Introduction: So far, there has been no closure grade system synthesizing morphological and microstructural features for large idiopathic macular holes (IMHs) treated by vitrectomy and internal limiting membrane (ILM) peeling. This study aimed to propose a concise one and explore its relevance with visual acuity and the related preoperative factors. Methods: Consecutive patients with large IMHs (minimum diameter >400 μm), undergoing vitrectomy and ILM peeling, obtaining primary closure and regularly followed-up were enrolled. Preoperative clinical charts and spectral-domain optical coherence tomography (SD-OCT) parameters were reviewed. SD-OCT images and best corrected visual acuity (BCVA) were assessed at 1, 4, and 10 months postoperatively. SD-OCT features at last visit were categorized by BCVA significance, and preoperative risk factors were analyzed. Results: Sixty-eight eyes from 64 patients were enrolled. The 10-month postoperative SD-OCT images were categorized into closure grade 1, 2, and 3 with successively decreased BCVA (p < 0.001). During early follow-up, part of grades 2 and 3 could evolve into the upper grade, respectively, but grade 3 could never evolve into grade 1 and exhibited the least satisfactory long-term BCVA. Binary logistic regression showed that large minimum linear diameter (MLD) was a risk factor for grade 3 occurrence (p < 0.001), with a cutoff value of 625.5 μm from the receiver operating characteristic curve for MLD predicting grade 3 occurrence (p = 0.001). Conclusion: Long-term closure status of large IMHs could be categorized into three grades with BCVA significance. Large horizontal MLD is a risk factor for occurrence of grade 3 closure with unsatisfactory visual recovery.

Types of macular hole closure could be appreciably affected by the design of surgical protocol and correlated with visual prognosis. Idiopathic macular holes (IMHs) treated by pars plana vitrectomy (PPV), internal limiting membrane (ILM) peeling, and intraocular tamponade obtain a closure rate of 91%–98% and median postoperative visual acuity of approximately 20/40 according to recent surgical studies [1]. The anatomical surgical outcomes were categorized into elevated/open, flat/open, and flat/closed by ophthalmoscopy [2], and U-type, V-type, and W-type by time-domain optical coherence tomography with visual significance [3]. In 2003, Kang et al. [4] indicated that type 1 closure (corresponding to U-type and V-type) exhibited larger extent of postoperative visual improvement than type 2 closure (corresponding to W-type), and recurrence of macular hole occurred only in the type 2 closure group. Therefore, type 1 closure seems to be more satisfactory.

To date, spectral-domain optical coherence tomography (SD-OCT) has improved the visualization of the fovea approaching the histologic level and demonstrates certain associations between postoperative visual acuity and indicators, including the central retinal thickness [5, 6], the integrity of the ellipsoid zone (EZ) and the external limiting membrane (ELM) [7, 8], and the presence of foveal lucency [9, 10]. Closure category systems including microstructural indicators besides morphology may better reflect the prognosis and accuracy of the surgery. However, these indicators are somewhat detailed for clinical applications, and the interrelationship of them has not been fully clarified. This study attempts to propose a concise closure grade system of large IMHs synthesizing morphological and microstructural features by SD-OCT, correlate it with postoperative visual outcome, and explore its evolution and preoperative predictors.

Consecutive patients with large IMHs who underwent PPV and ILM peeling in Beijing Tongren Hospital from August 2015 to February 2019, obtained primary IMH closure, and were regularly followed up 1, 4, and 10 months postoperatively were enrolled. Large IMHs referred to those with a minimum diameter >400 μm [11]. Primary closure meant the neurepithelium covering the retinal pigment epithelium with no exposure confirmed by the 1-month postoperative SD-OCT examination. Macular holes of other etiologies such as pathological myopia and trauma, treated with other techniques of the ILM, unclosed 1 month postoperatively, or reopened within follow-up were excluded, so were patients with ineligible images of SD-OCT, vision impair due to advanced cataract without treatment, or other fundus diseases during follow-up. This study adhered to the tenets of the Declaration of Helsinki and was approved by the Ethical Review Committee of Beijing Tongren Hospital, Capital Medical University. All the participants submitted written consents after explaining of the study protocol.

Preoperative medical charts of all the patients were reviewed, including age, gender, duration of symptoms, and best corrected visual acuity (BCVA), intraocular pressure (by Full Auto Tonometer TX-F; Canon Canada, Quebec), and axial length (from IOL Master Biometry, Carl Zeiss Meditec, Jena, Germany) a week before surgery.

The standard 3-port PPV with 23-gauge technique was conducted by a single surgeon (W.L.) under local anesthesia. A combined phacoemulsification and intraocular lens implantation were performed prior to the vitrectomy, if necessary. The ILM was removed with a diameter of approximately 2–3 optic disk diameters around the macular hole without staining, and the epiretinal membrane was also peeled, if presented. After adequate fluid-air exchange, long-lasting gas was injected or sterile air was retained in vitreous cavity. All the patients were asked to keep strict face-down position for one (air) or 2 weeks (gas) after surgery with conventional postoperative pharmaceutical protocol.

SD-OCT examination (Cirrus high-definition OCT; Carl Zeiss, Dublin, CA) was performed by experienced technicians. The fovea was scanned by the Macular Cube 512 × 128 and 5 Line Raster protocol both horizontally and vertically before and after surgery. Minimum linear diameter (MLD) was defined as the shortest distance between the neuroepithelial ends roughly parallel to the RPE, the basal diameter (BD) as the span of the two ends at the level of RPE, and the diameter hole index as MLD/BD. Instead of the mean central subfield thickness, foveal retinal thickness (FRT) was manually measured as the vertical distance between the lowest point of the inner neuroepithelial contour and the RPE in the foveola. Glial proliferation refers to the presence of a foveal homogenous hyper-reflective lesion [12]. Morphology and microstructure of the foveola were carefully inspected on every time point of follow-up. At last visit, the SD-OCT images were categorized into different closure grades with BCVA significance and correlated with preoperative factors.

Statistical analysis was performed by SPSS 22.0 (SPSS for Windows, Chicago, IL, USA). BCVA was converted to the logarithm of the minimum angle of resolution for analysis. Data were compared by appropriate tests accordingly. Binary logistic regression was performed to explore risk factors for certain closure grades, and the receiver operating characteristic (ROC) curve was drawn to find the cutoff value of the occurrence. A p < 0.05 was considered statistically significant.

A total of 68 eyes from 64 patients (10 males and 54 females with an average age of 63.6 years old) were enrolled. All the IMHs were larger than 400 μm in MLD and showed no vitreomacular traction. The BCVA of all eyes improved from 0.1 (0.05, 0.2) to 0.7 (0.6, 0.8) during the 10-month follow-up. At last visit, the closure status depicted by SD-OCT was categorized into three grades based on different BCVA and FRT (shown in Fig. 1 and Table 1).

  • 1.

    Grade 1 (11/68 eyes): the fovea revealed relatively normal contour with considerable FRT. The highly reflective layers corresponding to the ELM and EZ were clear and intact. Ten months postoperatively, this grade exhibited a mean FRT of 170.8 ± 26.4 μm (range: 129–210) with a mean BCVA of 0.8 (range: 0.6–1.0).

  • 2.

    Grade 2 (53/68 eyes): the fovea revealed considerable FRT, but various defects between the ELM and the RPE could be observed in the foveola. At last visit, this grade exhibited a mean FRT of 146.8 ± 40.0 μm (range: 53–268) with a median BCVA of 0.7 (range: 0.1–1.0) and could be divided into three subgrades as follows by shape of the defects:

  • 1)

    Grade 2A (20/53 eyes): a rectangle-like lucency of mild backscattering with a roof of the ELM and a foot of the RPE presented in the foveola, while the ELM and EZ were clear and continuous outside.

  • 2)

    Grade 2B (19/53 eyes): an irregular-shaped lucency of mild backscattering between the ELM and the RPE exhibited in the foveola, with ELM and EZ discontinuity outside but not filling the whole extension of the foveola.

  • 3)

    Grade 2C (14/53 eyes): there was discontinuity in the tissue between the ELM and the RPE in the foveola without a definite lucency.

  • 3.

    Grade 3 (4/68 eyes): the RPE in the foveola was covered with indistinguishable layers, and the FRT is rather thin. Relatively extensive defect of the ELM and EZ could be observed. Ten months postoperatively, this grade exhibited a median FRT of 77 μm (range: 53–123) with a median BCVA of 0.25 (range: 0.1–0.3).

Fig. 1.

SD-OCT images, descriptions, and number of cases of the three closure grades during follow-up.

Fig. 1.

SD-OCT images, descriptions, and number of cases of the three closure grades during follow-up.

Close modal
Table 1.

Comparisons of eyes with the three grades of closure

Grade 1 (N = 11)Grade 2 (N = 53)Grade 3 (N = 4)p value
Preoperative 
 Age, years 61.8±1.7 63.5±5.1 70 (61, 81) 0.19 
 Gender, M/F 1/10 9/44 1/3 0.72 
 DS, months 5.4±3.7 6 (3, 8) 9 (6, 17) 0.13 
 BCVA, LogMAR (Snellen) 1.0 (0.5, 1.3), 20/200 (20/400, 20/66) 0.8 (0.7, 1.3), 20/125 (20/400, 20/100) 1.5 (1.3, 1.7), 20/666 (20/1,000, 20/400) 0.01* 
 IOP, mm Hg 15.8±3.4 15.0 (13.0, 18.0) 16.0 (12.8, 17.0) 0.98 
 AXL, mm 23.0±0.8 23.2±0.9 22.7 (22.6, 23.1) 0.83 
 MLD, μm 609±134 581±143 746 (681, 861) 0.05a 
 Combined cataract surgery, Y/N/pseudophakic 10/1/0 48/5/0 3/0/1 0.39 
 Tamponade, gas/air 5/6 14/39 1/3 0.44 
Postoperative 
 10-month     
 BCVA, LogMAR (Snellen) 0.08±0.07, 20/25±20/25 0.2 (0.1, 0.2), 20/28 (20/33, 20/20) 0.6 (0.5, 0.9), 20/80 (20/200, 20/66) <0.001* 
 FRT, μm 171±26 147±40 77 (58, 112) 0.002* 
Grade 1 (N = 11)Grade 2 (N = 53)Grade 3 (N = 4)p value
Preoperative 
 Age, years 61.8±1.7 63.5±5.1 70 (61, 81) 0.19 
 Gender, M/F 1/10 9/44 1/3 0.72 
 DS, months 5.4±3.7 6 (3, 8) 9 (6, 17) 0.13 
 BCVA, LogMAR (Snellen) 1.0 (0.5, 1.3), 20/200 (20/400, 20/66) 0.8 (0.7, 1.3), 20/125 (20/400, 20/100) 1.5 (1.3, 1.7), 20/666 (20/1,000, 20/400) 0.01* 
 IOP, mm Hg 15.8±3.4 15.0 (13.0, 18.0) 16.0 (12.8, 17.0) 0.98 
 AXL, mm 23.0±0.8 23.2±0.9 22.7 (22.6, 23.1) 0.83 
 MLD, μm 609±134 581±143 746 (681, 861) 0.05a 
 Combined cataract surgery, Y/N/pseudophakic 10/1/0 48/5/0 3/0/1 0.39 
 Tamponade, gas/air 5/6 14/39 1/3 0.44 
Postoperative 
 10-month     
 BCVA, LogMAR (Snellen) 0.08±0.07, 20/25±20/25 0.2 (0.1, 0.2), 20/28 (20/33, 20/20) 0.6 (0.5, 0.9), 20/80 (20/200, 20/66) <0.001* 
 FRT, μm 171±26 147±40 77 (58, 112) 0.002* 

DS, duration of symptoms; BCVA, best corrected visual acuity; LogMAR, logarithm of the minimum angle of resolution; IOP, intraocular pressure; AXL, axial length; MLD, minimal linear diameter; FRT, foveal retinal thickness.

*p < 0.05.

aA critical p value.

At last visit, the FRT differed significantly among the three grades (p = 0.002, Kruskal-Wallis test) with grade 3 significantly thinner than grade 1 (p = 0.004, Mann-Whitney test) and grade 2 (p = 0.007, Mann-Whitney test), while grade 2 similar to grade 1 (p = 0.021, t test). Moreover, BCVA of the three grades differed significantly between one another (p < 0.001, Kruskal-Wallis test): P1 (grade 1 vs. grade 2) = 0.01, P2 (grade 1 vs. grade 3) = 0.003, P3 (grade 2 vs. grade 3) = 0.001, all by Mann-Whitney test. The three subgrades of grade 2 revealed no significant differences in either FRT or BCVA.

The three closure grades above also constituted the majority of the short-term closure status. One month postoperatively, 40 and 15 eyes exhibited closure grade 2 and 3, respectively, 3 presented bridge-like closure, 9 demonstrated glial proliferation, and the other exhibited cystoid edema. Four months after surgery, all eyes exhibited the three grades of closure except for 3 presenting glial proliferation.

No IMH obtained grade 1 closure within 1 month postoperatively. Grade 1 could only be evolved from grade 2 and remained stable once obtained. The three subgrades of grade 2 could convert among one another with relatively stable vision or evolve into grade 1 (7.5%, 3/40 from 1 month to 4 months; 18.0%, 9/50 from 4 months to 10 months) for a better vision. Grade 3 closure presented more frequent in the early stage postoperatively, the majority of which gradually evolved into different subgrades of grade 2 (46.7%, 7/15 from 1 month to 4 months; 75%, 9/12 from 4 months to 10 months), but could never evolve into grade 1. Half of the 4 eyes exhibited grade 3 closure at last visit obtained this grade 1 month after surgery and kept stable; the other half evolved from glial proliferation. No one of the total 68 eyes showed downgraded over the follow-up of 10 months.

Glial proliferation could be observed during the recovery process. From 1 month to 4 months after surgery, 3/9 eyes evolved from glial proliferation to grade 2 closure and 4/9 eyes evolved into grade 3. From 4 months to 10 months, the proportion of the evolution above were 2/3 and 1/3, respectively. Here presents one case (shown in Fig. 2) which demonstrated the evolution.

Fig. 2.

A typical case demonstrating the evolution of closure states. a A 70-year-old female with an IMH in the left eye. The MLD was 552 μm, and the BD was 963 μm. b One month after 23-gauge vitrectomy and ILM peeling, the hole was closed with homogeneous high backscattering, which was considered as glial proliferation, filling the foveola (asterisk). The ELM and EZ outside the glial tissue were defective to some degree (white arrows). c Four months after surgery, the glial proliferation evolved into closure grade 2A with a rectangle-like lucency of mild backscattering in the foveola (black arrow head). The ELM and EZ were clear and continuous outside the lucency (black arrows). d Closure grade 2A evolved into grade 1 with relatively smooth surface and appreciable retinal thickness similar to normal foveal contour. The highly reflective layers corresponding to the ELM and EZ were clear and continuous.

Fig. 2.

A typical case demonstrating the evolution of closure states. a A 70-year-old female with an IMH in the left eye. The MLD was 552 μm, and the BD was 963 μm. b One month after 23-gauge vitrectomy and ILM peeling, the hole was closed with homogeneous high backscattering, which was considered as glial proliferation, filling the foveola (asterisk). The ELM and EZ outside the glial tissue were defective to some degree (white arrows). c Four months after surgery, the glial proliferation evolved into closure grade 2A with a rectangle-like lucency of mild backscattering in the foveola (black arrow head). The ELM and EZ were clear and continuous outside the lucency (black arrows). d Closure grade 2A evolved into grade 1 with relatively smooth surface and appreciable retinal thickness similar to normal foveal contour. The highly reflective layers corresponding to the ELM and EZ were clear and continuous.

Close modal

Comparison of clinical charts among the three grades was listed in Table 1. Eyes exhibiting grade 3 closure showed significantly poorer BCVA and critically significantly larger MLD than those exhibiting the other two grades; eyes exhibiting grade 1 and 2 demonstrated similar characteristics. Different types of tamponade and whether combined cataract surgery was conducted exhibited no effect on closure grades.

A total of 16 eyes ever exhibited grade 3 during follow-up, demonstrating significantly thinner postoperative FRT (p < 0.001, t-test) and poorer postoperative BCVA (p < 0.001, Mann-Whitney test) at last visit (shown in Table 2), and none of them evolved into grade 1 closure. Preoperatively, these eyes exhibited longer duration of symptoms (p = 0.005, Mann-Whitney test), poorer preoperative BCVA (p = 0.035, Mann-Whitney test), and larger MLD (p = 0.002, t-test). Binary logistic regression showed that larger horizontal MLD (OR = 1.03, p = 0.01) and smaller horizontal diameter hole index (OR = 0.000, p = 0.045) were risk factors for occurrence of grade 3 (shown in Table 3).

Table 2.

Comparisons of eyes with or without grade 3 closure

Group 1 (N = 16)Group 2 (N = 52)p value
Age, years 65.1±6.9 63.2±4.9 0.22 
Duration of symptoms, months 6.5 (5.25, 12.0) 5 (3, 6) 0.005* 
Preoperative BCVA, LogMAR (Snellen) 1.3 (0.75, 1.3), 20/400 (20/400, 20/100) 0.82 (0.70, 1.11), 20/133 (20/250, 20/100) 0.035* 
IOP, mm Hg 16.25±3.0 15.6±3.3 0.48 
AXL, mm 23.2±0.89 23.1±0.84 0.82 
Horizontal MLD, μm 697.3±133.6 564.7±133.3 0.001* 
Horizontal BD, μm 1,228.8±179.9 1,114.1±252.8 0.10 
Horizontal DHI 0.57±0.10 0.51±0.08 0.01* 
Vertical MLD, μm 653.1±117.6 487.8±116.0 <0.001* 
Vertical BD, μm 1,189.1±157.8 1,006.2±214.6 0.002* 
Vertical DHI 0.55±0.08 0.48 (0.45, 0.55) 0.004* 
10-month postoperative BCVA, LogMAR (Snellen) 0.22 (0.22, 0.52), 20/33 (20/66, 20/33) 0.10 (0.10, 0.21), 20/20 (20/33, 20/20) <0.001* 
10-month postoperative FRT 99.9±30.2 161.3±33.1 <0.001* 
Group 1 (N = 16)Group 2 (N = 52)p value
Age, years 65.1±6.9 63.2±4.9 0.22 
Duration of symptoms, months 6.5 (5.25, 12.0) 5 (3, 6) 0.005* 
Preoperative BCVA, LogMAR (Snellen) 1.3 (0.75, 1.3), 20/400 (20/400, 20/100) 0.82 (0.70, 1.11), 20/133 (20/250, 20/100) 0.035* 
IOP, mm Hg 16.25±3.0 15.6±3.3 0.48 
AXL, mm 23.2±0.89 23.1±0.84 0.82 
Horizontal MLD, μm 697.3±133.6 564.7±133.3 0.001* 
Horizontal BD, μm 1,228.8±179.9 1,114.1±252.8 0.10 
Horizontal DHI 0.57±0.10 0.51±0.08 0.01* 
Vertical MLD, μm 653.1±117.6 487.8±116.0 <0.001* 
Vertical BD, μm 1,189.1±157.8 1,006.2±214.6 0.002* 
Vertical DHI 0.55±0.08 0.48 (0.45, 0.55) 0.004* 
10-month postoperative BCVA, LogMAR (Snellen) 0.22 (0.22, 0.52), 20/33 (20/66, 20/33) 0.10 (0.10, 0.21), 20/20 (20/33, 20/20) <0.001* 
10-month postoperative FRT 99.9±30.2 161.3±33.1 <0.001* 

Group 1: eyes that ever exhibited grade 3 closure during three times of follow-up. Group 2: eyes never exhibited grade 3 closure during follow-up.

BCVA, best corrected visual acuity; LogMAR, logarithm of the minimum angle of resolution; IOP, intraocular pressure; AXL, axial length; MLD, minimal linear diameter; BD, basal diameter; DHI, diameter hole index = MLD/BD; FRT, foveal retinal thickness.

*p < 0.05.

Table 3.

Variables estimated by binary logistic regression

BWaldp valueEXP(B)95% EXP(B)
lowerupper
X1 Duration of symptoms 0.195 3.802 0.051 0.823 0.676 1.001 
X2 Preoperative BCVA 1.748 1.600 0.206 0.174 0.012 2.614 
X3 Horizontal MLD 0.030 6.402 0.011* 1.030 1.007 1.054 
X4 Horizontal DHI −23.56 4.009 0.045* 0.000 0.000 0.607 
X5 Vertical MLD 0.017 0.230 0.632 0.984 0.919 1.052 
X6 Vertical BD 0.017 1.004 0.316 0.983 0.952 1.016 
X7 Vertical DHI 12.00 0.091 0.763 0.000 0.000 4.8E+28 
Constant 32.58 2.520 0.112 1.4E+14   
BWaldp valueEXP(B)95% EXP(B)
lowerupper
X1 Duration of symptoms 0.195 3.802 0.051 0.823 0.676 1.001 
X2 Preoperative BCVA 1.748 1.600 0.206 0.174 0.012 2.614 
X3 Horizontal MLD 0.030 6.402 0.011* 1.030 1.007 1.054 
X4 Horizontal DHI −23.56 4.009 0.045* 0.000 0.000 0.607 
X5 Vertical MLD 0.017 0.230 0.632 0.984 0.919 1.052 
X6 Vertical BD 0.017 1.004 0.316 0.983 0.952 1.016 
X7 Vertical DHI 12.00 0.091 0.763 0.000 0.000 4.8E+28 
Constant 32.58 2.520 0.112 1.4E+14   

BCVA, best corrected visual acuity; MLD, minimal linear diameter; BD, basal diameter; DHI, diameter hole index = MLD/BD.

*p < 0.05.

Logit (P) = 0.03 × 3 – 23.56 × 4 (χ2 = 38.22; p < 0.001).

The ROC curve showed a cutoff value of 625.5 μm with an area under the curve of 0.77 (95% confidence interval 0.64–0.90), a sensitivity of 76.5%, and a specificity of 72.6% for horizontal MLD predicting occurrence of grade 3 (p = 0.001, shown in Fig. 3), which meant IMHs larger than 625.5 μm had significantly higher risk for grade 3 occurrence after surgery.

Fig. 3.

Receiver operating characteristic curve of MLD predicting occurrence of grade 3 closure during follow-up of 10 months. The area under the curve was 0.77, with a cutoff value of 625.5 μm, a sensitivity of 76.5%, and a specificity of 72.6%.

Fig. 3.

Receiver operating characteristic curve of MLD predicting occurrence of grade 3 closure during follow-up of 10 months. The area under the curve was 0.77, with a cutoff value of 625.5 μm, a sensitivity of 76.5%, and a specificity of 72.6%.

Close modal

This retrospective study categorized three stable long-term closure grades of large IMHs undergoing PPV and ILM peeling, based not only on morphological features but also microstructural details by SD-OCT. The long-term BCVA decreased significantly by the sequence of grade 1, 2, and 3. The horizontal MLD was proved to be the preoperative risk factor for grade 3 closure, which demonstrated unsatisfactory visual recovery. Comparing to previous reports, this study focused on large IMHs, and all the features discussed here mainly presented in the foveola.

Grade 1 reveals intact layers corresponding to ELM and EZ covering the RPE with normal contour and retinal thickness in the foveola. The eyes exhibiting this grade also demonstrated better preoperative BCVA and relatively smaller size of IMH. This nearly flawless status needs healing of the inner and outer neuroepithelia, realignment of the photoreceptors, and functional recovery of various cells, thus exhibits the best postoperative vision and is difficult to obtain within short duration, especially in large IMHs.

Most previous studies indicated that the foveal lucency may have no negative effect on the long-term vision when comparing eyes with and without lucencies [9, 13‒16]. Both grade 1 and grade 3 closure have no lucencies, so results of this study are generally in accordance with the conclusion above: grade 2 closure exhibits lower long-term vision than grade 1 but higher vision than grade 3. Given the span of the photoreceptor deficiency in this grade is usually much smaller than that of the foveola, photoreceptors around the defects could still serve function [16]. Thus, though lower than that of grade 1, visual acuity of this grade is satisfactory. This defective grade suggests reconstruction and realignment of the photoreceptors, and the conversion among the three subgrades may demonstrate different states of this process. It also explains why the 12 eyes exhibiting this grade could evolve into grade 1 for better vision several months later in this series. The three subgrades actually not only shared similarly considerable FRT but also showed differences in the manifestation of outer retinal defects. They could evolve into one another, but still all presented at last visit so we subdivided them according to detailed differences. The differences among these three subgrades were not as remarkable as those among grade 1, 2, and 3, so for general usage, they could all be applied as grade 2.

Grade 3 reveals worst vision due to thinner FRT and disorder of the outer retina. This grade is somewhat similar to the V-type closure raised by Imai and colleagues [3]. They described this type as the RPE and choriocapillaris layers covered with moderately backscattering layers with a notch where the layer was extremely thin [3]. There have been few other reports discussing closure status similar to this grade. In this study, grade 3 closure exhibits rather thin FRT as above and showed undistinguishable layers of the outer retina. Once exhibiting grade 3, the eyes could never evolve into grade 1, so the long-term visual recovery would be relatively unsatisfactory. Eyes with larger size of IMH exhibited higher risk for grade 3 occurrence. It suggests that larger IMHs (especially larger than 625.5 μm in MLD) may demonstrate not only lower primary closure rate [17, 18] but also poorer structural and functional recovery.

Glial proliferation was conventionally described as the presence of homogeneous moderate to high reflective tissue at the outer fovea [19]. It was considered essential for MH healing because the reconstruction process of the outer retina, such as ELM and EZ, could be initiated by it [7, 20]. However, glial tissue may also replace some entire intraretinal layers and prevent the normal photoreceptor alignment, perhaps due to large hole diameter or advanced age [18], and lead to an unsatisfactory visual outcome [7, 19, 21]. There seems to be a balance between glial proliferation and photoreceptor restoration [7], which needs to be clarified with further research.

ELM and EZ are suggested to be significantly correlated to postoperative vision [7, 8, 20, 22], but they are inconvenient to quantify during clinical practice. FRT is a relatively intuitional indicator of foveal morphology; it has shown only moderate correlation with visual acuity according to previous studies with large sample size [3, 6, 23], however. The central vision attributes to not only the thickness of the foveal retina but also the microstructural and functional recovery of the photoreceptors. The closure grades raised in this study synthesized morphological and microstructural indicators and could reflect levels of the visual recovery. This system would be relatively more concise and applicable in clinical practice.

To the best of our knowledge, this study for the first time proposed a closure grade system, synthesizing morphological and microstructural features, which would be relatively precise and concise for clinical applications. The limitations of this study include the relatively higher female/male ratio and the small sample size of grade 3. They can hardly be avoided due to the retrospective nature. However, comparisons were still reliable because gender composition exhibited no significant difference among the 3 grades. Though only 4 eyes at last visit, a total of 16 eyes, which ever demonstrated grade 3, exhibited similar characteristics: poorer long-term vision because of larger size of the hole. Conventionally speaking, long-term recovery may refer to 1 year after surgery. However, visual recovery accomplishes mainly during the first 6 months postoperatively [21], and previous reports defined the surgery end points at around 9 months after surgery [3, 4]. We suppose it acceptable that the 10-month closure grade referred as “long-term.”

In conclusion, long-term closure status of large IMHs treated by PPV and ILM peeling could be categorized into grade 1, 2, and 3 based on morphological and microstructural features with corresponding BCVA decreasing successively. Under certain circumstances, early-stage grade 2 and grade 3 closure could evolve into its upper grade for better vision. Large horizontal MLD, especially larger than 625.5 μm, is a risk factor for occurrence of grade 3 closure and may lead to unsatisfactory long-term vision.

This study protocol was reviewed and approved by the Ethical Review Committee of Beijing Tongren Hospital, Capital Medical University, with approval number TRECKY2019-097. Written informed consent was obtained for participation in this study.

The authors have no conflicts of interest to declare.

There are no funding sources to disclose for this study.

Y.Y. drafted the manuscript based on data acquisition, analysis, and interpretation. Y.Y., J.Z.B., and L.W. conducted the conception and design of the work, realized the conceptualization of the manuscript, and reviewed the literature. W.Z. and W.J. participated in data acquisition, interpretation, and literature revision. L.W., J.Z.B., Q.B., and L.L. contributed to the analysis and interpretation of data, as well as critical review and synthesis of the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

1.
American Academy of Ophthalmology Retina/Vitreous Panel
.
Preferred practice Pattern® guidelines. Idiopathic macular hole
.
San Francisco, CA
:
American Academy of Ophthalmology
;
2014
.
2.
Tornambe
PE
,
Poliner
LS
,
Cohen
RG
.
Definition of macular hole surgery end points: elevated/open, flat/open, flat/closed
.
Retina
.
1998
;
18
(
3
):
286
7
. .
3.
Imai
M
,
Iijima
H
,
Gotoh
T
,
Tsukahara
S
.
Optical coherence tomography of successfully repaired idiopathic macular holes
.
Am J Ophthalmol
.
1999
;
128
(
5
):
621
7
. .
4.
Kang
SW
,
Ahn
K
,
Ham
DI
.
Types of macular hole closure and their clinical implications
.
Br J Ophthalmol
.
2003
;
87
(
8
):
1015
9
. .
5.
Takamura
Y
,
Tomomatsu
T
,
Matsumura
T
,
Arimura
S
,
Gozawa
M
,
Takihara
Y
, et al
.
Correlation between central retinal thickness after successful macular hole surgery and visual outcome
.
Jpn J Ophthalmol
.
2015
;
59
(
6
):
394
400
. .
6.
Apostolopoulos
MN
,
Koutsandrea
CN
,
Moschos
MN
,
Alonistiotis
DA
,
Papaspyrou
AE
,
Mallias
JA
, et al
.
Evaluation of successful macular hole surgery by optical coherence tomography and multifocal electroretinography
.
Am J Ophthalmol
.
2002
;
134
(
5
):
667
74
. .
7.
Wakabayashi
T
,
Fujiwara
M
,
Sakaguchi
H
,
Kusaka
S
,
Oshima
Y
.
Foveal microstructure and visual acuity in surgically closed macular holes: spectral-domain optical coherence tomographic analysis
.
Ophthalmology
.
2010
;
117
(
9
):
1815
24
. .
8.
Ooka
E
,
Mitamura
Y
,
Baba
T
,
Kitahashi
M
,
Oshitari
T
,
Yamamoto
S
.
Foveal microstructure on spectral-domain optical coherence tomographic images and visual function after macular hole surgery
.
Am J Ophthalmol
.
2011
;
152
(
2
):
283
90.e1
. .
9.
Grewal
DS
,
Reddy
V
,
Mahmoud
TH
.
Assessment of foveal microstructure and foveal lucencies using optical coherence tomography radial scans following macular hole surgery
.
Am J Ophthalmol
.
2015
;
160
(
5
):
990
9.e1
. .
10.
Mahmoud
TH
,
McCuen
NBW
.
Natural history of foveolar lucencies observed by optical coherence tomography after macular hole surgery
.
Retina
.
2007
;
27
(
1
):
95
100
. .
11.
Duker
JS
,
Kaiser
PK
,
Binder
S
,
de Smet
MD
,
Gaudric
A
,
Reichel
E
, et al
.
The international vitreomacular traction study group classification of vitreomacular adhesion, traction, and macular hole
.
Ophthalmology
.
2013
;
120
(
12
):
2611
9
. .
12.
Kitao
M
,
Wakabayashi
T
,
Nishida
K
,
Sakaguchi
H
,
Nishida
K
.
Long-term reconstruction of foveal microstructure and visual acuity after idiopathic macular hole repair: three-year follow-up study
.
Br J Ophthalmol
.
2019
;
103
(
2
):
238
44
. .
13.
Kawano
H
,
Uemura
A
,
Sakamoto
T
.
Incidence of outer foveal defect after macular hole surgery
.
Am J Ophthalmol
.
2011
;
151
(
2
):
318
22
. .
14.
Kang
SW
,
Lim
JW
,
Chung
SE
,
Yi
CH
.
Outer foveolar defect after surgery for idiopathic macular hole
.
Am J Ophthalmol
.
2010
;
150
(
4
):
551
7
. .
15.
Zarranz-Ventura
J
,
Ellabban
AA
,
Sim
DA
,
Keane
PA
,
Kirkpatrick
JN
,
Sallam
AAB
.
Prevalence of foveolar lucency with different gas tamponades in surgically closed macular holes assessed by spectral domain optical coherence tomography
.
Retina
.
2018
;
38
(
9
):
1699
706
. .
16.
Qi
B
,
Yu
Y
,
You
Q
,
Wang
Z
,
Wang
J
,
Liu
L
, et al
.
Evolution and visual outcomes of outer foveolar lucency after surgery for large idiopathic macular hole
.
Graefes Arch Clin Exp Ophthalmol
.
2020
;
258
(
10
):
2117
24
. .
17.
Ch’Ng
SW
,
Patton
N
,
Ahmed
M
,
Ivanova
T
,
Baumann
C
,
Charles
S
, et al
.
The manchester large macular hole study: is it time to reclassify large macular holes
.
Am J Ophthalmol
.
2018
;
195
:
36
42
. .
18.
Yu
YP
,
Liang
XD
,
Wang
ZY
,
Wang
J
,
Qi
BY
,
Liu
W
.
Correlation between diameter and primary closure rate by internal limiting membrane peeling and air tamponade in large idiopathic macular holes
.
Chin J Ophthalmol
.
2019
;
55
(
10
):
739
46
. .
19.
Oh
J
,
Yang
SM
,
Choi
YM
,
Kim
SW
,
Huh
K
.
Glial proliferation after vitrectomy for a macular hole: a spectral domain optical coherence tomography study
.
Graefes Arch Clin Exp Ophthalmol
.
2013
;
251
(
2
):
477
84
. .
20.
Chang
YC
,
Lin
WN
,
Chen
KJ
,
Wu
HJ
,
Lee
CL
,
Chen
CH
, et al
.
Correlation between the dynamic postoperative visual outcome and the restoration of foveal microstructures after macular hole surgery
.
Am J Ophthalmol
.
2015
;
160
(
1
):
100
6.e1
. .
21.
Kitao
M
,
Wakabayashi
T
,
Nishida
K
,
Sakaguchi
H
,
Nishida
K
.
Long-term reconstruction of foveal microstructure and visual acuity after idiopathic macular hole repair: three-year follow-up study
.
Br J Ophthalmol
.
2019
;
103
(
2
):
238
44
. .
22.
Geenen
C
,
Murphy
DC
,
Sandinha
MT
,
Rees
J
,
Steel
DHW
.
Significance of preoperative external limiting membrane height on visual prognosis in patients undergoing macular hole surgery
.
Retina
.
2019
;
39
(
7
):
1392
8
. .
23.
Ko
TH
,
Witkin
AJ
,
Fujimoto
JG
,
Chan
A
,
Rogers
AH
,
Baumal
CR
, et al
.
Ultrahigh-resolution optical coherence tomography of surgically closed macular holes
.
Arch Ophthalmol
.
2006
;
124
(
6
):
827
36
. .