Patient and operative details
A 68-year-old lady with normal tension glaucoma was noticed to have decreased visual acuity of the right eye from 0.5 to 0.1 during routine follow-up. Fundal examination revealed a stage IV full-thickness macular hole in the right eye. Spectral domain optical coherence tomography (OCT RS-3000, Nidek Inc, Gamagori, Aichi, Japan) confirmed the full-thickness nature of the macular hole and the basal diameter of the macular hole was 708 micron (Fig. 1), with the typical cystoid macular oedema over its edges. There was also posterior vitreous detachment.
She was arranged to undergo combined phacoemulsification, intraocular lens implantation, pars plana vitrectomy with 25 gauge plus transconjunctival system (Constellation®, Alcon Inc, Fort Worth, Texas, USA), ILM peeling after staining with MembraneBlue-Dual® (Dutch Ophthalmics Research Centre, Zuidland, The Netherlands) and gas tamponade with 12 % perfluoropropane. The surgical technique was similar to what we described previously [1], except the ILM staining was performed under Balance Salt Solution. In view of the chronicity and large size of the macular hole, the ILM flap technique described by Michalewska et al. was also adopted [5], but no attempt was made to cover the macular hole with the ILM flap. The ILM surgically removed was sent for histological analysis and its presence was confirmed. Postoperatively the patient was asked to adopt a face down posture for 2 weeks. She was prescribed a combination regime of eyedrops of dexamethasone and chloramphenicol (Gutt Nadexin 6 times per day) for 3 weeks. The postoperative progress was monitored with serial visual acuity and spectral domain OCT monitoring.
Results
The postoperative period was uneventful and our patient was compliant to posturing. After absorption of the gas bubble, there was a small persistent macular hole measuring 200 μm in diameter on OCT (Fig. 2a). The visual acuity of the right eye improved slightly to 0.15. The option of reoperation with further gas tamponade was discussed but our patient declined any further surgical intervention. The patient continued to be follow-up in our clinic.
During subsequent follow-up OCT was used to monitor her progress. At postoperative 6 months, new intraretinal cystic changes were observed on the edges of the macular hole (Fig. 2b). At postoperative 16 months, there was complete resolution of cystoid macular oedema, and a bridging membrane from the temporal edge leading to closure of the macular hole (Fig. 2c). At postoperative 26 months, her right eye visual acuity was found to improve to 0.4. OCT also showed good delineation of the external limiting membrane and ellipsoid zone, with the bridging membrane from the temporal edge remained in situ (Fig. 2d).
Discussion
The inverted ILM flap provides both Muller cell fragments to stimulate glial cell proliferation and a basement membrane substrate which acts as a scaffold for tissue proliferation. In the original inverted ILM technique described, the macular hole was covered with the inverted ILM flap [5]. Subsequently, Kuriyama et al. and Shin MK et al. also reported encouraging results in two case series employing a similar technique covering the macular holes with inverted ILM flaps [6, 7].
Whether it is required to cover or tuck the macular hole with the ILM flap remains controversial. According to the OCT images in the original study by Michalewska et al., the ILM flap might have been tucked inside the macular hole but subsequently studies described only covering of the macular hole with the ILM flap was necessary [5–7]. Surgical manipulation to tuck the ILM flap into the hole is not recommended as this can potentially damage the retinal pigmentary epithelium at the base of the macular hole. In the present case, no extra surgical manipulation was used to cover the macular hole with the ILM flap, and it might have led to the initial failure of complete macular hole closure. Although the ILM flap was not detected during the early OCT scans (Fig. 2a, b), the bridging membrane from the temporal edge visualized on later postoperative OCT scans (Fig. 2c, d) might have represented the ILM flap which has turned itself over and covered the hole during the course of observation leading to complete hole closure.
Late closure of idiopathic macular hole closure following surgery is a rare occurrence, although there were isolated reports of late closure of surgically repaired myopic macular holes and traumatic macular holes [8, 9]. Gross et al. reported two cases of reopening and spontaneous closure of previously repaired macular holes with corresponding OCT findings [10]. On serial OCT scans, intraretinal cystic changes on the edges of the macular hole with gradual resolution and spontaneous macular hole closure were observed. Similar observation was also reported by Inoue et al. that included six cases with spontaneous closure of primary macular hole [11]. However for idiopathic macular hole, this is the first reported case of late closure following vitrectomy with the inverted ILM flap technique. We also observed similar morphological changes in the present case over a course of 16 months.
