Article Text

Download PDFPDF
Medium-term outcomes of safe surgery system trabeculectomies
  1. J Gale1,
  2. A P Wells1,2
  1. 1
    Ophthalmology Department, Capital & Coast District Health Board, Wellington, New Zealand
  2. 2
    Capital Eye Specialists, Wellington, New Zealand
  1. Dr A P Wells, Ophthalmology Department, Wellington Hospital, Private Bag 7902, Wellington, New Zealand; twells{at}


Aim: To assess the safety and success of Safe Surgery System trabeculectomy beyond 3 years.

Methods: Consecutive case series of 39 eyes in 32 patients. Trabeculectomy was performed using fornix-based conjunctival flap, standard trabeculectomy punch, adjustable scleral flap sutures and antimetabolite treatment. Primary outcome of surgical failure was defined by two criteria: (A) need for further surgery, glaucoma medications or an intraocular pressure (IOP) >21 mm Hg during >10% of follow-up; or (B) IOP >15 mm Hg for >10% follow-up. A relatively aggressive regime of bleb needling and subconjunctival injections was used postoperatively in an attempt to reduce bleb fibrosis and failure. The mean follow-up was 42 months (range 25–55).

Results: The rate of surgical failure was 4.4% per eye-year for criterion A, and 8.0% per eye-year for criterion B. Complications were few and compared favourably with other published series.

Conclusion: The Safe Surgery System for trabeculectomy provides excellent IOP control both during the operation and in the short and medium term postoperatively, with few complications or surgical failures.

Statistics from

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.

The Safe Surgery System (SSS) for trabeculectomy is a relatively new technique developed at Moorfields Eye Hospital in the late 1990s, designed to mitigate risks of conventional trabeculectomy, and to maximise control of intraocular pressure (IOP).1 2 Standard approaches to trabeculectomy may result in variable filtration, have limited options for IOP control postoperatively and have complications related to frequent intraoperative hypotony. The SSS involves fornix-based conjunctival flaps, adjustable sutures which provide precise control of immediate postoperative IOP;3 minimisation of intraoperative hypotony, and a standardised punch to control trabeculectomy aperture.1 The short-term outcomes of SSS trabeculectomies that have been reported to date show precise IOP control postoperatively, few complications and excellent pressure control to 1 year.2 3

Longer-term results from SSS trabeculectomies have not yet been reported. Several long-term cohorts of conventional trabeculectomies with antimetabolite augmentation have been reported.47 Risk factors for surgical failure include Afro-Caribbean ethnicity, diabetes mellitus, high preoperative pressures, argon laser trabeculoplasty and some medications, and these patients benefit from antimetabolite use.4 8 9

Here we report the outcomes of 39 consecutive SSS trabeculectomies at 3 to 4 years of follow-up, for the first time. During follow-up, a relatively aggressive regime of subconjunctival injections with or without needling bleb revision was used to maintain failing blebs.


Patient selection

This was a prospective, consecutive, non-comparative case series of patients undergoing trabeculectomy alone for any indication between July 2002 and October 2003 under the glaucoma team in Wellington Hospital, New Zealand. The series included 40 trabeculectomies in eyes of 33 patients, but data from one patient could not be retrieved (n = 32 patients, 39 trabeculectomies). Patients having trabeculectomy combined with cataract surgery were excluded. Informed consent was obtained for all patients preoperatively. The Central Regional Ethics Committee approved the study.

Surgical technique

Trabeculectomy was performed by the glaucoma team consultant (APW), fellow or registrar with all procedures under direct consultant supervision. The approach was based on the previously described technique.3 10 Corneal traction suture was placed, and fornix-based conjunctival flap created. Sclera was cauterised and cleaned with a Tooke knife. Antimetabolites were used in every case, based on estimated risk of failure: either 5-fluorouracil 50 mg/ml or mitomycin C between 0.2 and 0.4 mg/ml. Antimetabolites were applied to a large posterior area of bared sclera with eight to ten 1×1×3 mm soaked sponges for 3–5 min, with the conjunctival wound edge kept free of antimetabolite by the assistant. The area was then washed thoroughly with 50 ml of balanced salt solution. A 3×4 mm half-thickness scleral flap was created hinged at the edge of clear cornea. Two four-throw adjustable 10-0 nylon sutures were preplaced tightly at the corners of the scleral flap as previously described.2 3 Paracentesis was made, and trabeculectomy was performed with a 0.25 mm Khaw punch (Duckworth and Kent, Baldock, UK), followed by peripheral iridectomy. The 10-0 nylon sutures were tightened, and if IOP could not be maintained intraoperatively, additional releasable sutures were placed on the anterior limbs of the scleral flap. The conjunctiva was closed with a 10-0 nylon purse-string suture on either side, and additional central mattress suture.

Postoperative management and follow-up

On the first postoperative day, all patients with an IOP greater than 15 mm Hg had transconjunctival adjustment of scleral sutures using Khaw suture adjustment forceps (Duckworth and Kent) to bring their IOP to 10–15 mm Hg, as previously described.3 All patients used guttae prednisolone acetate 1% six times daily for at least 4 weeks, and guttae chloramphenicol 0.5% four times daily for 2 weeks. Patients were followed up at day 1, weeks 1 and 3, months 1, 3, 6, 9 and 12, then approximately every 6 months, with additional visits as required.

At every visit, blebs were assessed for function and appearance. If adverse bleb features such as vascularity or fibrosis were present, subconjunctival injections of 5-fluorouracil 50 mg/mL and/or dexamethasone 4 mg/mL were given. If vascularity or fibrosis was accompanied by an unacceptable or increasing IOP, shallow or encapsulated blebs, then needling revision was performed with simultaneous subconjunctival injection of 5-fluorouracil, dexamethasone and viscoelastic (either at the slit lamp or, if likely to be poorly tolerated, in theatre under regional or general anaesthetic). In some cases, needling was repeated as often as monthly if signs of bleb failure and increased IOP persisted. Following either injection or needling, patients would take a tapering regimen of guttae prednisolone acetate 1% for several weeks.


The primary outcome measure of surgical failure was defined by the need for further surgery or glaucoma medications to maintain IOP control. Intraocular pressure control was defined by two criteria: (A) IOP consistently ⩽21 mm Hg; and (B) IOP consistently ⩽15 mm Hg. The lower IOP criterion of 15 mm Hg is often the target sought with a trabeculectomy, and none of our patients showed progression at this level. When IOP exceeded either criterion, it was still considered a surgical success if IOP was below criterion for more than 90% of follow-up months. Secondary outcomes that were recorded at each visit were IOP, visual acuity, medications, complications, interventions (including suture adjustment, subconjuntival injections or needling revision of failing blebs) and other ocular surgery.

Statistical methods

Microsoft Excel and SPSS software were used to organise and calculate data and create figures. Event rates were used to express outcomes, as these are more meaningful than simple proportions in case series with variable follow-up.11 The Student t test was used to determine statistically significant differences (α = 0.05).


Patient characteristics and follow-up

Fifteen of the 32 patients in the study were male. Their age and ethnicity are summarised in table 1. Eighteen of the 39 trabeculectomies were performed on right eyes, and preoperative details are summarised in table 2.

Table 1 Patient characteristics (n = 32)
Table 2 Baseline ocular characteristics (n = 39)

One patient died 25 months following her operation. Seven patients were followed up in other centres, and summarised follow-up information was collected by correspondence. The average duration of follow-up was 42 months (range 25 to 55) for 39 trabeculectomies, with a total of 136 eye-years of follow-up.

Trabeculectomy outcomes

Mean visual acuity (logMAR) at final follow-up was 0.50 (SD 0.75) and compared with baseline acuity was not significantly altered (p = 0.448). At baseline, four eyes had logMAR acuity ⩾1.0 (ie, 6/60 or worse) and 29 eyes had acuity ⩽0.18 (ie, 6/12 or better). At final follow-up, five eyes had acuity ⩾1.0, and 25 eyes had acuity ⩽0.18.

There were six eyes classified as surgical failures according to criterion A, which equated to a failure rate of 4.4% per eye-year of follow-up. Three patients failed because they required tube implants at 25, 36 and 41 months post-trabeculectomy, due to uncontrolled intraocular pressure despite needling revision. One patient met our failure criteria because she was restarted on guttae timolol 0.25% at 38 months post-trabeculectomy to maintain IOP below 15 mm Hg, but she never had IOP >16 mm Hg. Two further patients failed because they had pressure rises over 21 mm Hg in 12% of follow-up months, but were both controlled with needling bleb revision.

An additional five trabeculectomies were surgical failures according to criterion B, which was a total of 11 of 39 trabeculectomies or nine of 32 patients, and equated to a failure rate of 8.1% per eye-year of follow-up. These additional five patients had IOP >15 mm Hg in 10.5% to 39.0% of follow-up months, but all were controlled eventually with needling bleb revision, and all had acceptable IOP and functioning blebs before the end of follow-up.

The postoperative course of interventions, complications and failures are shown in table 3.

Table 3 Postoperative course (n = 39)

A low IOP of ⩽5 mm Hg occurred after 2 weeks in six eyes, and most occurred more than 6 months postoperatively. Two eyes had an IOP of 5 mm Hg within a week following needling intervention. The only complication of low IOP was a single transient choroidal detachment which resolved without consequence.

The intraocular pressure outcomes of the patients without tube implants are summarised in fig 1. The immediate postoperative IOP control of these patients has been reported previously.3 Patients with tube implants were excluded, and of the remaining patients, the mean intraocular pressure at 3 years was 12.4 (SD 3.2, range 8–24), and at final follow-up was 11.8 (SD 2.7, range 6–18).

Figure 1 Intraocular pressure during follow-up. D1, first postoperative day after suture adjustment if required; M6 to M54, months post-trabeculectomy; Pre, preoperative.


This case series provides the first data on outcomes of SSS trabeculectomies at 3 years and beyond. The data illustrate that the adapted surgical technique, combined with aggressive postoperative management, can result in low failure rates (4.4% per year by criterion A), very low complication rates and excellent IOP control. The majority of eyes (74%) had an IOP of 15 mm Hg or less for more than 90% of their follow-up without any glaucoma medications.

Within the scope of the study design, weaknesses included inclusion of high-risk patients, secondary glaucoma and three redo trabeculectomies, which have higher failure rates. On the other hand, most of the patients in the series had a low risk of failure. The study design cannot establish whether our consecutive case series was a representative sample of trabeculectomy patients in our community, and cannot determine the efficacy of bleb interventions for preventing surgical failure as compared with conservative treatment.

The longest previous follow-up on safe trabeculectomies was 1 year, when 91% of eyes had IOP <18 mm Hg and 61.4% <14 mm Hg at the 12-month visit, but glaucoma medications were restarted in 7%.2 In two large case series of conventional trabeculectomy with mitomycin C, 62% of phakic patients and 67% of pseudophakic patients maintained an IOP ⩽18 mm Hg at 3 years or were using two fewer medications than preoperatively to maintain this.4 5 At 2 years postoperatively, Scott et al reported 74.6% success for conventional trabeculectomies according to equivalent criteria to criterion A used in the present study, and 69.3% success according to the present criterion B.12 In a large cohort of trabeculectomies with up to 15 years of follow-up, 85% were reported to have maintained IOP under 21 mm Hg (but this included 20% of patients restarted on medical treatment).13 Using criterion B of the present study, Shigeeda reported 53% success at 5 years, and 45% success at 8 years.7 Our results compare favourably with all of these reports (87% at 3 years according to criterion A and 74% at 3 years according to criterion B).

All three patients who required tube implants had recognised preoperative risk factors for failure and received high-dose mitomycin C intraoperatively.8 9 One with primary open-angle glaucoma (POAG) had two previous laser trabeculoplasties and was minimally responsive to multiple medications, the second with POAG was of Afro-Caribbean origin and also resistant to multiple medications, and the third had iridocorneal endothelial syndrome. The patient who restarted timolol had POAG with low to moderate preoperative risk, and received intraoperative 5-fluorouracil. The seven other eyes that failed based on IOP during follow-up but who were controlled with needling had a range of risk factors for failure: two had secondary glaucoma, four had previous laser trabeculoplasties, one had a previous failed trabeculectomy in the contralateral eye, and two had previous uveitis.

Complications in the present series were few, and our results compare favourably to other series.2 13 After the initial 2 weeks, an IOP ⩽5 mm Hg occurred in six eyes, which in all cases resolved spontaneously without significant complications. Hypotony, defined as low IOP with secondary effects, occurred in only one eye with a transient choroidal detachment. There was no blebitis or endophthalmitis despite frequent postoperative needling and antimetabolite use.

When trabeculectomy blebs begin to fibrose or fail, management can either be to restart glaucoma medications and accept the imminent bleb failure or to resist bleb failure with antimetabolite injections or needling.2 Needling revision is a common technique used to salvage failing trabeculectomy blebs and restore function by allowing aqueous drainage beneath new areas of conjunctiva.14 15 When compared with acceptance of bleb failure and restarting glaucoma medications, needling has been described by some authors as having inferior immediate IOP control.15 However, the aims of needling are not immediate IOP control but to avoid bleb failure, bleb-related complications, re-operation, to reduce dependence on topical medications long term, and overall to enhance long-term success. One series of needling interventions reported that needling could result in IOP control without medications for 75% at 1 year and 52% at 3 years.14 Needling is repeatable, so the success could be increased with multiple attempts.

This series received relatively aggressive anti-fibrosis management of blebs, with the aim to create large, diffuse, free-draining blebs with normal vascularity. Using subconjunctival 5-fluorouracil and dexamethasone to reduce fibroblast activity, and needling to re-establish flow where fibrosis threatened to cause bleb failure, 36 of 39 eyes (92%) maintained functional blebs. Ocular hypotensive medications were avoided in case they reduced aqueous flow through the bleb, or caused conjunctival inflammation or hypervascularity. This represents a treatment philosophy for trabeculectomy patients: changing the focus from month-to-month IOP, to a focus on long-term bleb morphology and attempting to intervene on bleb signs before an anticipated IOP rise. With this approach, the majority (69%) of patients received at least one postoperative subconjunctival injection, and three patients (7.7%) received more than nine injections (one of whom later required tube surgery). Similarly, 43.6% received needling revision, and six patients (15.4%) received more than two (two of whom later required tube surgery). In patients who tolerate injections and needling at the slit lamp, this is easily undertaken in the clinic, but for some patients, needling can require a local or general anaesthetic.

The trabeculectomy technique is evolving, and the SSS technique has shown benefits in short-term IOP control and decreased perioperative risks and complications.2 3 Medium-term results from SSS trabeculectomies reported in this series demonstrate stable long-term IOP control and a low failure rate.



  • Funding: This work was supported by the Capital Vision Research Trust, but we received no direct funding for this research, and cannot identify any financial competing interests.

  • Competing interests: None.

  • Ethics approval: The Central Regional Ethics Committee approved the study.

  • Patient consent: Obtained.