Full thickness macular holes (MHs) are defects in the macula that involve its entire thickness. Macular hole surgeries have undergone several modifications over the years to facilitate hole closure and improve visual outcomes. Typical surgical management includes pars plana vitrectomy, gas tamponade, and internal limiting membrane (ILM) peeling [1]. Due to lower closure rates with ILM peeling, Michalewska et al. introduced the ILM flap procedure in which instead of peeling the entire ILM, a portion of the ILM is inverted to cover the hole and to induce gliosis [2]. Others have also reported that ILM flap technique was superior to ILM peeling in successful hole closure [3],[4]. Here, we present a case of MH reopening after an initial successful closure with an inverted ILM flap procedure.

A 72-year-old Hispanic male presented with decreased vision in the right eye, and on fundus exam and optical coherence tomography (OCT) imaging, was found to have a primary, full-thickness macular hole measuring 431 µm (Figure 1A). Initial uncorrected visual acuity in the right eye was 20/125, without improvement with pinhole. The patient underwent a 25-gauge pars plana vitrectomy. During posterior hyaloid detachment, pinpoint bleeding arose at a vitreous tuft, which was treated with endolaser to prevent retinal detachment in case of potential micro-breaks. The ILM was stained with indocyanine green, peeled around the hole. The ILM flap was trimmed, and inadvertently separated from the edge of the hole except for the inferotemporal edge. Fluid–air exchange was performed and the created flap was inverted to cover the hole (Figure 2A). Then gas–air exchange was performed with 14% C3F8. The patient postured face-down for one week. At week one, when the vitreous cavity was 60% gas-filled, the macular hole appeared closed. At postoperative week six, when there was only a residual gas bubble in the vitreous cavity, the macular hole appeared to reopen, with the folded ILM flap lying just below the fovea (Figure 1B and Figure 2). Then the patient’s visual acuity was 20/200. The patient underwent cataract surgery for brunescent cataract three months after vitrectomy. Six months after the first vitrectomy, the patient underwent a second vitrectomy in which the ILM flap was removed, gas–air exchange was performed with 14% C3F8, and the patient was again postured face-down for one week. The hole failed to close (Figure 1C); therefore, two months later, a third vitrectomy was performed, in which the ILM was peeled further to include the entire macula and, through subretinal injection of balanced salt solution, three subretinal blebs were created at the superotemporal, temporal, and inferotemporal macula using a 25/38-gauge subretinal needle (Figure 3). Face-down posturing was again done for one week. Postoperatively, the macular hole was shown to have successfully closed, and the best corrected visual acuity was at 20/80 one year after the last vitrectomy (Figure 1D).

Postoperative MH reopening have previously been reported, with instances occurring between 2 and 28 months [5],[6],[7]. The mechanisms behind MH reopening include, but are not limited to, iatrogenic retinal injury, impaired healing, and retinal elevation secondary to cystoid macular edema and epiretinal membrane growth and resulting tractional forces [5],[6],[8]. Here, we present another potential mechanism for MH reopening.

Since Michalewski et al.’s inverted ILM flap technique, inverted ILM flaps have been shown to further reduce MH recurrence [2],[4]. In Hu et al.’s 2019 study, myopic MH closure rate significantly improved from 66.7% to 100% following inverted ILM flap versus ILM peeling [4]. In our case, an inverted ILM flap may have contributed to the reopening of the hole. Intraoperatively, the ILM flap was separated from the macular hole edges except for the inferotemporal edge. Despite inverting and folding the ILM flap such that it covered the hole’s entire surface at the end of the surgery following air–fluid exchange, on the postoperative OCT, following MH reopening, and during the second pars plana vitrectomy, the ILM flap was observed to be folded over the retinal surface at and below the inferior edge of the hole. It is possible the ILM slipped downward during fluid–air exchange or in the early postoperative period; however, in our case, we believe that focal inferotemporal attachment of the flap may have resulted in the reopening of the hole. Considering the hole was 431 µm large, it would likely have closed without an inverted ILM flap. We hypothesize that as the fluid–gas interface rose and reached the lower MH edge, the surface tension of the fluid may have pulled the flap down and reopened the hole (Figure 4). This mode of ILM flap reopening, to the best of our knowledge, has not been observed previously.

This case demonstrates that inverted ILM flap may not be an entirely harmless procedure, and we suggest it be reserved for large holes. Additionally, if an inverted ILM flap procedure is pursued, we suggest creating the flap in such a way that its attachment is preserved at all edges of the macular hole or at least at the superior edge; this way, as the fluid–gas level rises, surface tension equally pulls all the edges or in the case of the attachment to the superior edge, that edge is pulled toward the center of the hole.

Given the high success rates of reoperation on previously reopened MH cases (as high as 80% per Reid et al.), we decided to proceed with further surgical correction [7]. After the macular hole reopened, the next pars plana vitrectomy failed due to retinal adherence to the retinal pigment epithelium (RPE), thereby preventing proper contraction and subsequent closure of the macular hole post-ILM peel. For the third (and successful) vitrectomy, the attempted subretinal injections were done to try to better facilitate hole closure after prior ILM peel failures, as per the literature [9] (Figure 3).

In conclusion, our case highlights that case selection and proper surgical technique are important in using inverted ILM flap for macular hole repair. Separation of the ILM flap from superior edges of the macular hole and attachment only to the inferior edge of the hole may result in reopening of the macular hole.