manually calculated as the area ratio of choroidal occupied within a circle of 1 mm radius centered on the foveal pit [15]. The average retinal thickness within the middle 1-mm diameter ring was defined as CFT. The results of OCTA were examined by an experienced ophthalmologist to prevent errors from the software.

Statistical methods

SPSS software (version 26, SPSS Corporation, Chicago, Illinois, U.S.) was used for statistical analysis. Continuous variables were expressed as means and standard deviation. Chi-square was used to compare the qualitative data (sex). The distribution of quantitative data was analyzed with non-parametric test, and the Bonferroni method was used as a post hoc analysis. In all ROP children, the correlation between foveal microvasculature, preterm status, visual acuity, as well as the correlation between FAZ area, CFT and VD were assessed using the Pearson correlation coefficient. Statistical significance was defined as a P value < 0.05, and it was defined as P value < 0.008 after Bonferroni correction.

Results

A total of 44 subjects participated in the study. There were 8 subjects (15 eyes) in Group A, including 4 boys and 4 girls, 14 subjects (26 eyes) in Group B, including 9 boys and 5 girls, 11 subjects (14 eyes) in Group C, including 8 boys and 3 girls, and 11 subjects (20 eyes) in Group D, including 7 boys and 4 girls. In Group A, 12 eyes (12/15, 80%) were diagnosed with plus disease, and the number of Group B was 19 eyes(19/26, 73.1%). There was no significance difference (p = 0.619).There was significant difference in gestational age (GA) and birth weight (BW) between ROP patients (including groups A, B and C) and controls (p < 0.001), but the post-hoc analysis did not show any significance from Group A to C (BW: A-B p = 0.855; A-C p = 0.849; B-C p = 0.973. GA: A-B p = 0.338; A-C p = 0.475; B-C p = 0.891), which ensured the comparability between the ROP groups. There was also significant difference of BCVA and refraction between ROP patients and controls (p < 0.001). According to the pairwise comparisons, the mean BCVA for group A, B, C was significantly worse than the eyes in group D (p < 0.001), and there was no significant difference between group A, B and C. The spherical equivalent (SE) for group C was significantly higher than that of group A, B, D (p = 0.019, p = 0.001, p = 0.017). There was no difference between all the groups with regard to gender, mean ages at the time of receiving OCTA examinations. The detailed baseline characteristics of the children in four groups are shown in Table 1.

Table 1.

Clinical characteristics of the study groups

	Group A	Group B	Group C	Group D	P value
Number of children	8	14	11	11
Number of eyes	15	26	14	20
Sex M/F	4/4	9/5	8/3	7/4	0.798
Age (years)	5.09 ± 0.34	5.04 ± 1.08	5.50 ± 1.86	5.55 ± 0.50	0.078
GA (weeks)	29.04 ± 1.59	29.93 ± 2.02	30.04 ± 1.90	38.84 ± 1.17	<0.001 *
BW (g)	1342.00 ± 342.05	1354.38 ± 371.82	1393.36 ± 325.5	3434.5 ± 373.14	<0.001 *
BCVA (logMAR)	0.11 ± 0.07	0.12 ± 0.14	0.16 ± 0.13	0.03 ± 0.05	<0.001 *
SE (diopters)	0.62 ± 1.43	0.01 ± 1.28	1.38 ± 0.74	0.66 ± 1.78	0.007 *

Results are expressed as mean ± standard deviation for continuous parameters. Kruskal-Wallis test was used to compare, *: p < 0.05. Group A: patients treated with intravitreal Conbercept; Group B: patients treated with intravitreal Ranibizumab; Group C: patients who had spontaneously regressed retinopathy of prematurity; Group D: term born healthy children

Abbreviation: GA: gestational age; BW: birth weight; BCVA: best-corrected visual acuity; SE: spherical equivalent

The pairwise comparison showed that the FAZ area of the group A, B, C was significantly lower than that of group D (p = 0.004, p = 0.006, p = 0.012). The CFT of group B, C was significantly higher than that of the group D (p = 0.043, p = 0.037), while there was no significant difference in the CFT between group A and group D. The mean VD-SCP (foveal) and VD-DCP (foveal) of group A, B, C were higher than group D, though there was no statistical difference existed. Meanwhile, the VD-SCP (parafoveal) of the group A, B was significantly lower than that of group C (p = 0.024, p = 0.014). The detailed OCT and OCTA parameters of each group were shown in Table 2. Representative images from the patients of each group are demonstrated in Fig. 1. And the boxplots reflecting the distribution of FAZ, CFT, VD-SCP, VD-DCP in each group are showed in Fig. 2.

Table 2.

Comparison of optical coherence tomography angiography parameters between the groups

	Group A	Group B	Group C	Group D	P value
FAZ area (mm2)	0.14 ± 0.05	0.16 ± 0.07	0.16 ± 0.08	0.24 ± 0.11	0.009 *
CFT (µm)	247.33 ± 7.56	257.46 ± 15.50	261.07 ± 27.33	247.5 ± 18.88	0.037 *
VD-SCP (foveal) (%)	28.40 ± 4.86	26.70 ± 5.98	32.16 ± 8.48	24.43 ± 9.09	0.062
VD-SCP (parafoveal) (%)	48.83 ± 3.52	48.24 ± 5.23	52.63 ± 4.26	50.80 ± 4.88	0.026 *
VD-DCP (foveal) (%)	40.06 ± 7.16	41.11 ± 7.69	46.19 ± 8.98	39.17 ± 9.13	0.119
VD-DCP (parafoveal) (%)	47.47 ± 5.68	50.72 ± 4.22	50.17 ± 3.90	49.90 ± 5.71	0.210
Choroid VD (%)	61.53 ± 5.89	63.03 ± 4.87	64.45 ± 5.60	64.80 ± 4.96	0.127

Results are expressed as mean ± standard deviation for continuous parameters. Kruskal-Wallis test was used to compare, *: p < 0.05. Group A: patients treated with intravitreal Conbercept; Group B: patients treated with intravitreal Ranibizumab; Group C: patients who had spontaneously regressed retinopathy of prematurity; Group D: term born healthy children

Abbreviation: FAZ: foveal avascular zone, CFT: central foveal thickness, VD: vessel density, SCP: superficial capillary plexus, DCP: deep capillary plexus

Fig. 1.

Representative optical coherence tomography angiography (OCTA) images of superficial capillary plexus (SCP), deep capillary plexus (DCP) and foveal avascular zone (FAZ) outlines in each group. Conbercept: children with ROP treated with Conbercept; Ranibizumab: children with ROP treated with Ranibizumab; SR ROP: children with a history of spontaneously regressed ROP without any ocular treatment; Control: full-term and age matched healthy children

Fig. 2.

The boxplots of the distribution of FAZ, CFT, VD-SCP (foveal), VD-DCP (foveal) in 4 groups. (a): the distribution of FAZ area in each group: group A, B, C was significantly lower than that of group D (p = 0.004, p = 0.006, p = 0.012). (b): the distribution of CFT in each group, group A was significantly lower than that of the group B, C (p = 0.044, p = 0.036), and group D was also significantly lower than that of the group B, C (p = 0.043, p = 0.037). (c): the distribution of VD-SCP (foveal) in each group. (d): the distribution of VD-DCP (foveal) in each group. Group A: children with ROP treated with Conbercept; B: children with ROP treated with Ranibizumab; C: children with a history of spontaneously regressed ROP without any ocular treatment; D: full-term and age matched healthy children

The correlation analysis between ROP children revealed that GA and BW were positively correlated with VD-DCP (foveal) (r = 0.291, p = 0.031; r = 0.356, p = 0.008). There was no significant correlation between BCVA and OCT/OCTA parameters. The FAZ area was negatively correlated with CFT, VD-SCP (foveal) and VD-DCP (foveal) (r=-0.507, p < 0.001; r=-0.472, p < 0.001; r=-0.453, p = 0.001), while it was positively correlated with the choroidal VD (r = 0.336, p = 0.012). The CFT was significantly correlated with both VD-SCP (foveal) and VD-DCP (foveal) (r = 0.393, p = 0.003; r = 0.437, p = 0.001). The correlation analysis data are summarized in Table 3.

Table 3.

Univariate linear regression result of functional and structural parameters in ROP children

	GA (weeks)	BW (g)	BCVA (logMAR)	FAZ area (mm)	CFT (µm)
FAZ area (mm)	-0.065,0.64	-0.177,0.20	-0.144,0.30	-	-0.507,<0.001*
CFT (µm)	-0.080,0.56	0.027,0.85	0.044,0.75	-0.507,<0.001*	-
VD-SCP (foveal) (%)	0.072,0.60	0.067,0.63	0.070,0.61	-0.472,<0.001*	0.393,0.003*
VD-SCP (parafoveal) (%)	0.130,0.34	0.050,0.72	-0.041,0.77	-0.050,0.72	0.044,0.75
VD-DCP (foveal) (%)	0.291,0.03*	0.356,0.008*	0.188,0.17	-0.453,0.001*	0.437,0.001*
VD-DCP (parafoveal) (%)	0.077,0.58	-0.027,0.85	0.152,0.27	0.035,0.80	0.188,0.17
Choroid VD (%)	-0.240,0.08	-0.250,0.08	0.179,0.19	0.336,0.01*	0.159,0.25

Results are expressed as Pearson correlation coefficient, p value.*p < 0.05

Abbreviation: GA: gestational age; BW: birth weight; BCVA best-corrected visual acuity: logMAR logarithm of the minimum angle of resolution; FAZ: foveal avascular zone; CFT: central foveal thickness VD: vascular density; SCP: superficial capillary plexus; DCP: deep capillary plexus

Discussion

Using OCTA technology, this study compares structural and vascular density parameters among ROP-affected eyes treated with Conbercept or Ranibizumab, spontaneously regressed lesions, and those of age-matched, full-term healthy children. Additionally, we are the first to evaluate the long-term effects of two anti-VEGF drugs on retinal structure and microcirculation. Our findings reveal a significant reduction in FAZ area and an increased CFT in children with ROP. Interestingly, the choice of anti-VEGF treatment minimally influences FAZ area, while Conbercept demonstrates a comparable CFT to that of healthy children. Furthermore, we observe a negative correlation between FAZ area and foveal VD, as well as CFT.

At present, there is no unified conclusion about the relationship between BCVA and macula structure in children with ROP. Villegas et al. similarly noted that none of the structural factors significantly correlated with BCVA [16]. Balasubramanian et al. found that worse BCVA was associated with a smaller FAZ in a cross-sectional study of 32 preterm children’s eyes [17], while Chen et al. identified higher VD and inner retinal thickness were linked to suboptimal visual acuity [18]. Stoica et al. discovered that better VA correlated with higher BW and GA at birth, as well as a thinner CFT [19]. Embryogenesis demonstrates that the development of the fovea externa is relatively independent from the foveal pit [20], Consequently, despite structural disparities in FAZ and CFT in ROP eyes, visual acuity development may not be significantly affected, aligning with our observations.

Blood vessels in the human optic disc region first appear around week 14–15 [21], and the vascular plexus encircling the central avascular zone begins to form around fetal week 25. The central area of the macula is never vascularized [22, 23]. In contrast, foveal retinal tissue continues to mature for several years after birth. Dubis et al. found that morphologic development of the fovea may be completed by around 17 months [24]. Therefore, alterations in the local microenvironment during this period can influence the development of both macular tissue and the FAZ region. VEGF plays a crucial role in the growth of retinal vessels towards the fovea section. Local hypoxia generated from ganglion cell differentiation propels angioblast migration at the vascular front by promoting local VEGF synthesis within astrocyte precursor cells. This process leads to capillary expansion into the previously hypoxic retina, following the template of astrocytes, thereby alleviating local hypoxia [21]. Simultaneously, the foveal center expresses more anti-angiogenic molecules than the perifoveal and peripheral retinal areas, mainly pigment epithelium-derived factor (PEDF), brain natriuretic peptide, and others. The balance between these factors is crucial to maintaining the FAZ [25]. Consequently, during FAZ and foveal pit formation, the increase in VEGF and aforementioned anti-angiogenic molecules in ROP-afflicted eyes may disrupt this balance, resulting in FAZ reduction or disappearance, elucidating the significant reduction in FAZ area associated with ROP. However, our present study found no significant difference in FAZ area between groups A, B, C, demonstrating that the choice of drug may minimally affect FAZ area. One plausible explanation is that FAZ reduction may have occurred prior to anti-VEGF agent administration, with FAZ differences primarily stemming from the ROP pathology itself rather than treatment choice. Peripheral vascular morphological anomalies confirmed through FFA yielded a similar result, showing no significant difference between ROP neonates who received intravitreal Conbercept or Ranibizumab [26].

A smaller FAZ is more likely to coincide with a thicker CFT, as previous studies have shown. Asli Vural et al. reported that FAZ was negatively correlated with CFT (r=-0.27, p = 0.03) [27]. Bowl et al. has also found that the mean values of the CFT were highest in the spontaneously regressed ROP group, intermediate in the premature children without ROP group, and lowest in the age-matched term-born group [28]. Our study also revealed that the CFT of ROP children treated with Ranibizumab and spontaneously regressed was higher than that of age matched healthy children. A decrease in the migration ability of retinal neurons in the inner layer of the fovea primarily contributes to increased retinal thickness [29]. The foveal pit forms through the centrifugal migration of inner retinal neurons away from the fovea and the centripetal migration of cone cell nuclei towards the center [30]. It is believed that FAZ section is more elastic and malleable than the vascularized retina around it under the influence of intraocular pressure [31]. The presence of capillaries mechanically interferes with the centrifugal displacement of ganglion cell layer and inner nuclear layer cells, thus a smaller FAZ limits foveal pit widening, subsequently increasing foveal thickness [29]. Our study also showed the relationship between FAZ and CFT, as the CFT of group B, C was higher than group D. Interestingly, our study further discovered no significant difference in CFT between ROP treated with Conbercept and age-matched healthy children. While there has been relatively limited research on the use of Conbercept for treating ROP outside of China, a wealth of literature has demonstrated the effectiveness and safety of Conbercept in ROP treatment [32, 33]. Moreover, 0.15 mg Conbercept has already been confirmed as an effective therapeutic choice for ROP in China [34]. We speculate that Conbercept may impact CFT through alternative pathways, although further longitudinal research is needed to confirm this conclusion.

Takagi et al. discovered that vessel density at the fovea in the ROP group was significantly higher than in the control group (35.7 vs. 30.1%, p < 0.01) [35]. Chen et al. also discovered increased foveal VD (SCP and DCP) in intravitreal Bevacizumab-treated preterm children in comparison to healthy controls [18]. Asli Vural et al. found smaller FAZ, and increased central foveal VD in all eyes with ROP regardless of treatment options [27]. In our study, we also found that the mean foveal VD-SCP and VD-DCP in groups A, B, and C were higher than those in group D, though not significantly different. We speculate that the lack of statistical significance may be attributed to the relatively small sample size. Nevertheless, the trend of quantitative parameters remained consistent with the previous studies mentioned above. Furthermore, we observed a strong negative correlation between FAZ area and foveal VD at both the superficial capillary plexus (SCP) and deep capillary plexus (DCP). We suspect there are two possible reasons for this phenomenon: First, an increase in VEGF concentration might lead to a compensatory elevation of retinal blood vessel density in the foveal region. Second, the physiological displacement of the inner retina might be incomplete due to the smaller avascular area in ROP.

In terms of choroidal VD, Rezar-Dreindl et al. conducted a cross-sectional study involving 15 children (30 eyes) with a history of ROP, which revealed no statistically significant differences in choroidal VD between the ROP and control groups [36]. Our study produced similar findings; choriocapillary vascular density (VD) did not exhibit an association with the choice of treatment or drug, indicating that the choriocapillaris may not undergo significant alterations during the course of the disease. Nonetheless, diverse conclusions have been drawn regarding the vascularization of other layers within the choroidal tissue. Lavric et al. identified that children with ROP exhibited a markedly reduced choroidal vascularity index, a metric defined as the proportion of luminal areas to the total sub-foveal choroidal area, signifying compromised choroidal vascularity [37]. A diminished choroidal vascularity index may imply a reduction in oxygen delivery to the outer retina, and a thinner choroidal thickness was also associated with worse VA [38]. As of now, studies examining the relationship between the choroid and ROP remain relatively limited, necessitating further research to elucidate the role of the choroid in ROP development.

Our study does have some limitations. Firstly, we did not include children who had undergone laser treatment or those with a history of prematurity without ROP. It’s plausible that prematurity itself or laser treatment may have contributed to the retinal vascular and structural changes outlined above. Secondly, all the children in our study were of Chinese descent. It is important to acknowledge that retinal vascular or structural parameters may vary based on ethnicity. We have not been able to find a possible explanation for the influence of Conbercept in CFT and the highest SE in group C in the relevant studies. Finally, the sample size of this study is relatively small, necessitating expansion for more robust validation.

In conclusion, we leveraged OCTA technologies to investigate the foveal structural and microvascular characteristics of ROP-affected children (treated with Conbercept, Ranibizumab, and those with spontaneously regressed cases) alongside age-matched children. OCTA, as a non-invasive tool, provides a clearer understanding of macular microstructural and microvascular abnormalities in children with ROP, including reduced FAZ and increased CFT, as well as correlations between OCTA parameters and demographic data. Additionally, our study affirms that the group treated with Conbercept demonstrated comparable CFT to that of healthy children. However, further research is imperative to gain deeper insights into the developmental processes of retinal vessels and structures in ROP. This will aid in discerning the advantages and drawbacks of various treatment modalities.

Abbreviations

ROP

Retinopathy of prematurity

OCTA

Optical coherence tomography angiography

SD-OCT

Spectral domain-optical coherence tomography

CFT

Central foveal thickness

FAZ

Foveal avascular zone

SCP/DCP

Superficial/deep capillary plexus

VD

Choroidal vessel density

FFA

Fundus fluorescein angiography

BCVA

Best-corrected visual acuity

GA

Gestational age

BW

Birth weight

Author contributions

W. L., L. G. and Y. C. contributed to the conception of the work and the case acquisition. Y. Ch., X. D. and H. X. did the data collection and data analysis. D. L., M. Z. and X. Z. contributed to the literature search. W. L, L. G. and Y. C. written the original draft preparation. X. S. and J. L. edited the manuscript. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Key R&D Program of China, No. 2020YFC2008200, and the National Natural Science Foundation of China (Grant No. 81970815).

Data availability

The datasets used and analyzed in the course of the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethical Committee of Peking University People’s Hospital.It was conducted according to the guidelines of the Declaration of Helsinki. Informed consent was obtained from all participants’ guardians.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.