Prediction of spontaneous closure of traumatic macular hole with spectral domain optical coherence tomography

It has been known that some traumatic macular holes can close spontaneously. However, knowledge about the types of macular hole that can close spontaneously is limited. In this retrospective study, we investigated patients with traumatic macular hole who were followed-up for at least 6 months without any surgical intervention. Clinical data and spectral domain optical coherence tomography (SD-OCT) images were compared between groups with and without macular hole closure. Overall, 27 eyes were included. Spontaneous closure of macular hole was observed in 10 (37.0%) eyes. The holes with spontaneous closure had smaller minimum diameter (244.9 ± 114.4 vs. 523.9 ± 320.0 μm, p = 0.007) and less intraretinal cysts (10% vs. 76.5%, p = 0.001) compared to the holes that did not close spontaneously. The area under the curve of receiver operative characteristic was 0.812 and 0.832 for minimum diameter of macular hole and presence of intraretinal cysts respectively. Multivariate logistic regression showed that the presence of intraretinal cysts was an independent predictive factor for closure of macular holes. The group with spontaneous macular hole closure had a high chance of visual improvement. Our study suggests that the absence of intraretinal cysts on SD-OCT can predict spontaneous closure of traumatic macular hole.

The minimum diameter of traumatic macular holes with spontaneous closure (244.9 ± 114.4 μ m) was smaller compared to those that did not close spontaneously (523.9 ± 320.0 um) (p = 0.007, Mann-Whitney test). The incidence of intraretinal cysts around macular hole was less in macular holes which closed spontaneously (1/10, 10.0%) compared to those that did not (13/17, 76.5%) (p = 0.001, Fisher's exact test). No statistically significant difference was found in age, gender, time from trauma to OCT scan,  (Table 1). Multivariate logistic regression found that presence of intraretinal cysts was the only independent factor predicting spontaneous closure of macular hole (p = 0.005). The area under the receiver operating characteristic curve was 0.832 (95% confidence interval 0.668-0.997) and 0.812 (95% confidence interval 0.649-0.974) respectively, for the presence of intraretinal cysts and minimum diameter of macular hole (Fig. 3).

Discussion
In this study, we found that 37.0% of the traumatic macular holes closed spontaneously over 6 months of follow up. Macular holes that closed spontaneously had an overall small minimum diameter and few intraretinal cysts. Multiple logistic regression showed that absence of intraretinal cysts was the only independent predictive factor for spontaneous closure of traumatic macular hole. Presence or absence of intraretinal cysts also had larger area under the ROC curve than minimum diameter of macular hole. Spontaneous closure of traumatic macular hole was associated with a better visual outcome compared to the macular holes that did not close.
Spontaneous closure of traumatic macular hole has been reported earlier in case reports [5][6][7] . There are only a few case series, which studied the natural history of traumatic macular hole.  of traumatic macular hole after 1-to 6-month follow-up 8 . The rate of spontaneous closure of traumatic macular hole in our cases (37.0%) is comparable to the previous reports.
Our study identified two factors associated with spontaneous closure of traumatic macular holes. The first factor is small minimum diameter of the macular hole at baseline. It has been suggested earlier that small macular holes may resolve without any intervention [9][10][11][12] . Small macular defect may allow easy migration of glial cells. The other associated factor was absence of intraretinal cysts. In the earlier published case reports and case series, absence of intraretinal cysts has not been reported as a factor for closure of macular hole. However, most instances where traumatic macular holes closed spontaneously were not accompanied by intraretinal cysts on OCT [11][12][13][14][15][16][17][18][19][20][21][22][23] , although there are also a few cases with spontaneous closure in the presence of intraretinal cysts [24][25][26] . The presence of intraretinal cysts may be an indicator of internal limiting membrane traction, which may activate Müller cells and cause accumulation of extracellular fluid in the retina. Besides these two factors, young age has been reported to be associated with spontaneous closure of macular hole 20 . However, there was also a report of a 55-year old patient with traumatic macular hole that closed spontaneously 8 . In this case-control study, we did not find significant difference in age between the two groups. Since most of the published information has been from case reports or case series, there is insufficient evidence to conclude that young age is associated with spontaneous closure of macular holes.
Since significant proportion of traumatic macular hole close spontaneously with improvement in vision, it has been recommended that traumatic macular hole should be followed up and observed for at least several months before surgery to avoid unnecessary risk of surgical complications 3 . On the other hand, it has been reported that for idiopathic macular holes early surgery can result in a better functional outcome 27 . In this study we did not compare the functional outcomes after spontaneously closure and early surgical intervention. Further studies are needed to answer this question.
We recognize the limitations of this study. Firstly, the time from trauma to OCT scan is not consistent among patients. Although, there was no statistically significant difference between both groups, it may still be a confounding factor because of the limited sample size. Secondly, this is a retrospective study and the follow-up schedule was not consistent. OCT scan was not performed on a regularl basis, therefore we closed not closed p could not investigate the changes in the macular hole size or number of cysts during follow-up. Further prospective studies with larger sample size and regular OCT scan are needed. In summary, more than one third of traumatic macular holes closed spontaneously in our study. Spontaneous closure of macular holes was associated with smaller minimum diameter of macular holes and absence of intraretinal cysts. Absence of intraretinal cyst is more valuable in predicting the outcome of traumatic macular hole.

Methods
This retrospective study was approved by the Institutional Review Board of Joint Shantou International Eye Center (JSIEC), Shantou University and the Chinese University of Hong Kong. Informed consent was waived because of the retrospective nature of this study. The medical records database of JSIEC from July 2008, when we started using spectral domain OCT, to September 2014 was searched. The records of the consecutive patients who had undergone spectral domain OCT examination after closed globe injury were reviewed. The inclusion criteria were: (1) history of blunt ocular trauma; (2) traumatic macular hole confirmed by spectral domain OCT; (3) no vitreoretinal surgery performed within 6 months after diagnosis of macular hole; (4) follow up for 6 months.
The following examinations were performed: best-corrected visual acuity (BCVA), non-contact tonometry, slit-lamp biomicroscopy of anterior segment and retina, spectral domain OCT. Visual acuity was measured with the Chinese standard logarithmic visual chart and converted to Logarithm of minimal angle of resolution (LogMAR) units. A BCVA of finger counting (FC) was converted to 2.0 LogMAR and hand motion (HM) was converted to 2.3 LogMAR as suggested in the literature 28 .
Spectral domain OCT examination was performed using the Topcon 3D OCT-1000 (Topcon Corporation, Tokyo, Japan). Macula was scanned with the 6 *6 mm protocol, which consisted of 128 B scans. The axial and transverse resolution was 6-μ m and 20-μ m respectively. The minimum and base diameters of macular hole were measured using the caliper function of Topcon 3D OCT-1000 software. The retinal thickness at 9 ETDRS region was measured automatically. Accompanied conditions including choroidal rupture, retinal detachment, retinal atrophy, intraretinal cyst, vitreomacular traction, and epiretinal membrane were also recorded. Intraretinal cyst was defined as minimally reflective space within the neurosensory retina 29 . These parameters were obtained at baseline when traumatic macular hole was diagnosed. At the follow-up visits, status of macular hole closure was determined using OCT.
The main outcome measures were closure of macular hole at the last follow-up. Single variant analysis, such as Fisher exact test, Mann-Whitney test was used to compare the baseline characters, including age, gender, BCVA and spectral domain OCT data between the groups of macular hole with and without spontaneous closure. The parameters with p < 0.1 were included in binary multivariate logistic regression. P value < 0.05 was considered as statistical significance. Receiver operating characteristic (ROC) curves were derived for variables indicating good model discrimination, assessed by the area under the curve: values close to 1.00 represent good discrimination, whereas values close to 0.500 indicate that the model does not discriminate to any greater extent than that achieved by random allocation. The SPSS software (version 16.0, SPSS Inc. Chicago, IL) was used for statistical analysis.