Fleur Goezinne Department of Ophthalmology University Hospital Maastricht PO Box 5800 P. Debyelaan 25 6202 AZ Maastricht The Netherlands Tel: +31 43 387 5346 Fax: +31 43 387 5343 Email: firstname.lastname@example.org
Purpose: The preoperative and intraoperative clinical variables associated with redetachment and/or a poor visual outcome following scleral buckling (SB) surgery for rhegmatogenous retinal detachment (RRD) have mainly been studied after a short follow-up. This study aimed to analyse long-term effects by following patients for at least 6 months.
Methods: In a retrospective survey we evaluated the data of 436 eyes that underwent SB surgery. Postoperative data were collected at 3-month intervals.
Results: After a mean follow-up period of 51 months, anatomic reattachment was achieved in 76% after one SB procedure, with a final reattachment rate of 97% after additional vitreoretinal procedures. In total, 104 eyes developed redetachment during follow-up. After more than 6 and 12 months of follow-up, 32 eyes (7%) and 20 eyes (5%), respectively, developed redetachment. Multivariate regression analysis showed that recurrent redetachment and more than 7 days of visual field loss were significant predictors for a poor postoperative visual outcome at 12 months. A cumulative size of the tear of more than three disc diameters was a significant predictor of recurrent RRD.
Conclusion: Conventional SB surgery is a reliable procedure in a selected group of eyes with primary RRD. However, in eyes with a retinal tear with a cumulative size of more than three disc diameters, a primary vitrectomy should be considered. Taking into account that 7% of eyes developed redetachment after 6 months, a longer follow-up period seems necessary to evaluate the anatomical and visual outcomes after SB surgery.
To make a better preoperative selection and identify which patients have good operative results with SB surgery, we evaluated what preoperative and intraoperative clinical variables are associated with an increased risk for redetachment and/or a poor visual outcome after a minimal follow-up of 6 months in a consecutive series of eyes that underwent SB for a primary RRD.
Materials and Methods
We reviewed the medical records of all patients who underwent primary SB surgery without vitrectomy for RRD. All patients who were included in the study were operated on by the same surgeon (E.L.H.). Between January 1997 and October 2004, all primary clear media non-traumatic, non-uveitic retinal detachments that had limited proliferative vitreoretinopathy (PVR) underwent SB surgery, except three eyes that had a pneumatic retinopexy. None of the eyes that presented during the aforementioned period underwent a PPV. Seventeen eyes with a follow-up period of < 6 months were excluded. Eyes with a history of uveitis or trauma and eyes with PVR grade C2 and higher were excluded, as described earlier (La Heij et al. 2000, 2001). We also excluded eyes with a complicated RRD caused by vitreous haemorrhage obscuring fundus details. The study was conducted in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. When present, PVR was graded according to the Classification of the Retina Society Terminology Committee (Retina Society Terminology Committee 1983).
All patients were operated on by a single surgeon. The standard procedure was a SB technique where all patients received an encircling band. The decision of a radial or segmental buckle was made based on the size and location of the retinal tear. Drainage of subretinal fluid was performed when judged necessary by the surgeon. Cryocoagulation was performed in almost all cases and intravitreal gas tamponade was used only for superior RRDs.
The following preoperative clinical characteristics of the study patients were collected for statistical analysis: age, sex, preoperative visual acuity, prior intraocular surgery, the number of detached quadrants of the retina, whether or not the central area of the macula (foveal region) was involved in the detachment, the presence and grade of PVR, the type and localization of the retinal break, the number of breaks and their cumulative size in disc diameters (DD). The type of break was noted as an atrophic hole, a horseshoe tear, a combination of both or an ora-dialysis. Preoperative and postoperative Snellen visual acuities (also with pinhole correction) and preoperative and postoperative Goldmann applanation tonometry were recorded. During slit-lamp examination, the presence of a pseudophakic lens or intravitreal pigment was recorded. Funduscopy was performed using indirect binocular ophthalmoscopy in combination with a Goldmann three-mirror contact lens (without scleral depression) and with a panfundus contact lens (Supersquad 160°, Volk Optical Inc., OH, USA). Fundus drawings were made of the RRD in clock hours. By carefully interviewing the patient, we determined the approximate time of onset of the detachment. The duration of the detachment before surgery was categorized into the following groups: 0–7 days and/or more than 7 days, as described by (Diederen et al. 2007).
Surgery reports were reviewed to determine details of the surgical procedure used in repairing the detachment. The following variables relating to the buckling surgery were noted: whether a radial or segmental silicone buckle was placed, if subretinal fluid drainage was performed, whether cryocoagulation had been applied, and if air or intravitreal sulphur hexafluoride (SF6) gas had been used to create an internal retinal tamponade. All intraoperative complications, such as scleral perforation, vitreous incarceration or retinal incarceration, were noted.
After the SB surgery, we noted whether patients developed a recurrent RRD and after what time, the cause of the redetachment and the subsequent clinical procedures. Anatomical success was defined either as a complete reattachment of the retina or as a stable situation with an attached retina and confined pockets of subretinal fluid. Recurrent retinal detachment was categorized as being attributable to vitreous traction with PVR, new or missed retinal breaks, persistent and/or progressive reaccumulation of subretinal fluid, or progressive ‘leakage’ of subretinal fluid at the peripheral edge of the buckle caused by lack of indentation. If the redetachment occurred within 6 months of the SB procedure, this was defined as ‘early redetachment’.
After the SB operation, we collected data at 3-month intervals. After 12 months, many patients returned to their referring ophthalmologist. Therefore, follow-up data were also collected from the referring ophthalmologist. Data on anatomical success were obtained by one of the authors (F.G.) calling and interviewing patients if they had not returned for follow-up examination either at our department or at the practice of the referring ophthalmologist. We called the patients to ask whether their visual function was not worsened, and if they had noticed any signs or symptoms of a (re)detachment. And if this was the case, we asked the name of the ophthalmologist to whom they went to seek help. Subsequently, we asked the patient for permission to obtain their data from these ophthalmologists. Statistical analysis was performed using the spss software, version 13.0 (SPSS Inc., Chicago, IL, USA). Snellen visual acuities were converted to a logarithmic scale (LogMAR, i.e. the logarithm of the minimal angle of resolution), as described earlier (Rosser et al. 2001). For this study, a poor visual outcome was defined as a visual acuity of lower than 0.1 Snellen visual acuity 12 months postoperatively. Missing data on covariables as well as on outcome variables (if this outcome was measured on at least three different occasions) were imputed using a multiple imputation technique, as described by van Buuren et al. (1999), using stata 8 software (Stata Corporation, College Station, TX, USA). Comparisons between preoperative and postoperative visual acuities were made using the Wilcoxon signed rank test. Univariate analysis was performed with the χ2or Fisher’s exact test to determine which of the preoperative and intraoperative clinical variables were associated with a recurrent RRD or a poor visual outcome. A stepwise forward multiple logistic regression analysis was used for the following preoperative variables to determine the strongest predictors of a poor visual outcome using a probability for visual acuity < 0.1 or recurrent RRD: age older than 70 years; pseudophakia; reported days of visual field loss; type of retinal break; three or more breaks; cumulative break size more than three DD; extent of total detachment; presence of PVR B or C; and macular involvement. A p-value of ≤ 0.05 was considered significant.
A total of 436 eyes (263 men and 173 women, 60.3% versus 39.7%) of 424 patients were included in this study. These patients had a mean age of 58.2 years (range 15–90 years). Mean follow-up period was 50.6 months (range 6–120 months). Of the 436 operated eyes, 356 patients (81.7%) were aged 70 years or less and 80 (18.3%) were older than 70 years. There were 339 phakic eyes (77.8%) and 97 pseudophakic eyes (22.2%). Preoperatively, 252 eyes (57.8%) had a detached macula and 184 eyes (42.2%) had an attached macula.
Anatomical success after one SB procedure was achieved in 76.1% (n = 332) of eyes, and the retina was ultimately attached after multiple vitreoretinal procedures in 424 (97.2%) eyes. A Kaplan–Meier survival curve of time versus redetachment is shown in Fig. 1. Of the 104 eyes that developed redetachment after SB surgery, 100 eyes underwent subsequent vitreoretinal procedures to reattach the retina. In four of the 104 eyes, the patient and/or the surgeon decided to perform no further operations because of the bad prognosis for the eye or because of the poor physical condition of the patient. This decision was made in nine (2%) of the 436 eyes, including five eyes that underwent additional vitreoretinal procedures.
Of the 104 eyes (23.9%) that developed redetachment, data on how long after the SB surgery the redetachment occurred were available in 100 eyes, while data on the timing and cause of redetachment were available in 87 eyes (Table 1). Of all 100 eyes, redetachment occurred after a mean period of 54.4 weeks (median 12 weeks, range 0 days to 516 weeks). Within the first 3 months after the primary SB operation, 55 (12.6%) eyes had redetachment. At 6 months, 68/436 eyes (15.6%) had developed redetachment. At 1 year after the SB procedure, 80/436 eyes (18.3%) had developed a recurrent RRD. Twenty of the 100 eyes (4.6% of total operated eyes) developed redetachment more than 1 year after the SB operation. Details concerning the timing and causes of redetachment are shown in Table 1.
Table 1. Cause of redetachment after the first scleral buckling procedure during the follow-up period.
Time period after the first operation
Cause of redetachment
Total cumulative n (cumulative % of total)
Total n (cumulative % of redetachment and cause)
PVR/ traction n
Missed/ new breaks n
Persisting SRF n
Of the 104 eyes (23.9%) that developed redetachment during the study period, data on the time elapsed between SB surgery and redetachment occurrence were available in 100 eyes.
*Leakage is defined as the presence of accumulated subretinal fluid at the peripheral edge of the buckle because of lack of indentation.
†Other causes of redetachment such as removal of the scleral buckle in two patients after 1 year, or not noted or cause unknown.
6 months–1 year
> 1 year
We called the patients to ask whether their visual function was not worsened, and if they had noticed any signs or symptoms of a (re)detachment. With this method, we found that only one patient underwent retinal redetachment surgery in another hospital. Eight other patients who stated that their visual acuity had changed were seen in our hospital. However, none of these eyes had redetachment.
Twenty-nine of the 97 pseudophakic eyes developed redetachment (30%). No aphakic eyes were included in the current series. In three eyes no further operations were performed because of the bad prognosis for the eye or because of the poor physical condition of the patient. Twenty-six eyes underwent subsequent vitrectomies; in 20 eyes, one vitrectomy with gas tamponade (n = 6) or oil tamponade (n = 14) was performed. In six eyes two vitrectomies were performed: in five eyes vitrectomy was performed with oil tamponade, and one eye first underwent vitrectomy with gas, followed by a second vitrectomy with oil tamponade. In one eye a total of three vitrectomies with oil tamponade were performed to reattach the retina. Of these 26 eyes that underwent subsequent surgery, the retina was ultimately attached in 100% (26/26). However, two eyes with attached retinas were enucleated because of severe pain.
At the end of the follow-up, 14% of patients had a visual acuity of hand movement or worse. Two hundred and twenty-five eyes (52%) had 20/20 to 20/40 vision, 148 eyes (34%) had 20/50 to 20/200 vision and 63 eyes (14%) had a visual acuity less than 20/200. Thirty-seven of the group with visual acuity of 20/200 or less had redetachment (59%). In nine eyes (14%), the patients and/or the surgeon decided to perform no further operations because of the bad prognosis for the eye or because of the poor physical condition of the patient. In the subgroup of 63 eyes with a visual acuity of 20/200 or less, the causes of this low visual acuity were as follows: eight patients developed severe cataract; two patients had a vascular disease (one case of anterior ischaemic optic neuropathy and one case of central retinal vein occlusion); one patient developed a corneal decompensation; two patients developed amblyopia; one developed a phitisis bulbi; one had an epiretinal membrane; and one developed neovascular glaucoma. Because of the retrospective nature of the study and the fact that the referring ophthalmologists only provided us with visual acuity results, data on the cause of the low visual acuity in the other 37 eyes were not available.
Of all the preoperative, intraoperative and postoperative clinical variables that were evaluated to assess their odds ratio (OR) for a poor postoperative visual outcome by univariate analysis, a preoperative Snellen visual acuity of < 0.1, more than 7 days of reported visual field loss, more than three quadrants of retinal detachment, PVR B or C, macular involvement and redetachment were found to be significant risk factors (Table 2). After multivariate analysis, more than 7 days of reported visual field loss and recurrent retinal detachment were found to be significant risk factors for a poor visual outcome (OR 2.3, p = 0.006 and OR 5.0, p < 0.001, respectively).
Table 2. Clinical variables associated with a poor visual acuity (< 0.1 Snellen) at 12 months postoperative by univariate and multivariate analysis.
Of all the preoperative and intraoperative clinical variables that were evaluated to assess their OR for redetachment by univariate analysis, a cumulative break size of more than three DD, more than three quadrants of retinal detachment and PVR grade B/C were found to be significant risk factors (Table 3). After multivariate analysis, only the size of the break was significantly associated with the occurrence of redetachment (p = 0.009) (Table 3). Further stepwise forward multiple logistic regression analysis of these variables and the time of redetachment showed a significant association between a cumulative size of the break of more than three DD and early redetachment (p = 0.007, data not shown in Table 3).
Table 3. Clinical variables associated with redetachment by univariate and multivariate analysis.
Complications occurred during the SB procedure in 21 (4.8%) of the 436 eyes: five eyes (1.1%) had an inadvertent transscleral perforation; a haemorrhage occurred during the transscleral puncture to drain the subretinal fluid in five (1.1%) eyes; five (1.1%) eyes had a retinal incarceration; and in six (1.4%) eyes the surgery was complicated by a choroidal detachment. Seven of these 21 eyes (33.3%) developed redetachment during the follow-up period. Statistical analysis did not reveal a significant association between these complications and the incidence of redetachment.
Postoperatively, the scleral buckle or encircling band had to be removed in six eyes. This was because of intolerable pain in four eyes, infection in one eye and exposure of the buckle in one eye. Two (33%) of these six eyes developed redetachment during follow-up.
After more than 6 and 12 months of follow-up, 32 eyes (7%) and 20 eyes (5%), respectively, developed redetachment. Multivariate regression analysis showed that recurrent redetachment and more than 7 days of visual field loss were significant predictors for a poor postoperative visual outcome at 12 months. A cumulative size of the tear of more than three DD was a significant predictor of recurrent RRD. The predictors of poor visual outcome found in the current study correspond to the risk factors reported in earlier studies (Boberg-Ans et al. 2006; Burton & Lambert 1978; Diederen et al. 2007; Ross & Kozy 1998); however, they differ from the findings of a recently published study by Salicone et al. (2006b) stating that macular involvement was the main risk factor for a poor visual outcome after SB surgery. The discrepancy between the other studies and Salicone et al.’s may be caused by differences in the amount of follow-up time after SB surgery that had elapsed when the visual acuity data were obtained. Several studies indicate that visual recovery after SB surgery increases over time (Abouzeid & Wolfensberger 2006; Liem et al. 1994; Sasoh et al. 2005; Wolfensberger & Gonvers 2002). Salicone et al. (2006b) based their conclusions mainly on postoperative visual acuity data obtained after a follow-up of 2 months, whereas in our study conclusions were based on visual acuity data obtained at 12 months.
Following our patients for a longer time period revealed that ultimately 23.9% of the eyes developed redetachment. Six months after SB surgery, 8.3% developed ‘late’ redetachment, and after 12 months another 4.6% of the eyes developed a recurrent RRD. Foster & Meyers (2002 found recurrent RRD after 1 year in 2.2% (10/453) of eyes. Few studies have been published with a relatively long mean postoperative follow-up period (Table 4) (Foster & Meyers 2002; Greven et al. 1999; Kreissig et al. 1992, 1995; Sasoh et al. 2005; Schwartz et al. 2002). Schwartz et al. (2002) had a follow-up period of 20 yearsand only included in their analysis those patients who responded to a follow-up call after 20 years. Moreover, their data were not retrieved from referring ophthalmologists or by interviewing patients, as in the current study. The rate of late redetachment as reported by Foster & Meyers, Greven et al. and Kreissig et al. ranged from 2.2% to 6.5%, which is comparable to our study (Foster & Meyers 2002; Greven et al. 1999; Kreissig et al. 1992) (Table 4). In the current study, we identified a cumulative retinal break size of more than three DD as a significant risk factor for recurrent RRD. These findings are in agreement with earlier studies (Grizzard et al. 1994; Hassan et al. 2002; La Heij et al. 2000; Ross & Kozy 1998; Salicone et al. 2006b). In our study, a large retinal break size was also a significant risk factor for early redetachment (within 6 months of the SB procedure). Kreissig et al. (1992, 1995) also made a distinction between early and late redetachment. Similarly to our series, these authors also observed that early redetachment (up to 6 months after SB surgery) was caused mainly by PVR, while in half of the cases a late redetachment was caused by new breaks. Although PVR is also observed in patients with late recurrent RRDs, we believe that vitreous base traction is probably the main causative factor in these cases and that the associated PVR may well be a secondary phenomenon (Foster & Meyers 2002; Kreissig et al. 1992).
Table 4. Summary of studies on patients treated with scleral buckling for rhegmatogenous retinal detachment with a long follow-up period.
Number of patients (n)
FUP: mean (range)
Detached after one surgery
Detached after 1 year
FUP, follow-up period; SB, scleral buckle; NN, not noted.
20/20–20/40 in 52%, 20/50–20/200 in 34%, 20/200>14%.
Operated by the same surgeon
Sasoh et al.*
10 years (30.9% of total)
20/15–20/20 76% in macula-on group, 20/15–20/20 50% in macula-off group
Only patients with 10 years FUP; number of surgeons NN
Foster et al.†
8.5 years (69–140 months); only 10 late redetachment
Study of 10 eyes with late redetachment; more than one surgeon
Schwartz et al.‡
20/40 median 20 years visual acuity
More than one surgeon
Greven et al.§
6–78 months (mean 29 months)
20/20 in 32%, 20/25 in 18%, 20/30 in 25%, 20/40 in 18%, 20/50 in 7%
Asymptomatic retinal detachment;operated by two surgeons
Kreissig et al.¶
20/30 in group with macula attached, 20/100 in group with macula partially detached and 20/400 in group with complete detachment
Number of surgeons NN
Kreissig et al.**
Amount of surgeon NN
The occurrence of late redetachment is an argument in favour of following patients for at least 6 months after the SB procedure. Also, visual acuity may still improve up to several months after SB surgery. Using foveal densitometry, Liem et al. (1994) showed that recovery of foveal cone photopigments after macular detachment may take several months. In addition, subfoveal fluid may persist subclinically for several months following the SB procedure, which may explain slow visual recovery in some patients (Benson et al. 2006; Hagimura et al. 2002; Lecleire-Collet et al. 2006). On the other hand, a disadvantage of a longer postoperative follow-up period may be that visual acuity can decrease in elderly patients because of cataract or macular problems. Therefore, visual acuity results were analysed in the current study after a fixed time period of 12 months following the primary operation.
Le Rouic et al. (2003) investigated the indication of buckle removal. These authors found an overall redetachment rate of 8.8% in a series of 90 eyes from which they removed scleral buckles. They reported redetachment rates of 16% in eyes from which a solid silicone buckle was removed and 7% in eyes where a silicone sponge was removed (Le Rouic et al. 2003). In the current study, we report on only six eyes that underwent SB removal, which is in contrast to the larger series of Le Rouic et al. (2003). Two of these six eyes developed redetachment after the removal of a silicone sponge. In one of these two eyes redetachment was caused by a perforation that occurred during removal in an eye with a staphyloma. No clear cause was found for the other patient who developed redetachment.
PPV has gained widespread popularity in the treatment of RRD and several studies have compared SB and/or PPV in eyes with a fresh RRD (Afrashi et al. 2004; Ahmadieh et al. 2005; Heimann et al. 2005; Salicone et al. 2006a; Saw et al. 2006; Sharma et al. 2005). For pseudophakic retinal detachment, a meta-analysis by Arya et al. (2006) revealed that PPV is more likely to achieve a favourable anatomical and visual outcome than conventional SB alone. However, evidence-based analysis of uncomplicated RRD did not show a difference in retinal reattachment or final VA between the SB and PPV groups (Saw et al. 2006). An argument in favour of performing a PPV instead of SB surgery is the fact that peripheral breaks may be missed preoperatively, especially in pseudophakic eyes (Christensen & Villumsen 2005). In the current study, in which 22% of the eyes were pseudophakic, missing breaks were not significantly associated with a poor visual or functional outcome.
In conclusion, conventional SB surgery is a reliable procedure in selected eyes with a primary RRD. A PPV should be considered in eyes with a retinal tear with a cumulative size of more than three DD. Because a high percentage of eyes may develop redetachment after 6 months, and because further visual recovery over time is still possible, a longer follow-up period after SB surgery should be taken into account in studies that aim to predict the impact of macular detachment on visual outcome or recovery.