Article Text

Download PDFPDF

The silent enemy: a review of cataract in relation to glaucoma and trabeculectomy surgery
  1. Rashmi G Mathew,
  2. Ian E Murdoch
  1. Moorfields Eye Hospital, London, UK
  1. Correspondence to Rashmi G Mathew, Moorfields Eye Hospital, City Road, London EC1V 2PD, UK; rashmi.mathew{at}


Glaucoma is one of the leading causes of blindness worldwide. In the UK, the estimated prevalence of open angle glaucoma ranges from 0.3% in the 40s to 3.3% in the 70s. It is one of the leading causes for blind registration in the UK, second only to macular degeneration. Trabeculectomy remains the mainstay of surgical treatment. In this review, the authors look at the evidence and reasons for cataract formation after trabeculectomy surgery and examine the evidence surrounding bleb failure after cataract extraction. The review highlights that the reasons for cataract formation in those undergoing filtration surgery are poorly understood, and more research needs to be undertaken in this area.

  • Cataract
  • trabeculectomy
  • bleb failure
  • optic nerve
  • epidemiology
  • treatment other
  • field of vision
View Full Text

Statistics from


Glaucoma is one of the leading causes of blindness worldwide. In the UK, the estimated prevalence of open angle glaucoma ranges from 0.3% in the 40s to 3.3% in the 70s. The incidence ranges from 30 to 181 per 100 000 person-years for ages 50 and 70, respectively. Currently, about 500 000 people are estimated to suffer from glaucoma, of which up to 67% are undetected.1 It is one of the leading causes for blind registration in the UK, second only to macular degeneration.2

There are several treatment options for glaucoma, all of which are designed to halt visual loss by lowering intraocular pressure (IOP). Topical medications remain the mainstay. Other treatments include laser trabeculoplasty and filtration surgery. Despite a decline in the number of trabeculectomies performed in the UK, it still remains the operation of choice in the surgical management of glaucoma. In England and Wales, 18 676 trabeculectomies took place in 1995 compared with 5289 in 2004.3 This decline has largely coincided with the introduction of new topical medications, particularly the prostaglandin analogues. In the last 5 years, our own surgical audit at Moorfields and, anecdotally, that of others has shown a small recovery in the number of operations but not to the volume of the pre-prostaglandin analogue era. In contrast to this, surgical management remains first line in the developing world where follow-up and availability of antiglaucoma medication is unreliable.

Trabeculectomy is not without its risks and complications; in addition to failure there are rarer complications that are sight threatening, in particular blebitis, and those that trouble patients such as ptosis and bleb dysaesthesia. Cataract is a common and recognised consequence of trabeculectomy. This is the subject of our review.

Cataract and glaucoma

Little is known about the rate of cataractogenesis and aetiological factors surrounding cataract formation in the glaucoma population. It has been suggested that cataract may be a result of either the glaucomatous process or its treatment or both. A problem in the literature on this subject is the lack of agreed diagnostic criteria for either presence of cataract or cataract progression. Many studies have gone for a pragmatic outcome of cataract surgery, which is open to bias. Others have used lens assessment (usually Lens Opacities Classification System (LOCS)) and defined changes within this. This method is open to measurement error.

Cataract formation in the normal population

There have been several population-based cohort studies looking at cataract incidence and progression in normal subjects. Two such recent studies are the Latino Eye Study and Barbados Incidence Study of Eye Diseases (BISED).4 5 The Latino Eye Study, looked at the incidence and progression of cataracts of 4658 adults, with a mean age of 54.7 years, over a 4-year period and found the incidence of cataract surgery to be 1.48%.4 The 4-year incidence for nuclear lens opacities was 10.2% and the progression rate was 3.7%. BISED looked at 2793 subjects, with a mean age of 56.0 years and found that the 9-year incidence of nuclear lens opacities was 42% and the progression of pre-existing nuclear lens opacities was 17.8%.5 Both studies used LOCS II grading system. The Longitudinal Study of Cataract looked at 764 normal subjects, with a mean age of 63.4 years and found an 8% incidence of new nuclear lens opacities during 5 years and, using a LOCS III grading system, almost 50% of existing opacities had progressed.6 Thus, the population studies quoted above show a large variation in incidence and progression of nuclear sclerotic cataracts. The incidence ranges from 10% to 42%, with a progression ranging from 3.7% to 50%, with up to 9 years follow-up. It should be noted that the populations studied are very different.

Cataract formation in the glaucoma population

Table 1 shows data available on cataract formation in the glaucoma population. It can be seen that the incidence of cataract formation in those with glaucoma ranges 6–23% in untreated glaucoma patients, this increases to 7–39% with up to 7 years follow-up on medical therapy and is markedly enhanced to 20–52% with up to 7 years follow-up by surgical therapy.

Table 1

Incident cataract reported in longitudinal studies by therapy over period of follow-up

Elevated IOP

Shaffer and Rosenthal found no difference in the prevalence of cataract in 5156 normal patients and 245 patients with open-angle glaucoma.19 The Beaver Dam Eye Study (BDES) reported a RR of 2.6 for cataract surgery after 5 years in those with elevated IOP.20 The Blue Mountain Eye Study (BMES) found a similar risk.13 Nuclear cataract was greater in eyes with elevated IOP at baseline than cortical or posterior subcapsular cataract (PSC) (39.1% vs 11.7% and 3.0%, respectively). Elevated IOP was suggested as a possible mechanism for lens damage in both these studies. However, BISED found only a borderline relationship with elevated IOP (>21 mm Hg) and nuclear lens opacities (RR, 1.4; p=0.07).10 Thus, a small association may exist between elevated IOP and cataract formation.

Topical IOP lowering agents

Topical IOP lowering agents have been reported to increase the risk of cataract. In 1975, Levine reported cataract development (defined as loss of 2+ lines of acuity and myopic shift greater than −0.5 D) in eyes treated with pilocarpine with no such change in the fellow eyes without such therapy.21 Anticholinesterases have also been particularly reported in relation to cataract formation.22 In both instances however, the miosis may aggravate pre-existing lens opacity and confuse case ascertainment. BISED found that 33.9% of patients using topical IOP lowering medications developed nuclear opacities after 4 years, compared with 7.5% of persons with IOP <21 mm Hg and on no topical treatment.10 Among participants with IOP >21 mm Hg, those receiving treatment had a fivefold RR in cataract formation (RR, 5.0; 95% CI 1.7 to 15.1) versus those who were not treated. BDES reported a borderline increased RR of PSC (p=0.09) among participants taking topical β-blockers.20

In BMES, participants with elevated IOP, not using glaucoma medication had an increased risk of nuclear cataract in age and gender analyses, respectively (OR, 2.03 (95% CI 1.06 to 3.91) and OR, 2.07 (95% CI 1.16 to 3.71)).13 Glaucoma medication was associated with non-significant increased adjusted odds for incident nuclear opacities (OR, 1.90 (0.92–3.92)). Incident nuclear opacities were greater in eyes treated with glaucoma medications at baseline than cortical cataract or PSC (39.1% vs 6.6% and 4.5%, respectively).

In the ocular hypertension treatment study (OHTS), 7.6% versus 5.6% underwent cataract extraction in the treated and untreated groups, respectively (HR 1.56; 95% CI 1.05 to 2.29). In the medication group, participants who underwent cataract extraction had higher mean baseline IOP compared with those who did not (25.9 mm Hg±2.6 SD and 24.7 mm Hg±2.7 SD, p=0.004). In the observation group, no difference in mean baseline IOP was detected between participants who did or did not undergo cataract extraction (p=0.40).7

Thus, medical treatment for cataract appears to have an increased risk of cataract formation, particularly nuclear sclerosis, this is summarised in table 1.

Trabeculectomy (surgical treatment)

Following trabeculectomy surgery, the risk of requiring cataract surgery is reported as between 20% and 52% up to 7 years postoperatively (see table 1). In the Collaborative Normal Tension Glaucoma Study (CNGTS), the incidence of cataract in the treated group (38%) was significantly higher than in the control group (14%, p=0.001).8 Among treated subjects developing cataract, nearly three-quarters had undergone filtration surgery, whereas only 29% were on medical therapy. Results from the Collaborative Initial Glaucoma Treatment Study (CIGTS) were similar.11 Of the 607 randomised to surgery or medical therapy, they found that the probability of cataract extraction after initial trabeculectomy was eight times greater in the initial surgery group. By 5 years however, the rate had reduced to threefold higher in the initial surgical treatment group compared with the medical therapy group. The authors concluded that surgery placed a patient at an increased risk of cataract for a moderate amount of time, after which other factors such as ageing played a more important role. The study also showed an increased risk of cataract extraction with myopia, pseudoexfoliative glaucoma and diabetes.

In a recently reported study, Palanca-Capistrano et al found that 50% of eyes had subsequent cataract extraction in a randomised trial of trabeculectomy surgery with 5-fluorouracil (5-FU) or mitomycin-C (MMC).16

Costa et al looked at visual loss in the early postoperative period of 508 trabeculectomised eyes. They found that 8.3% showed loss of visual acuity after the first 3 months. Of these, 38% were due to progressive lens opacity.23

Thus, trabeculectomy surgery carries a clear increased risk of cataract formation. The evidence for non-penetrating surgery is less clear. O'Brart and associates performed two studies comparing viscocanalostomy with augmented trabeculectomy.24 25 In the first study, the risk of requiring cataract extraction was significantly higher in the trabeculectomy group (p<0.05).24 However in the second study, they found no difference in the proportion requiring subsequent cataract surgery in either group.25

Why cataract?

The reasons for cataract formation after trabeculectomy surgery are poorly understood. Many hypotheses have been proposed.

Surgical peripheral iridectomy/iris manipulation

de Barros and associates believe that the formation of a surgical peripheral iridectomy (PI) may be a contributing factor to the formation of cataract.26 They advocate avoiding a PI in cases not predisposed to shallow anterior chambers. In their case–control study, they studied 43 patients with PI and 32 without PI and found that cases with PI had significantly more postoperative inflammation (p=0.018 in phacotrabeculectomy group, p=0.038 in trabeculectomy group). They concluded that trabeculectomy performed without a PI was just as effective in lowering IOP, as those with a PI and significantly reduced the degree of postoperative inflammation.

Use of antimetabolites and postoperative medication

Antimetabolites have also been inculpated in the development of cataract. Some authors have postulated that MMC may increase cataract formation by direct toxicity to the lens. Robin and associates conducted a prospective randomised controlled study looking at 300 eyes undergoing trabeculectomy.27 They found cataract to be the most frequent complication (18.1%). Their study showed that the duration of application of MMC influenced cataract formation, with eyes undergoing trabeculectomy with 0.2 mg/ml MMC applied for 4 min having the most significant risk of cataract development. A Cochrane review of MMC for trabeculectomy found a RR of 1.8 (1.00–3.22) for cataract formation in those receiving MMC compared with no antimetabolite.28

The use of β radiation instead of antimetabolite in the developing world has been shown to substantially reduce the risk of surgical failure after glaucoma surgery. However, those receiving β radiation had a higher incidence of operable cataract at 1 year (3.2% compared with 0.8% in the placebo group (p=0.01)).29

The same effect does not seem to be present for 5-FU. In the Singapore 5-FU Trabeculectomy study, Wong et al compared intraoperative 5-FU with placebo, in eyes undergoing primary trabeculectomy in a randomised controlled trial.15 Within 3 years, 51.9% of patients had undergone cataract extraction (50% 5-FU group, 54% placebo group).

Other factors related to surgery should also be considered. Postoperatively these patients are on high-dose topical steroids for an average of 3 months. They may also receive multiple subconjunctival steroid and 5-FU injections in the postoperative period. The cataract subtype formed posttrabeculectomy is generally nuclear sclerotic and steroid use is most frequently associated with PSC cataract. The issue of hypotony is discussed in the section ‘Postoperative shallow or flat anterior chamber’.

Thus, MMC and β radiation but not 5-FU augmented trabeculectomies have been found to be associated with an increased risk of cataract formation.

Postoperative shallow or flat anterior chamber

Tornqvist and Drolsum followed 277 trabeculectomised eyes for up to 10 years.14 They found that 50% of eyes with early postoperative hypotony and flat anterior chambers developed cataracts after 5 years. Costa and associates studied the cause of loss of visual acuity post-trabeculectomy in the early postoperative period.23 They also found that shallow anterior chamber (p=0.0003) correlated with the development of lens opacification. Supporting this, a Korean study with 70 patients showed that 50% of eyes undergoing trabeculectomy surgery, had evidence of cataract progression, if the anterior chamber was flat postoperatively.30 In most studies, however, the reported prevalence of shallow or flat anterior chamber following trabeculectomy is much lower than the prevalence of cataract formation.


Popovic and Sjostrand investigated the mean IOP in those who developed cataract following trabeculectomy and those who did not.31 While there was a small difference (17.7 vs 19.6 mm Hg), this was well within random variation. Vesti and Raitta reported postoperative hypotony of ≤5 mm Hg for at least 5 days to be a risk factor for cataract formation.32 This could, of course, have been associated with a shallow anterior chamber.

Other factors

Trauma at the time of surgery will cause lens opacity, usually focal and cortical at the site of trauma in the first instance. This was more widely reported in the past when surgical equipment and techniques were less advanced and is infrequently reported now. A wide range of studies have reported increased incidence of cataract formation in those with secondary glaucoma, particularly pseudoexfoliation.

Problem of bleb failure after cataract surgery

The literature reports 10–61% of trabeculectomies failing at 12–36 months post-cataract surgery (see table 2). There are numerous studies looking at IOP control and bleb failure postcataract extraction. Interpretation of the results is, however, hampered by very few having a control group, different study methods, definitions of failure and patient groups. To our knowledge, there are only two case–control studies (see table 2). Swamynathan et al performed a case–control study with 2-year follow-up and found that 24% of trabeculectomies failed that underwent cataract extraction, compared with 7% failure rate in those that did not have cataract surgery.33 In contrast to this, Park et al also performed a case–control study, looking at 80 trabeculectomised eyes, 40 of whom underwent cataract extraction and found no difference in failure rate between the groups at 1 year.36 Although 13 patients failed in the cataract extraction group, five of the thirteen failures, had already happened before cataract extraction. Nine failed in the non-cataract surgery group. In those who had cataract surgery, 5 of the 13 failures, had already failed before cataract extraction. Seven eyes underwent sphincterotomies and three eyes underwent anterior vitrectomies during cataract surgery; neither of these factors was found to have a bearing on postoperative IOP control. Although in both studies, subjects were reported as matched, a matched analysis was not employed.

Table 2

Longitudinal studies of eyes with trabeculectomy surgery looking at the impact of subsequent cataract surgery on success

Chen et al looked at 115 patients with functioning filtration blebs undergoing cataract surgery (58 extracapsular cataract extraction (ECCE); 57 phacoemulsification of the lens).37 Thirty per cent of patients required additional antiglaucoma medication and/or needling postcataract surgery and 10% required re-operation (an overall failure of 40% at 21 months). They found being less than 51 years old, having a preoperative IOP greater than 10 mm Hg, intraoperative iris manipulation and early postoperative IOP greater than 25 mm Hg were associated with poorer postoperative IOP control postcataract extraction. In addition, cataract extraction performed 6 or more months after initial trabeculectomy was possibly protective against failure (RR, 0.5; 95% CI 0.2 to 1.1; p=0.087). Regression analysis showed that trabeculectomies with ECCE were more likely to fail than with phacoemulsification (RR, 3.0; 95% CI 0.7 to 12.8 failure in ECCE vs phaco group). A limitation of this study is that no control group was enrolled to assess the baseline rate of bleb failure in this population. They reported that intraoperative iris manipulation during cataract surgery was associated with an increased risk of bleb failure (p=0.019).

Manoj and colleagues also compared ECCE and phacoemulsification in patients with pre-existing filtering blebs. The IOP increased significantly (1.9 mm Hg) in the ECCE group, but remained unchanged in the phacoemulsification group (21 eyes).39

Ehnrooth et al found that in eyes classified as a complete success after trabeculectomy, 59% remained so after cataract extraction.17 The proportion of failures was double in those undergoing cataract surgery (61% vs 31%), with mean follow-up of 25.3 months. However, 14 trabeculectomies had already failed prior to cataract surgery in the cataract extraction group. In a further study, 49 eyes undergoing cataract surgery with functioning filtration blebs, 83.6% and 68.2% remained a complete success at 1 year and 2 years, respectively.35 Higher IOP prior to cataract surgery was again associated with bleb failures and the need for additional antiglaucoma medication. In a further ultrasound biomicroscopy study looking at bleb morphology before and after cataract surgery, an IOP greater than 10 mm Hg, a bleb with an invisible route under the scleral flap and stronger intrableb reflectivity before phacoemulsification were associated with postoperative trabeculectomy failure.40

Thus, a majority of the literature suggests an increase in bleb failure following cataract surgery, extracapsular surgery having a higher toll than phacoemulsification.

Phacotrabeculectomy versus trabeculectomy

Belluci et al compared 100 trabeculectomies with 200 phacotrabeculectomies, and found that trabeculectomy alone resulted in a larger decrease in IOP than the combined procedure.41 At 1 year, IOP was lower in the trabeculectomy alone group, than in the combined group (15.2±5.1 mm Hg vs 18.7±7.3 mm Hg, p<0.01). Also, the number of filtering blebs was lower after the combined procedure. They concluded that although much of the literature reports good short-term results of combined procedures on IOP, these are misleading. IOP decrease is seen after routine phacoemulsification in patients without trabeculectomy in the short term; this is probably due to a prolonged low-grade inflammation.

Caprioli et al compared 40 cases undergoing phacotrabeculectomy with 40 cases of trabeculectomy alone.42 The authors also reported that mean IOP was lower in the trabeculectomy alone group at 1 year (13.6±3.8 mm Hg vs 15.4±3.3 mm Hg, p=0.028). The mean reduction in IOP at 1 year was 10.3±7.6 mm Hg in the trabeculectomy only group compared with 6.8±5.5 mm Hg in the phacotrabeculectomy group.

From the limited studies with sufficient longitudinal follow-up, it appears that trabeculectomy alone, rather than a combined procedure has an improved IOP lowering effect.


Reviewing the literature does provide clarity on some issues surrounding this topic. There are two key problems. First, the incidence of cataract seems to increase with a diagnosis of glaucoma, medical therapy for glaucoma and filtration surgery. The incidence is significantly higher in those with filtration surgery than in those treated with topical medication. Thus, it is important that all surgeons specifically mention cataract development as a risk when counselling for trabeculectomy surgery. The cause of this lens change remains uncertain and is probably multifactorial.

Second, cataract surgery in eyes with functioning filtration blebs is associated with an increased risk of bleb failure. The risk of bleb failure seems to be increased if the patient undergoes ECCE in the presence of a functioning bleb or phacotrabeculectomy. Cataract surgery in patients requiring filtration surgery remains an issue of contention. The jury is still divided as to whether performing cataract surgery before or after filtration surgery reduces the likelihood of bleb failure.

In the developing world the problem is even more evident. Our therapy for glaucoma is extremely limited and the existence of dual pathology demands a strong evidence-based therapeutic strategy in order to prevent blindness.

There is much important research to undertake to determine the aetiology and risk factors for cataract formation in glaucoma patients undergoing medical and surgical therapy. In addition, the prevention of bleb failure in relation to cataract surgery needs further research. Cataract should not remain the silent enemy of successful trabeculectomy surgery.


View Abstract


  • Competing interests None.

  • 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.