LAUSR

The mechanism of closure of a macular hole (MH) is not yet fully understood. Spontaneous closure in nontraumatic cases is possible, but quite rare.1 Alternatively, vitrectomy and gas tamponade, with or without internal limiting membrane (ILM) peeling, achieve anatomic success in 80% to 95% of cases. During recent years, different techniques have been attempted to improve the closure rate of MH and bring it closer to 100%. The addition of autologous serum does not appear to alter the prognosis of surgery,1 while the addition of platelet-rich plasma seems to work better,2 even in challenging cases, such as myopic MH. More recently, the ILM flap technique was introduced for cases of MH that pose a high risk of persistence, such as myopic, chronic, or large holes. Initially, many surgeons thought that it would be necessary to put the ILM inside the hole to create a kind of plug. Subsequently, the pioneers of this technique stated that it would be sufficient to cover the hole with a single ILM layer, and they proposed performing the peeling only on the temporal side.3 An MH can be considered a small laceration in the weakest point of the retina, the fovea. The formation of a blood clot is the first step in the repair of a skin wound, which creates a closed environment, within which the repair mechanisms can act. Similarly, there is some evidence that anything that covers the MH and can separate the vitreous cavity from the intraretinal and subretinal space could facilitate its closure. This might explain why a single layer of ILM covering a myopic MH makes such a significant difference in the rate of closure of myopic MH.4 Moreover, the ILM flap was found to be displaced or not present in all MHs that failed to close. For the same reason, platelet-rich plasma is likely to work better than autologous serum that does not stick at the edges of the MH. We could even speculate that the main role of a tamponade might be the formation of a cover over the MH that can promote the maintenance of the homeostatic conditions necessary for the healing process. Silicone oil has been found to be an effective tamponade for the treatment of MH. This is surprising, as it is a lasting, but weaker, tamponade compared with gas. A possible explanation for its positive effect in promoting MH closure is that it creates a kind of bursa inside the edge of the MH and, therefore, compartmentalizes chemotactic and nutrient substances that may support its closure. The ILM flap also acts as substrate for a secondary intention wound-healing mechanism. When the hole is large, the edges are stiff, or the retina is not elastic enough, the ILM plug can support the closure without complete reapposition of the edges. In some cases, this step is followed by further anatomical and functional improvements during follow-up: the plug is slowly displaced toward the inner retina while the external limiting membrane (ELM) and the ellipsoid zone start reappearing between the retinal pigment epithelium and the hyperreflective ILM plug.4 This appears to be a regenerative process that was thought to be impossible in the retina. There are different hypotheses for this phenomenon. In large long-lasting MHs, the ELM is not detectable on the edges of the hole. The ELM is formed by the junctional complexes of Müller cells on the inner segment of the photoreceptors. Nevertheless, even if the ELM is not present, the nucleus of both cells involved in ELM formation—the photoreceptors and Müller cells— may still be alive and have the potential to regenerate their distal part when under optimal homeostatic conditions. This offers a potential explanation for the regeneration of the outer retinal layers that tend to migrate into their original position. At this point, the ILM plug seems to behave as the scar tissue of a skin wound: it is displaced on the surface to enable the complete restoration of the underlying layers.4 This mechanism suggests the importance of the preservation of an intact retinal pigment epithelium layer underneath the hole. In addition to this mechanism, a subset of Müller cells in the mature retina was found to possess neural progenitor properties.5 These cells might be activated during the healing process and act as tissue stem cells to promote the restoration of the outer retinal layers. However, the potential of Müller cells as a source of progenitor cells appears to be age related. The activation of these cells in response to retinal injuries becomes progressively confined to the retinal periphery.5 Thus, it is not clear whether this mechanism can play a role in the repair of an MH in human adults. The function of the ILM plug was also considered to be related to the proliferation of fibroblasts. If this was true, we would expect an increase in hyperreflective material on optical coherence tomographic images during the first postoperative period, as seen for epiretinal membranes. However, the contrary was observed: the ILM plug tends to contract over time and can even disappear in some cases. More recently, neurosensory, retinal free flap transplantation was proposed when no ILM was available to cover the hole.6 This can offer an excellent solution in selected cases. The authors in this particular study6 speculated that the retinal free flap may retain some function, as best-corrected visual acuity and retinal sensitivity improved after surgery. It seems highly unlikely that a retinal free flap could conduct a signal to the optic nerve, as the retinal circulation is interrupted. Previous studies have demonstrated that a central artery occlusion lasting longer than 4 hours causes profound and VIEWPOINT