From the Institute of Ophthalmology (AS, PS, AM, AB, SDD, EB), Catholic University of Rome; and the Ophthalmology Unit (ACT, FBP, SDD), San Filippo Neri Hospital, Rome, Italy.
The authors have no financial or proprietary interest in the materials presented herein.
Address correspondence to Alessandra C. Tiberti, MD, PhD, Ophthalmology Unit, San Filippo Neri Hospital, Via G. Martinotti, 20, 00165 Rome, Italy. E-mail: firstname.lastname@example.org
Idiopathic macular hole is a relatively frequent retinal disease characterized by a full-thickness anatomic defect in the fovea, leading to visual function decrease, metamorphopsia, and central scotoma. Idiopathic macular holes usually occur in elderly patients and are more frequent in women.
Although the natural history of idiopathic macular hole usually leads to progression through four clinical stages from an impending lesion to a full-thickness defect involving all retinal layers, spontaneous macular hole closure has been reported as an infrequent event.1 The exact mechanism of spontaneous macular hole closure is still unclear, but four different hypotheses have been proposed: complete vitreous detachment over the fovea releasing the anteroposterior tractional forces, formation of an epiretinal membrane resulting in hole shrinkage and closure, glial cell proliferation at the base of the hole, and growth of retinal tissue bridging over the hole.2
Since optical coherence tomography (OCT) was introduced into clinical practice, important insights into idiopathic macular hole development and progression have been provided. OCT currently represents an essential tool in defining idiopathic macular hole stage and size and in evaluating abnormalities in the vitreofoveal interface. Spectral-domain OCT (SD-OCT) is the newest generation OCT system that offers improved visualization of retinal architecture and morphology compared with time-domain OCT and is particularly sensitive for visualizing the inner and outer segments of the photoreceptors and the internal and external limiting membranes.
We present SD-OCT findings in three patients with stage II, III, and IV idiopathic macular hole showing spontaneous closure due to different hypothetical mechanisms.
A 64-year-old woman presented with a complaint of bilateral progressive vision decrease over the previous 6 months. The best-corrected initial visual acuity (BCVA) was 20/100 in her right eye and 20/70 in her left eye. Binocular stereoscopic slit-lamp ophthalmoscopy with Super Field lens (Volk Optical Inc., Mentor, OH) revealed a full-thickness macular hole in both eyes. SD-OCT (Cirrus HD-OCT; Carl Zeiss Meditec, Inc., Dublin, CA) showed a full-thickness stage IV macular hole in her right eye and a full-thickness stage II macular hole in her left eye (Fig. 1A). The basal and minimal diameters were 400 and 237 μm in the right eye and 392 and 338 μm in the left eye.
Figure 1. (A) Spectral-domain optical coherence tomography (SD-OCT) of case 1 at presentation showed that the separation of the vitreous from the macula in the left eye was not complete, with adhesions of the posterior hyaloid to the roof of the macular hole and a perifoveal posterior vitreous detachment. Cystoid spaces on the level of the inner nuclear layer were visible; the basal and minimal diameters were 392 and 338 μm, respectively. (B) SD-OCT of case 1 at 1 week after prone positioning showed the posterior hyaloid was completely detached from the retinal layers in the macula; retinal tissue bridging is evident beginning from the inner layers. (C) SD-OCT of case 1 at 1 year of follow-up showed the foveal architecture was completely restored.
Vitrectomy with internal limiting membrane peeling and fluid–air and air–gas exchange with 20% sulfur hexafluoride was then performed in the right eye. Strict prone positioning was recommended after surgery. At 1 week of follow-up, fundus examination showed a closed macular hole in the right eye and also unexpectedly in the left eye. SD-OCT confirmed the macular hole closure in both eyes (Fig. 1B). Final BCVA at last follow-up 1 year after surgery was 20/30 in the right eye and 20/20 in the left eye (Fig. 1C).
A 76-year-old man was referred for decreased vision and metamorphopsia in the right eye. BCVA in the right eye was 20/100. Fundus examination revealed a stage II macular hole associated with an epiretinal membrane and a posterior hyaloid detachment. SD-OCT disclosed a full-thickness retinal defect in the foveal area with perilesional cystoid spaces and a slight epiretinal membrane (Fig. 2A). The basal and minimal diameters were 250 and 237 μm, respectively.
Figure 2. (A) Spectral-domain optical coherence tomography (SD-OCT) of case 2 at presentation showed full-thickness retinal defect in the foveal area associated with perilesional cystoid spaces and a slight epiretinal membrane. The basal and minimal diameters were 250 and 237 μm, respectively. (B) SD-OCT of case 2 at 1 year after presentation showed a posterior hyaloid detachment in the macula and the slight epiretinal membrane attached to the neuroepithelial surface. The retinal defect was closed in the inner and outer layers; central cystoid spaces were present. (C) SD-OCT of case 2 at 1 year and 3 months after diagnosis revealed a normal foveal configuration and an epiretinal membrane without surface wrinkling; minor irregularities in the line of the photoreceptors were evident.
Surgical internal limiting membrane peeling was then proposed to the patient, who refused for personal reasons. Six months later, BCVA, fundus examination, and SD-OCT were substantially unchanged.
One year after the initial visit, the patient returned reporting improved vision in the right eye over the previous 2 months. BCVA was 20/40. The macular hole was no longer evident on fundus biomicroscopy. SD-OCT showed that the retinal defect had closed in the inner and outer layers but a central cystoid space was still evident. An epiretinal membrane could also be well visualized at this stage (Fig. 2B). Three months later, BCVA remained stable at 20/40 and SD-OCT revealed a normal foveal configuration and a persisting epiretinal membrane (Fig. 2C).
A 67-year-old woman was referred for decreased vision and metamorphopsia in the right eye beginning 4 months previously. BCVA in the right eye was 20/50. Fundus examination and SD-OCT revealed a stage III full-thickness macular hole associated with an epiretinal membrane (Fig. 3A) with the posterior hyaloid still attached to the optic disc. The basal and minimal diameters were 690 and 203 μm, respectively. Surgery was proposed to the patient, who asked to delay vitrectomy for a few months.
Figure 3. (A) Spectral-domain optical coherence tomography (SD-OCT) of case 3 at presentation showed a full-thickness retinal defect in the foveal area with perilesional cystoid spaces and an epiretinal membrane are evident. The basal and minimal diameters were 690 and 203 μm, respectively. (B) SD-OCT of case 3 at 5 months after presentation revealed a closure of the macular hole beginning from the outer layers, reproducing the shape of a lamellar macular hole. (C) SD-OCT of case 3 at 6 months after presentation showed that the foveal architecture was completely restored and the epiretinal membrane was still evident.
On preoperative assessment 5 months later, the patient noticed a spontaneous increase of vision. BCVA was 20/20 and fundus examination showed a normal aspect of the fovea. SD-OCT revealed a closure of the macular hole beginning from the outer layers, reproducing the shape of a lamellar macular hole (Fig. 3B). One month later, the foveal architecture on SD-OCT was completely restored and the epiretinal membrane was still evident (Fig. 3C).
The pathogenetic mechanisms of macular hole are still unclear and hypotheses involving degenerative and tractional factors have been formulated. According to tractional theory, the development of macular hole is a consequence of either anteroposterior forces exerted by vitreous on the fovea or tangential traction due to cellular proliferation along the vitreomacular interface.3 If a spontaneous complete separation of the vitreous occurs in the early stages of macular hole formation, before the development of tangential forces, release of anteroposterior foveal traction may allow macular hole repair via glial cell proliferation through the gap.
This mechanism of early anteroposterior traction release followed by glial cell proliferation to bridge the retinal dehiscence may have occurred in the first case we described; spontaneous closure of the macular hole indeed occurred after separation of the vitreous from the macula in the early stages of the disease. SD-OCT showed the posterior hyaloid detachment and the following complete restoration of retinal architecture starting from the inner retina and then progressing into the outer layers. In this case, the complete posterior hyaloid detachment occurred during prone positioning following surgery to repair the macular hole in the fellow eye. Therefore, gravitational forces are probably involved in causing the complete posterior vitreous detachment. Michalewska et al. described a similar event after surgery in the fellow eye and prone positioning.4
A different mechanism seems to be implicated in the spontaneous idiopathic macular hole closure in cases 2 and 3. SD-OCT in both cases revealed the presence of an epiretinal membrane adherent to the retinal surface before and after the idiopathic macular hole closure. Anteroposterior traction was not evident and therefore its release did not seem to be implicated in the spontaneous closure of the hole. Smiddy reported a similar case of spontaneous closure of a stage IV idiopathic macular hole in a patient with preexisting posterior vitreous detachment and epiretinal membrane, and hypothesized a role for epiretinal membrane contraction in narrowing the gap and allowing the glial tissue bridging.5 In the cases we described, a similar progression of events may have occurred.
Previous time-domain OCT studies have documented several cases of spontaneous macular hole closure; however, SD-OCT enables more exact visualization of the retinal layers, especially photoreceptor layer and external limiting membrane, leading to additional details on the closure mechanisms and on possible anatomic indicators of a good visual recovery. SD-OCT allowed a good visualization of the posterior hyaloid detachment in case 1 (Fig. 1), leading to the macular hole closure; in cases 2 (Fig. 2) and 3 (Fig. 3), the high-resolution images obtained with SD-OCT enabled a well-defined visualization of bridging retinal tissue across the hole (case 2) and glial cell proliferation at the base of the hole (case 3). Furthermore, thanks to better axial resolution, SD-OCT helps in the detection of possible external limiting membrane and photoreceptor layer defect, justifying a poor functional outcome. Cases 1 and 3 showed a good recovery of the photoreceptor layer and a smooth and continuous external limiting membrane, and the final visual acuity was considerably high. However, SD-OCT imaging in case 2 displayed multiple interruptions of the external limiting membrane and a poorly arranged photoreceptor layer even after the hole closure. This can explain the lack of a good functional recovery.
The analysis of spontaneous macular hole, feasible with SD-OCT, may have therapeutic implications. In early stages, with evident anteroposterior tractional forces and the absence of epiretinal membrane proliferation, a surgical approach could consider only the removal of the posterior hyaloid and internal limiting membrane peeling may be unnecessary. In more advanced stages, the assumption of centrifugal tangential forces requires their release by surgical removal of the internal limiting membrane.
Furthermore, the common condition in our series was the small diamter of the macular hole (range: 203 to 338 μm). Reviewing previously reported cases of spontaneous closure, a small diameter of idiopathic macular hole, regardless of stage, seems to be the pivotal factor to determine the spontaneous closure because a small opening could easily be restored by glial cell bridging, especially after posterior hyaloid traction release.
Imaging technologies can be effective in identifying mechanisms of idiopathic macular hole spontaneous closure. Idiopathic macular hole closure seems to depend on the presence of a proximate surface and a small hole diameter.
- Ezra E, Gregor ZJMoorfields Macular Hole Study Group Report No. 1. Surgery for idiopathic full-thickness macular hole: two-year results of a randomized clinical trial comparing natural history, vitrectomy, and vitrectomy plus autologous serum: Moorfields Macular Hole Study Group Report no. 1. Arch Ophthalmol. 2004;122:224–236. doi:10.1001/archopht.122.2.224 [CrossRef]
- Takahashi H, Kishi S. Optical coherence tomography images of spontaneous macular hole closure. Am J Ophthalmol. 1999;128:519–520. doi:10.1016/S0002-9394(99)00173-7 [CrossRef]
- Johnson MW, Van Newkirk MR, Meyer KA. Perifoveal vitreous detachment is the primary pathogenic event in idiopathic macular hole formation. Arch Ophthalmol. 2001;119:215–222.
- Michalewska Z, Cisiecki S, Sikorski B, et al. Spontaneous closure of stage III and IV idiopathic full-thickness macular holes: a two-case report. Graefes Arch Clin Exp Ophthalmol. 2008;246:99–104. doi:10.1007/s00417-007-0647-9 [CrossRef]
- Smiddy WE. Spontaneous macular hole closure with appearance of epiretinal membrane: implications for therapy. Ophthalmic Surg Lasers Imaging. 2008;39:237–238. doi:10.3928/15428877-20080501-18 [CrossRef]