EFFECT OF INTERNAL LIMITING MEMBRANE PEELING ON THE DEVELOPMENT OF EPIRETINAL MEMBRANE AFTER PARS PLANA VITRECTOMY FOR PRIMARY RHEGMATOGENOUS RETINAL DETACHMENT KI YUP NAM, MD,*† JUNG YEUL KIM, MD* Purpose: To investigate the difference in the occurrence of postoperative epiretinal membranes (ERMs) in vitrectomy for rhegmatogenous retinal detachment with and without peeling of the internal limiting membrane (ILM). Methods: The medical records of the 135 patients, who underwent vitrectomy for primary rhegmatogenous retinal detachment from November 2007 to August 2011, were analyzed retrospectively. Of the subjects, 70 patients underwent ILM peeling during the surgery and 65 did not. The best-corrected visual acuity, fundus photograph, and optical coherence tomography were collected 3, 6, and 12 months postoperatively. The relationship between ILM peeling and the preoperative findings of rhegmatogenous retinal detachment and development of a postoperative ERM was analyzed. Results: No ERM occurred in the ILM peeling group, whereas an ERM occurred in 14 of 65 patients who underwent vitrectomy without ILM peeling (21.5%). This difference was significant (P , 0.001). The occurrence of a postoperative ERM was not significantly correlated with other preoperative factors. In the macular-on group, the overall mean bestcorrected visual acuity was better in the ILM peeling group and was significantly higher 12 months postoperatively (P = 0.03). Conclusion: Internal limiting membrane peeling seems to prevent the occurrence of a postoperative ERM in patients with primary rhegmatogenous retinal detachment. RETINA 35:880–885, 2015
A
and we retrospectively analyzed the correlation between ILM peeling and the occurrence of a postoperative ERM in patients who underwent vitrectomy for rhegmatogenous RD. An ERM is semitransparent membranous tissue that forms at the vitreoretinal interface of the macula and can causes abnormalities of macular function. Roth and Foos1 postulated that an idiopathic ERM proliferates as retinal tissue-derived glial cells escape from microdefects in the ILM that occur during posterior vitreous detachment and migrate to the surface of the retina. Another theory attributes the pathogenesis of an ERM to the growth and fibrous metaplasia of the
n epiretinal membrane (ERM) is one of the most common complications of rhegmatogenous retinal detachment (RD) after a vitrectomy. In an ERM that occurs after rhegmatogenous RD, unlike the pathogenesis of an idiopathic ERM, the retinal pigment epithelial cells migrate to the vitreous cavity through the retinal break and settle on the macular surface, forming the membrane. Therefore, we postulated that removing the macular internal limiting membrane (ILM) during a vitrectomy for rhegmatogenous RD should help to remove any remaining posterior vitreous cells and other precursors of the ERM present on the retinal surface, such as the retinal pigment epithelial cells,
880
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
vitreous cells that remain on the retina surface after posterior vitreous detachment. The ERM that occurs after retinal reattachment surgery for treating rhegmatogenous RD is known as macular pucker. Various factors cause visual loss after the surgical treatment of rhegmatogenous RD, and an ERM is one of the most common causes of visual loss.2,3 The reported prevalence of ERM is 3% to 8.5% after scleral buckling and 6.1% to 12.8% after vitrectomy.4–9 The pathogenesis of ERM that occurs after rhegmatogenous RD, unlike the idiopathic ERM, that the retinal pigment epithelial cells migrate to the vitreous cavity through the retinal break and settle on the macular surface, forming the membrane.10 Also, the ILM, which corresponds to the basement membrane of the Müller cell, is located at the innermost layer of the retina and can be play a role of scaffold on various cells that develops ERM. Therefore, we postulated that removing the macular ILM during a vitrectomy for rhegmatogenous RD could help to remove any remaining posterior vitreous cells and other precursors of the ERM present on the retinal surface, such as the retinal pigment epithelial cells. This would prevent the occurrence of a postoperative ERM, as the ILM acts as a scaffold for cell growth, as the basement membrane of the Müller cells. Accordingly, we investigated the correlation between ILM peeling and the occurrence of a postoperative ERM in patients who underwent vitrectomy for rhegmatogenous RD. Methods We analyzed retrospectively the medical records of 135 patients who underwent vitrectomy for primary rhegmatogenous RD from November 2007 to August 2011. Institutional Review Board (IRB)/Ethics Committee approval was obtained, and the study adhered to the tenets of the Declaration of Helsinki. Patients were From the *Department of Ophthalmology, Chungnam National University College of Medicine, Daejeon, Republic of Korea; and †Department of Ophthalmology, Kosin University Gospel Hospital, Busan, Republic of Korea. None of the authors have any financial/conflicting interests to disclose. Involved in design of study (K.Y.N., J.Y.K.); conduct of study (K.Y.N., J.Y.K.); collection of data (K.Y.N.); analysis and interpretation of data (K.Y.N., J.Y.K.); and critical review of article (J.Y.K.). The protocol was approved by the Institutional Review Board of Chungnam National University Hospital, and the study adhered to the tenets of the Declaration of Helsinki. Operation was performed after obtaining the approval of patients for the procedure with informed consent. Reprint requests: Jung Yeul Kim, MD, Department of Ophthalmology, Chungnam National University Hospital, 640 Daesa-dong, Jung-gu, Daejeon 301-721, South Korea; e-mail: [email protected]
881
excluded if they had a previous ERM, history of trauma or uveitis, proliferative vitreoretinopathy, silicone oil filling after the vitrectomy, and follow-up for ,12 months. We evaluated ERMs during the initial RD diagnosis with optical coherence tomography (OCT; Stratus OCT3; Carl Zeiss Meditec, Dublin, CA) in cases of macular-on RD or shallow macular-off RD. If OCT examination was unavailable, the operator investigated ERM before and after perfluorocarbon liquid (PFCL) injection during surgery under a surgical microscope. If an ERM was found, the operator removed the ERM and peeled the ILM, and these cases were excluded. The surgery was conducted by a single surgeon (J.Y.K.). The operator experienced some severe cases of postoperative ERM after primary vitrectomy for rhegmatogenous RD in the early phase. Therefore, the operator peeled the ILM routinely for about 2 years, until July 2009. Internal limiting membrane peeling was not continued thereafter, and its value for the prevention of ERM development is therefore unclear. A 20-gauge vitrectomy was performed under general or retrobulbar anesthesia. After the conjunctival incision, scleral incisions were made 4.0 mm and 3.5 mm from the corneoscleral limbus in phakic and psudophakic/aphakic eyes, respectively, in the inferotemporal, superonasal, and superotemporal sclera. The surgeon performed combined operation (pars plana vitrectomy + phacoemulsification). Surgery was performed using a vitrectomy machine (ACCURUS Surgical System; Alcon Laboratories, Fort Worth, TX). A wide-field visualization system (Mini Quad XL; Volk Optical, Mentor, OH) was used. If there was no posterior vitreous detachment, as verified with intravitreal triamcinolone injection, posterior vitreous detachment was induced using a vitreous cutter, intraocular forceps, and pick. To remove the ILM, it was stained with indocyanine green (Diagnogreen; Daiichi Sankyo, Tokyo, Japan) and then removed using forceps. In cases with macular detachment, the ILM was stained and the area near the vascular arcade filled with perfluorocarbon to flatten the macula before ILM peeling. After fluid-gas exchange, endolaser retinopexy was performed, and perfluoropropane (C3F8) or sulfur hexafluoride (SF6) was injected. The surgery was finished after suturing the sclera and conjunctiva. In all cases, cryopexy was not performed. The subjects were divided into two groups according to ILM peeling. Their medical records were reviewed for age when diagnosed with rhegmatogenous RD, gender, best-corrected visual acuity (BCVA), macular involvement, the location, number, and shape of the break(s), RD extent, and gas type.
882
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 5
The visual acuity was recorded 3 months, 6 months, and 12 months postoperatively, and the occurrence of an ERM was confirmed using OCT. The break location was classified as upper, lower, and both upper and lower sites. The breaks were classified into single and multiple breaks. The extent of RD was classified as #1/4, 1/4 to 1/2, 1/2 to 3/4, and 3/4 to entire. An ERM was defined as a case in which the macula was covered by a highly reflective band on OCT. In patients in whom a postoperative ERM developed, it was removed surgically if the patients had a BCVA , 20/40 or complained of metamorphopsia. To analyze the results, the Snellen visual acuity was converted into the logMAR visual acuity. The statistical analysis was conducted using SPSS (version 17.0). The chi-square test was used to compare the occurrence of an ERM between the groups with and without ILM peeling. An independent t-test was used to compare the visual acuities. Patients who underwent surgery to remove an ERM were followed in the same manner as if they did not undergo the surgery, and the visual acuity at each follow-up after ERM removal was defined as the BCVA measured just before the ERM removal surgery. The non-ILM peeling group was subdivided into groups with and without a postoperative ERM, and the differences in various factors were compared between the two groups; the independent t-test was used to analyze age and BCVA at diagnosis and the chi-square test to analyze macular involvement, break location, break number, extent of RD, and gas type.
Results The study enrolled 135 eyes of 135 patients: 70 who underwent ILM peeling and 65 patients who did not. At the time of diagnosis of rhegmatogenous RD, the 2 groups did not differ in terms of age, BCVA, break shape, number, or location, and RD extent (Table 1). One hundred and two patients had phakic eyes, and the preoperative lens grade by nucleus sclerosis was similar in the 2 groups (P = 0.491). Cataract removal was performed in all except one case. At the 12-month follow-up, no ERM was observed in the patients who had ILM peeling during vitrectomy, whereas an ERM was observed in 14 of 65 patients who did not undergo ILM peeling during the vitrectomy (21.5%). This difference was significant (P , 0.001). For the patients who had surgery to remove an ERM, the BCVA measured immediately before surgery was used as the visual acuity at each follow-up
Table 1. Baseline Characteristics in ILM Peeling Group and Non-ILM Peeling Group
Age (year) Sex Male Female Macula on/off On Off Break position Upper Lower Both Break number Single Multiple Break shape Horse shoe tear Atrophic hole Others BCVA (logMAR) at RD diagnosis Extent of RD ,1/4 1/4–1/2 .1/2 Gas C3F8 SF6
ILM Peeling (n = 70)
Non-ILM Peeling (n = 65)
48.2 (±17.8)
47.9 (±19.0)
42 (60.0%) 28 (40.0%)
34 (52.3%) 31 (47.7%)
41 (58.6%) 29 (41.4%)
35 (53.8%) 30 (46.2%)
40 (57.1%) 21 (30.0%) 9 (12.9%)
39 (60.0%) 23 (35.4%) 3 (4.6%)
47 (67.1%) 23 (32.9%)
37 (56.9%) 28 (43.1%)
37 (52.9%)
36 (55.4%)
P 0.919* 0.390† 0.606† 0.232†
0.287† 0.338† 23 (32.9%) 25 (38.5%) 10 (14.2%) 4 (6.1%) 0.76 (±0.84) 0.85 (±0.87)
0.547* 0.199†
25 (35.7%) 30 (42.9%) 15 (21.4%)
16 (24.6%) 35 (53.8%) 14 (21.6%)
60 (85.8%) 10 (14.2%)
58 (89.2%) 7 (10.8%)
0.610†
*Independent t-test. †Chi-square test.
after ERM removal for the analysis. No significant difference in the mean visual acuity was found between the groups with and without ILM peeling. However, when the macula-on group was analyzed, the mean visual acuity was better in the ILM peeling group than in the non-ILM peeling group, and a significant difference was found at the 12-month follow-up (P = 0.03) (Figures 1 and 2). For the 65 patients who did not undergo ILM peeling, the differences in various factors were analyzed between the patients with and without a postoperative ERM. No significant difference in the age, BCVA, macular involvement, break location, number, or shape, RD extent, and gas type was found between the 2 groups (Table 2). Of the 14 patients who developed an ERM, it was diagnosed within 3 months in 12 patients (85.7%) and between 3 and 6 months in 2 (14.3%). It took a mean of 9.2 (range, 2.1–25.7) weeks to confirm the ERM after RD. Of the 14 patients who developed an ERM, 10 (71.4%) required surgery to remove it. There was
883
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
Table 2. Factor Analysis Between ERM Group and NonERM Group in Non-ILM Peeling Patients ERM Group (n = 14)
Fig. 1. Visual acuity changes in all patients (135 patients) after primary vitrectomy for rhegmatogenous RD. Error bars represent mean (SD).
a mean of 22 (range, 5.6–51.7) weeks between RD and removal of the ERM. There was one case of recurred RD in each group. There were no other obvious complications of vitrectomy and ILM peeling in this study. Discussion An ERM is one of the most common complications of rhegmatogenous RD after a vitrectomy, and ILM may play a role of scaffold on various cells that develops ERM. The ILM is composed of the basement membrane of Muller cells, proteoglycans, and type IV collagen and is an important structure in the formation of the vitreoretinal interface.11 Astrocytes and myofibroblasts migrate to the surface of the ILM and invade the deep layer, thereby causing an ERM.12 The ERM
Fig. 2. Visual acuity changes in macular-on patients (76 patients) after primary vitrectomy for rhegmatogenous RD. Error bars represent mean (SD) (median visual acuity at month 3, 6, and 12 was 0.22, 0.09, and 0, respectively, in ILM peeling group and 0.22, 0.09, and 0.09, respectively, in non-ILM peeling group). *P = 0.027 by independent t-test.
Non-ERM Group (n = 51)
P
Age, years 51.6 (±20.3) 46.9 (±18.7) 0.420* Sex 0.375† Male 9 (64.3%) 25 (49.0%) Female 5 (35.7%) 26 (51.0%) Macula on-off 1.000† On 8 (57.1%) 27 (52.9%) Off 6 (42.9%) 24 (46.2%) Break position 0.765† Upper 9 (64.3%) 30 (58.8%) Lower 4 (28.6%) 19 (37.3%) Both 1 (7.1%) 2 (3.9%) Break number 0.561† Single 7 (50.0%) 30 (58.8%) Multiple 7 (50.0%) 21 (41.2%) Break shape 0.066† Atrophic hole 4 (28.6%) 32 (62.7%) Horse shoe tear 8 (57.1%) 17 (33.3%) Others 2 (14.3%) 2 (4.0%) BCVA (logMAR) 0.93 (±0.97) 0.83 (±0.85) 0.547* at RD diagnosis Extent of RD 0.199† ,1/4 2 (14.3%) 14 (27.5%) 1/4–1/2 7 (50.0%) 28 (54.9%) .1/2 5 (35.7%) 9 (17.5%) Gas 1.000† C3F8 13 (92.9%) 45 (88.2%) SF6 1 (7.1%) 6 (11.8%) *Independent t-test. †Chi-square test.
that occurs after retinal reattachment surgery for treating rhegmatogenous RD is caused by retinal pigment epithelial cells that migrate to the surface of the posterior pole retina by diffusing into the vitreous cavity through the break or through fibrosis and the proliferation of precursor cells that remain on the retinal surface after the vitrectomy.10 Therefore, removal of the ILM can remove the ERM precursor cells from the retinal surface. In addition, in cases of severe bullous RD, a mild ERM might not be diagnosed in the fundus examination or with OCT. Therefore, an ERM that was not diagnosed before the surgery would also be removed by removal of the ILM. These provide a rationale for ILM peeling as a method of preventing the occurrence of a postoperative ERM in vitrectomy for rhegmatogenous RD. In our series, ERM developed in 14 of the 135 patients (10.4%). However, the occurrence rate was 21.5% in the patients who did not undergo ILM peeling, which was higher than the reported rate of 6.1% to 12.8%. This difference is likely attributable to the fact that we considered an ERM to be present if
884
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 5
a thin membrane was observed on the macula with OCT, even if it did not influence the visual acuity and was not obvious in the fundus examination.8–10 The authors investigated the effects of the postoperative ERM on visual acuity. The mean BCVA was compared between the groups with and without ILM peeling 3 months, 6 months, and 12 months postoperatively. In the patients who underwent surgery to remove an ERM that occurred early after RD surgery with a BCVA , 0.5, it was impossible to compare the visual acuity between the two groups because it recovered postoperatively. Therefore, they were assumed to have been followed as although they had not undergone surgery, and BCVA measured just before the ERM removal surgery was used for analysis. For example, when an ERM occurred within 3 months of the RD surgery and was then removed, the BCVA measured immediately before ERM removal was used as the visual acuity at 3 months, 6 months, and 12 months after RD surgery. The visual acuity data were replaced based on the assumption that visual acuity generally worsens over time after the formation of an ERM. As a result, the overall visual acuity was better in the group with ILM peeling than that without ILM peeling, although the difference was not significant. However, when the macula-on group was analyzed, the mean visual acuity was better in the ILM peeling group than in the non-ILM peeling group, and a significant difference was found at the 12-month follow-up (P = 0.03). This showed that the visual acuity was affected by the ERM that occurred after primary vitrectomy for rhegmatogenous RD, implying that removal of the ILM results in better visual acuity by preventing the occurrence of a postoperative ERM compared with patients who do not undergo ILM peeling. Preoperative cataract grade and cataract surgery might affect visual acuity. However, preoperative cataract grade was similar between the two groups, and cataract surgery was performed in nearly all phakic patients to shave the peripheral vitreous more easily and to avoid reoperation for postoperative cataract progression. We therefore hypothesize that postoperative cataracts and cataract surgery did not affect visual acuity. We analyzed the time of ERM development in this study. Epiretinal membrane occurred in 12 of the 14 patients (85.7%) within 3 months of RD surgery. Of the 14 patients who developed an ERM, it was removed in the 10 patients (71.4%) who had a BCVA , 0.5. This was similar to the result obtained by Martínez-Castillo et al,9 who observed ERM in 32.1% and 57.1% of patients at 1 month and 3 months, respectively, after vitrectomy for rhegmatogenous RD, and that removal of the ERM was required in 22 of the
28 patients (78.5%) who developed a postoperative ERM due to significant visual loss. Therefore, postoperative ERM tend to occur within 3 months of vitrectomy for rhegmatogenous RD. There were some reports about risk factors associated ERM development after rhegmatogenous RD surgery. A study reported a significant correlation between the postoperative ERM after scleral buckling for rhegmatogenous RD with preoperative macular detachment, low visual acuity, and proliferative vitreoretinopathy, and reoperation. However, MartínezCastillo et al reported no significant correlation in the occurrence of a postoperative ERM with macular detachment and visual acuity.2,4,7,9,13 In this study, to investigate the risk factors associated with postoperative ERM occurrence, we compared various factors between the ERM group and the non-ERM group in non-ILM peeling patients. There was no significant correlation between ERM development and age, BCVA at initial diagnosis, macular involvement, break location, number, shape, RD extent, or gas type in non-ILM peeling patients. However, the percentage of RD . 1/2 was higher in the ERM group than in the non-ERM group (35.7%:17.6%, P = 0.199) (Table 2). This may be because retinal pigment epithelium cells disperse more easily through a larger break or larger RD. Our results were not statistically significant; however, with a higher powered study that included more patients, this correlation between larger RD and ERM may become apparent. In a prospective study of patients who required silicone oil filling and underwent retinectomy during a vitrectomy due to accompanying proliferative vitreoretinopathy upon diagnosis of RD, Odrobina et al reported that retinal pucker occurred in 17.8% of the non-ILM peeling group (33 patients), but no retinal pucker occurred in the ILM peeling group (51 patients). Although there were a few cases with fibrous proliferation in the patients who underwent ILM peeling, its progression was limited to the margin of ILM peeling.11 However, our study is meaningful in that it enrolled patients with primary rhegmatogenous RD, who are expected to have a relatively good postoperative prognosis in the case of successful retinal adhesion. In this study, ILM peeling during vitrectomy for rhegmatogenous RD was shown to help to prevention of the formation of a postoperative ERM. No postoperative ERM was found in the group with ILM peeling, whereas an ERM occurred in 21.5% of the nonpeeling group. Although this result was obtained at the 12-month follow-up, it was considered clinically meaningful because most studies report that ERMs occur within 3 months after a vitrectomy for rhegmatogenous RD.
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
Recurrent RD after primary vitrectomy may affect postoperative visual acuity and ERM development. We only observed one case of recurrent RD in each group, and neither was associated with proliferative vitreoretinopathy or macular detachment. The safety of ILM peeling when there is no macular lesion is controversial. Complications related to ILM peeling include retinal hemorrhage, eccentric scotoma, retinal edema, iatrogenic punctuate chorioretinopathy, dissociated nerve fiber layer, vitreous hemorrhage, and subretinal hemorrhage. In addition, stain toxicity has been reported.14–21 However, we observed no definite clinical complication caused by ILM peeling. In addition, when the mean BCVA was compared between the groups with and without ILM peeling, no significant difference was found. However, within the macular-on group, the overall mean BCVA was better in the group with ILM peeling. Therefore, it seems that visual acuity-related complications caused by ILM peeling were insignificant. Several studies reported that complications related to ILM peeling were mainly associated with mechanical trauma to the retinal tissues, and the safe concentration of the dye is well known. Therefore, with improvement in the ILM manipulation technique and the proper staining technique, the rate of complications could be reduced. This study has some limitations. First of all, this study is retrospective, and the data were collected from limited medical records. Second, preoperative OCT examination was not available to all RD patients. And in cases of macular-off RD, ERM was investigated by operator during vitrectomy under surgical microscope; thus, very thin ERM might be missed. Finally, the number of cases might not be enough to analyze the factors related with ERM development in non-ILM peeling group. In conclusion, ILM peeling during vitrectomy for primary rhegmatogenous RD is useful for preventing postoperative ERM formation and provides a better visual prognosis. However, we have to consider whether ILM peeling for all patients is beneficial. This should be confirmed in a further long-term prospective study with a greater number of patients. Key words: rhegmatogenous retinal detachment, pars plana vitrectomy, internal limiting membrane peeling, epiretinal membrane. References 1. Roth AM, Foos RY. Surface wrinkling retinopathy in eyes enucleated at autopsy. Trans Am Acad Ophthalmol Otolaryngol 1971;75:1047–1058. 2. Lobes LA Jr, Burton TC. The incidence of macular pucker after retinal reattachment surgery. Am J Ophthalmol 1978; 85:72–77.
885
3. Wilkinson CP. Visual results following scleral buckling for retinal detachments sparing the macula. Retina 1981;1: 113–116. 4. Tanenbaum HL, Schepens CL, Elzeneiny I, et al. Macular pucker following retinal detachment surgery. Arch Ophthalmol 1970;83:286–293. 5. De Bustros S, Thompson JT, Michels RG, et al. Vitrectomy for idiopathic epiretinal membranes causing macular pucker. Br J Ophthalmol 1988;72:692–695. 6. De Bustros S, Rice TA, Michels RG, et al. Vitrectomy for macular pucker. Use after treatment of retinal tears or retinal detachment. Arch Ophthalmol 1988;106:758–760. 7. Uemura A, Ideta H, Nagasaki H, et al. Macular pucker after retinal detachment surgery. Ophthalmic Surg 1992;23: 116–119. 8. Heo MS, Kim HW, Lee JE, et al. The clinical features of macular pucker formation after pars plana vitrectomy for primary rhegmatogenous retinal detachment repair. Korean J Ophthalmol 2012;26:355–361. 9. Martínez-Castillo V, Boixadera A, Distéfano L, et al. Epiretinal membrane after pars plana vitrectomy for primary pseudophakic or aphakic rhegmatogenous retinal detachment: incidence and outcomes. Retina 2012;32:1350–1355. 10. Katira RC, Zamani M, Berinstein DM, et al. Incidence and characteristics of macular pucker formation after primary retinal detachment repair by pars plana vitrectomy alone. Retina 2008;28:744–748. 11. Odrobina D, Bednarski M, Cisiecki S, et al. Internal limiting membrane peeling as prophylaxis of macular pucker formation in eyes undergoing retinectomy for severe proliferative vitreoretinopathy. Retina 2012;32:226–231. 12. Moris R, Kuhn F. Surgical treatment of macular surface disorders. World Atlas Series of Ophthalmic Surgery. Vol 4. Panama City, Panama: Highlights of Ophthalmology; 1998: 58–64. 13. Hagler WS, Aturaliya U. Macular puckers after retinal detachment surgery. Br J Ophthalmol 1971;55:451–457. 14. Haritoglou C, Gass CA, Schaumberger M, et al. Macular changes after peeling of the internal limiting membrane in macular hole surgery. Am J Ophthalmol 2001;132: 363–368. 15. Haritoglou C, Gass CA, Schaumberger M, et al. Long-term follow-up after macular hole surgery with internal limiting membrane peeling. Am J Ophthalmol 2002;134:661–666. 16. Mester V, Kuhn F. Internal limiting membrane removal in the management of full-thickness macular holes. Am J Ophthalmol 2000;129:769–777. 17. Van De Moere A, Stalmans P. Anatomical and visual outcome of macular hole surgery with infracyanine green-assisted peeling of the internal limiting membrane, endodrainage, and silicone oil tamponade. Am J Ophthalmol 2003;136:879–887. 18. Karacorlu M, Karacorlu S, Ozdemir H. Iatrogenic punctate chorioretinopathy after internal limiting membrane peeling. Am J Ophthalmol 2003;135:178–182. 19. Mitamura Y, Ohtsuka K. Relationship of dissociated optic nerve fiber layer appearance to internal limiting membrane peeling. Ophthalmology 2005;112:1766–1770. 20. Kwok AK, Leung DY, Hon C, et al. Vision threatening vitreous haemorrhage after internal limiting membrane peeling in macular surgeries. Br J Ophthalmol 2002;86:1449–1450. 21. Nakata K, Ohji M, Ikuno Y, et al. Sub-retinal hemorrhage during internal limiting membrane peeling for a macular hole. Graefes Arch Clin Exp Ophthalmol 2003;241:582–584.
A
and we retrospectively analyzed the correlation between ILM peeling and the occurrence of a postoperative ERM in patients who underwent vitrectomy for rhegmatogenous RD. An ERM is semitransparent membranous tissue that forms at the vitreoretinal interface of the macula and can causes abnormalities of macular function. Roth and Foos1 postulated that an idiopathic ERM proliferates as retinal tissue-derived glial cells escape from microdefects in the ILM that occur during posterior vitreous detachment and migrate to the surface of the retina. Another theory attributes the pathogenesis of an ERM to the growth and fibrous metaplasia of the
n epiretinal membrane (ERM) is one of the most common complications of rhegmatogenous retinal detachment (RD) after a vitrectomy. In an ERM that occurs after rhegmatogenous RD, unlike the pathogenesis of an idiopathic ERM, the retinal pigment epithelial cells migrate to the vitreous cavity through the retinal break and settle on the macular surface, forming the membrane. Therefore, we postulated that removing the macular internal limiting membrane (ILM) during a vitrectomy for rhegmatogenous RD should help to remove any remaining posterior vitreous cells and other precursors of the ERM present on the retinal surface, such as the retinal pigment epithelial cells,
880
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
vitreous cells that remain on the retina surface after posterior vitreous detachment. The ERM that occurs after retinal reattachment surgery for treating rhegmatogenous RD is known as macular pucker. Various factors cause visual loss after the surgical treatment of rhegmatogenous RD, and an ERM is one of the most common causes of visual loss.2,3 The reported prevalence of ERM is 3% to 8.5% after scleral buckling and 6.1% to 12.8% after vitrectomy.4–9 The pathogenesis of ERM that occurs after rhegmatogenous RD, unlike the idiopathic ERM, that the retinal pigment epithelial cells migrate to the vitreous cavity through the retinal break and settle on the macular surface, forming the membrane.10 Also, the ILM, which corresponds to the basement membrane of the Müller cell, is located at the innermost layer of the retina and can be play a role of scaffold on various cells that develops ERM. Therefore, we postulated that removing the macular ILM during a vitrectomy for rhegmatogenous RD could help to remove any remaining posterior vitreous cells and other precursors of the ERM present on the retinal surface, such as the retinal pigment epithelial cells. This would prevent the occurrence of a postoperative ERM, as the ILM acts as a scaffold for cell growth, as the basement membrane of the Müller cells. Accordingly, we investigated the correlation between ILM peeling and the occurrence of a postoperative ERM in patients who underwent vitrectomy for rhegmatogenous RD. Methods We analyzed retrospectively the medical records of 135 patients who underwent vitrectomy for primary rhegmatogenous RD from November 2007 to August 2011. Institutional Review Board (IRB)/Ethics Committee approval was obtained, and the study adhered to the tenets of the Declaration of Helsinki. Patients were From the *Department of Ophthalmology, Chungnam National University College of Medicine, Daejeon, Republic of Korea; and †Department of Ophthalmology, Kosin University Gospel Hospital, Busan, Republic of Korea. None of the authors have any financial/conflicting interests to disclose. Involved in design of study (K.Y.N., J.Y.K.); conduct of study (K.Y.N., J.Y.K.); collection of data (K.Y.N.); analysis and interpretation of data (K.Y.N., J.Y.K.); and critical review of article (J.Y.K.). The protocol was approved by the Institutional Review Board of Chungnam National University Hospital, and the study adhered to the tenets of the Declaration of Helsinki. Operation was performed after obtaining the approval of patients for the procedure with informed consent. Reprint requests: Jung Yeul Kim, MD, Department of Ophthalmology, Chungnam National University Hospital, 640 Daesa-dong, Jung-gu, Daejeon 301-721, South Korea; e-mail: [email protected]
881
excluded if they had a previous ERM, history of trauma or uveitis, proliferative vitreoretinopathy, silicone oil filling after the vitrectomy, and follow-up for ,12 months. We evaluated ERMs during the initial RD diagnosis with optical coherence tomography (OCT; Stratus OCT3; Carl Zeiss Meditec, Dublin, CA) in cases of macular-on RD or shallow macular-off RD. If OCT examination was unavailable, the operator investigated ERM before and after perfluorocarbon liquid (PFCL) injection during surgery under a surgical microscope. If an ERM was found, the operator removed the ERM and peeled the ILM, and these cases were excluded. The surgery was conducted by a single surgeon (J.Y.K.). The operator experienced some severe cases of postoperative ERM after primary vitrectomy for rhegmatogenous RD in the early phase. Therefore, the operator peeled the ILM routinely for about 2 years, until July 2009. Internal limiting membrane peeling was not continued thereafter, and its value for the prevention of ERM development is therefore unclear. A 20-gauge vitrectomy was performed under general or retrobulbar anesthesia. After the conjunctival incision, scleral incisions were made 4.0 mm and 3.5 mm from the corneoscleral limbus in phakic and psudophakic/aphakic eyes, respectively, in the inferotemporal, superonasal, and superotemporal sclera. The surgeon performed combined operation (pars plana vitrectomy + phacoemulsification). Surgery was performed using a vitrectomy machine (ACCURUS Surgical System; Alcon Laboratories, Fort Worth, TX). A wide-field visualization system (Mini Quad XL; Volk Optical, Mentor, OH) was used. If there was no posterior vitreous detachment, as verified with intravitreal triamcinolone injection, posterior vitreous detachment was induced using a vitreous cutter, intraocular forceps, and pick. To remove the ILM, it was stained with indocyanine green (Diagnogreen; Daiichi Sankyo, Tokyo, Japan) and then removed using forceps. In cases with macular detachment, the ILM was stained and the area near the vascular arcade filled with perfluorocarbon to flatten the macula before ILM peeling. After fluid-gas exchange, endolaser retinopexy was performed, and perfluoropropane (C3F8) or sulfur hexafluoride (SF6) was injected. The surgery was finished after suturing the sclera and conjunctiva. In all cases, cryopexy was not performed. The subjects were divided into two groups according to ILM peeling. Their medical records were reviewed for age when diagnosed with rhegmatogenous RD, gender, best-corrected visual acuity (BCVA), macular involvement, the location, number, and shape of the break(s), RD extent, and gas type.
882
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 5
The visual acuity was recorded 3 months, 6 months, and 12 months postoperatively, and the occurrence of an ERM was confirmed using OCT. The break location was classified as upper, lower, and both upper and lower sites. The breaks were classified into single and multiple breaks. The extent of RD was classified as #1/4, 1/4 to 1/2, 1/2 to 3/4, and 3/4 to entire. An ERM was defined as a case in which the macula was covered by a highly reflective band on OCT. In patients in whom a postoperative ERM developed, it was removed surgically if the patients had a BCVA , 20/40 or complained of metamorphopsia. To analyze the results, the Snellen visual acuity was converted into the logMAR visual acuity. The statistical analysis was conducted using SPSS (version 17.0). The chi-square test was used to compare the occurrence of an ERM between the groups with and without ILM peeling. An independent t-test was used to compare the visual acuities. Patients who underwent surgery to remove an ERM were followed in the same manner as if they did not undergo the surgery, and the visual acuity at each follow-up after ERM removal was defined as the BCVA measured just before the ERM removal surgery. The non-ILM peeling group was subdivided into groups with and without a postoperative ERM, and the differences in various factors were compared between the two groups; the independent t-test was used to analyze age and BCVA at diagnosis and the chi-square test to analyze macular involvement, break location, break number, extent of RD, and gas type.
Results The study enrolled 135 eyes of 135 patients: 70 who underwent ILM peeling and 65 patients who did not. At the time of diagnosis of rhegmatogenous RD, the 2 groups did not differ in terms of age, BCVA, break shape, number, or location, and RD extent (Table 1). One hundred and two patients had phakic eyes, and the preoperative lens grade by nucleus sclerosis was similar in the 2 groups (P = 0.491). Cataract removal was performed in all except one case. At the 12-month follow-up, no ERM was observed in the patients who had ILM peeling during vitrectomy, whereas an ERM was observed in 14 of 65 patients who did not undergo ILM peeling during the vitrectomy (21.5%). This difference was significant (P , 0.001). For the patients who had surgery to remove an ERM, the BCVA measured immediately before surgery was used as the visual acuity at each follow-up
Table 1. Baseline Characteristics in ILM Peeling Group and Non-ILM Peeling Group
Age (year) Sex Male Female Macula on/off On Off Break position Upper Lower Both Break number Single Multiple Break shape Horse shoe tear Atrophic hole Others BCVA (logMAR) at RD diagnosis Extent of RD ,1/4 1/4–1/2 .1/2 Gas C3F8 SF6
ILM Peeling (n = 70)
Non-ILM Peeling (n = 65)
48.2 (±17.8)
47.9 (±19.0)
42 (60.0%) 28 (40.0%)
34 (52.3%) 31 (47.7%)
41 (58.6%) 29 (41.4%)
35 (53.8%) 30 (46.2%)
40 (57.1%) 21 (30.0%) 9 (12.9%)
39 (60.0%) 23 (35.4%) 3 (4.6%)
47 (67.1%) 23 (32.9%)
37 (56.9%) 28 (43.1%)
37 (52.9%)
36 (55.4%)
P 0.919* 0.390† 0.606† 0.232†
0.287† 0.338† 23 (32.9%) 25 (38.5%) 10 (14.2%) 4 (6.1%) 0.76 (±0.84) 0.85 (±0.87)
0.547* 0.199†
25 (35.7%) 30 (42.9%) 15 (21.4%)
16 (24.6%) 35 (53.8%) 14 (21.6%)
60 (85.8%) 10 (14.2%)
58 (89.2%) 7 (10.8%)
0.610†
*Independent t-test. †Chi-square test.
after ERM removal for the analysis. No significant difference in the mean visual acuity was found between the groups with and without ILM peeling. However, when the macula-on group was analyzed, the mean visual acuity was better in the ILM peeling group than in the non-ILM peeling group, and a significant difference was found at the 12-month follow-up (P = 0.03) (Figures 1 and 2). For the 65 patients who did not undergo ILM peeling, the differences in various factors were analyzed between the patients with and without a postoperative ERM. No significant difference in the age, BCVA, macular involvement, break location, number, or shape, RD extent, and gas type was found between the 2 groups (Table 2). Of the 14 patients who developed an ERM, it was diagnosed within 3 months in 12 patients (85.7%) and between 3 and 6 months in 2 (14.3%). It took a mean of 9.2 (range, 2.1–25.7) weeks to confirm the ERM after RD. Of the 14 patients who developed an ERM, 10 (71.4%) required surgery to remove it. There was
883
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
Table 2. Factor Analysis Between ERM Group and NonERM Group in Non-ILM Peeling Patients ERM Group (n = 14)
Fig. 1. Visual acuity changes in all patients (135 patients) after primary vitrectomy for rhegmatogenous RD. Error bars represent mean (SD).
a mean of 22 (range, 5.6–51.7) weeks between RD and removal of the ERM. There was one case of recurred RD in each group. There were no other obvious complications of vitrectomy and ILM peeling in this study. Discussion An ERM is one of the most common complications of rhegmatogenous RD after a vitrectomy, and ILM may play a role of scaffold on various cells that develops ERM. The ILM is composed of the basement membrane of Muller cells, proteoglycans, and type IV collagen and is an important structure in the formation of the vitreoretinal interface.11 Astrocytes and myofibroblasts migrate to the surface of the ILM and invade the deep layer, thereby causing an ERM.12 The ERM
Fig. 2. Visual acuity changes in macular-on patients (76 patients) after primary vitrectomy for rhegmatogenous RD. Error bars represent mean (SD) (median visual acuity at month 3, 6, and 12 was 0.22, 0.09, and 0, respectively, in ILM peeling group and 0.22, 0.09, and 0.09, respectively, in non-ILM peeling group). *P = 0.027 by independent t-test.
Non-ERM Group (n = 51)
P
Age, years 51.6 (±20.3) 46.9 (±18.7) 0.420* Sex 0.375† Male 9 (64.3%) 25 (49.0%) Female 5 (35.7%) 26 (51.0%) Macula on-off 1.000† On 8 (57.1%) 27 (52.9%) Off 6 (42.9%) 24 (46.2%) Break position 0.765† Upper 9 (64.3%) 30 (58.8%) Lower 4 (28.6%) 19 (37.3%) Both 1 (7.1%) 2 (3.9%) Break number 0.561† Single 7 (50.0%) 30 (58.8%) Multiple 7 (50.0%) 21 (41.2%) Break shape 0.066† Atrophic hole 4 (28.6%) 32 (62.7%) Horse shoe tear 8 (57.1%) 17 (33.3%) Others 2 (14.3%) 2 (4.0%) BCVA (logMAR) 0.93 (±0.97) 0.83 (±0.85) 0.547* at RD diagnosis Extent of RD 0.199† ,1/4 2 (14.3%) 14 (27.5%) 1/4–1/2 7 (50.0%) 28 (54.9%) .1/2 5 (35.7%) 9 (17.5%) Gas 1.000† C3F8 13 (92.9%) 45 (88.2%) SF6 1 (7.1%) 6 (11.8%) *Independent t-test. †Chi-square test.
that occurs after retinal reattachment surgery for treating rhegmatogenous RD is caused by retinal pigment epithelial cells that migrate to the surface of the posterior pole retina by diffusing into the vitreous cavity through the break or through fibrosis and the proliferation of precursor cells that remain on the retinal surface after the vitrectomy.10 Therefore, removal of the ILM can remove the ERM precursor cells from the retinal surface. In addition, in cases of severe bullous RD, a mild ERM might not be diagnosed in the fundus examination or with OCT. Therefore, an ERM that was not diagnosed before the surgery would also be removed by removal of the ILM. These provide a rationale for ILM peeling as a method of preventing the occurrence of a postoperative ERM in vitrectomy for rhegmatogenous RD. In our series, ERM developed in 14 of the 135 patients (10.4%). However, the occurrence rate was 21.5% in the patients who did not undergo ILM peeling, which was higher than the reported rate of 6.1% to 12.8%. This difference is likely attributable to the fact that we considered an ERM to be present if
884
RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 2015 VOLUME 35 NUMBER 5
a thin membrane was observed on the macula with OCT, even if it did not influence the visual acuity and was not obvious in the fundus examination.8–10 The authors investigated the effects of the postoperative ERM on visual acuity. The mean BCVA was compared between the groups with and without ILM peeling 3 months, 6 months, and 12 months postoperatively. In the patients who underwent surgery to remove an ERM that occurred early after RD surgery with a BCVA , 0.5, it was impossible to compare the visual acuity between the two groups because it recovered postoperatively. Therefore, they were assumed to have been followed as although they had not undergone surgery, and BCVA measured just before the ERM removal surgery was used for analysis. For example, when an ERM occurred within 3 months of the RD surgery and was then removed, the BCVA measured immediately before ERM removal was used as the visual acuity at 3 months, 6 months, and 12 months after RD surgery. The visual acuity data were replaced based on the assumption that visual acuity generally worsens over time after the formation of an ERM. As a result, the overall visual acuity was better in the group with ILM peeling than that without ILM peeling, although the difference was not significant. However, when the macula-on group was analyzed, the mean visual acuity was better in the ILM peeling group than in the non-ILM peeling group, and a significant difference was found at the 12-month follow-up (P = 0.03). This showed that the visual acuity was affected by the ERM that occurred after primary vitrectomy for rhegmatogenous RD, implying that removal of the ILM results in better visual acuity by preventing the occurrence of a postoperative ERM compared with patients who do not undergo ILM peeling. Preoperative cataract grade and cataract surgery might affect visual acuity. However, preoperative cataract grade was similar between the two groups, and cataract surgery was performed in nearly all phakic patients to shave the peripheral vitreous more easily and to avoid reoperation for postoperative cataract progression. We therefore hypothesize that postoperative cataracts and cataract surgery did not affect visual acuity. We analyzed the time of ERM development in this study. Epiretinal membrane occurred in 12 of the 14 patients (85.7%) within 3 months of RD surgery. Of the 14 patients who developed an ERM, it was removed in the 10 patients (71.4%) who had a BCVA , 0.5. This was similar to the result obtained by Martínez-Castillo et al,9 who observed ERM in 32.1% and 57.1% of patients at 1 month and 3 months, respectively, after vitrectomy for rhegmatogenous RD, and that removal of the ERM was required in 22 of the
28 patients (78.5%) who developed a postoperative ERM due to significant visual loss. Therefore, postoperative ERM tend to occur within 3 months of vitrectomy for rhegmatogenous RD. There were some reports about risk factors associated ERM development after rhegmatogenous RD surgery. A study reported a significant correlation between the postoperative ERM after scleral buckling for rhegmatogenous RD with preoperative macular detachment, low visual acuity, and proliferative vitreoretinopathy, and reoperation. However, MartínezCastillo et al reported no significant correlation in the occurrence of a postoperative ERM with macular detachment and visual acuity.2,4,7,9,13 In this study, to investigate the risk factors associated with postoperative ERM occurrence, we compared various factors between the ERM group and the non-ERM group in non-ILM peeling patients. There was no significant correlation between ERM development and age, BCVA at initial diagnosis, macular involvement, break location, number, shape, RD extent, or gas type in non-ILM peeling patients. However, the percentage of RD . 1/2 was higher in the ERM group than in the non-ERM group (35.7%:17.6%, P = 0.199) (Table 2). This may be because retinal pigment epithelium cells disperse more easily through a larger break or larger RD. Our results were not statistically significant; however, with a higher powered study that included more patients, this correlation between larger RD and ERM may become apparent. In a prospective study of patients who required silicone oil filling and underwent retinectomy during a vitrectomy due to accompanying proliferative vitreoretinopathy upon diagnosis of RD, Odrobina et al reported that retinal pucker occurred in 17.8% of the non-ILM peeling group (33 patients), but no retinal pucker occurred in the ILM peeling group (51 patients). Although there were a few cases with fibrous proliferation in the patients who underwent ILM peeling, its progression was limited to the margin of ILM peeling.11 However, our study is meaningful in that it enrolled patients with primary rhegmatogenous RD, who are expected to have a relatively good postoperative prognosis in the case of successful retinal adhesion. In this study, ILM peeling during vitrectomy for rhegmatogenous RD was shown to help to prevention of the formation of a postoperative ERM. No postoperative ERM was found in the group with ILM peeling, whereas an ERM occurred in 21.5% of the nonpeeling group. Although this result was obtained at the 12-month follow-up, it was considered clinically meaningful because most studies report that ERMs occur within 3 months after a vitrectomy for rhegmatogenous RD.
ILM PEELING DURING PPV IN RHEGMATOGENOUS RD NAM AND KIM
Recurrent RD after primary vitrectomy may affect postoperative visual acuity and ERM development. We only observed one case of recurrent RD in each group, and neither was associated with proliferative vitreoretinopathy or macular detachment. The safety of ILM peeling when there is no macular lesion is controversial. Complications related to ILM peeling include retinal hemorrhage, eccentric scotoma, retinal edema, iatrogenic punctuate chorioretinopathy, dissociated nerve fiber layer, vitreous hemorrhage, and subretinal hemorrhage. In addition, stain toxicity has been reported.14–21 However, we observed no definite clinical complication caused by ILM peeling. In addition, when the mean BCVA was compared between the groups with and without ILM peeling, no significant difference was found. However, within the macular-on group, the overall mean BCVA was better in the group with ILM peeling. Therefore, it seems that visual acuity-related complications caused by ILM peeling were insignificant. Several studies reported that complications related to ILM peeling were mainly associated with mechanical trauma to the retinal tissues, and the safe concentration of the dye is well known. Therefore, with improvement in the ILM manipulation technique and the proper staining technique, the rate of complications could be reduced. This study has some limitations. First of all, this study is retrospective, and the data were collected from limited medical records. Second, preoperative OCT examination was not available to all RD patients. And in cases of macular-off RD, ERM was investigated by operator during vitrectomy under surgical microscope; thus, very thin ERM might be missed. Finally, the number of cases might not be enough to analyze the factors related with ERM development in non-ILM peeling group. In conclusion, ILM peeling during vitrectomy for primary rhegmatogenous RD is useful for preventing postoperative ERM formation and provides a better visual prognosis. However, we have to consider whether ILM peeling for all patients is beneficial. This should be confirmed in a further long-term prospective study with a greater number of patients. Key words: rhegmatogenous retinal detachment, pars plana vitrectomy, internal limiting membrane peeling, epiretinal membrane. References 1. Roth AM, Foos RY. Surface wrinkling retinopathy in eyes enucleated at autopsy. Trans Am Acad Ophthalmol Otolaryngol 1971;75:1047–1058. 2. Lobes LA Jr, Burton TC. The incidence of macular pucker after retinal reattachment surgery. Am J Ophthalmol 1978; 85:72–77.
885
3. Wilkinson CP. Visual results following scleral buckling for retinal detachments sparing the macula. Retina 1981;1: 113–116. 4. Tanenbaum HL, Schepens CL, Elzeneiny I, et al. Macular pucker following retinal detachment surgery. Arch Ophthalmol 1970;83:286–293. 5. De Bustros S, Thompson JT, Michels RG, et al. Vitrectomy for idiopathic epiretinal membranes causing macular pucker. Br J Ophthalmol 1988;72:692–695. 6. De Bustros S, Rice TA, Michels RG, et al. Vitrectomy for macular pucker. Use after treatment of retinal tears or retinal detachment. Arch Ophthalmol 1988;106:758–760. 7. Uemura A, Ideta H, Nagasaki H, et al. Macular pucker after retinal detachment surgery. Ophthalmic Surg 1992;23: 116–119. 8. Heo MS, Kim HW, Lee JE, et al. The clinical features of macular pucker formation after pars plana vitrectomy for primary rhegmatogenous retinal detachment repair. Korean J Ophthalmol 2012;26:355–361. 9. Martínez-Castillo V, Boixadera A, Distéfano L, et al. Epiretinal membrane after pars plana vitrectomy for primary pseudophakic or aphakic rhegmatogenous retinal detachment: incidence and outcomes. Retina 2012;32:1350–1355. 10. Katira RC, Zamani M, Berinstein DM, et al. Incidence and characteristics of macular pucker formation after primary retinal detachment repair by pars plana vitrectomy alone. Retina 2008;28:744–748. 11. Odrobina D, Bednarski M, Cisiecki S, et al. Internal limiting membrane peeling as prophylaxis of macular pucker formation in eyes undergoing retinectomy for severe proliferative vitreoretinopathy. Retina 2012;32:226–231. 12. Moris R, Kuhn F. Surgical treatment of macular surface disorders. World Atlas Series of Ophthalmic Surgery. Vol 4. Panama City, Panama: Highlights of Ophthalmology; 1998: 58–64. 13. Hagler WS, Aturaliya U. Macular puckers after retinal detachment surgery. Br J Ophthalmol 1971;55:451–457. 14. Haritoglou C, Gass CA, Schaumberger M, et al. Macular changes after peeling of the internal limiting membrane in macular hole surgery. Am J Ophthalmol 2001;132: 363–368. 15. Haritoglou C, Gass CA, Schaumberger M, et al. Long-term follow-up after macular hole surgery with internal limiting membrane peeling. Am J Ophthalmol 2002;134:661–666. 16. Mester V, Kuhn F. Internal limiting membrane removal in the management of full-thickness macular holes. Am J Ophthalmol 2000;129:769–777. 17. Van De Moere A, Stalmans P. Anatomical and visual outcome of macular hole surgery with infracyanine green-assisted peeling of the internal limiting membrane, endodrainage, and silicone oil tamponade. Am J Ophthalmol 2003;136:879–887. 18. Karacorlu M, Karacorlu S, Ozdemir H. Iatrogenic punctate chorioretinopathy after internal limiting membrane peeling. Am J Ophthalmol 2003;135:178–182. 19. Mitamura Y, Ohtsuka K. Relationship of dissociated optic nerve fiber layer appearance to internal limiting membrane peeling. Ophthalmology 2005;112:1766–1770. 20. Kwok AK, Leung DY, Hon C, et al. Vision threatening vitreous haemorrhage after internal limiting membrane peeling in macular surgeries. Br J Ophthalmol 2002;86:1449–1450. 21. Nakata K, Ohji M, Ikuno Y, et al. Sub-retinal hemorrhage during internal limiting membrane peeling for a macular hole. Graefes Arch Clin Exp Ophthalmol 2003;241:582–584.
