The Observation of Peripapillary Hyperreflective Ovoid Mass-Like Structures in Adolescents by Swept-Source Optical Coherence Tomography Angiography

Abstract

Purpose

To present the swept-source optical coherence tomography angiography (SS-OCTA) characteristics of peripapillary hyperreflective ovoid mass-like structures (PHOMS) in healthy adolescents and evaluate the effects of PHOMS on peripapillary structural and microvascular parameters.

Methods

A total of 218 eyes of 117 healthy adolescents were included. SS-OCTA images centered on the optic nerve head (ONH) were captured on eyes included. Structural characteristics of the ONH were compared between adolescents with or without PHOMS. After a propensity score matching was conducted, the alterations of peripapillary structures and capillaries in adolescents with PHOMS were analyzed.

Results

PHOMS were found in 33.5% of the adolescent eyes. OCTA showed blood flow signals within all PHOMS. Large ONH tilt angle and small cup-to-disc ratio were independent risk factors for PHOMS (P < 0.001), and they were closely related to the size of PHOMS (P < 0.01). Thickened peripapillary retinal nerve fiber layer and elevated radial peripapillary capillary vessel density were observed in adolescents with PHOMS, especially in those with large PHOMS (all P < 0.05).

Conclusions

PHOMS, complex structures containing nerve fibers and vascular networks, were common in adolescents. The presence of PHOMS in healthy individuals was closely related to the structural features of ONH. PHOMS were usually accompanied by congested and dilated nerve fibers and radial peripapillary capillaries.

Translational Relevance

This study provides new insights into the composition of PHOMS and their effects on peripapillary structures and capillaries.

Introduction

The emergence of enhanced depth imaging optical coherence tomography (EDI-OCT) has dramatically increased imaging penetration depth, allowing clear visualization of optic nerve head (ONH) structures.^1,2 In a recently published consensus, the Optic Disc Drusen Studies Consortium first explicitly stated that peripapillary hyperreflective ovoid mass-like structures (PHOMS) were distinctly different entities from optic disc drusen (ODD).³ On cross-sectional images, they appear as oval structures with similar hyperreflectivity to the retinal nerve fiber layer (RNFL), and elevate the surrounding retinal layers. Subsequently, PHOMS have been reported in multitudinous ocular diseases, such as myopic/tilted optic discs, ODD, idiopathic intracranial hypertension, nonarteritic anterior ischemic optic neuropathy (NAION), optic neuritis (ON), and retinal vein occlusion.^4–11 PHOMS are believed to be a nonspecific marker of axoplasmic stasis.^10,12

In children, PHOMS showed a high incidence and were the most frequent reason for pseudopapilloedema.^13,14 The prevalence of PHOMS varied between 53.6% and 67.3% in children referred for suspected papilledema.^14,15 Behrens et al.¹⁶ conducted a cohort study, and observed PHOMS in 8.9% of healthy adolescents aged 11–12 years. However, it's unclear whether PHOMS are pathological alterations or nonpathological structures. Exploration of the internal structure of PHOMS may provide us with more information.

The internal composition of PHOMS has been a matter of great interest. PHOMS were speculated to correspond to a lateral bulge of nerve fibers above the Bruch membrane in previous investigations.^3,10,12 Recently, the presence of blood flow signals within PHOMS has been reported in several cases.^17–20 However, exactly how PHOMS develop remains unclear and whether they are pathological is still debated. According to previous studies, PHOMS were possible anatomic risk factors for NAION and retinal vein occlusion,^21–23 which indicated the potential influence of PHOMS on retinal microvasculature.

Swept-source optical coherence tomography angiography (SS-OCTA), which can provide excellent visualization of retinal capillary structures,² is an excellent tool for observing the structure of PHOMS and assessing their effects on surrounding tissue. In this study, we described the SS-OCTA findings of PHOMS in adolescents without ODD for the first time. In addition, we evaluated the risk factors for PHOMS formation and the effects of PHOMS presence on peripapillary structural and microvascular parameters.

Methods

Participants

The Ethics Committee of Renmin Hospital of Wuhan University granted approval for this cross-sectional study (WDRY2023-K025), which adhered to the principles of the Declaration of Helsinki. Before the examinations, informed consent was acquired from the parents or legal guardians of the adolescents.

This prospective cross-sectional study consecutively enrolled healthy adolescents at our hospital from June 2023 to February 2024. Each participant experienced a thorough ophthalmic examination, including assessment of best-corrected visual acuity (BCVA), intraocular pressure, slit-lamp biomicroscopy, axial length (AL), refractive error and SS-OCT/OCTA examinations. Spherical equivalent (SE) was calculated as the sphere plus half a cylinder. The following were the inclusion criteria: age between 8 and 18 years, BCVA of 20/25 or better, intraocular pressure < 21 mm Hg and AL < 26mm. Participants with true disc edema, ODD, myopic-related fundus pathological changes, glaucoma, intraocular media opacity, vitreoretinal diseases, ophthalmic surgery history, or systemic diseases were excluded from this study. Images with obvious artifacts or poor quality were also excluded.

SS-OCT and OCTA Image Analysis

SS-OCT and SS-OCTA examinations were performed using a commercial device (VG100; SVision Imaging, Ltd., Luoyang, China). A 6 × 6 mm² scan (512 A-scans × 512 B-scans) centered on the ONH of each eye was captured. Every B-scan was averaged after being scanned four times. The device had an eye-tracking tool to reduce eye-motion artifacts.

The mean and vertical cup-to-disc ratio (CDR), the cup area and volume, the area of RIM and optic disc were measured in a glaucoma analysis mode. The RNFL thickness and radial peripapillary capillaries (RPC) vessel density (VD) of two concentric ring regions with a diameter of 2 mm (papillary) and 4 mm (peripapillary) were evaluated using the built-in program (Figs. 1C, 1D). According to Garway-Heath map,²⁴ the peripapillary region was subdivided into eight sectors, including nasal superior, nasal inferior, inferior nasal, inferior temporal (IT), temporal inferior (TI), temporal superior (TS), superior temporal (ST), and superior nasal (SN).

Figure 1.

ONH and peripapillary structural and microvascular parameter measurement. (A) The ratio between the longest diameter (yellow line) and the shortest diameter (red line) of the optic disc on en face image was defined as the ONH tilt ratio. The ONH rotation referred to the angle between the long axis of the optic disc and the line that was 90° from the line connecting the optic disc center and the fovea. (B) The BMO distance was evaluated between both sides of BMO in B-scan images (yellow line). ONH plane (blue line) was defined as the line connecting the nasal BMO and the innermost margin of externally oblique border tissue. ONH tilt angle was defined as the angle between BMO plane and ONH plane. (C, D) The RNFL thickness and radial peripapillary capillaries vessel density of two concentric ring regions with a diameter of 2 mm (papillary) and 4 mm (peripapillary) were evaluated. (E) OCT angiography showed that the radial peripapillary capillary, located in the retinal nerve fiber layer, traveled radially around the optic disc along the main arteries and veins.

Manual Measurement of ONH Parameters

On the en face image, the longest and shortest diameters of the optic disc were measured using an embedded tool, and the ratio between them was described as the ONH tilt ratio (Fig. 1A).²⁵ The ONH rotation degree referred to the angle between the longest axis of the optic disc and the line that was 90° from the line connecting the optic disc center and the fovea (Fig 1A).²⁶ The Bruch's membrane opening (BMO) distance and ONH tilt angle were measured on the horizontal B-scan across the center of the discs. The BMO plane was defined as the line joining two sides of the BMO, and the ONH plane as the line joining the nasal BMO and the innermost margin of the externally oblique border tissue. The angle between the two planes was regarded as the ONH tilt angle (Fig. 1B).²⁶ The largest cross-sectional area of PHOMS on horizontal B-scans was measured to reflect the PHOMS size.

Statistical Analysis

The statistical software SPSS ver. 26.0 (IBM-SPSS Inc., Chicago, IL) was applied for all analyses. The normality of the variables was evaluated using the Kolmogorov-Smirnov test. The χ² test was used for the comparison of categorical data, whereas the independent sample t test or Mann-Whitney U test was used for continuous parameters. Univariate and multivariate logistic regression analysis were performed to explore the risk factors related to PHOMS. The correlation between PHOMS size and ONH parameters was evaluated by Spearman correlation analysis. To obtain a control group with characteristics comparable to those of PHOMS group, a 1:1 propensity score matching was carried out with a match tolerance of 0.02 based on sex, age, AL, and SE. P < 0.05 was considered statistically significant.

Results

Demographic and Ocular Characteristics

A total of 218 eyes of 117 individuals were included and PHOMS appeared in 73 eyes (33.5%). Table 1 displayed the two groups’ demographic and ocular characteristics. The mean ages were 11.1 ± 1.6 years in PHOMS group and 11.0 ± 1.7 years in no PHOMS group. Between the two groups, there was no statistic difference in terms of age, sex distribution, or BCVA (all P > 0.05). Significantly lower SE and longer AL were seen in PHOMS group as compared to no PHOMS group (all P < 0.05). For ONH parameters, PHOMS group exhibited significantly smaller mean CDR, vertical CDR, cup area, cup volume, and larger RIM area (all P < 0.05). In comparison to no PHOMS group, PHOMS group showed more tilted and rotating optic discs (P < 0.05). The optic disc area was comparable between the two groups (P = 0.779).

Table 1.

Demographics and Ocular Characteristics of Participants With or Without PHOMS

	PHOMS	No PHOMS	P Value
No. of eyes	73	145
Age (years)	11.1 ± 1.6	11.0 ± 1.7	0.493
Sex			0.535
Male	30	66
Female	43	79
BCVA ≥ 20/25 (%)	100	100	1.000
Refractive error (SE, D)	−2.35 ± 1.79	−1.51 ± 1.74	<0.001
Axial length (mm)	24.71 ± 0.79	24.41 ± 0.95	0.002
ONH parameters
Mean CDR	0.41 ± 0.14	0.51 ± 0.14	<0.001
Vertical CDR	0.36 ± 0.12	0.46 ± 0.13	<0.001
Cup area (mm2)	0.39 ± 0.28	0.59 ± 0.36	<0.001
Cup volume (mm3)	0.33 ± 0.37	0.59 ± 0.53	<0.001
RIM area (mm2)	1.58 ± 0.20	1.40 ± 0.24	<0.001
Optic disc area (mm2)	1.98 ± 0.30	1.99 ± 0.33	0.779
BMO distance (µm)	1686.37 ± 234.29	1652.28 ± 157.26	0.014
ONH tilt ratio	1.25 ± 0.13	1.28±0.99	0.015
ONH tilt angle	7.62° ± 2.92°	4.56° ± 3.86°	<0.001
ONH rotation degree	19.00° ± 14.89°	15.10° ± 15.28°	0.022

Structural and Microvascular Features of PHOMS

PHOMS were located primarily on the nasal quadrant of the ONH, sometimes extending to the superior or inferior quadrant, but were not observed on the temporal quadrant in our study. Continuous vertical OCT B-scans revealed that PHOMS seemed to be connected to the nerve fibers entering the ONH (Supplemental Movie S1). Figure 2 demonstrated the en face OCT and OCTA B-scan images of PHOMS in different sizes. On en face OCT, PHOMS were characterized by hyperreflective sector structures with clear boundaries (Fig. 2). They appeared as slightly elevated structures with slightly uneven internal signal. On horizontal and vertical OCTA B-scans, the internal blood flow signals were observed within all PHOMS. After adjusting layer of PHOMS, the vascular complex was visualized on en face OCTA in the position corresponding to PHOMS (Fig. 3). It traveled radially around the optic disc, which was similar to the RPC.

Figure 2.

En face structural OCT, horizontal and vertical OCTA B-scan images of PHOMS in different sizes. (A, D, G) The en face structural OCT image showed a hyperreflective, slightly elevated ring structure on the nasal side of the optic disc (yellow arrow heads). (B, E, H) On the horizontal OCTA B-scans, blood flow signals within PHOMS were visualized on the nasal side of the optic disc (yellow arrow heads). (C, F, I) The vertical OCTA B-scans displayed the presence of blood flow signals within PHOMS above the Bruch's membrane (yellow arrow heads).

Figure 3.

En face structural OCT, en face OCTA and OCTA B-scan image of the manually adjusted layer of PHOMS. (A) The en face structural OCT image showed a hyperreflective, slightly elevated ring structure on the nasal side of the optic disc (yellow arrow heads). (B) On the en face OCTA image, the vascular complex was visualized in the corresponding position (yellow arrow heads). (C) The OCTA B-scan across the nasal side of the optic disc displayed the presence of blood flow within this hyperreflective lesion (yellow arrow heads).

Independent Factors Associated With PHOMS

The logistic regression was utilized to assess the relationship between several ocular characteristics and the presence of PHOMS (Table 2). According to univariate regression analysis, PHOMS were found to be associated with increased AL (P = 0.024), lower SE (P = 0.002), smaller CDR (P < 0.001), and a larger ONH tilt angle (P < 0.001). Further multivariate regression analyses indicated that CDR and ONH tilt angle were independent factors related to PHOMS presence (P < 0.001). The correlation between ocular variables and PHOMS size was shown in Figure 4. The maximum cross-sectional area of PHOMS exhibited a positive relationship with ONH tilt angle and a negative correlation with mean CDR, vertical CDR, cup area, cup volume, and optic disc area (all P < 0.05). The area of PHOMS was independent of age, SE, AL and BMO distance (all P > 0.05).

Table 2.

Univariate and Multivariate Logistic Regression Analyses of PHOMS

	Univariable Model			Multivariable Model
Variables	OR	95% CI	P Value	OR	95% CI	P Value
Age per 1 y increase	1.043	0.881–1.234	0.626
Sex female vs. male	0.835	0.473–1.476	0.535
AL per 1 mm increase	1.445	1.050–1.988	0.024
SE per 1 D increase	0.772	0.658–0.907	0.002
CDR per 0.1 increase	0.597	0.480–0.742	<0.001	0.618	0.489–0.780	<0.001
Optic disc area per 1 mm2 increase	0.880	0.360–2.149	0.778
BMO distance per 1µm increase	1.001	0.999–1.003	0.205
ONH tilt ratio per 1 increase	0.936	0.616–1.423	0.755
ONH tilt angle per 1° increase	1.272	1.161–1.395	<0.001	1.260	1.144–1.387	<0.001
ONH rotation per 1° increase	1.016	0.998–1.035	0.078

Figure 4.

Correlation analysis between area of PHOMS and ONH structural parameters. The area of PHOMS was positively correlated with ONH tilt angle and negatively correlated with mean CDR, vertical CDR, cup area, cup volume and optic disc area.

Comparison of RNFL Thickness and RPC Vessel Density

To assess the effect of PHOMS presence on peripapillary structures and capillaries, we performed a propensity score matching based on age, sex, SE and AL to reduce the confuse factor disturbance. After that, 70 eyes from PHOMS group and 70 eyes from no PHOMS group were matched. When comparing the RNFL thickness in the papillary and peripapillary regions between the matched two groups, significant differences were observed (P < 0.001) (Table 3). Among eight subregions of the peripapillary region, participants with PHOMS had thicker RNFL in the IT, TI, TS, ST, and SN quadrants (all P < 0.05). Regarding the RPC, PHOMS group exhibited higher VD than no PHOMS group in both papillary and peripapillary regions (P < 0.05). In the TI, TS, and ST quadrants from the peripapillary regions, participants with PHOMS had higher VD (all P < 0.05).

Table 3.

Comparison of RNFL Thickness and Vessel Density of RPC Between Participants With or Without PHOMS

	PHOMS (n =70)	No PHOMS (n = 70)	P Value
RNFL thickness (µm)
Papillary	467.73 ± 62.61	413.08 ± 44.59	<0.001
Peripapillary	132.86 ± 17.42	122.03 ± 11.72	<0.001
NS	94.91 ± 30.14	89.59 ± 24.64	0.222
NI	75.43 ± 20.68	74.79 ± 20.20	0.934
IN	143.29 ± 30.93	135.69 ± 26.95	0.305
IT	203.71 ± 39.67	185.01 ± 27.72	0.002
TI	120.82 ± 27.22	108.84 ± 29.55	0.005
TS	112.84 ± 24.74	105.09 ± 21.77	0.036
ST	193.42 ± 31.61	170.75 ± 24.57	<0.001
SN	161.01 ± 37.01	141.84 ± 29.68	0.006
RPC vessel density
Papillary	90.83% ± 3.63%	89.26% ± 4.90%	0.030
Peripapillary	57.57% ± 4.26%	55.73% ± 4.89%	0.023
NS	42.10% ± 13.01%	41.90% ± 14.11%	0.901
NI	32.73% ± 9.62%	33.60% ± 12.31%	0.643
IN	57.46% ± 10.58%	57.54% ± 10.62%	0.963
IT	80.325 ± 1.01%	78.22% ± 7.27%	0.065
TI	59.42% ± 11.75%	54.81% ± 13.54%	0.033
TS	58.14% ± 11.07%	52.68% ± 12.83%	0.008
ST	82.02% ± 6.48%	78.44% ± 6.86%	0.002
SN	66.04% ± 11.63%	64.51% ± 12.06%	0.611

In further exploration, we divided the PHOMS group into three subgroups based on the area of PHOMS, including the small PHOMS group (area ≤ 0.04 mm²), the medium PHOMS group (0.04 mm² < area ≤ 0.07 mm²), and the large PHOMS group (area > 0.07 mm²). In the further comparison between the no PHOMS group and the PHOMS subgroups of different sizes, there were significant differences in RNFL thickness and RPC VD in some sectors of the peripapillary region, mainly on the temporal side (all P < 0.05) (Table 4). We observed that the significant differences among the four groups mainly stemmed from the differences between the large PHOMS group and the no PHOMS group. Specifically, peripapillary RNFL thickness and RPC VD were significantly increased in the large PHOMS group. In contrast, no significant differences were found between the small PHOMS and no PHOMS groups (all P > 0.05).

Table 4.

Comparison of RNFL Thickness and RPC Vessel Density Between Participants With PHOMS of Different Sizes

	(A) No PHOMS (n = 70)	(B) Small PHOMS (n = 23)	(C) Medium PHOMS (n = 24)	(D) Large PHOMS (n = 23)	P Value	Post-Hoc
RNFL thickness (µm)
Papillary	413.08 ± 44.59	439.56 ± 37.15	457.35± 49.57	506.73 ± 76.22	<0.001	A<C/D, B<D
Peripapillary	122.03 ± 11.72	128.28 ± 9.17	130.31± 14.56	140.11 ± 23.70	<0.001	A< D
NS	89.59 ± 24.64	95.58 ± 21.06	96.89± 30.62	92.19 ± 37.59	0.415	/
NI	74.79 ± 20.20	74.78 ± 17.43	76.95± 22.41	74.50 ± 22.55	0.979	/
IN	135.69 ± 26.95	143.57 ± 21.96	141.20 ± 29.99	145.19 ± 39.61	0.593	/
IT	185.01 ± 27.72	200.58 ± 28.46	190.34 ± 28.53	220.78 ± 52.58	0.001	A< D
TI	108.84 ± 29.55	110.85 ± 23.61	114.39 ± 23.12	137.50 ± 27.70	<0.001	A/B< D
TS	105.09 ± 21.77	107.85 ± 20.34	105.88 ± 16.66	125.11 ± 30.13	0.017	A< D
ST	170.75 ± 24.57	182.74 ± 22.66	188.58 ± 26.19	209.16 ± 38.73	<0.001	A<C/D
SN	141.84 ± 29.68	148.37 ± 26.47	165.82 ± 40.40	168.64 ± 40.43	0.009	A< D
RPC vessel density (%)
Papillary	89.26% ± 4.90%	91.28% ± 3.28%	91.04% ± 2.97%	90.16% ± 4.54%	0.133	/
Peripapillary	55.73% ± 4.89%	57.70% ± 3.50%	57.21% ± 4.26%	57.80% ± 5.05%	0.130	/
NS	41.90% ± 14.11%	44.93% ± 9.77%	42.13% ± 13.71%	39.25% ± 14.90%	0.538	/
NI	33.60% ±12.31%	34.40% ± 9.02%	33.17% ± 10.74%	30.62% ± 8.98%	0.657	/
IN	57.54% ± 10.62%	60.41% ± 7.53%	57.31% ± 10.88%	54.65% ± 12.39%	0.332	/
IT	78.22% ± 7.27%	80.52% ± 5.73%	79.08% ± 5.86%	81.40% ± 6.46%	0.185	/
TI	54.81% ± 13.54%	55.34% ± 12.88%	58.30% ± 9.81%	64.67% ± 10.89%	0.011	A/B<D
TS	52.68% ± 12.83%	56.26% ± 10.98%	55.64% ± 9.70%	62.63% ± 11.55%	0.008	A/C<D
ST	78.44% ± 6.86%	80.41% ± 7.35%	82.68% ± 4.76%	82.92% ± 7.08%	0.009	A<C/D
SN	64.51% ± 12.06%	64.69% ± 9.87%	66.77% ± 12.77%	66.61% ± 12.39%	0.718	/

Discussion

The high resolution, fast scan rate, and non-invasive features of SS-OCTA make it an advantageous tool for studying the fundus structure.²⁷ In our research, we presented the structural OCT and OCTA characteristics of PHOMS by SS-OCTA, which provided convincing evidence for revealing the composition of PHOMS and their influence on the papillary structure and microvasculature.

In our study, PHOMS appeared in approximately one third of the adolescent eyes, which was higher than previously reported in European regions. For instance, the incidence of PHOMS was reported as 8.9% among Copenhagen child aged 11-12 and 25.8% among Danish myopic children.^28,29 The different age distributions, epidemiology of myopia, and ONH morphology may contribute to variations in PHOMS prevalence across ethnicities. First, the different age ranges of the participants may influence the results, because previous studies have shown that PHOMS may be acquired structures because of the displacement of myopia during adolescence.³⁰ Then, the prevalence of myopia and tilted optic discs are more common among Asian adolescents than among Caucasians,^31–33 which are important risk factors of PHOMS. In our study, the average AL was 24.51 mm, which was much higher than that of Copenhagen children.²⁸ In addition, the SS-OCTA device we use provides deeper scan depth and higher resolution than spectral-domain OCT, which may increase the sensitivity of PHOMS recognition.

Since PHOMS were described, their composition has been debated. The histopathology evidence supported the theory that PHOMS arose from lateral bulging of nerve fibers.^10,12 The histopathology of some entities, including papilledema and optic disc drusen, have shown a peripapillary bulge formed by distended and vacuolated nerve fibers, which corresponds with the morphology and location of PHOMS as detected on OCT.^{10,12,34–37} The fibers are observed to be constipated with large degenerating mitochondria that are unable to move distally because of axoplasmic stasis.³⁴ In addition, small dilated veins and capillaries permeate the bulge,³⁵ which suggests that there are vascular structures inside PHOMS. However, some scholars argued that the protruding nerve fibers may be specimen folding that occurred during the preparation of histopathological sections.³⁸ Therefore in vivo imaging was important to demonstrate the interior structure of PHOMS and their relationship with surrounding tissues. Through a continuous and intensive scanning of the optic discs by SS-OCT, we observed that PHOMS seemed to be connected to the retinal nerve fibers entering the optic disc. What's more, the blood flow signals were observed within all PHOMS, which emphasized that vascularization was one of the characteristic structural features of PHOMS. Although vascular complexes within PHOMS have been previously observed in several cases such as tilted disc syndrome, ODD and NAION,^17–20 the origin of the vascular complexes remains unclear. In this study, we observed that the morphology of vascular complex on en face OCTA was similar to that of RPC, which was characterized by straight capillaries parallel to each other.^39,40 Therefore our results supported that PHOMS corresponded to herniation of the RPC accompanying the protrusion of nerve fibers.

Apart from the internal composition, we also explored the risk factors for PHOMS formation. Compared to eyes without PHOMS, PHOMS eyes had higher degree of myopic refractive error, longer AL, and larger ONH tilt angles. Increased ONH tilt significantly elevated the risk of PHOMS. Our findings supported the opinion of Lyu et al.⁶ that PHOMS were associated with tilted discs due to myopic shift in children. Tilted optic disc may produce compressive stress on the axons and impact axoplasmic transport in optic nerve fibers.^5,6 Besides tilted optic disc, we identified a new risk factor for PHOMS in adolescents: a smaller CDR. More than one million axons of RGC are crowded as they pass through the optic disc with a small CDR,⁴¹ which will result in a relatively slow axoplasmic flow. Therefore, we should pay more attention to adolescents with tilted optic disc and small CDR clinically.

Since PHOMS are common findings in adolescents, it is of great importance to find out whether they can cause secondary impairment of the optic nerve fibers. In our research, it was observed that adolescents with PHOMS had significantly thicker RNFL compared with no PHOMS group, especially on the temporal quadrant. Further subgroup analyses showed that the significant differences originated mainly between the large PHOMS group and the no PHOMS group. This suggests that the size of PHOMS is an important factor affecting RNFL and we should pay attention to the development of PHOMS. Similar results were also reported by Lee et al.⁴² They revealed that large buried ODD, actually PHOMS, was associated with increased temporal RNFL thickness.⁴² However, the mechanism explaining this discovery remains unclear. Previous studies have shown that the initial manifestation is axonal swelling with accumulation of mitochondria and other organelles when axonal transport is disrupted.^43,44 We speculate that the thickening RNFL may be temporary result of impaired axoplasmic transport with accumulation of organelles, or a compensatory increase of mitochondrial biogenesis, or a combination of them. In addition, retinal dragging toward the temporal in tilted optic disc may also play a role in this change. What calls for special attention is that prolonged axoplasmic flow blockage may cause damage to nerve fibers, resulting in impaired optic nerve function.^41,45 This might be a relatively lengthy process in adolescents with PHOMS. In long-term follow-up of patients with tilted disc syndrome and in observation of adults with high myopia, PHOMS were reported to be related to visual field defects.^5,46 Therefore long-term monitoring of RNFL and visual field alteration in individuals with PHOMS is necessary in the future.

RPC is distributed in RNFL and supplies it with metabolic and nutritional support.³⁹ Consistent with RNFL thickness, eyes with PHOMS, especially those with large PHOMS, had increased VD of RPC in papillary and peripapillary region, which reflected the dilatation of capillaries. Previous histopathology studies also have shown dilated veins and capillaries within the optic nerve tissue and PHOMS,¹² which supports our findings. Different with us, Ahn et al.¹⁵ observed decreased VD of the optic disc area in large PHOMS eyes. This difference might be attributed to the fact that some patients included in their study have coexisting buried ODD, which affects the peripapillary VD.⁴⁷ However, it was intriguing that the significant difference in VD appeared specifically in the temporal region rather than the nasal area, where PHOMS were present. We speculated that the temporal VD increase may reflect a compensatory mechanism. Contrary to eyes with large PHOMS, RPC was unaffected in eyes with small PHOMS. This reminds us to be concerned about how to prevent PHOMS from increasing in size in the future.

We acknowledge that several limitations exist in our study. The relatively small sample size was the main limitation. Second, the single center and cross-sectional design were unable to reveal the causal connection between PHOMS and the structural and microvascular alterations. Third, PHOMS were three-dimensional structures, and volume instead of area may be a more accurate indicator of PHOMS size. Identifying and quantifying the true volume of PHOMS through machine learning is an important direction for future research. Finally, future studies combining advanced imaging with histopathological correlation and long-term follow-up are needed to reveal the pathophysiology of PHOMS.

In summary, we found that PHOMS were common structures in adolescents. They may represent herniation of axons and capillaries into the peripapillary retina. The structural features of ONH played a crucial role in PHOMS formation. Adolescents with large PHOMS had significantly thicker RNFL and higher RPC VD. Long-term follow-up of PHOMS is required to make it clear that whether they are pathological changes.