To review outcomes of vitrectomy plus or minus scleral buckling for retinal detachment (RD) attributable to peripheral break(s) with noncausal macular hole, plus or minus internal limiting membrane (ILM) peeling.
Retrospective chart review.
Forty-nine consecutive patients from March 1, 1998 to March 31, 2009 with RD attributable to peripheral break and macular hole were treated by vitrectomy. Five had no scleral buckle placed. Forty-three underwent ILM peeling.
The main outcome measures were retinal reattachment, macular hole status, and vision. Final retinal reattachment rate was 95.9% (47/49) with 1 and 100% with 2 operations. Final macular hole closure rate was 39/43 with ILM peeling (90.7%), and 2/6 without (33.3%, P value = .0041). Mean final acuity was 20/120, and 20/100 if the macular hole was closed.
Vitrectomy, plus or minus scleral buckle, with ILM peeling is effective for repair of RD with macular hole. ILM peeling can increase the rate of macular hole closure.
A coexistent macular hole (MH) is seen in approximately 1% to 4% of cases of rhegmatogenous retinal detachment (RRD). Prior to the introduction of vitrectomy, repair of the retinal detachment (RD) was accomplished with scleral buckling, and it was determined that the MH could be left untreated. Later, vitrectomy and scleral buckling combined with gas injection was found to be useful for repair of the RRD, but the status of the MH remained of secondary importance. Once it became established that idiopathic MH could be repaired, more interest developed in trying to repair both problems. There are few published series that discuss the use of vitrectomy and gas injection to treat RRD + MH simultaneously. In 1 series, the single-surgery retinal reattachment rate was 78% and the MH closure rate was 31%. A second, small series showed closure of the MH in 7 of 8 cases, but surgery to repair the MH was done after repair of the RRD. A third, recent series showed successful simultaneous retinal reattachment and hole closure in 5 of 5 cases.
In recent years, the use of internal limiting membrane (ILM) peeling as an adjunct for surgical repair of idiopathic MHs has improved closure rates. Since approximately 1996 our group has peeled ILM for repair of most idiopathic MH, and as well in the setting of RRD + MH. We undertook a review of all patients in our practice who had RRD with a coexistent, noncausal MH (RRD + MH) to look at anatomic and visual outcomes, and to examine whether ILM peeling increased MH closure rates. This appears to be the largest series reported to date of patients with RRD + MH; no larger studies were found on computerized search using PubMed.
Patients and Methods
A surgical database was queried for patients with RRD and MH. The charts of patients with a concurrent diagnosis of RRD and MH who underwent vitrectomy between March 1, 1998 and March 31, 2009 were retrospectively reviewed. This review included patients who had an MH noted either preoperatively or discovered at the time of vitrectomy to repair the RRD. Some patients had either an overhanging RD or media problems that precluded a preoperative view of the macula. Patients with a history of previous vitrectomy or scleral buckling were included. Exclusion criteria included more than 6 clock hours stage C anterior or stage C posterior proliferative vitreoretinopathy (PVR), myopic MHs with associated RD without peripheral retinal break, and MH found after successful retinal reattachment by scleral buckling.
Data collected included preoperative vision, macular status, surgical approach for primary repair, type of scleral buckle used (if done), gas for intraocular tamponade, and duration of face-down positioning. Vision and anatomic status of both the peripheral retina and macula were documented postoperatively. The need for repeat surgery or in-office treatment following initial repair was also recorded.
In all patients, the primary intent was to repair the RD, but an attempt to repair the MH was also made. Therefore, all patients had a long-acting gas bubble and a recommended period of face-down positioning. All patients underwent 3-port 20-gauge vitrectomy, and nearly all had placement of an encircling element, usually a #41 or #42 band. Any prior encircling bands were left untouched. ILM peeling was performed in most surgeries, and no stains were used. Bimanual dissection with an illuminated pick was commonly used. Subretinal fluid was drained through the MH in most cases. Gas tamponade was primarily with SF6, in some cases C 3 F 8 was used, and room air was not used. C 3 F 8 was used with more complex detachments, such as those having multiple or posterior breaks and/or PVR, as well as when compliance with face-down positioning was unlikely. Strict postoperative face-down positioning was planned, usually for 7 days. Differences in outcome between patients with ILM peeling and those without ILM peeling were compared with the Fisher exact test of proportions.
Forty-nine consecutive patients met the criteria for this study between 1998 and 2009. The average age was 60 (range 15-88), and there were 31 men (63.3%) and 18 women (36.7%). Five had a history of blunt trauma, which was felt to be the cause of their MH. Nine had a history of previous retinal surgery: 2 with RD treated with scleral buckle, 1 vitrectomy with scleral buckle for PVR, 2 MH surgeries, 2 macular pucker surgeries, 1 vitrectomy for retained lens fragments, and 1 vitrectomy for endophthalmitis. Forty patients had no history of prior retinal or vitreous surgery. Eight patients (16.3%) had macula-on RD, and 41 (83.7%) had macula-off RD.
Encircling scleral buckling was performed in 40 patients, and a preexisting encircling element was present in 4 cases. A #41 band was used in 30 patients, a #42 band in 6 patients, a #4050 band in 5 patients, a #240 band in 2 patients, and an unknown element in 1 patient. Five patients had vitrectomy without an encircling element, and all had successful retinal reattachment. Endolaser was used for retinopexy in 47 of 49 cases, with cryotherapy used in the other 2. The MH was not treated with cryotherapy or laser in any case. ( Supplemental videos show bimanual ILM peeling, ILM peeling from a detached retina, and use of Rice pick, available online at AJO.com ).
For gas tamponade, 20% or 25% sulfur hexafluoride (SF 6 ) was used in 34 of 49 patients (69.4%), and 10% to 18% perfluoropropane (C 3 F 8 ) was used in 15 patients (30.6%) ( Table 1 ). The number of days of face-down positioning was documented in 33 of 49 patients (67.3%), and the average was 6.3 days (range 3 to 8 days). Four patients required a reoperation for RD. Of these 4, 1 had tamponade with 20% SF 6 , 2 with C 3 F 8 , and 1 with silicone oil.
|Gas Type||MH Outcome a P = .08||RD Outcome b P = .09|
|Perfluoropropane (C 3 F 8 )||11/15 (73.3%) closed||13/15 (86.7%) attached on first surgery|
|Sulfur hexafluoride (SF 6 )||30/34 (88.2%) closed||34/34 (100%) attached on first surgery|
In 47 of 49 patients (95.9%), the retina was attached after 1 surgery. Two of these re-detached and were later reattached. The remaining 2 of 49 patients had successful retinal repair: 1 after the first reoperation and the second after 2 reoperations including silicone oil tamponade. All 49 patients (100%) had the retina attached at study completion.
The overall MH closure rate after 1 surgery was 83.7% (41/49) after 1 surgery and 85.7% (42/49) after 2 surgeries. The MH closure rate among patients with ILM peeling was 90.7% (39/43), compared to 33.3% (2/6) without ILM peeling ( P value = .0041; Table 2 ). Eight patients had macula-on preoperative status, and 100% of these patients had closure of the MH and improvement in visual acuity.
|Follow-up and MH Closure Rate||ILM Peeling N = 43||No ILM Peeling N = 6||Total N = 49|
|Average follow-up, months||8.1||5.7||7.8|
|MH closure after initial surgery||39 (90.7%)||2 (33.3%)||41 (83.7%)|
|P < .01|
Visual acuity improved in 47 of 49 patients (95.9%). In the 2 patients whose vision did not improve, 1 had a suprachoroidal hemorrhage during initial surgery and possible optic nerve damage from elevated IOP postoperatively; and the other refused reoperation to attempt MH closure. Both were macula-off preoperatively. The average postoperative visual acuity for all 49 patients was 0.75 logMAR (20/120 on the Snellen scale).
Rhegmatogenous retinal detachment attributable to a peripheral break (or breaks) with a concomitant noncausal MH (RRD + MH) occurs in approximately 1% to 3% of spontaneous rhegmatogenous RDs. This entity should be differentiated from RRD caused by MH, which occurs in patients with very high myopia and progresses from the posterior pole to the periphery. Multiple reports describe treatment of posterior RD caused by MH in myopic eyes with posterior staphyloma. In this form of RD closure of the MH is essential for RD repair, while in RRD + MH the repair of the retinal detachment does not require closure of the MH. Patients with RD attributable to myopic MH were excluded from this study.
Until the 1980s, the primary surgical goal for RRD + MH was to repair the RD by closing the peripheral breaks, with no attempt made to repair the MH. Later, PPV was combined with scleral buckle and gas injection, but again no specific attempt was made to repair the MH. The outcome of successful surgery was retinal reattachment with a residual MH and a central scotoma, and anatomic retinal reattachment was seen in most patients. After the initial reports of successful closure of idiopathic MHs, interest developed in the postoperative status of the MH in surgically treated RRD + MH. O’Driscoll and associates analyzed results of 23 patients with RRD + MH who underwent pars plana vitrectomy with scleral buckle. In this series, no effort was made to peel the ILM and postoperative positioning was done as needed to repair the RD, not specifically aimed at hole closure. MH closure was achieved in 31% of patients, which was similar to our experience in eyes without ILM peeling. Retinal reattachment was achieved initially in 78% of eyes, and ultimately in 87%.
ILM peeling as an adjunct to MH repair became popular in the 1990s as reports indicated that it led to improved success rates. Campochiaro and associates suggested that the ILM affects the vitreoretinal interface, contributing to tractional forces and the formation of MHs. In 2000, Brooks reported that patients undergoing ILM peeling had better closure rates: 100% in 116 eyes with ILM peeling vs 82% in 44 eyes without ILM removal. Of the holes successfully repaired without ILM peeling, 25% reopened, vs zero in the group with ILM peeling. Yoshida and Kishi also found that ILM peeling reduces MH recurrence, possibly because it likely prevents formation and contraction of epiretinal membranes.
The results in our series suggest that ILM peeling during repair of RRD + MH is more successful for closing the MH than vitrectomy without ILM peeling. Of 43 patients who underwent ILM peeling, 91% had closure of the MH, vs 33% ( P < .01) of those who did not undergo ILM peeling. Compared to other studies that looked at ILM peeling for MH with RRD, this series had favorable results and a larger number of patients. Kiné and associates found that ILM peeling for repair of MH after the initial vitrectomy for RRD repair was successful: 7 of their patients had RD with MH, and 6 of 7 (85.7%) had MH closure after the first MH surgery. Singh found in a small prospective series that either combined or sequential surgery could be successful in the situation of RRD + MH. In this series, the ILM was peeled in each case after staining with trypan blue, and 5 of 5 holes were closed with a single procedure in combination with successful retinal reattachment.
ILM peeling is technically challenging because the membrane is barely visible and complete removal is hard to assess. Indocyanine green (ICG) dye staining can aid in visualization of the membrane, but this has been associated with cytotoxicity. Our group does not use ICG dye for idiopathic MH because of the potential toxicity, and felt this risk would be even greater in the presence of a detached retina. Trypan blue was not available in the U.S. during this time period. The glassy appearance of the ILM and the whitish color change of the retina where the ILM has been peeled are appreciable, even when the retina is detached. Perfluoro-octane (PFO) was not used to peel the ILM and stabilize the retina because of concerns that PFO would pass through the hole into the subretinal space and present serious problems.
ILM peeling is particularly difficult with a detached retina since the membrane is hard to grasp and separate from the mobile retina. We found a bimanual technique using a forceps and illuminated pick to be very helpful in allowing counter-traction to be applied. We also performed drainage in most cases through the MH, and did not create a posterior retinotomy. There was no negative effect on anatomic outcome. It is possible that stretching of the hole during drainage had a deleterious effect on visual outcome, but we suspect the macula-off status in most cases had a greater negative effect. Visual acuity outcomes in this series were not as good as those typically found with successful primary MH repair. The average postoperative VA for patients in this series with successful MH closure for those with macula-off RRD was 0.75 logMAR (20/110 Snellen), vs 0.65 logMAR (20/90 Snellen) for the patients with macula-on detachment.
Patients were instructed to position strictly face-down for 7 days, on average, in our study. Early in the treatment of idiopathic MHs, this was our typical recommendation. After ILM peeling became common, the need for lengthy face-down positioning appeared lessened, and our current usual recommendation is 4 days. In the setting of RRD + MH, a longer period may not be necessary, but seems rational. The choice of gas for tamponade was driven by both ocular and patient issues, so patients with more complex detachments and/or likely poor compliance were more likely to get C 3 F 8 gas. The choice of gas did not appear to significantly affect outcomes for either the detachment or hole. Either gas used was effective for repairing both the detachment and hole, and any differences in outcomes were more related to the clinical situation than the gas chosen.
In patients with RRD + MH, closure of the MH is an added visual benefit, but repair of the RRD is of primary importance. In our series, the single-surgery reattachment rate of 96% and final reattachment rate of 100% compares favorably with other reports of RRD treated with vitrectomy and SB. O’Driscoll’s series had a lower initial and final retinal reattachment rate but included a higher percentage of PVR cases, and was from a slightly earlier era. We did have successful reattachment in all 5 cases treated with vitrectomy and no scleral buckling, but we believe in most cases the addition of an encircling scleral buckle is beneficial.
This series suggests that ILM peeling should be considered during primary repair of RRD + MH, as we have found that it improves MH closure rates. Neither staged surgery nor ILM-staining dyes are needed, as it is possible to visualize the ILM and retina where peeled away.
The authors indicate no funding support. Ed Ryan is the recipient of consulting fees from Alcon. Sundeep Dev indicates an unpaid speaking engagement for Genentech. David Williams has been a consultant for Genentech, Merck, and Surmodics. Herbert Cantrill received grant support for a Genentech investigator sponsored trial. Involved in conception and design (E.R.); analysis and interpretation (E.R., C.B.); writing the article (E.R., C.B.); critical revision of the article (E.R., C.B.); final approval of the article (E.R., C.B.); data collection (E.R., R.M., S.D., S.B., D.W., H.C.); provision of materials, patients, or resources (E.R.); statistical expertise (C.B.); literature search (E.R., C.B.); and administrative, technical, or logistical support (E.R., C.B.). The research was approved by Allina Institutional Review Board, Minneapolis, Minnesota.
The authors thank Julianne Enloe, VitreoRetinal Surgery, for her logistical support in compiling the paperwork for this manuscript.
Use of Rise pick for ILM peeling.Movie 2
Bimanual ILM peeling from a detached retina.Movie 3
Bimanual ILM peeling with Rice pick, then forceps.Movie 4
Bimanual ILM peeling with forceps.
Supplemental Material available at AJO.com .