Correlations between Visual Performance and Chorioretinal Variables after Vitrectomy for the Idiopathic Macular Hole

Purpose To investigate the relationships between visual function and the retinal and choroid microstructure in idiopathic macular hole patients after surgery. Methods A prospective study investigated changes in the fundus structure and visual function in 16 macular hole patients before and after surgery. Patients' best-corrected visual acuity (BCVA) and retinal sensitivity were measured by an EDTRS visual chart and microperimetry (MP1), respectively. The thickness of the retina and the blood supply to the retina and superficial choroid were detected by OCTA, and the choroidal capillary blood flow density was analysed with MATLAB. The thickness of the choroid and the aperture size of the macular hole were detected by Heidelberg OCT. Results Compared with before surgery and one month after surgery, the BCVA (3 months: 0.47 ± 0.27, before: 1.02 ± 0.22, 1 month: 0.66 ± 0.27, and P < 0.05) and the central sensitivity of the retina (3 months: 14.88 ± 2.87 dB, before: 8.76 ± 3.27 dB, 1 month: 12.22 ± 3.30 dB, and P < 0.05) were significantly improved three months after surgery. The change in BCVA was significantly correlated with the basal diameter (r = 0.677 and P = 0.004), the minimum diameter (r = 0.585 and P = 0.017), the macular hole cystoid height area index (r = −0.618 and P = 0.011), the central macular hole index (r = −0.727 and P = 0.001), the peripheral macular hole index (r = −0.758 and P = 0.001), the central tractional hole index (r = −0.717 and P = 0.002), the peripheral tractional hole index (r = −0.725 and P = 0.001), and changes in the peripheral blood vessel density of the choroid capillary layer (r = 0.585 and P = 0.0017). The change in central retinal sensitivity was correlated with the change in the superficial foveal avascular zone (FAZ; r = 0.520 and P = 0.039), change in the retinal superficial peripheral blood flow density (r = −0.503 and P = 0.047), change in the deep FAZ (r = 0.599 and P = 0.014), and change in the retinal deep peripheral blood flow density (r = −0.601 and P = 0.014). Conclusions The morphology of the macular hole as well as changes to the retinal and choroidal microstructure contributes to the recovery of visual function after surgery.


Introduction
Te idiopathic macular hole (IMH) seriously afects central vision [1,2]. Surgery to close the hole can improve the patient's vision [3]. Studies have reported that the rate of successful macular hole treatment is as high as 85-100% through vitrectomy [4]. Although the macular hole anatomical closure rate is very high, the corresponding improvements in visual acuity are not ideal. Improvements in visual acuity after macular hole surgery may be related to changes to the hole morphology and fundus retinal and choroidal microstructure [4][5][6][7]. Although the macular hole may achieve good anatomical closure after surgery, diferences in postoperative visual function are observed.
Evidence suggests that at least half of all macular hole patients still have low vision [8,9]. Even individuals with similar preoperative vision exhibit signifcant diferences in visual function after surgery, and it is difcult to explain the diferences between visual function recovery and anatomical recovery.
In addition to the assessment of visual acuity, microperimetry (MP1) [10][11][12] can be used to analyse retinal sensitivity to assess visual function. OCT [2,[13][14][15][16] allows for analysis of the morphology of a macular hole and its derivative indices (maximum basal diameter, maximum basal diameter, and macular hole cystoid height). Furthermore, retinal thickness and inner or outer retinal capillary zones can be easily observed with OCTA [16][17][18][19][20][21]; these measures can be used to evaluate the changes in the deep or superfcial retinal capillaries and the foveal avascular zone (FAZ) as well as the relationships between these changes and visual function after macular hole surgery. Reibaldi et al. [5] argued that the choroidal thickness of macular hole eyes is signifcantly lower than that of fellow eyes (i.e., the patient's healthy eye) and normal eyes, and thus, the recovery of vision might be related to changes in the choroidal thickness. Kim [6] argued that the choroidal thickness might be the protective factor for fnal BCVA and changes in vision after surgery, with patients with higher choroidal thickness in the macular hole eye achieving better postoperative vision.
In patients with the macular hole, visual function and the morphology of the retina and choroid are diferent from normal eyes [4][5][6]. Although visual function and morphology can approach that of normal eyes after surgery to correct the hole, the relationships between visual function and the retinal and choroid microstructure in patients with the idiopathic macular hole require further exploration. Trough OCT, OCTA, and MP-1, this study comprehensively analysed the changes in visual function and the retinal and choroidal morphology to investigate the relationships between these factors.

Subjects.
Te sample for this study comprised 16 patients who were diagnosed with the idiopathic full-thickness macular hole in the Wenzhou Medical University Eye Hospital from June 2018 to July 2019. All patients with the macular hole underwent vitrectomy combined with internal limiting membrane peeling. All patients agreed to participate and provided written informed consent. Te inclusion criteria were as follows: diagnosed with the idiopathic full-thickness macular hole (Gass classifcation [18]); spherical equivalent refraction −3.0O-+3.00D; no history of eye trauma or surgery; no history of hypertension, diabetes, or other serious systemic diseases; high-quality retinal and choroid images able to be obtained from the patient and the patient is able to cooperate with the examination. Tis study was performed in accordance with the Helsinki Declaration and was approved by the hospital's ethics committee (KYK[2018]52).

Visual Function.
Patients underwent comprehensive ophthalmologic examinations, including BCVA measured through EDTRS visual charts, a slit-lamp examination, indirect ophthalmoscopy, and intraocular pressure (IOP) measurement by using a noncontact tonometer (KT-500 Japan) at the preoperative and postoperative visits. Retinal sensitivity was measured by an MP-1 microperimetry examination [10][11][12]: the retinal sensitivity is determined within 10°of the gaze point, 0-4°is the sensitivity of the central retina, and 4-10°is the sensitivity of the peripheral retina.

Choroid.
OCT images were collected from 9 a.m. to 11 a.m.; clear and full-thickness choroidal images [2,[13][14][15][16] were obtained using a Heidelberg OCT (EDI mode). Ten, the mean values of the minimum diameter and base diameter of the macular hole in the horizontal and vertical scanning directions were calculated. Te thickness of the choroid was measured by two independent researchers, and the average value of the two measurements was used for analysis. Te measurement positions were the subfoveal choroidal thickness (SFCT) and eight positions in the superior, inferior, nasal, and temporal directions, approximately 1 mm and 3 mm below the fovea ( Figure 1). Te minimum diameter and basal diameter in the horizontal and vertical direction were determined, and the mean value was calculated. Te total cystic height within 1 mm diameter of the hole centre was measured by the Heidelberg OCT, and the macular hole cystoid height area index (MCHAI) (total cystoid height/basal diameter), macular hole index (MHI) (average height/basal diameter), diameter hole index (DHI) (minimum diameter/basal diameter), and tractional hole index (THI) (average height/minimum diameter) were calculated. Te peripheral blood vessel density of the choroid capillary layer was measured by OCTA ( Figure 2).

Retina.
Te condition of the retina [16][17][18][19][20][21] was examined with an OCTA (Figures 3 and 2), and the blood fow density of the superfcial retinal capillaries (ILM: internal limiting membrane to IPL: inner plexiform layer), the blood fow areas of the FAZ, and the full retinal thickness were measured by scanning a 3 * 3 mm 2 area of the macula. Ten, the retinal superfcial or deep capillaries (IPL: inner plexiform payer to OPL: outer plexiform layer), the FAZ, and the blood vessel density of the choroid capillary layer (within the 0-0.6 mm, 0.6-1.5 mm, 1.5-2.0 mm, 2.0-2.5 mm, and 0.6-2.5 mm diameter areas) were measured by MATLAB, and OCTA image analysis software developed in the OCT laboratory ( Figure 4). Finally, the blood vessel density of the choroid capillary was calculated.
2.6. Statistical Analysis. Te data are expressed as the mean ± standard deviation (SD). Te diferences between the three groups were analysed by ANOVA (an independent samples T-test was used for variance heterogeneity). Pearson correlation analysis was performed to examine the relationships between the variables.P < 0.05 was considered statistically signifcant. SPSS 25 software was used for all statistical analyses (SPSS, IBM, Chicago, IL, USA).

Visual Function.
Te BCVA of the macular hole eyes gradually improved after surgery; the improvement in these eyes was statistically signifcant, P < 0.05. Te BCVA of fellow eyes (nonafected eyes of the macular hole patients) exhibited no statistically signifcant diference compared to normal eyes, P > 0.05. Te retinal sensitivity of the macular hole eyes (0°-4°) gradually increased after surgery; these changes were also statistically signifcant, P < 0.05. Tere were no signifcant diferences between the macular hole eyes three months after surgery relative to the fellow eyes and the normal eyes, P > 0.05. Tere were no signifcant diferences in peripheral retinal sensitivity (4°-10°) among the groups, P < 0.05. Tese fndings are presented in Table 1.

Retina.
Te preoperative retinal thickness of the hole eyes was statistically higher than the postoperative retinal thickness of the hole eyes, the retinal thickness of the fellow eyes, and the retinal thickness of the normal eyes, P < 0.05; however, there were no signifcant diferences between the postoperative retinal thickness of the hole eyes, the retinal thickness of the fellow eyes, and the retinal thickness of the normal eyes, P > 0.05. Te preoperative central retinal blood density of the superfcial and deep capillaries was statistically lower than the postoperative central retinal blood density, P < 0.05. Te FAZ of the hole eyes in both the superfcial and deep layers was signifcantly lower than the fellow eyes and the normal eyes, P < 0.05. Tese results are presented in Table 2.

Choroid.
Te preoperative superfcial blood fow areas of the choroid capillary layer in the hole eyes were statistically lower than the postoperative superfcial blood fow areas of the hole eyes, the superfcial blood fow areas of the fellow eyes, and the superfcial blood fow areas of the normal eyes, P < 0.05; however, there were no statistically signifcant diferences in the postoperative superfcial blood fow areas of the hole eyes, the superfcial blood fow areas of the fellow eyes, and the superfcial blood fow areas of the normal eyes, P > 0.05. Te peripheral choroidal blood vessel superfcial capillary density of the hole eyes was signifcantly lower than that of the fellow eyes and normal eyes, P < 0.05. Te preoperative subfoveal choroidal thickness of the hole eyes was signifcantly lower than the postoperative subfoveal choroidal thickness of the hole eyes, the fellow eyes, and the normal eyes, P < 0.05; however, there were no statistically signifcant diferences in the postoperative subfoveal choroidal thickness of the hole eyes, the fellow eyes, and the normal eyes, P > 0.05 ( Table 3).

Correlation
Analysis. Te change in BCVA was signifcantly correlated with the basal diameter (r � 0.677 and P � 0.004), the minimum diameter (r � 0.585 and P � 0.017), the MCHAI (r � −0.618 and P � 0.011), the central MHI (r � −0.727 and P � 0.001), the peripheral MHI (r � −0.758 and P � 0.001), the central THI (r � −0.717 and P � 0.002), the peripheral THI (r � −0.725 and P � 0.001), and the change in the peripheral blood vessel density of the choroid capillary layer (r � 0.585 and P � 0.0017). Te change in the central retinal sensitivity was correlated with the change in the superfcial FAZ (r � 0.520 and P � 0.039), the change in the retinal superfcial peripheral blood fow density (r � −0.503, P � 0.047), the change in the deep FAZ (r � 0.599 and P � 0.014), and the change in the retinal deep peripheral blood fow density (r � −0.601 and P � 0.014), as shown in Table 4. Te change in the peripheral blood vessel density of the choroid capillary layer was correlated with the change in the basal diameter (r � 0.592 and P � 0.016), the CMHI (r � −0.548 and P � 0.028), and the PMHI (r � −0.531 and v � 0.034), as shown in Figure 5.

Discussion
Tis study found that BCVA and central retinal sensitivity of the macular hole gradually improved after vitrectomy combined with internal limiting membrane stripping and tamponade. Chen et al. [10][11][12] also confrmed that the visual function of the macular hole eye was improved after surgery. Te results of the current study also indicated that the thickness of the retina in the central and peripheral zones gradually decreased with closure of the hole, and the cyst within 1 mm diameter of the hole centre disappeared after the hole closed. Te decrease in the retinal thickness is consistent with the research of Imamura [22], with a decrease in the retinal thickness and the disappearance of the cyst appearing to contribute to the improvement in visual function. Moreover, the superfcial blood fow areas of the choroid capillary layer, the choroidal blood vessel superfcial density capillary, and the choroidal thickness exhibited signifcant improvements after macular hole surgery. Te studies of Reibaldi et al. [5] and Kim et al. [6] also found that the choroidal thickness of the hole eyes was lower than that of the fellow eyes and normal eyes. Te improvements in visual function and fundus microstructure in the retina and choroid of the hole eyes after macular hole surgery resulted in function and microstructure characteristics that were close to those of the fellow eyes and normal eyes. Te correlation analysis indicated that the postoperative change in BCVA was correlated with the morphology of the macular hole and its derivative indices. Te ∆BCVA was negative as it is calculated as the three-month BCVA minus the preoperative BCVA. Te smaller basal diameter, smaller minimum diameter, larger MCHAI, larger central or peripheral MHI, larger central or peripheral THI, and smaller peripheral blood vessel density of the choroid capillary layer all contributed to the improvement in BCVA. Macular holes seriously damage vision, and patients with macular holes with a smaller basal diameter and smaller minimum diameter may exhibit greater improvements in visual function following surgery. Tis is consistent with the research of Venkatesh et al. [2]. Te presence of retinal cysts is indicative of greater anteroposterior tractional forces and taller macular holes. Te anteroposterior traction forces could be represented by the macular hole height and nasal and temporal arm lengths, and the diameter hole index evaluates the distance between the edges of the hole. Te MHI, THI,    and MCHAI could help analyse the anteroposterior tractional forces caused by cysts, and the lower index might obtain better restoration of visual function.
Tickening of the retina in the macular hole may be attributed to the anteroposterior and tangential traction caused by incomplete vitreoretinal detachment, the presence of the vitreous cortex and epimacular glial tissues, or retinal swelling with subretinal fuid components [1][2][3][4]. Te MHI includes both the horizontal and vertical dimensions of the macular hole, representing the putative tangential and anteroposterior vitreomacular traction or retinal hydration; a higher MHI value indicates a smaller basal diameter and a larger height [2]; a higher MCHAI and THI also represent smaller horizontal dimensions and a larger height. Te effusion separates the retinal neurosensory layer from the retinal pigment epithelium, and the oxygen transmission from the choroidal capillaries to the deep retina is blocked by the efusion, resulting in hypoxia of the retinal neurosensory layer. After surgery, the efusion disappears and the hypoxia of the retina is ameliorated, thus improving visual function. Tus, a patient with a macular hole with a smaller basal diameter and greater height might achieve better improvement in visual function after surgery.
Te retinal thickness was decreased in the central region, upper and lower nasal regions, and temporal region after surgery. Tinning of the retina in the macular hole mainly occurs in the nerve fbre layer and ganglion cell layer. Ozdemir et al. [23] found that the nerve fbre layer was signifcantly thinner in the temporal, upper, and lower sides, which might be caused by mechanical damage due to internal limiting membrane stripping. After surgery, the retinal cells are afected by local infammation, microcirculation ischaemia and stretching, or the toxicity of macular stain; this can lead to damage to the nerve fbre layer and ganglion cell layer, resulting in a reduction in retinal thickness.
Te results of this study indicated that the change in central retinal sensitivity was correlated with changes in the superfcial or deep FAZ and the retinal superfcial or deep peripheral blood fow density. After the macular hole is closed, the FAZ reduces. Te main reason for the decrease in the FAZ is the surgical stripping of the inner limiting membrane, which improves the closure rate of the hole. After surgery, the retinal cells migrate to fll the gap caused by the macular hole, the tissue increases in the macular hole, and the FAZ decreases [7,[24][25][26]. Tus, it can be inferred that a greater reduction in the FAZ relates to better recovery of the central retinal tissue and a greater improvement in visual function.
Te central superfcial retinal blood fow density was signifcantly increased, and the peripheral superfcial retinal blood fow density was decreased one month after surgery. Tis suggests that the month after surgery is the critical period for closure of the macular hole. Te morphology and function of the central retina recovered gradually; the peripheral retinal blood fow slowly migrated to the central part of the retina, contributing to the morphological and functional recovery of the macular hole. Tus, the peripheral retinal blood fow decreased, and the central blood fow increased. Te decrease in the FAZ also refects this phenomenon. Although the retinal blood fow increased from one month to three months after surgery, this diference was not statistically signifcant, presumably, because the hole was basically closed one month after surgery and the morphology tended to be stable. Te retinal blood fow also tended to be stable; thus, there was no need for peripheral blood fow support.
Tis study revealed that the subfoveal, temporal, and superior choroidal thickness were greatest, followed by the inferior choroidal thickness; the nasal choroidal thickness was the thinnest. Tis is consistent with previous studies [5][6][7][8]. It is speculated that the region where the nasal optic papilla is located has fewer choroidal vessels than the other regions. Ozdemir et al. [23] showed that the choroidal thickness became thinner after surgery. However, some studies have found only signifcant subfoveal choroidal thickness thinning, while other parts of the choroidal thickness exhibit little statistical change after surgery [17,27]. Although the relationship between thinning of the choroid and the change in visual function is unclear, most studies supported thickening of the choroid after surgery [7,13,17,24,[26][27][28][29]. Zhang et al. [13] reported that the subfoveal choroidal thickness is lower in macular hole eyes than normal eyes, that the change in the choroidal blood fow occurs before complete formation of the macular hole, and that thinning of the choroidal thickness might be caused by the macular hole. Aras et al. [30] and Morgan and Schatz [31] suggested that choroidal hypoperfusion is the result of pore formation rather than the cause of pore formation. Zeng et al. [32] reported that thinning of the choroidal thickness in the macular hole is not limited to the fovea but occurs in all 6mm-diameter regions of the posterior pole. Tis suggests that changes in the choroidal blood fow might be one reason for the occurrence of macular holes. Tis study found no relationship between thinning of the choroidal thickness and changes in visual function; however, ∆BCVA was correlated with a change in the peripheral blood vessel density of the choroid capillary layer, and the morphology of the macular hole was correlated with changes in the peripheral blood vessel density of the choroid capillary layer but not with thinning of the choroidal thickness. It is speculated that the peripheral blood fow density of the choroid capillary layer may be more sensitive to changes in the choroidal thickness supporting the retinal vessels. A greater decrease in the peripheral blood vessel density may support more blood fow to the retina, thus contributing to better recovery of visual function.
Tere are several limitations of this study that should be noted. First, the number of cases was limited. Te study recruited more than 40 individuals with the idiopathic macular hole, but most of them were excluded due to the ft and poor image quality from the optical examinations. Second, the central retina and choroid could not be examined directly due to the macular hole; thus, the peripheral changes in the retina and choroid were examined. Tird, because of the high loss of participants, this study could only analyse changes three months after surgery rather than the trend in visual recovery. BCVA can also represent recovery of visual function to a certain extent.

Conclusion
Te recovery of visual function is afected by many factions after macular hole surgery. Te morphology of the macular hole as well as changes in the retinal and choroidal microstructure all contribute to the recovery of visual function after surgery. Te MHI, THI, and MCHAI could help analyse the anteroposterior tractional forces caused by cysts, and the lower index might obtain better restoration of visual function. Macular holes with a smaller basal diameter and greater height might obtain more blood fow support from the choroid, especially from the peripheral blood vessel of the choroid, allowing better recovery and greater improvements in visual function.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author (Dan Xu) upon request.

Ethical Approval
Tis study was consistent with the Helsinki Declaration and approved by the Wenzhou Medical University Eye Hospital Ethics Committee (KYK[2018]52).

Consent
Informed consent for participation was obtained from all the patients.