Commentary: Retinal changes after internal limiting membrane peeling.

Abstract

The internal limiting membrane (ILM) is the innermost of the 10 retinal layers. It is a transparent membrane measuring nearly 10 microns and is formed by the basement membrane of the Müller cells. It is peeled as a standard treatment of full-thickness macular hole (FTMH). However, many reports have shown that peeling the ILM may actually damage the retina. Wollensak et al.[1] showed that ILM removal reduces the mean strength of the central retina by almost half. Histological studies performed on human and primate eyes have shown that ILM peeling causes damage to the processes and end feet of the Müller cells. A selective delay has been noted in the B-wave recovery of focal macular electroretinogram in eyes that have undergone ILM peeling. Several anatomical changes have been described after ILM peeling.[2] Swelling of the arcuate nerve fiber layer (SANFL): It was first described by Clark et al.[3] as an early postoperative change. It is seen as hypoautofluorescent arcuate striae in the macular region on fundus autofluorescence (FAF) and infrared (IR) imaging; however, it cannot be detected on either biomicroscopic fundus examination or color photograph. It is seen as a hyperreflectant swelling on spectral domain optical coherence tomography (SD-OCT). It usually disappears gradually over a period of 2–3 months. It does not seem to affect the best-corrected visual acuity (BCVA) or retinal sensitivity. Dissociated optic nerve fiber layer (DONFL): It was first described by Tadayoni et al.[4] in 2001. It is also known as “concentric macular dark spots (CMDS)” or “inner retinal dimpling”.[2] It gives the macula a ‘‘moth-eaten’’ appearance due to the presence of small depressions on the retinal contour situated along the course of retinal nerve fibers on the en face scan. SD-OCT scans show that the depth of these dimples is limited to the RNFL only. Liu et al. found three distribution patterns, namely, papillomacular bundle-dominated, scattered, and temporal raphe-spared patters.[5] Ye et al.[6] showed that DONFL appeared within two months of surgery. The number progressed during the next four months and then slowed down. The number did not change after 12 months since the surgery. It has not been found to have association with either BCVA or retinal sensitivity. Steel et al.[7] showed that the use of a diamond-dusted membrane scraper to initiate the peel is a risk factor for the formation of DONFL. The extent of DONFL has been seen to be positively associated with the amount of cellular debris on the retinal side of the peeled ILM.[8] The perioperative optical coherence tomography (OCT) for ophthalmic surgery (PIONEER) study showed that the immediate increase in the inner retinal layer thickness on intraoperative OCT following ILM peeling was associated with the development of DONFL, while the technique used for peeling did not.[9] Foveal displacement: Several authors have reported a nasal shift of the fovea, that is, shortening of the foveopapillary distance (FPD). Ishida et al.[10] reported that the displacement was greater in the temporal retina compared to the nasal retina. They further reported that the degree of the retinal displacement significantly correlated with the maximum base diameter of the FTMH. However, Nair et al.[11] reported that a temporal foveal shift may be seen in around 25% of the eyes, especially those with a longer temporal segment. Ohta et al. reported that the steepness of the foveal slope increased after ILM peeling in an asymmetrical pattern as the nasal side became significantly steeper than the temporal one.[12] None of these findings seem to affect the BCVA or retinal sensitivity. Retinal thinning: Several studies have shown that the macular volume, especially that of the temporal macula, shows a progressive reduction.[2] It has been proposed that hemi-temporal ILM peeling is as effective as conventional 360-degree peeling for the treatment of FTMH. Additionally, it is associated with less retinal displacement, lesser DONFL formation, lower rates of subjective metamorphopsia, faster visual rehabilitation, and greater safety.[13–15] We congratulate the authors for their study showing that the outcomes of papillomacular bundle sparing ILM peeling is comparable to conventional ILM peeling in terms of hole closure rates and visual gain, with the advantage of less retinal damage for the treatment of idiopathic FTMH ≤400 microns.[16] Studies with larger sample size and larger macular hole size will help validate the surgical technique.