Characteristics of Retina, Choroid, and Retrobulbar Blood Flow in Myopes With Posterior Staphyloma Using Ultra-Widefield OCTA and CDI

Abstract

Purpose

To explore the characteristics of retinal, choroidal, and retrobulbar blood flows in myopic patients with posterior staphyloma (PS) using ultra-widefield optical coherence tomography angiography (UWF-OCTA) and Color Doppler imaging (CDI).

Methods

The retrospective cross-sectional study enrolled 134 adults with high myopia from the Eye and ENT Hospital of Fudan University from December 2021 to September 2022. After propensity score matching, 45 eyes of 30 patients and 45 eyes of 32 patients were included in the PS and non-PS (NPS) groups, respectively. Retinal, choroidal, and retrobulbar blood flow parameters were obtained from UWF-OCTA and CDI. Parameters were compared between the PS and NPS groups, as well as among different PS types.

Results

There were no statistical differences in age, sex, axial length, spherical equivalent, intraocular pressure, or best-corrected visual acuity between the PS and NPS groups or in the different types of PS. The PS group had lower fovea, parafovea, and perifovea choroidal vessel volume (CVV) compared to the NPS group (all P < 0.05). The wide and narrow macular types of PS had significantly thinner choroidal thickness (CHT), lower CVV, and lower choroidal stromal volume (CSV) than other groups (all P < 0.01). PS occurrence was correlated with the decrease in CVV and posterior ciliary artery end-diastolic velocity (PCA-EDV), and the decrease in PCA-EDV was an associated factor for PS. The wide and narrow macular types of PS were correlated with the decrease of CHT, CVV, and CSV; the decrease in perifovea CHT was an associated factor for wide and narrow macular types of PS.

Conclusions

Change in blood flow of the PCA was associated with PS occurrence. Reduced CHT in the temporal perifoveal region was a factor associated with both wide and narrow macular types of PS. Further longitudinal study will help to investigate the causal relationship between PS and changes in blood flow.

Posterior staphyloma (PS), a characteristic hallmark of pathological myopia, is defined as the localized outward bulging of the scleral wall with a curvature radius smaller than that of the surrounding area.¹ Previous studies have indicated that PS is a critical risk factor for the progression of myopic maculopathy² and is significantly associated with decreased visual acuity.³ With the rapid increase in the incidence of myopia and the aging of the population, the incidence of PS is expected to increase sharply in the future.

Despite the importance of PS, its specific pathogenic mechanisms have not yet been elucidated. For a long time, the sclera was considered the primary tissue involved in PS. More recently, however, the role of the choroid in the development of myopia has been increasingly recognized. Ohno-Matsui and Jonas⁴ reported that the choroid may be the primary tissue affected during PS development. Early research found that PS is associated with choroidal thinning and reduced choroidal perfusion.⁵ However, due to the susceptibility of choroidal thickness (CHT) to various factors, relying solely on a single parameter to assess PS-related choroidal changes may not produce accurate results.^6,7 Later studies have reported that the novel choroidal quantitative parameter choroidal vascularity index (CVI) is more stable than CHT, with less influence from physiological factors, allowing for a more accurate quantification of PS-related changes.⁶

With the flourishing development of optical coherence tomography (OCT) imaging technology, a new generation of ultra-widefield OCT (UWF-OCT) has shown significant improvements in scanning range, depth, speed, and resolution and provides detailed characteristics of the changes in various layers of the fundus. Some researchers have proposed that UWF-OCT could replace three-dimensional magnetic resonance imaging (3D-MRI) for assessing changes in PS, offering new insights into its pathogenesis.⁸ Using the latest generation of UWF-OCTA combined with Color Doppler imaging (CDI), this study aimed to comprehensively evaluate the characteristics and correlations of retinal, choroidal, and retrobulbar blood flow changes in patients with PS.

Methods

Study Participants

This study enrolled 134 adults with high myopia from the Eye and ENT Hospital of Fudan University between December 2021 and September 2022. The inclusion criteria were (1) age ≥ 18 years; (2) spherical equivalent (SE) of both eyes < −8.0 diopters (D) or axial length (AL) > 26.5 mm; and (3) intraocular pressure (IOP) < 21 mm Hg. The exclusion criteria were as follows: (1) poor cooperation or inability to comply with procedures, such as saccadic movements or nystagmus and poor gaze stability; (2) ocular media opacity, such as cataracts; (3) severe retinal conditions without pathological myopia, such as choroidal neovascularization and age-related macular degeneration; (4) history of ocular surgeries; and (5) systemic diseases affecting blood circulation, such as hypertension and diabetes. The study received approval from the Ethics Review Committee of the Eye and ENT Hospital of Fudan University (ChiCTR1800017594). All procedures were conducted in accordance with the tenets of the Declaration of Helsinki, and participants provided informed consent before their examinations.

Study Design

To mitigate the potential impact of AL on the results, the study adopted a case–control design with 1:1 matching of AL. Matching of AL between the control group and the case group was performed using propensity scores, with a caliper value of 0.2 ×bSD (logit propensity score). Following a previously reported method for diagnosing PS based on OCT images,⁹ we used the latest 96 radial UWF-OCTA scan lines for diagnosing PS. According to the classification method proposed by Ohno-Matsui,¹⁰ PS was classified into the following categories: I wide, macular staphyloma; II narrow, macular staphyloma; III peripapillary staphyloma; IV nasal staphyloma; V inferior staphyloma and VI others. Considering the unbalanced sample size of the different types, we categorized all types except type I and type II into single “other types” in the research. The diagnostic criteria for PS were as follows: (1) the choroid gradually thins from the periphery to the edge of the staphyloma and then thickens toward the posterior pole; (2) the sclera protrudes inward at the edge of the staphyloma; and (3) the sclera behind the edge of the staphyloma shifts backward (Fig. 1).

Figure 1.

UWF-OCTA images of different PS types and NPS patients. (A, B) Left-eye images of a 35-year-old male patient with wide macular PS (AL = 29.95 mm, SE = −13.625 D). (C, D) Left-eye images of a 31-year-old female patient with narrow macular PS (AL = 30.91 mm, SE = −12.5 D). (E, F) Right-eye images of a 44-year-old male patient with other type of PS (AL = 27.35 mm, SE = −11.0 D). (G, H) Right-eye images of a 28-year-old female patient without PS (AL = 27.73 mm, SE = −10.0 D). The white arrows indicate the edge of PS. The choroid is thinnest at the edge of PS, gradually thickening from the edge to the periphery; to the posterior part of the fundus, the sclera protrudes inward at the edge of PS, and the posterior sclera shifts backward.

Ophthalmic Examination

All enrolled patients underwent comprehensive ophthalmic examinations in addition to UWF-OCTA and CDI. Anterior segment examinations were performed using a slit-lamp biomicroscope (BQ900; Haag-Streit AG, Köniz, Switzerland). AL values were acquired from three consecutive measurements and averaged using an IOLMaster 500 (Carl Zeiss Meditec AG, Jena, Germany). IOP was monitored using a TX-10 Tonometer (Canon, Tokyo, Japan). Experienced optometrists evaluated the manifest refraction in cycloplegia using a comprehensive refractometer (RT-5100; Nidek, Aichi, Japan).

Color Doppler Imaging Analysis

The retrobulbar blood flow was evaluated in vessels including the central retinal artery (CRA), posterior ciliary artery (PCA), and ophthalmic artery (OA) using an ATL HDI 5000 CDI system (Philips, Amsterdam, the Netherlands). The examination procedures were consistent with those described in previous studies.^11,12 Retrobulbar parameters including peak systolic velocity (PSV), end diastolic velocity (EDV), and pulsatile index (PI) were acquired. The resistance index (RI) was calculated using the following formula:

$$\begin{array}{c}\hfill RI=\frac{PSV-EDV}{PSV}.\end{array}$$

Ultra-Widefield Optical Coherence Tomography Angiography

UWF-OCTA (VG200D; SVision Imaging, Ltd., Luo Yang, China) examinations were conducted by the same experienced ophthalmologist from 13:00 to 16:00 hours under constant illumination (15 lux) to minimize the influence of daily rhythms and light on the outcomes. UWF-OCTA images were acquired using a 1050-nm swept-source laser with a speed of 200,000 A-scans per second, axial resolution of 5 µm, lateral resolution of 15 µm, and a maximum scanning depth of 12 mm. The 96 radial UWF-OCTA scan centered on the fovea was performed to determine the presence and type of PS. Choroidal vascular parameters were measured using UWF-OCTA, with each scan consisting of 1280 A-scans and 1034 B-scans, repeated twice and averaged. The scanning range was 21 mm × 21 mm, centered on the fovea, and an automatic eye-movement correction program was set. The built-in algorithm automatically segmented the retinal superficial vascular complex (from 5 µm above the inner limiting membrane to two-thirds inward of the ganglion cell layer and inner plexiform layer), deep vascular complex (from one-third outward of the ganglion cell layer and inner plexiform layer to 25 µm below the lower margin of the inner nuclear layer), and choroidal layer. The built-in software automatically calculated the superficial vessel density (SVD), deep vessel density (DVD), CHT, CVV, and CVI (Fig. 2). SVD and DVD were calculated as the percentage area occupied by superficial and deep retinal blood vessels within the region of interest, determined from binarized OCTA images shown in Figures 2A and 2B. CHT was determined directly from the segmented choroidal layer. Similarly, CVV and CVI were derived from the segmented choroidal vessels. In Figures 2C and 2D, the dark and light pixels represent CVI and CVV values at corresponding positions, with higher values resulting in brighter images. The calculation methods for these parameters have been described in previous literature.^12,13 Subsequently, the choroidal stromal volume (CSV) was calculated using the following formula:

$$\begin{array}{c}\hfill CSV=\frac{CVV-CVI\times CVV}{CVI}\end{array}$$

Figure 2.

Automatically calculated retinal and choroidal vascular parameters of PS patients with the Early Treatment Diabetic Retinopathy Study (EDTRS) grid. (A) Vessel density of the superficial vascular complex with the EDTRS grid. (B) Vessel density of the deep vascular complex. (C) CVI with the EDTRS grid. (D) CVV with the EDTRS grid.

Statistical Analysis

Data processing was performed using SAS 9.4 (SAS Institute, Cary, NC, USA). Normality of each variable was confirmed using the Shapiro–Wilk test. A linear mixed model analysis with inter-eye correlation adjustment was performed to compare SVD, DVD, CHT, choroidal vessels, and retrobulbar blood flow between the PS and non-PS (NPS) groups and different PS types. The Tukey–Kramer test was used for post hoc analysis, if a variable showed significant difference among PS subgroups. Single-factor and stepwise logistic regression analyses were conducted to identify associated factors for the occurrence of PS and the wide and narrow macular type of PS. The threshold for statistical significance was P < 0.05.

Results

Clinical Characteristics of Patients

A total of 268 eyes of 134 highly myopic patients were recruited for this study. After 1:1 matching of axial length based on propensity scores, successful matches were achieved in 90 eyes of 55 patients, with a male-to-female ratio of 2:3. This included 45 eyes of 30 patients in the PS group and 45 eyes of 32 patients in the NPS group. Twenty-six eyes from 19 patients had wide macular PS, five eyes from four patients had narrow macular PS, and 14 eyes from 11 patients had other types of PS. There were no statistically significant differences in age, sex, AL, SE, IOP, and best-corrected visual acuity (BCVA) between the PS and NPS groups or among the different types of PS (all P > 0.05) (Tables 1 and 2).

Table 1.

Clinical Characteristics of PS and NPS Patients

Parameter	NPS	PS	P
Patients/eyes, n	32/45	30/45	—
Male:female ratio, n	14:18	9:21	0.199
Age (y), mean ± SD	26.88 ± 4.87	28.47 ± 6.21	0.216
SE (D), mean ± SD	−9.38 ± 3.75	−10.15 ± 2.24	0.240
BCVA (logMAR), mean ± SD	0.001 ± 0.007	0.003 ± .016	0.407
IOP (mm Hg), mean ± SD	15.26 ± 2.38	15.02 ± 2.36	0.635
AL (mm), mean ± SD	27.88 ± 0.75	27.86 ± 0.74	0.901

Table 2.

Clinical Characteristics of Different Types of PS

Parameter	Wide Macular PS	Narrow Macular PS	Other PS	P
Patients/eyes, n	19/26	4/5	11/14	—
Male:female ratio, n	7:12	1:3	6:5	0.481
Age (y), mean ± SD	29.579 ± 7.105	27 ± 4.690	28.182 ± 5.600	0.716
SE (D), mean ± SD	−10.303 ± 2.068	−11.525 ± 2.171	−9.366 ± 2.414	0.132
BCVA (logMAR), mean ± SD	0.002 ± 0.009	0 ± 0	0.007 ± 0.026	0.568
IOP (mm Hg), mean ± SD	14.715 ± 2.407	15.72 ± 1.834	15.35 ± 2.473	0.508
AL (mm), mean ± SD	28.04 ± 0.786	27.886 ± 0.739	27.531 ± 0.538	0.116

Comparison of Retinal, Choroidal, and Retrobulbar Vascular Parameters Between PS Groups

The differences in OCTA and CDI parameters between the PS and NPS groups are shown in Table 3. Compared to the NPS group, the PS group had lower central fovea, parafovea, and perifovea CVV (NPS group vs. PS group: fovea CVV, 0.062 ± 0.015 mm³ vs. 0.055 ± 0.015 mm³; parafovea CVV, 0.499 ± 0.110 mm³ vs. 0.451 ± 0.106 mm³; perifovea CVV, 1.68 ± 0.326 mm³ vs. 1.499 ± 0.356 mm³; all P < 0.05). However, there were no significant differences in SVD, DVD, CHT, CSV, CVI, or any other retrobulbar vascular parameters between the two groups (all P > 0.05) (Table 3).

Table 3.

Retinal, Choroidal, and Retrobulbar Vascular Parameter Comparisons in the PS and NPS Groups

Parameter	NPS, Mean ± SD	PS, Mean ± SD	P
Retinal vessels
Fovea SVD (%)	33.435 ± 8.553	35.818 ± 13.261	0.315
Fovea DVD (%)	40.729 ± 7.766	39.328 ± 9.622	0.445
Parafovea SVD (%)	72.502 ± 5.248	73.392 ± 5.339	0.424
Parafovea DVD (%)	64.486 ± 4.187	63.482 ± 5.174	0.297
Perifovea SVD (%)	75.527 ± 4.92	76.594 ± 5.038	0.315
Perifovea DVD (%)	63.189 ± 4.305	61.688 ± 6.502	0.199
Choroidal vessels
Fovea CHT (µm)	208.304 ± 61.103	188.397 ± 59.276	0.122
Fovea CVV (mm3)	0.062 ± 0.015	0.055 ± 0.015	0.042
Fovea CSV (mm3)	0.101 ± 0.036	0.113 ± 0.134	0.551
Fovea CVI (%)	0.392 ± 0.071	0.379 ± 0.096	0.472
Parafovea CHT (µm)	209.281 ± 57.275	191.825 ± 57.646	0.155
Parafovea CVV (mm3)	0.499 ± 0.11	0.451 ± 0.106	0.039
Parafovea CSV (mm3)	0.803 ± 0.266	0.756 ± 0.277	0.412
Parafovea CVI (%)	0.394 ± 0.064	0.387 ± 0.067	0.618
Perifovea CHT (µm)	215.383 ± 51.972	197.895 ± 56.886	0.133
Perifovea CVV (mm3)	1.68 ± 0.326	1.499 ± 0.356	0.014
Perifovea CSV (mm3)	2.796 ± 0.786	2.589 ± 0.951	0.262
Perifovea CVI (%)	0.382 ± 0.05	0.389 ± 0.139	0.753
Retrobulbar vessels
CRA
PSV (cm/s)	9.456 ± 2.046	9.473 ± 1.651	0.964
EDV (cm/s)	1.693 ± 0.411	1.609 ± 0.369	0.310
PI	2.148 ± 0.425	2.251 ± 0.514	0.307
RI	0.818 ± 0.04	0.826 ± 0.042	0.324
Nasal PCA
PSV (cm/s)	8.707 ± 1.793	9.536 ± 6.218	0.393
EDV (cm/s)	2.131 ± 0.597	2.024 ± 0.724	0.445
PI	1.656 ± 0.362	1.781 ± 0.382	0.116
RI	0.754 ± 0.053	0.768 ± 0.061	0.250
Temporal PCA
PSV (cm/s)	8.393 ± 1.936	8.438 ± 1.97	0.915
EDV (cm/s)	2.06 ± 0.549	1.818 ± 0.305	0.011
PI	1.655 ± 0.387	1.759 ± 0.424	0.222
RI	0.748 ± 0.068	0.775 ± 0.063	0.053
OA
PSV (cm/s)	31.511 ± 7.582	30.86 ± 9.947	0.728
EDV (cm/s)	5.687 ± 1.879	5.92 ± 2.07	0.579
PI	2.184 ± 0.429	2.146 ± 0.408	0.669
RI	0.816 ± 0.051	0.806 ± 0.045	0.370

Bold terms marked P value smaller than 0.05.

Single-factor variance analysis revealed significant differences in the CHT, CVV, and CSV among the PS types (all P < 0.05) (Table 4). Post hoc multiple comparisons further revealed that, compared to other types of PS, the wide and narrow macular types of PS had significantly thinner CHT, lower CVV, and lower CSV in the central fovea, parafovea, and perifovea (P < 0.01). However, there were no significant differences in any parameter between the wide and narrow macular types of PS (P > 0.05) (Table 4).

Table 4.

Retinal, Choroidal, and Retrobulbar Vascular Parameter Comparisons in Different Types of PS

Parameter	Wide Macular PS, Mean ± SD	Narrow Macular PS, Mean ± SD	Other PS, Mean ± SD	P *	P †	P ‡	P §
Retinal vessels
Fovea SVD (%)	34.71 ± 11.12	34.84 ± 11.95	38.22 ± 17.45	0.733	—	—	—
Fovea DVD (%)	36.64 ± 6.59	39.72 ± 13.86	44.19 ± 11.43	0.062	—	—	—
Parafovea SVD (%)	73.51 ± 4.19	72.87 ± 7.09	73.36 ± 6.87	0.953	—	—	—
Parafovea DVD (%)	63.33 ± 4.43	60.99 ± 11.22	64.66 ± 3.19	0.333	—	—	—
Perifovea SVD (%)	75.94 ± 4.48	77.56 ± 4.54	77.46 ± 6.25	0.645	—	—	—
Perifovea DVD (%)	62.18 ± 5.77	59.64 ± 11.05	61.52 ± 6.27	0.727	—	—	—
Choroidal vessels
Fovea CHT (µm)	174.98 ± 49.78	143.4 ± 44.89	229.39 ± 59.66	0.003	0.466	0.010	0.010
Fovea CVV (mm3)	0.05 ± 0.02	0.05 ± 0.01	0.06 ± 0.02	0.063	—	—	—
Fovea CSV (mm3)	0.12 ± 0.17	0.06 ± 0.03	0.12 ± 0.04	0.638	—	—	—
Fovea CVI (%)	0.38 ± 0.1	0.45 ± 0.09	0.36 ± 0.09	0.218	—	—	—
Parafovea CHT (µm)	178.79 ± 46.89	144.97 ± 46.59	232.78 ± 57.98	0.002	0.381	0.008	0.006
Parafovea CVV (mm3)	0.44 ± 0.1	0.36 ± 0.06	0.52 ± 0.1	0.006	0.217	0.043	0.008
Parafovea CSV (mm3)	0.69 ± 0.23	0.54 ± 0.21	0.95 ± 0.28	0.002	0.418	0.007	0.007
Parafovea CVI (%)	0.4 ± 0.07	0.41 ± 0.05	0.36 ± 0.07	0.206	—	—	—
Perifovea CHT (µm)	185.55 ± 46.16	146.83 ± 36.44	239.06 ± 57.57	0.001	0.252	0.007	0.003
Perifovea CVV (mm3)	1.43 ± 0.31	1.17 ± 0.23	1.75 ± 0.33	0.001	0.187	0.008	0.002
Perifovea CSV (mm3)	2.38 ± 0.88	1.87 ± 0.48	3.24 ± 0.87	0.003	0.477	0.011	0.011
Perifovea CVI (%)	0.41 ± 0.18	0.39 ± 0.04	0.36 ± 0.05	0.561	—	—	—
Retrobulbar vessels
CRA
PSV (cm/s)	9.78 ± 1.69	9.78 ± 1.53	8.79 ± 1.51	0.181	—	—	—
EDV (cm/s)	1.6 ± 0.26	1.78 ± 0.88	1.56 ± 0.28	0.553	—	—	—
PI	2.27 ± 0.5	2 ± 0.53	2.3 ± 0.54	0.496	—	—	—
RI	0.83 ± 0.04	0.82 ± 0.07	0.82 ± 0.03	0.718	—	—	—
Nasal PCA
PSV (cm/s)	10.48 ± 7.96	9.48 ± 1.52	7.8 ± 1.82	0.453	—	—	—
EDV (cm/s)	2.13 ± 0.9	1.8 ± 0.46	1.91 ± 0.3	0.499	—	—	—
PI	1.83 ± 0.38	1.75 ± 0.42	1.7 ± 0.38	0.593	—	—	—
RI	0.78 ± 0.05	0.8 ± 0.07	0.75 ± 0.07	0.165	—	—	—
Temporal PCA
PSV (cm/s)	8.34 ± 2.09	10.34 ± 1.82	7.95 ± 1.43	0.067	—	—	—
EDV (cm/s)	1.78 ± 0.35	1.82 ± 0.26	1.89 ± 0.23	0.479	—	—	—
PI	1.79 ± 0.51	1.79 ± 0.17	1.69 ± 0.31	0.715	—	—	—
RI	0.78 ± 0.07	0.82 ± 0.03	0.76 ± 0.05	0.211	—	—	—
OA
PSV (cm/s)	29.84 ± 11.44	35.24 ± 9.17	31.2 ± 6.9	0.586	—	—	—
EDV (cm/s)	5.9 ± 2.08	6.74 ± 3.59	5.66 ± 1.38	0.646	—	—	—
PI	2.09 ± 0.35	2.51 ± 0.75	2.12 ± 0.31	0.100	—	—	—
RI	0.8 ± 0.05	0.81 ± 0.06	0.82 ± 0.04	0.517	—	—	—

^*Results of one-way ANOVA.

^†Results for wide macular PS versus narrow macular PS.

^‡Results for wide macular PS versus other types.

^§Results for narrow macular type versus other types.

Bold terms marked P value smaller than 0.05.

Relationship Between Posterior Staphyloma Occurrence and Ocular Vascular Parameters

Single-factor logistic regression analysis showed that decreases of CVV in the central fovea (odds ratio [OR] = 8.73 × 10⁻¹⁴), parafovea (OR = 0.016), and perifovea (OR = 0.208), as well as a decrease in posterior ciliary artery end-diastolic velocity (PCA-EDV) (OR = −1.323), were correlated with PS occurrence (all P < 0.05). Considering the potential correlations between variables, stepwise logistic regression analysis revealed that only a decrease in PCA-EDV was an associated factor for PS occurrence (OR = 0.266, P < 0.05) (Table 5).

Table 5.

Logistic Regression of PS Factors

	Single-Factor Analysis				MultiFactor Analysis
Parameter	β	OR	SD	P	β	OR	SD	P
Temporal PCA-EDV (cm/s)	−1.323	0.266	0.554	0.017	−1.323	0.266	0.554	0.017
Fovea CVV (mm3)	−30.069	8.73e-14	14.950	0.044	—	—	—	—
Parafovea CVV (mm3)	−4.164	0.016	2.050	0.042	—	—	—	—
Perifovea CVV (mm3)	−1.569	0.208	0.656	0.017	—	—	—	—

Bold terms marked P value smaller than 0.05.

Relationship Between Different PS Types and Ocular Vascular Parameters

Based on the above results, further analysis of the relationship between different PS types and ocular vascular parameters was performed. Single-factor logistic regression analysis showed that, compared with other types of PS, decrease of fovea CHT, and decreases of CHT, CVV, CSV in parafovea and perifovea area were correlated with the wide and narrow macular types of PS (respective OR values were shown in Table 6; all P < 0.05). Stepwise logistic regression analysis showed that, compared to other types of PS, only a decrease in perifovea CHT was an associated factor for the occurrence of both wide macular PS (OR = 0.978) and narrow macular PS (OR = 0.959; all P < 0.05) (Table 6).

Table 6.

Logistic Regression for Factors of Wide and Narrow Macular PS

	Single-Factor Analysis				Multifactor Analysis
Parameter	β	OR	SD	P	β	OR	SD	P
Wide macular PS*
Fovea CHT (µm)	−0.0194	0.981	0.007	0.009	—	—	—	—
Parafovea CHT (µm)	−0.021	0.979	0.008	0.009	—	—	—	—
Parafovea CVV (mm3)	−8.656	<0.001	3.845	0.024	—	—	—	—
Parafovea CSV (mm3)	−4.248	0.014	1.603	0.008	—	—	—	—
Perifovea CHT (µm)	−0.022	0.978	0.008	0.008	−0.022	0.978	0.008	0.008
Perifovea CVV (mm3)	−3.358	0.035	1.267	0.008	—	—	—	—
Perifovea CSV (mm3)	−1.394	0.248	0.524	0.008	—	—	—	—
Narrow macular PS*
Fovea CHT (µm)	−0.033	0.968	0.013	0.01	—	—	—	—
Parafovea CHT (µm)	−0.004	0.964	0.014	0.008	—	—	—	—
Parafovea CVV (mm3)	−19.493	<0.001	7.765	0.012	—	—	—	—
Parafovea CSV (mm3)	−7.408	<0.001	2.855	0.01	—	—	—	—
Perifovea CHT (µm)	−0.042	0.959	0.015	0.005	−0.042	0.959	0.015	0.005
Perifovea CVV (mm3)	−6.666	0.001	2.428	0.006	—	—	—	—
Perifovea CSV (mm3)	−2.04	0.13	0.741	0.006	—	—	—	—

^*Compared with other types of PS.

Discussion

The PS is a characteristic feature of pathological myopia. Curtin et al.¹⁴ diagnosed and classified PS by using stereoscopic fundus examination glasses in 1977. However, these methods lack objectivity and are susceptible to the effects of refractive media. Subsequently, B-scan ultrasound was used for PS diagnosis due to its convenience and insusceptibility to refractive media. However, both methods are limited to two-dimensional imaging, which cannot encompass the entire PS area. In 2011, Moriyama et al.¹ applied 3D-MRI to image the entire eyeball and analyzed its shape, thereby providing complete coverage of the PS region. However, the complexity and cost of 3D-MRI, coupled with its relatively low spatial resolution, limit its application for PS screening, especially in detecting subtle changes associated with mild PS. With the rapid development of OCT technology, UWF-OCT has emerged as a valuable tool for PS diagnosis and classification. In this study, the latest UWF-OCTA and CDI systems were used to comprehensively evaluate the changes in ocular blood flow in patients with PS and provide new insights into the pathogenesis of PS.

Currently, there is no effective treatment for PS, and its pathogenesis remains controversial. Numerous studies have suggested that PS occurrence is closely related to fundus morphological alterations accompanied by reduced blood flow.^15,16 Additionally, AL and age are major risk factors for PS development.¹⁷ However, in cases with significant differences in axial length, it is difficult to determine whether changes in choroidal structure and blood flow parameters are due to axial factors or PS itself. This study employed propensity score matching to achieve 1:1 matching of axial length and good matching of age, further enabling assessment of the independent association between PS and changes in ocular vascular parameters. This approach allowed for a more accurate evaluation of the characteristic ocular blood flow changes associated with PS.

This study aimed to investigate whether PS and different types of PS are associated with changes in the retinal microvascular density, choroidal vasculature, and retrobulbar blood flow using the latest UWF-OCTA imaging systems. The results revealed that a decrease in the EDV of the PCA was an associated factor for the occurrence of PS. Furthermore, compared with other types of PS, a reduction in perifoveal CHT was identified as an associated factor for both wide and narrow macular types of PS.

Blood supply to the eyes is primarily provided by the OA. After entering the orbit, the OA is divided into the CRA and short PCA. The CRA supplies the inner five layers of the retina, whereas the short PCA mainly nourishes the outer five layers of the retina and choroid. In the early stages, researchers discovered that the arteries and veins behind the eyeball could be observed in highly myopic eyes by using indocyanine green angiography (ICGA). Moreover, changes in the spatial distribution of the vessels behind the eyeball and within the sclera have been observed in highly myopic eyes with PS.^18,19 In 2012, Ohno-Matsui et al.²⁰ utilized swept-source OCT to observe changes in the vascular structures within the sclera and behind the eyeball in pathologic myopia. Through verification with ICGA, they discovered temporal displacement of the entrance of the ciliary artery toward the temporal edge of the PS. Our study found that a reduction in the EDV of the temporal PCA was an associated factor of PS. We deduced that the displacement of the ciliary artery may reduce its velocity in the end diastolic stage, further leading to a proportional reduction of choroidal blood flow and thinning of CHT. This finding is consistent with the previous literature.⁵ Moreover, CHT thinning weakens the buffering effect of pressure, potentially causing a more direct and significant pressure load on the sclera through IOP, leading to the development of PS.⁴ However, it is currently uncertain whether a change in blood flow in the temporal PCA is an accompanying phenomenon in the process of PS or a consequence. Further investigation of this causal relationship is required in longitudinal studies.

In recent years, the hypoxia theory has gained attention in myopia-related research. Wu et al.²¹ suggested that the hypoxia signal was a key regulatory factor in the remodeling of the scleral extracellular matrix during myopia progression, promoting the differentiation of choroidal fibroblasts into myofibroblasts. Medications targeting hypoxia-induced factor-1α expression were able to slow down experimental myopia progression without affecting normal eye growth. Zhou et al.²² found that increasing choroidal blood flow in myopic guinea pigs through drug administration inhibited myopia development. Based on these findings, we hypothesized that a reduction in choroidal blood flow, leading to decreased oxygen levels and an inadequate nutrient supply to the scleral tissue, triggers the transformation of fibroblasts, further altering the biomechanical properties of the sclera and promoting axial elongation, thus contributing to the development of myopia. This concept aligns with the hypothesis that the thinning of the choroid, leading to a diminished buffering effect on pressure, may result in more direct and substantial pressure on the eyeball wall, thus facilitating PS development. Ohno-Matsui and Jonas⁴ proposed that the choroid was primarily involved in the formation of PS for similar reasons. Thinning choroidal tissue reduces the buffering effect on pressure, resulting in more direct and greater pressure on the eyeball wall and promoting PS formation.

In this study, a thinner perifoveal CHT was identified as an associated factor for both the wide and narrow macular types of PS. A potential reason is that the margins of wide and narrow macular PS are more likely to be located in the perifoveal region, which may explain their association with perifoveal choroidal thinning. In addition, some studies have reported that PS can also occur in non-highly myopic eyes. Xu et al.²³ found that PS can occur in nonmyopic patients with retinal pigmentosa. The PS boundaries corresponded to the margin between the degenerative and relatively normal areas. We speculate that the presence of pressure differences in the local tissues of the fundus, rather than choroidal thinning, may be an important factor influencing PS. This hypothesis requires further clinical and basic validation.

This study has some limitations. First, this was a cross-sectional study, and further long-term longitudinal observations are required to explore the relationship between PS and changes in ocular blood flow. Second, despite the significant improvements in depth and range provided by UWF-OCTA, imaging may not cover the PS beyond the scan range, necessitating continued follow-up observations in future studies.

Conclusions

Change in blood flow of the PCA was identified as an associated factor for PS. In addition, reduced CHT in the temporal perifoveal region was recognized as an associated factor for both wide and narrow macular types of PS. OCTA and CDI facilitate the comprehensive assessment of ocular vascular alterations in patients with PS.