Etiologies and Clinical Characteristics of Patients with Macular Hole: A 8-Years Single-Center Retrospective Study

10 Background ： To investigate the etiologies and clinical characteristics of full-thickness macular 11 hole (FTMH) patients at Shanxi eye hospital of North China. 12 Methods ： Patients diagnosed with FTMH and treated with surgery from 2012 to 2020 were 13 included, and the etiologies and clinical features of different types of MHs were analysed in the 8-years 14 cross sectional retrospective study. Multivariate correlation analysis was used to predict the related 15 factors affecting baseline vision. 16 Results ： A total of 752 cases (776 eyes) were analysed. The top three causes of MH were idiopathic 17 (IMH , 64.4%), myopic (MMH , 21.1%) and traumatic (TMH , 3.7%). Among these three causes ’ 18 groups, there were significant differences in sex distribution, age, and baseline BCVA. Female was 19 predominated in IMH and MMH, while it was the opposite in TMH. The age of onset in IMH was older 20 than MMH and TMH. The baseline Logarithm of the Minimum Angle of Resolution (logMAR) 21 best-corrected visual acuity (BCVA) in IMH (Z=8.9, p<0.001) and Others group (Z=4.0, p<0.001) were 22 significantly better than in MMH. In IMH, female patients had younger age, shorter axial length, and 23 poorer baseline BCVA than male, while in MMH there were no significant differences between sexes. 24 Multivariate correlation analysis showed that the smaller hole diameter of IMH, detachment and younger age in TMH, may resulted in better baseline Conclusions in Therefore myopia is helpful for early detection and timely treatment.

group, IMH, traumatic MH (TMH), and myopic MH (MMH) accounted for 87.1%, 5.4% and 2.0% of 40 the total MH. 6 Previous studies 5 7 8 on clinical features of MMH and TMH have been relatively rare 41 compared to IMH, and the results have shown that IMH and MMH had higher incidence in female, 42 while TMH was more common in young male. The studies [9][10][11] found the factors affecting baseline 43 vision included hole size in IMH and MMH with retinal detachment (RD), no TMH-related studies 44 were reported. 45 It is well known that East Asia has the largest number of myopia people in the world and the 46 prevalence of high myopia is up to 6.8-21.6%, while it is 1-4% in general. 12 This discrepancy may 47 result in different composition ratios of diverse MH types in the published articles and there was little 48 literature comparing the clinical features of these types together. 4 Up to date, we have no found any 49 relevant study in China. Therefore the aim of this study was to explore the etiologies and clinical 50 characteristics of MH and to analyse the related factors affecting baseline vision at Shanxi eye hospital 51 of North China. 52

53
The inclusion and exclusion criteria of MH patients 54 This study was conducted at Shanxi eye hospital, which is the only tertiary eye hospital in Shanxi 55 province of North China, and approved by the ethnic committee of Shanxi eye hospital. The study 56 protocol adhered to the tenets of the Declaration of Helsinki. FTMH patients admitted and performed 57 surgery in our hospital from October 2012 to October 2020 were included. According to the etiologies, 58 they were classified into four groups: IMH, MMH, TMH and Others. The inclusion criteria of MMH 59 were defined as refractive status＞-6.00DS or axial length (AL) ≥26.0mm, accompanied with or without 60 RD. If with RD, macular hole was the only one and the extent did not exceed the retinal vascular arch. 61 The cases with ocular trauma history were included in TMH no matter visual loss immediately or lately. 62 Others group included all other recorded causes, such as vitrectomy, the history of DR, retinal vein 63 occlusion, glaucoma, laser photocoagulation and intraocular injection, etc. The cases were excluded 64 when they had peripheral RD caused by peripheral retinal degeneration or refractive media opacity. 65 The priority order of enrollment was Others>TMH>MMH>IMH. 66

82
In total 776 eyes (752 cases) were enrolled. The eyes were classified into four groups, as shown in 83 Figure 1  The clinical features of IMH, MMH, TMH and Others groups were shown in Table 1. In terms of 93 male-to-female ratio of affecting eyes, male had higher proportion in TMH, while in other groups 94 female were more common. IMH and Others had better VA than MMH. Age of onset was arranged in 95 order of: IMH>MMH>TMH, of which MMH was 6.5 years younger than IMH, and Others group was 96 comparable with MMH in age. Meanwhile, IMH and MMH had different incidence in different age 97 ranges, as shown in Figure 2. Before the sixth decade, the incidence of MMH was higher than that of 98  Figure 2 The incidence of IMH and MMH in different age groups 103 To compare the differences of epidemiological characteristics between sexes (Table 2), we found 104 that female had significantly younger age (Z=3.5, p<0.001), worse baseline VA (Z=2.6, p=0.010) and 105 shorter AL (Z=7.0, p＜0.001) than male, but there were no differences in diameter of hole and duration 106 in IMH. While there were no differences between sexes in MMH and TMH. For MMH, patients with 107 RD had longer AL (Z=-2.3, p=0.021), shorter duration (Z=3.2, p=0.01) and worse VA (Z=-6.9, p<0.001) 108 than without RD. After comparing the cases of IMH and MMH with or without RD (Figure 3), no 109 difference was found in the baseline BCVA between IMH and MMH without RD. 110  Factors related to preoperative VA among MHs are shown in Table 3. The factors affecting VA in 117 IMH were sex, hole size and duration, but the VA only differed between the duration ≤1month and 118 3-6 month group. After adjusting for possible confounding factors, only hole size was significantly 119 associated with VA (r=0.386, p<0.001, 95%CI 0.313 to 0.463). Moreover, there was significant 120 correlation between duration and hole size in IMH (r=0.303, p<0.001, 95%CI 0.223 to 0.387). 121 Spearman correlation analysis showed that both AL and with/without RD were correlated with VA in 122 MMH, while after controlling for possible confusing factors, with RD was the only factor correlated 123 with worse VA (r=-0.491, p<0.001, 95%CI -0.590 to -0.374). The multivariate correlation analysis of 124 AL, hole size, age and VA in TMH showed that there was a significant correlation only between age and 125 VA (r=0.446, p=0.025, 95%CI 0.051 to 0.714). 126

129
The present study using 8-years cases analysed the etiologies and epidemiological characteristics of 130 MH, and focused on comparing the discrepancies of age and sex proportion in IMH, MMH and TMH, as 131 well as the different factors affecting baseline VA among them. 132 In previous epidemiological investigations of Norway and Australia, 6 9 the proportion of IMH was 133 larger than this study, accounted for 85.9% and 87.1% respectively, and male-to-female ratio was 1:2.2 134 and 1:2, but in other retrospective clinical reports concerning surgical patients, sex ratio was comparable 135 to our data nearly 1:4. 14 15 Also in the Norway and Australia studies, 6 9 MMH only accounted for 1-2%, 136 while 21.1% in our study, the reason for this discrepancy probably is that we included MMH with 137 retinal detachment (MHRD), about 57.9% of all MMH. But even though MHRD were discarded, 138 MMH still accounted for nearly 10%, higher than previous reports. The higher prevalence of myopia in 139 East Asia like China attributed to the result of different MH proportion. 12 16 In previous studies 9 on 140 clinical surgical patients, TMH represented 3% of MH, which is consistent with our study, but for those 141 studies 17 including all TMH cases, TMH accounted for 5-8.2%, the difference for those who less than 142 24 years old with diameter of hole <0.2DD have more chance of achieving spontaneous closure. 17 143 Therefore, we only included surgery patients with a significant decrease in VA or a trend of gradual 144 enlargement of the hole during follow-up. TMH was called the second largest MH, 7 however in 145 consideration of the result of MH proportions in our study and spontaneous closure in TMH, the 146 accuracy of the above study need to be further verified. 147 IMH was older than MMH and TMH, the distinctions of the three types of MH at the age of onset 148 may be related to their underlying pathogenesis. Both IMH and MMH are complications during the 149 process of posterior vitreous detachment (PVD) which is the consequence of the interaction between 150 vitreous liquefaction and progressive weakening of the vitreoretinal adhesion. 18 In general, the 151 posterior vitreous cortex initially detaches at the paramacular area and extends to the perifoveal area 152 and then to the optical disc, finally a complete PVD develops, and this inevitable process changes with 153 age. [18][19][20] IMH is caused by vitreomacular traction (VMT) which is characterized by aberrant PVD and 154 accompanied by anatomic distortion of the foveal, whereas secondary MH is caused by other 155 pathological characteristics other than VMT. 1 The axial elongation and the formation of posterior 156 scleral staphyloma in high myopia accelerate the vitreous liquefaction and its instability, which results 157 in abnormal PVD that has more likely to develop MH, and the greater degree of refraction and the 158 longer AL, the earlier the PVD occurs. 21 22 Although the axial elongation contributes to the earlier 7 addition to the effect of PVD on formation of FTMH, lower concentration of collagen, protein and 161 hyaluronic acid can prompt the MH development. 23 162 The exact mechanism of TMH following blunt trauma is still controversial, it is generally believed 163 that the blunt trauma leads to foveal tissue loss caused by anteroposterior vitreous traction on the fovea. A 164 sudden decrease in the globe's anterior-posterior diameter causes a equatorial expansion of globe, 165 resulting in horizontal and tangential forces and splitting of the retinal layers at the fovea. 7 24 25 While 166 Rossi et al 25 found TMH could also occur in non-vitreous eyes, it revealed that damp shockwaves was 167 also responsible for trauma-related retinal lesions. Accidental high-power laser MH is caused mainly 168 by the rapid photo-thermal damage or photodisruptive mechanism. 26 169 With respect to the age of IMH, the results of various epidemiological investigations were 170 inconsistent, roughly between 56.2-70.2 years. 6 9 27 A respective study 4 of different types of MHs has 171 described that the mean age of MMH was 42 years, younger than our study. The onset of both IMH and 172 MMH changed with age, which is consistent with the changing of PVD. The area of vitreous macular 173 adhesion gradually decreases after 30 years, the stress acting on the foveal could be increased with 174 decreasing of adhesion area, and the incidence of partial PVD with sustained PVD peaks in the sixth 175 decade. 28 Therefore, the onset of IMH is about 60 years old. PVD studies on MMH have shown that it 176 occurred earlier, 19 22 but the exact time is not known. In our study, the age of MMH was 6.5 years 177 younger than IMH, which may indicate the PVD of high myopia occurred almost 6.5 years earlier than 178 without myopia. Ali et al 27 revealed that age was an independent risk factor of IMH, yet both MMH 179 and IMH showed a gradual increase in the proportions of cases with age in our study, so it may also be 180 an important risk factor for the occurrence of MMH. TMH is more common in young male, since 181 ocular trauma mostly occurred in sports or work-related accidents. 7 24 182 Regarding the onset in different sexes, female had higher incidence and younger age than male in 183 IMH and MMH, although there was no statistically significant difference in MMH age. In female, the 184 decreased estrogen affects the connective tissue, which causes the acceleration of vitreous liquefaction, 185 making it earlier in PVD and more quickly in declining of vitreomacular adhesion area, ultimately 186 leading to more and earlier onset in female. [29][30][31] Previous studies 9 15 on IMH found male had greater AL 187 than female, and no difference in baseline VA between sexes. The reason for the worse baseline VA in 188 female in our study could be that the average hole diameter was larger than male, though this difference 189 had no statistically significant. Similarly, Steel et al 32 also noted female tended to have larger size of 190 eyes without RD in MMH, besides, the longer AL the more probability to be accompanied by RD. The 205 results of previous studies have shown that the occurrence of RD was related to the AL. 12

222
Our data demonstrated that the most common causes of MH were IMH, MMH and TMH, MMH 223 accounted for 21.1%, higher than previous studies. Different pathogenesis of the three types of MH 224 makes it significant differences in age of onset, sex distribution and vision. Not only age, female was also 225 concerned about the risk factor of IMH and MMH. The MMH was nearly 6.5 years earlier than IMH.