Hole Diameter ratio (HDR) for prediction of anatomical outcomes in stage III and IV idiopathic macular holes

Summary: In this study, we reviewed 101 eyes with stage III or IV IMHs treated by PPV combined with ILM peeling and room air tamponade, took the ratio of minimum diameter to maximum diameter (hole diameter ratio, HDR) as the predictive factor and found HDR<0.6(“A” type IMH) can be a predictive factor for a good anatomical outcome. Purpose: To identify hole diameter ratio (HDR) as the predictive factor for unclosed stage III and IV idiopathic macular holes (IMHs) by the use of preoperative optical coherence tomography (OCT). Methods: 101 eyes with stage III or IV IMH were included in this retrospective case series study. (cid:0) All cases were treated with vitrectomy combined with ILM peeling and room air tamponade. The MH minimum and maximum diameter, the ratio of the minimum to the maximum diameter (which dened as HDR) were performed. Results: 81 eyes (80.2%) got a Type I closure after surgery (group A). Postoperative unclosed MHs were found in 20 eyes (19.8%) (group B). The minimal diameter of IMHs (703.6±116.1um VS 597.6±120.1um, P<0.01) and HDR (0.6±0.1 VS 0.5±0.1, P=0.01) were both signicantly different between two groups. The cutoff value of the receiver operating characteristic curve was 0.6. The IMH with the HDR that less than 0.6 were dened as “A” type IMH and the others were dened as “H” type IMH. The closure rate of two type IMHs were 90.2% and 65.0% respectively. It had signicant differences. (P=0.002) Conclusions: HDR<0.6(“A” type IMH) can be a predictive factor for a good anatomical outcome in stage III and IV IMHs after vitrectomy combined with LM peeling and room air tamponade.


Background
Idiopathic macular hole (IMH) is one of the most often causes of poor vision. The estimated incidence of IMH ranged from 0.3-0.8% in the population. 1,2 It may cause a small dark spot in the central vision. MH may be related with high myopia or ocular trauma, but the reason of most MHs is unknown (idiopathic). 3 Based on the optical coherence tomography (OCT) images of macular region, IMHs can be divided into 4 stages, stage III or IV MH was taken as refractory MH for the low close rate after surgery. 4 Pars plana vitrectomy (PPV) is the most popular way to treat full-thickness MH at present. 5 Internal limiting membrane (ILM) peeling has been proved to increase the closure rates of IMHs. 6 More than 90% of IMHs were closed after PPV combined with ILM peeling and inert gas or room air tamponade. [7][8][9][10] The initial closure rates of IMHs ranged from 75.6-100%. [11][12][13][14][15] However, there are still some IMHs cannot be healed after surgery, especially in stage III and IV IMHs. The reason is still unknown.
In clinical practice, most unclosed IMHs were stage III and IV IMHs, especially when PPV combined with the short-term gas-room air. Therefore, the preoperative factors that affect the closure of stage III and IV IMHs are particularly important. It can help prompting surgeons to choose different operation methods before operation. Several studies have demonstrated the indexes according preoperative OCT features to predict the anatomic results after surgical interventions for MHs. While quantity of these studies and sizes of their samples were relatively small and none of them were aimed specially at stage III or IV IMHs which are the hardest IMHs to be healed. These parameters are also complex and di cult to calculate. [16][17][18][19] It's not that t for the busy clinical work to make a quick judgement.
In this study, we retrospectively reviewed the records of 101 eyes with stage III and IV IMHs and took the ratio of minimum diameter to maximum diameter (we named it as hole diameter ratio, HDR) as the predictive factor to predict the anatomical outcomes. All of them were treated with PPV combined with ILM peeling and room air tamponade.

Methods
This study followed the tenets of the Declaration of Helsinki and was approved by the Institutional Patients less than fty years old, or with a stage I and II IMH 4 , or high myopia, or history of vitrectomy, or any other fundus diseases will be excluded.
The data of age, sex, preoperative lens status, refractive error, axial length, preoperative best-corrected visual acuity (BCVA), and OCT ndings were recoded. The IMHs' minimum and maximum diameter, HDR, were measured and calculated OCT (Carl Zeiss, Dublin, CA, USA) was performed pre-and post-operation. The OCT images were reviewed and analyzed independently by two separate investigators (Y.Q. and YP.Y.). Any discrepancies were resolved by discussion with the principal investigator (W.L.) until a consensus was reached. Statistical analysis was performed using SPSS for Windows (version 23.0; IBM-SPSS, Chicago, IL). The results were expressed as the mean ± standard deviation (SD). Differences in the incidence rates of OCT ndings between patients with or without unclosed MH were analyzed by corrected Chi-square test. The measurements of BCVA were converted into the logMAR (logarithm of the minimum angle of resolution).
Mean age, duration of symptoms, axial length, BCVA, intraocular pressure and refractive error were analyzed by independent samples t-test. The discriminating power of HDR to predict closure rate after initial surgery was estimated by calculating the area under the receiver operating characteristic curve, with the area larger than 0.6 considered not determined by chance. Normal distributions of the data were checked before selecting the statistical analysis methods. A P-value < 0.05 was considered statistically signi cant.  BCVA: best corrected visual acuity; IOP: intra-ocular pressure;

Comparison of the Preoperative Clinical Characters of the Two Groups
Minimum diameter of IMH was also analyzed. The mean minimum diameter was 597.6 ± 120.1um in group A and 703.6 ± 116.1um in group B. It had signi cant differences between two groups (t=-3.6, P < 0.01).

Analysis of HDR in Two Groups
The mean HDR was 0.5 ± 0.1 in group A and 0.6 ± 0.1 in group B. It was signi cantly different in two groups (t=-1.9 P = 0.03).
The area under the receiver operating characteristic curve was 0.7 (95% con dence interval 0.6-0.8) for HDR with a cutoff value of 0.6, a sensitivity of 75.0%, and a speci city of 67.9% for predicting closure rate after initial surgery (P = 0.01, shown as Fig. 1).
For the cutoff value was 0.6 (which got from the receiver operating characteristic curve), we de ned IMHs with the HDR that less than 0.6 as "A" type IMHs and IMHs with the HDR that more than or equal to 0.6 as "H" type IMHs. There were 55 "A" type IMHs (67.9%) in group A and 6 "A" type IMHs (30.0%) in group B, and 26 "H" type IMHs (32.1%) in group A and 14 "H" type IMHs (70.0%) in group B. Two types of IMHs were signi cantly different in two groups (χ 2 = 9.6, P = 0.002).
The mean diameter had signi cant differences between "A" and "H" type IMHs (567.0 ± 108. 2   Comparison of closure rate of two types of IMHs There were 61 eyes with "A" type IMHs and 40 eyes with "H" type IMHs. In eyes with "A" type IMHs, 55 of 61 eyes got closed MHs following one surgical procedure, and the anatomical success rate was 90.2%. While in eyes with "H" type IMHs, only 26 of 40 eyes got closed MHs, and the anatomical success rate was 65.0%. The closure rates of two types of IMHs were signi cantly different.(P < 0.01)

Discussion
In this study, we identi ed HDR as a predictive factor for unclosed IMHs after PPV combined with ILM peeling in stage III and IV IMHs and found the IMH with HDR more than 0.6 ("H" type IMH) was hard to heal.
The relationships between preoperative OCT measurements and prognosis were revealed for the rst time by Ip et al. at 2002. 16 By comparing the anatomical reduction rate and recurrence rate of the IMH after operation, they found that IMH with a diameter under 400um had a higher recovery rate. Kusuhara et al. 18 proposed the concept of macular hole index (MHI, ratio of hole height to base diameter of hole) and found that MHI signi cantly correlated with the postoperative BCVA. While, most of the current studies associated with MHI are limited to the IMHs under stage III. Hole form factor (HFF) 20 de ned as the ratio of the sum of the lengths of the left and right oblique sides to the base diameters of the IMH. Ullrich et al. 21 found in patients with HFF > 0.9 the macular hole was closed following one surgical procedure, whereas in eyes with HFF < 0.5 anatomical success rate was only 67%. Haritoglou et al. 22 believed that higher HFF correlated with better postoperative visual outcome and the correlation coe cient is 0.36.
Ruiz-Moreno et al. 19 proposed tractional hole index (THI, ratio of height to minimum diameter) and diameter hole index (DHI, ratio of minimum diameter to baseline diameter). They found THI correlated signi cantly with postoperative best spectacle corrected visual acuity, but DHI did not. Wakely L et al. 23 measured macular hole inner opening diameter, minimum linear diameter, base diameter and macular hole height and calculated the MHI and FHI. They found base diameter, macular hole inner opening and minimum linear diameter could be used to predict anatomical and/or functional success rate of macular hole surgery. Preoperative base diameter is the most useful variable in this regard. Mingwei Zhao et al. 24 proposed a method named macular hole closure index (MHCI) to predict anatomical outcome after IMH surgery. MHCI was calculated as (M + N)/BASE based on the preoperative OCT status. M and N were the curve lengths of the detached photoreceptor arms, and BASE was the length of the retinal pigment epithelial layer (RPE layer) detaching from the photoreceptors. Some studies used central sub eld retinal thickness (CSRT) to predict anatomical results of MH surgery. While these results were still contradictory. 25,26 All these indexes can predict the anatomical or functional outcomes of IMH surgery in certain degrees. While they were either complex in calculation or lack of convincingness for obtained from a small sample. It is known to all that stage III and IV IMHs are more di cult to heal 4 . While few articles have been devoted to factors that affecting the healing of stage III and IV IMHs. In this study, we reviewed the records of 101 eyes treated by PPV combined with ILM peeling and room air tamponade to determine the anatomical predictive factors for stage III and IV IMHs and tried to nd an easy way to evaluate it before surgery.
In this study, there were no signi cant differences in age, duration of symptoms, preoperative BCVA, intraocular pressure (IOP) and axial length between patients with and without unclosed IMHs. While the minimal diameter(P < 0.01) and the HDR (P = 0.01) are signi cantly different in two groups. Which means the minimum diameter and HDR are two important factors for closure of stage III and IV IMHs.
Diameter is an important factor to in uence the result of IMH surgery. It has been reported that IMH with a minimum diameter under 400um had a higher recovery rate. 6 Wu Liu et al. 11 found the diameter more than 677um is a risk factors for stage III and IV IMH that underwent 23-gauge vitrectomy, ILM peeling, and air tamponade. In this study, the mean minimum diameter was 597.57.09 ± 120.14um in eyes with closed stage III or IV IMHs and 703.60 ± 116.13um in eyes with unclosed stage III or IV IMHs (P < 0.01). Which means the minimum diameter is indeed an important factor for postoperative anatomical outcomes in stage III and IV IMHs. The result in these cases is consistent with previous literature.
Based on past experiences, we found the forms were different between eyes with closed and unclosed stage III/IV IMH, thus we tried to nd a method to predict the closure of stage III and IV IMHs though the IMH's shape before surgery and hoped that this method is simple and easy to use in clinical work.
HDR (hole diameter ratio) is de ned as minimum diameter to maximum diameter in this study. It re ects the magnitude of pulling force in the tangential direction of the IMH. The mean HDR was 0.5 ± 0.1 in eyes with closed IMHs and 0.6 ± 0.1 in eyes with unclosed IMHs. It was signi cantly different in two groups (P = 0.03). Ratio of minimum diameter to maximum diameter has been mentioned as diameter hole index (DHI) in previous studies. While in those study, no relationship was found between the ratio and anatomical/functional outcome. We think the reasons for the difference conclusions may be as follow: First, the objects were different. All stages of IMHs were included in previous studies, while only stage III and IV IMHs were included in this study. Stage III and IV IMHs are refractory IMHs. The diameters are larger and the pulling force in the tangential direction may play a more important role in these IMHs (In stage III/IV IMHs, posterior vitreous detachment happens, which means the tangential force become the main force to IMH). Therefore, HDR may be more suitable for evaluating stage III/IV macular holes. Second, sample size is different. In previous studies, the sample size was mostly below 50, which was relatively small. The sample size of this study is more than 100, and the conclusion is relatively more accurate. Also, the tamponades, the surgeon and others can all be factors that in uence the conclusions.
As HDR with a cutoff value of 0.6 in these cases, we chose 0.6 as a dividing line to de ning different types of IMHs. The IMHs with the HDR that less than 0.6 were de ned as "A" type IMHs. The IMHs with the HDR that more than or equal to 0.6 was de ned as "H" type IMHs. We found the mean diameter had signi cant differences between "A" and "H" type IMHs (P < 0.01) in this study, which means "H" type IMHs are with larger diameter. (Table 2) This indicates that the HDR itself contains the element of diameter.
The closure rate of stage III/IV IMHs is 90.2% in "A" type IMHs and only 65.0% in "H" type IMHs. It had signi cant differences in type of IMHs (P = 0.002). Which means the anatomical results can be well predicted according HDR. We classi ed IMHs according to HDR and named them according to the shape of IMHs. In this way, the possible prognosis can be judged simply and quickly according to the types of IMH before operation. An "A" type stage III and IV IMH with small diameter is relatively easy to close. While for large IMH with a "H" type shape (HDR greater-than-equal 0.6), we recommend that more methods should be used during surgery. (Fig. 2,3) The limitations of the current study included as follow: First, it is a retrospective case series study; Second, we only analyzed the relationship between HDR and anatomical outcomes, the functional outcomes need further discuss.
In conclusion, HDR < 0.6("A" type IMH) can be a predictive factor for a good anatomical outcome in stage III and IV IMHs after vitrectomy combined with LM peeling and air tamponade. It can be easily calculated and intuitively noticed from OCT and may help doctors and patients predicting the surgical results of stage III and IV MHs. Abbreviations HDR: hole diameter ratio; PPV: pars plana vitrectomy; ILM: internal limiting membrane; IMH: idiopathic macular hole; BCVA: best-corrected visual acuity; OCT: optical coherence tomography; IOL: intra ocular lens; SD: standard deviation; logMAR: logarithm of the minimum angle of resolution.

Declarations
Ethics approval and consent to participate All of the patients gave a written informed consent and the study was conducted in accordance with the tenets of Declaration of Helsinki and approved by the Ethics Committee of Beijing Tongren Hospital (No. 2017-056-02). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This is a 63-year-old man with an "A" type IMH. The minimal diameter was 697um and maximum diameter was 1656um. The diameter of his MH was "large" but the HDR was less than 0.6. The MH was closed 1 month after PPV combined with ILM peeling and room air tamponade.