Article Text

Download PDFPDF

Original article
Surgical failure following primary retinal detachment surgery by vitrectomy: risk factors and functional outcomes
  1. Louisa Wickham1,
  2. Gregory O Ho-Yen1,
  3. Catey Bunce1,
  4. David Wong2,
  5. David G Charteris1
  1. 1Vitreoretinal Department, Moorfields Eye Hospital, London, UK
  2. 2Vitreoretinal Department, The Royal Liverpool University Hospital, Liverpool, UK
  1. Correspondence to Louisa Wickham, Vitreoretinal Department, Moorfields Eye Hospital, City Road, London EC1V 2PD, UK; louisa.wickham{at}


Aim To identify preoperative features associated with surgical failure following vitrectomy using data collected in a large, prospective randomised controlled trial. Outcomes of patients who redetached were then examined in more detail.

Methods 615 patients were analysed as part of an randomised controlled trial investigating the use of 5-fluorouracil and low-molecular-weight heparin. Treatment status had no effect on success rates and did not therefore form part of the analyses. Failure was defined as retinal redetachment within 6 months of primary vitrectomy. Univariate logistic regression analysis was used to assess association between failure and putative risk factors (age, pathological myopia, intraocular pressure, vitreous haemorrhage, previous lens extraction, uveitis, number of retinal quadrants detached, number and distribution of retinal breaks, and grade C proliferative vitreoretinopathy (PVR)). Additional characteristics of patients were then elucidated including number of operations required to achieve retinal reattachment, surgical techniques used and final logMAR visual acuity.

Results 96 patients (15.6%) redetached following surgery, and 37 failed due to PVR. Surgical failure was associated with number of retinal quadrants detached (OR per increase, 1.69 (1.33 to 2.15) p<0.001) and grade C PVR (OR 3.98 (1.47 to 10.73) p=0.006). Inferior breaks were not identified as a risk factor (p=0.602). Repeat retinal detachment surgery showed a trend towards reduced visual acuity at 6 months providing PVR did not develop. PVR resulted in a significant deterioration in visual acuity.

Conclusions The extent of retinal detachment and preoperative PVR are risk factors for surgical failure following vitrectomy for primary retinal detachment. PVR was again confirmed as the major factor influencing visual outcomes.

  • Retinal detachment
  • vitrectomy
  • failure
  • proliferative vitreoretinopathy
  • risk factors
  • retina
  • treatment surgery

Statistics from


Reported primary success rates for rhegmatogenous retinal detachment repair range from 64 to 91%.1–3 Retinal redetachment following initial surgery may occur secondary to development of proliferative vitreoretinopathy (PVR), failure of treatment, for example missed or partially treated breaks, or development of new retinal breaks. The consequences of failure as a result of PVR on visual function are well documented, with only 11–25% of patients achieving a visual acuity of 20/100,4 5 but outcomes in patients without PVR are less well described. Closer examination of patients who require further surgery may help to determine factors associated with an increased risk of failure as well as the effect of redetachment on visual function.

Consistent evidence regarding the significance of risk factors for failure of primary retinal detachment repair is difficult to find. Studies investigating risk factors show a number of differences relating to the different study designs, inclusion criteria and the surgical techniques employed. Furthermore the majority of studies are retrospective. Reported preoperative characteristics that may be associated with an increased risk of surgical failure include duration of symptoms, extent of retinal detachment and involvement of inferior quadrants, an absence of detectable retinal breaks, high myopia and hypotony.6–8 In addition, risk factors associated with the development of PVR include aphakia, uveitis, vitreous haemorrhage, presentation with PVR, raised intraocular pressure and increasing age.9–11

The identification at initial presentation of patients at high risk of failure could potentially allow surgeons to tailor their management to the individual patient. This analysis aimed to identify preoperative features of retinal detachment associated with surgical failure using data collected in a large prospective randomised controlled trial investigating the use of 5-fluorouracil and low-molecular-weight heparin in patients undergoing primary vitrectomy. In addition, the outcomes of patients who redetached following their initial surgery are examined in more detail.


Ethics approval was granted by the Moorfields Local Research Ethics Committee (CHAD 1005) and adhered to the tenets of the Helsinki declaration. Data were handled in accordance with the Data Protection Act and were recorded in a Microsoft Access database.

Six hundred and forty-one patients were recruited from two specialised vitreoretinal units between February 2001 and January 2005 as described previously.12 In brief, patients presenting with a primary rhegmatogenous retinal detachment requiring repair by vitrectomy and intraocular gas tamponade were recruited to the trial and randomly allocated to the treatment or placebo group. The leading indications for primary vitrectomy surgery included retinal detachment with a posterior vitreous detachment owing to superior posterior, multiple, or tractional retinal breaks, pseudophakia and/or inadequate view of the fundus—for example secondary to vitreous haemorrhage. Eyes without a posterior vitreous separation and/or limited peripheral detachment would normally undergo scleral buckling surgery. The primary outcome measure for this trial was retinal reattachment following primary vitrectomy without any reoperations at 6 months. Patients were classed as a ‘failure’ if retinal redetachment was observed within the 6-month follow-up period. Treatment status (with or without adjuncts) was found to have no effect on success rates and did not therefore form part of the analyses.12

Additional information on the subsequent management and surgical outcomes of patients who failed primary surgery were collected on patients recruited at Moorfields Eye Hospital (553 of a total of 641 patients). A subanalysis was undertaken on this patient cohort to further elucidate the effects of surgical failure on final visual acuity.

Statistical analysis

The risk factors analysed were those previously demonstrated to be associated with retinal detachment failure or PVR development and which are in addition readily available at the preoperative situation.10 11 13 The analysis was focused on those features which could be measured precisely. Factors considered were age, duration of symptoms, high myopia (defined as a spherical equivalent of −6 dioptres or more), preoperative intraocular pressure, preoperative vitreous haemorrhage, previous lens extraction, preoperative uveitis, number of retinal quadrants detached, number and distribution of retinal breaks, and presence of PVR (grade C) according to the Retinal Society classification. Risk factors were cross-tabulated against failure and univariate logistic regression analysis conducted to assess evidence of any association between each putative risk factor and failure.


Of the 641 patients recruited, 615 patients completed the trial, and 96 (15.6%) patients redetached following primary retinal detachment surgery.12

Univariate logistic regression analysis showed evidence of an association between surgical failure and number of retinal quadrants detached and the presence of grade C PVR (table 1).

Table 1

Univariate logistic regression analysis of possible risk factors for anatomical failure following primary surgery

The position of retinal breaks at presentation was recorded in 512 patients. Inferior retinal breaks, in detached or flat retina, were present in 204 patients at presentation and in 31 of the failures; however a univariate logistic regression analysis did not identify inferior breaks as a risk factor for failure in either detached (p=0.602) or attached retina (p=0.656). In 16 patients, inferior breaks were also treated with a segmental scleral buckle at the time of vitrectomy, but this did not appear to increase the success of surgery (table 2), although the numbers are small.

Table 2

Failure rate in patients presenting with inferior retinal breaks

Pseudophakia was not associated with failure of retinal detachment surgery (p=0.093). The number of pseudophakic patients presenting with a retinal detachment and their subsequent surgical success rates are given in table 3.

Table 3

Surgical outcome according to lens status

Of the 553 patients recruited at Moorfields Eye Hospital, 79 patients (14%) redetached following primary retinal detachment surgery, and of these 31 failed due to PVR. A further seven patients subsequently developed PVR following their second operation for retinal detachment (figure 1). Four patients elected not to have further surgery following initial failure. The majority of patients who failed following surgery did so within the first 8 weeks, and this appeared to be unrelated to whether PVR was the cause of failure (figure 2). The presence of PVR at the time of failures was associated with the requirement for more than one reoperation (χ2=4.02, p=0.045); the total number of retinal detachment operations required to achieve anatomical success is summarised in figure 1.

Figure 1

Flow chart of cause of failure and subsequent management in patients who failed primary retinal detachment surgery at Moorfields Eye Hospital (MEH). PVR, proliferative vitreoretinopathy.

Figure 2

Time to failure following primary retinal detachment repair. PVR, proliferative vitreoretinopathy.

Surgical techniques used following primary failure to achieve retinal reattachment are shown in table 4. In patients who presented with grade C PVR, a number of additional surgical techniques were used to reattach the retina, including heavy liquid (N=11), membrane peeling (N=13) and retinectomy (N=11). Silicone oil was used as a tamponade agent in 25 (86%) patients when PVR was present compared with only seven (15%) patients in whom PVR was absent (table 4).

Table 4

Techniques used to achieve retinal reattachment following failure of primary surgery

In patients who had successful primary surgery, the presence of PVR at presentation resulted in a difference in the median vision of 2 logMAR lines (median logMAR visual acuity of 0.3 in the group without PVR versus 0.7 in the PVR group) at 6 months (figure 3). In the failure group, repeat retinal detachment surgery showed only a minimal trend towards reduced visual acuity at 6 months providing that PVR did not develop. With PVR development, however, repeat reattachment procedures were associated with significant deterioration in visual acuity following each procedure (figure 3). Failure of the second surgery was observed in 27 patients, and of these 21 had silicone oil injected at their next operation. In 13 patients, silicone oil was still in situ 6 months following their last surgery. In this group of patients who required three or more procedures to achieve anatomical success, the visual outcomes were limited regardless of whether silicone oil had been removed (figure 3).

Figure 3

Effect of number of retinal detachment operations and the development of proliferative vitreoretinopathy (PVR) on visual acuity at 6 months. LogMAR visual acuities are displayed as median and IQR (IQ1, IQ3). In the group that underwent three operations to achieve retinal reattachment following development of PVR, acuities are subdivided into those with silicone oil in situ and those without silicone oil. so, silicone oil.

In macula involving retinal redetachments, initial surgical success resulted in a reduction in the median visual acuity of 1 line on the logMAR chart at 6 months compared with macula-sparing retinal redetachments. If the macular subsequently detached again, there was a much greater impact on final acuity at 6 months (figure 4).

Figure 4

Effect of recurrence of macular involving retinal redetachment on final visual acuity at 6 months. Visual acuities are presented as median and IQR (IQ1, IQ3). Five patients had three macular-involving retinal redetachments; in this group, the median and IQR were the same (logMAR acuity 1.8).


In this study, using univariate analysis, we identified risk factors for redetachment following vitrectomy and gas tamponade that have been noted previously such as the number of retinal quadrants involved and presence of preoperative grade C PVR.6–8 However, other previously reported risk factors such as the presence of inferior retinal breaks were not found to be associated with surgical failure.14

Retinal detachment extent can be considered a surrogate marker for detachment duration (which may be reported inaccurately in many cases). It is notable, however, that detachment extent was demonstrated to be a risk factor independent of reported duration or PVR presence. Nevertheless, such cases may have greater retinal glial upregulation, RPE cell dispersion and may have PVR development less than grade C.15 Modifying surgical management may therefore be appropriate—for example, with longer acting intraocular gases.

Previous lens extraction was not associated with surgical failure in this study, and a number of other studies have also shown similar success rates in pseudophakic and phakic patients with or without the use of additional scleral buckling.8 16 Halberstadt, however, did show that pseudophakia was associated with worse outcomes in retinal detachments involving three or more quadrants,16 and previous lens extraction has also been shown to be a risk factor in the development of PVR.11 Previously reported associations between lens surgery and surgical failure may be related to the observation that pseudophakic and aphakic patients often develop small, sometimes multiple, anterior breaks. Although the use of a scleral buckle with encircling band or improved visualisation of anterior retina with modern viewing systems may address this, it is still possible that some small anterior breaks may be missed at the time of surgery or that pseudophakic eyes have a greater propensity to develop new retinal breaks postoperatively.

In this study, the presence of inferior breaks did not increase the risk of failure following vitrectomy and gas tamponade. Furthermore, the addition of a scleral buckle at the time of surgery did not appear to decrease the risk of subsequent failure, although the number of cases treated by vitrectomy and a buckle is relatively small (N=16). This supports the findings of other studies which have also failed to show any advantage in combining vitrectomy with scleral buckling over vitrectomy alone in the treatment of inferior retinal breaks.17

This study demonstrates that if PVR was absent at the time of failure, visual acuity was relatively well preserved, even in those who underwent three operations. However, if PVR developed, visual outcomes were significantly worse compared with those without PVR, and successive redetachment appeared to have a much greater impact on visual function. Poor visual outcomes following the development of PVR are well reported,4 5 and are likely to be due to neural and non-neural retinal changes as observed in retinectomy specimens from patients with PVR.15 Poor final visual acuity may also be related, in some cases, to other sequelae of the PVR process and its management—for example, postoperative hypotony, high intraocular pressure following use of silicone oil tamponade and the well-documented loss of vision seen in some eyes after silicone oil removal.18

Macula redetachment appeared to have a cumulative effect in reducing final visual acuity at 6 months, with a single macula detachment appearing to have less effect on final visual acuity than subsequent detachments (figure 4). The potential for neural recovery at the macula following retinal detachment has been shown in adult rhesus monkeys, although persistent outer segment abnormalities were frequently present in otherwise well-regenerated areas.19 It has also been demonstrated experimentally that rapid retinal reattachment can stop and reverse the retinal degeneration associated with detachment.20 Clinically, visual recovery following repair of a macula involving retinal detachment is thought to be dependent on a number of factors including the degree of retinal elevation, the duration of detachment, patient age and preoperative visual acuity.21 22 The cumulative effect of macula detachment would suggest that close observation of patients in the postoperative period, as the gas tamponade resorbs, could allow early intervention and potentially improve clinical outcomes.

In this study, final visual acuity was measured 6 months following the last retinal detachment procedure; however, previous work has demonstrated that visual recovery following reattachment may continue for many years. Kusaka documented an improvement in visual acuity of two lines or more in 53% (17/32) patients over 5 years when compared with the visual acuity measured 3 months following surgery. In his study, long-term improvement was correlated with younger age, myopia of less than 5 dioptres, and duration of detachment of less than 30 days.23 Some of this visual recovery may be explained by recent reports that subclinical subretinal fluid, demonstrated on OCT, may persist for many months following retinal reattachment and is associated with poor visual acuity postoperatively.24 Visual improvement may also be due to the gradual repair of the cellular pathology associated with neural retinal detachment—for example, outer segment regeneration and neural remodelling.25 The additive effects of repeated macula detachment may contribute to a diminishing capacity for neural recovery.

This study is limited by the factors available for precise measurement. It could not evaluate those factors that are difficult to measure such as surgical skill. However, primary success rates achieved in this study were greater than those reported in the National Survey by vitreoretinal specialists suggesting that surgery was performed by competent surgeons and that grade of surgeon was not the cause of surgical failure.26 A further subgroup analysis on the relative merits of different surgical techniques at the time of redo surgery was also not possible owing to the small numbers in each group.

In this group of patients who failed primary retinal detachment surgery, we observed an association between the extent of retinal detachment and the preoperative presence of grade C PVR and surgical failure. We found no evidence to suggest that inferior retinal breaks influenced surgical success, or that patients with inferior retinal breaks require a different surgical approach. We also demonstrated that, in the absence of PVR, good visual outcomes may still be achieved despite initial involvement of the macula or macular redetachment. The development of PVR was again confirmed as the major factor influencing visual outcomes.



  • Linked article 190314.

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the Moorfields Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles