Purpose of glaucoma treatment and surgery

The overall purpose of glaucoma treatment is to maintain visual function and related quality of life at a sustainable cost. Glaucoma surgery aims for a prolonged IOP reduction without complications. Generally, medical treatment is considered first, followed by laser. Surgery is considered later because medical treatment is sufficient to control IOP in most patients; the inconvenience of lifelong eye drops is considered a smaller imposition than risks of surgery. Often medical treatment is thought to be ‘sufficient’. Many glaucoma patients are managed by general ophthalmologists who may not perform glaucoma surgery regularly. These factors all favor later surgery.

Although the aim of surgery is usually to preserve vision by lowering IOP, sometimes the primary aim may be to reduce pain in a poorly sighted eye, to correct an anatomic abnormality, e.g. peripheral iridotomy/iridectomy, to relieve pupil block, lens extraction in primary angle closure, removal of subluxated lens in secondary angle closure, or vitrectomy in aqueous misdirection.

These goals can be achieved by proper patient selection and choice of the most appropriate technique. As surgery always involves some risk, a careful assessment of risk versus the potential benefit is needed; this amounts to the risk of not operating. This must be considered and discussed in full. The factors that determine the level of treatment aggression depend on the severity of disease at presentation (i.e. the location and extent of visual field defects), the patient’s life expectancy, and the degree of IOP elevation. Other factors that play a role include the type of glaucoma, unilateral or bilateral disease, the family history, and a previous history of intraocular or conjunctival surgery.

Medical versus surgical treatment

While medical and laser treatment are usually the first lines of therapy, there are some situations where early surgery is advisable. Surgery is the principal treatment for angle closure, developmental glaucomas, and a number of secondary glaucomas. Surgery is recommended for POAG in three circumstances (Box 34.1):

Fig. 34.2 Greyscale portion of three sequential Humphrey 24-2 visual fields (SITA-standard algorithm) performed over a 2-year period showing progression of a paracentral visual field defect that was consistent on repeat testing, in a patient with treated IOP levels consistently in the normal range. Because of the central nature of this defect, the patient noted a positive scotoma. The defect stabilized following trabeculectomy surgery. In this patient’s case central vision was reduced due to the position of the defect, illustrating the high risk of visual loss with this type of defect and hence the argument for earlier surgery.

In the developing world, medication is often unaffordable, and surgery may be indicated as an opportunity to prevent progression with one intervention alone. However, the barrier to treatment of glaucoma in the developing world is often lack of awareness or understanding of the nature of the condition or that preventive treatment is important. Few therefore opt for treatment even when it is available or affordable.

The term maximal-tolerable medical therapy (MTMT) is often used to indicate the level of medical therapy above which there is likely to be no further benefit in terms of IOP-lowering, or above which the patient’s ability to tolerate the regimen is likely to diminish because of increasing side effects. The term is of limited usefulness: first it takes no account of an individual patient’s ability to sustain the maximal level of medication. This term usually denotes the maximal level of medication that a patient can tolerate at any one time, which is often higher than the level tolerated in the longer term. Second, the implication, perhaps unintended, is that it is in the patient’s best interest to take as many medications as possible before even considering surgery. This is inaccurate for the following reasons:

When and on whom to operate

Surgery is the mainstay of treatment for a number of types of glaucoma including primary angle closure, secondary, and developmental glaucomas. Surgery is indicated for POAG in the situations described above, especially in those in whom POAG is advanced at the time of presentation, and especially if the patient is younger. On the other hand, it might be wrong to deny surgery to the better eye of a very elderly patient with borderline IOP control if the fellow eye has recently lost vision from glaucoma under similar circumstances. As above, surgery may be offered earlier in the developing world because of lack of availability or affordability of medications, and lack of regular long-term access to medical care as in remote communities.

On which eye to operate first?

Usually the decision is simple: the eye with the higher IOP and the faster progressing visual field. However, it is not uncommon in advanced glaucoma to operate on the eye with better vision in order to preserve central vision, rather than an eye in which fixation has already been lost. This requires a leap of faith on the part of the patient, who must understand fully the reasons to operate on the better eye first. This is especially true in developing world situations, where patients may fail to understand the reasoning and permanence of visual loss in the worse eye, and the risk to the better eye. Patients undergoing surgery in the better eye must understand the potential for visual blurring for some weeks as the eye recovers, as well as the longer-term risks and benefits.

If surgery is required in both eyes, and surgery to the more severely affected eye has been successful, the patient may have more confidence and be better prepared for surgery to the better sighted eye. This risk of surgery has to be balanced against the risk of visual loss due to glaucoma in an only eye. Meticulous surgery erring on the side of safety is essential.

Which procedure first?

Traditionally trabeculectomy has been the procedure of choice in patients who have not had prior glaucoma surgery. This has been challenged in a number of ways. Advocates of both non-penetrating glaucoma surgical procedures (NPGS) and GDDs promote these as better options to trabeculectomy. Both cite safety issues. GDDs do not rely on aqueous drainage to the perilimbal subconjunctival space, whereas NPGS produce less limbal drainage than trabeculectomy. Bleb-related infection is therefore less likely with both. The risk of hypotony is less with NPGS. Contact lens wear is also safer after both than after trabeculectomy. Advocates of GDDs also claim better efficacy. Recently a large scale randomized trial has supported the use of GDDs15. While there have also been randomized trials comparing deep sclerectomy with trabeculectomy, they have been relatively small in size, one suggesting a significantly lower IOP was reported with trabeculectomy18 while another reported a slightly lower but statistically insignificant mean IOP with trabeculectomy19. As neither study used MMC for either procedure, nor was an implant (e.g. collagen) used for deep sclerectomy, it is questionable whether they are representative of current surgical methodology.

Primary open angle glaucoma

Surgery is offered in POAG if there is advanced glaucoma (e.g. fixation-threatening visual field defects) at presentation, especially in younger patients, or if glaucoma is progressing despite 2–3 classes of drugs or after laser trabeculoplasty. Other indications include poor adherence to the prescribed medical regimen or inability to tolerate medical therapy. The United Kingdom National Institute for Health and Clinical Excellence (NICE) guidelines on management of primary open angle glaucoma recommend that surgery is offered to those with advanced glaucoma (MD < –12 dB or sensitivity < 15 dB in the central degrees in both hemifields) at presentation20, as these patients are at highest risk of loss of central vision from glaucoma.

Surgery achieves lower mean IOP and less diurnal fluctuations than medical treatment. The Collaborative Initial Glaucoma Treatment Study (CIGTS) group recently reported that, 9 years after randomization, patients with higher visual field loss at baseline (–10 dB) had less visual field loss over time when treated with initial surgery as compared with initial medical treatment, except in patients with diabetes21,22.

However, glaucoma surgery differs from other types of intraocular surgery in that there is no immediate tangible visual benefit. The aim of surgery is preventive. Initially the vision may be worse after surgery and in some cases, e.g. after a ligated GDD, the IOP may be higher initially than before surgery, at least for the initial postoperative period; this is evident from early results of the Ahmed Baerveldt Comparison Study (ABC), comparing16,17 the Ahmed glaucoma valve with the Baerveldt glaucoma implant. In the early postoperative period, the patients with the occluded non-valved Baerveldt implant had significantly higher IOP levels than those with the valved Ahmed valve. Patients must be aware of this possibility and that it does not imply failure. Usually these factors will not influence the decision to proceed with surgery, but are important to manage patient expectations once the decision to operate is made.

The risks of a slight visual reduction, change in refraction, ocular surface discomfort, and cataract must be discussed, as must the necessity for frequent outpatient visits during the first 3–4 weeks after surgery, as these may influence the decision to operate. The patient must also be made aware of the low but potentially serious risk of visual loss from hypotony or infection.

Ultimately, surgical trials with visual fields and patient-reported outcome measures as endpoints would give a clearer indication of the relative benefits of surgery, but such trials are costly. Most patients prefer to avoid surgery if possible, though interestingly a recent conjoint analysis of factors most import to patients found they were less concerned about the mode of treatment than the risk of visual loss23. In the CIGTS study, the impact on quality of life was similar in both medically and surgically treated arms of the study 5 years after initial treatment24.

The choice as to the best surgical procedure is becoming less clear. The most widely used procedures are trabeculectomy, GDDs, and various non-penetrating procedures. Trabeculectomy drains aqueous from the anterior chamber via a guarded sclerostomy or peripheral posterior keratectomy to the limbal subconjunctival space. Non-penetrating procedures do the same while maintaining a barrier between the anterior chamber and external filtration site. Some argue that non-penetrating procedures restore flow via Schlemm’s canal to the collector vessels. Most GDDs drain aqueous via a silicone tube from the anterior chamber to a site remote from the limbus, usually the equatorial subconjunctival space, again to a drainage bleb. In GDDs, the size of the bleb is determined by the size of the end-plate.

In general trabeculectomy and deep sclerectomy procedures require healthy unoperated superior conjunctiva with an intact blood–aqueous barrier, whereas GDDs are more versatile and may be used where there has been previous incisional surgery (e.g. sclera buckle, previous trabeculectomy), conjunctival disease (e.g. chronic inflammation, atopy, trauma, aniridia, chemical burns), scleral thinning (e.g. rheumatoid), chronic breakdown in blood aqueous barrier (e.g. rubeotic glaucoma), and anterior segment disorders (e.g. fibrous ingrowth, ICE syndrome, aphakia, and penetrating keratoplasty).

In other situations where there is higher risk of infection (e.g. significant posterior lid margin disease), or where an anterior bleb is not desirable (e.g. committed soft contact lens wearer), trabeculectomy is better avoided and a GDD may be the preferred option. For many years, trabeculectomy has been the standard surgical procedure in previously unoperated eyes in primary glaucomas, whereas shunts have been preferred for several secondary and refractory glaucomas or where further ocular surgery is planned (e.g. penetrating keratoplasty, vitreoretinal surgery, etc). Had it not been for the advent of the newer ocular hypotensive drugs, such as prostaglandin agonists, one might speculate that trabeculectomy would have been displaced by non-penetrating procedures performed relatively early in the disease process. However, as surgery is still performed relatively late, in patients with significant visual field defects, many surgeons argue that definitive IOP control is required, tipping the balance in favor of the greater perceived efficacy of trabeculectomy. However, randomized trial evidence that the Baerveldt 101-350 implant (Abbott Medical Optics, Abbott Park, Illinois, USA) is able to achieve IOP levels in the low teens25 was the first to substantiate a growing body of surgical experience15,17 that contradicts the popular belief that GDD (glaucoma drainage devices) cannot achieve low target IOP levels in the average patient, and as a result shunts are beginning to displace trabeculectomy in some centers.

The safety and efficacy of trabeculectomy surgery, both as a primary treatment21,26 and in higher risk eyes7,15,27, has been reported in a number of trials and a number of risk factors for trabeculectomy failure have been identified. These include type of glaucoma, race, young age, previous filtration surgery, pseudophakia, other prior surgery (penetrating keratoplasty, vitreoretinal surgery), anterior segment disease (fibrous/epithelial ingrowth, ICE syndrome), ocular surface disease (e.g. atopy, rosacea), and postoperative inflammation. The fluorouracil filtering surgery study, performed in the late 1980s investigated the potential benefits of pharmacological augmentation of trabeculectomy with an intensive regimen of 5FU. While 5FU almost doubled trabeculectomy success, the study highlighted the poor likelihood of trabeculectomy success in high failure risk eyes7. After 5 years’ follow-up, 51% of 105 5FU eyes and 74% of 108 control eyes were classified as failures (P < 0.001) (Fig. 34.3). 5FU augmentation clearly improved success, but did not provide the answer in higher risk cases. The use of 5FU was rapidly superseded by MMC, which improved the success rate in these cases, albeit at a higher risk of complications from hypotony and infection. Randomized clinical trials comparing MMC with 5FU trabeculectomy have reported better efficacy in terms of IOP control with MMC28, but there has been no study of an equivalent size and follow-up, until the Tube Versus Trabeculectomy study (TVT).

Fig. 34.3 The Fluouracil Filtering Surgery Study: the poor success rate of trabeculectomy, either with or without 5-fluorouracil at 5 years in eyes that had previous failed filtration surgery or prior cataract surgery.

Redrawn from Five-year follow up of the Fluouracil Filtering Surgery Study, AJO 1996:349–66.

The TVT study was a randomized prospective surgical trial comparing trabeculectomy at the superior limbus augmented with a 4-minute application of 0.4 mg/ml of MMC with supero-temporal insertion of a Baerveldt 101-350 anterior chamber aqueous shunt. Unlike the 5FU filtering surgery study and previous randomized clinical trials of GDDs, the TVT investigators excluded patients with many higher failure risk factors such as aphakia and a number of secondary glaucomas including iris neovascularization, uveitis, ICE syndrome, and epithelial downgrowth. As a result the largest diagnostic groups in the 212 patients enrolled from 17 clinical centers were POAG (81%), chronic angle closure (8%), and exfoliative glaucoma (4%). The entry criteria permitted enrolment of patients who had either a previous failed filtration surgery, or prior cataract surgery. In theory this might have resulted in a high proportion of patients in whom the only prior surgery was clear corneal phacoemulsification. In reality, only 12% of patients enrolled were in this category, whereas 56% had failed filtration surgery prior to enrollment and the remainder comprised a group of patients who either had prior extracapsular or intracapsular cataract surgery or had scleral tunnel phacoemulsification with previous violation of the conjunctiva.

At three years both trabeculectomy with MMC and Baerveldt 350 tube surgery showed similar reductions in IOP (mean IOP 13.3 ± 6.8 mmHg in the trabeculectomy and 13.0 ± 4.9 mmHg in GDD group, P = 0.78) with similar numbers of glaucoma medications (1.1 ± 1.5 in trabeculectomy and 1.3 ± 1.3 in tube group, P = 0.30) (Fig. 34.4A). However the cumulative probability of failure was higher in the trabeculectomy group as compared with the GDD group (30.7% vs. 15.1%, P = 0.01).While more patients in the trabeculectomy group failed from inadequate IOP reduction, the same is also true of those failing from hypotony and the overall impression is that the Baerveldt GDD produces more predictable pressure control at 3 years than trabeculectomy i.e. both fewer high and fewer low IOP failures (Fig. 34.4B).

Fig. 34.4 After 3 years of follow-up in the tube versus trabeculectomy study there was little difference in IOP control between the two study groups (A), but there were more failures in the MMC trabulectomy group than the tube group, both from low IOP and from high IOP (B), demonstrating a greater predictability in the outcome of tubes versus trabeculectomies at 3 years.

Redrawn from Gedde SJ, Schiffman JC, Feuer WJ, et al. Three-year follow-up of the Tube Versus Trabeculectomy Study. AJO 2009;148:670–84.

The TVT investigators also stratified enrolled patients according to perceived failure risk:

The relative efficacy profile of the two procedures was anticipated to differ between these strata, but surprisingly at 3 years this was not so. An identical trend was reported in each stratum, i.e. a slightly higher complete success (i.e. IOP <22 mmHg and at least 20% drop in IOP off medications) in the trabeculectomy group, which was outweighed by a greater qualified success (the same as above, but with ocular hypotensive medications) in the tube group, and a greater failure rate in the trabeculectomy group. Overall trabeculectomy patients were significantly more likely to be free of ocular hypotensive medications, although only 40% of trabeculectomy patients achieved this. Overall the TVT study results suggest better predictability for the Baerveldt GDD than MMC-augmented trabeculectomy: while mean IOP levels and numbers of glaucoma medications were similar at 3 years, the trabeculectomy group had more failures, from both high and low IOPs (see Fig. 34.4). There was a higher incidence of early postoperative complications in the trabeculectomy group, but significantly the difference between the two groups was largely due to early conjunctival wound leak and hyphema. There was no difference in the rate of vision loss from complications in the two study groups.

With improved surgical technique, complications that appeared to be relatively frequent after the introduction of MMC, such as bleb-related endophthalmitis8,29, bleb-dysesthesia12, and hypotony maculopathy11 have become less of a concern, though a lingering worry is the influence of trabeculectomy on cataract creation27,30 and the detrimental influence of subsequent cataract surgery on long-term trabeculectomy function31. When considering a GDD, these must be balanced against the 5–10% risk of diplopia17,32 and possible longer-term risk of corneal damage and erosion33–35.

Primary angle closure

The aim of treatment in primary angle closure is often to relieve pupil block (iridotomy or lens extraction), either in order to prevent or treat angle closure, to alter iris configuration (iridoplasty in plateau iris syndrome), or to correct forward movement of the lens–iris diaphragm (e.g. vitrectomy in aqueous misdirection). Ultimately all are intended to prevent the development or progression of glaucomatous optic neuropathy.

Most patients with occludable angles or angle closure in clinical practice have not had an acute episode, and are rather more challenging to diagnose than those with acute primary angle closure. Careful slit-lamp examination and especially gonioscopy is essential to detect those at risk and to differentiate pupil block from other mechanisms of angle closure, such as plateau iris or forward movement of the lens–iris diaphragm. As the level of slit-lamp and room illumination has a large influence on angle configuration, perform gonioscopy with minimal illumination, especially when attempting to diagnose an occludable angle in a patient without angle closure. Generally, in a patient without IOP elevation or symptoms of intermittent angle closure (blurring of vision and colored rings around lights late at night or early morning) or evidence of glaucomatous optic neuropathy, the indication for laser iridotomy is an occludable angle, diagnosed as either failure to visualize trabecular meshwork through at least 180°, or the presence of irido-trabecular contact. Anterior segment optical coherence tomography is helpful to differentiate occludable from non-occludable angle narrowing (Fig. 34.5), having the advantage that it can be performed in the dark.

Fig. 34.5 Anterior segment optical coherence tomography is useful for visualizing the width of the irido-corneal angle in eyes with narrow angles. In the example shown, the pupil is small because of pilocarpine treatment. The convexity of the iris surface can be clearly seen, indicating a degree of pupillary block. Despite the pilocarpine treatment, the drainage angle is extremely narrow and laser iridotomy would likely result in significant widening.

Pupil block is treated with laser iridotomy although, for patients with visually significant cataract, lens extraction may be considered first as this alone may open the angle. Occasionally, in a patient with angle closure and persistent narrowing despite laser iridotomy, removal of a clear lens may be indicated.

After relief of pupil block, persistent anterior synechiae (chronic primary angle closure) or trabecular meshwork damage, from an episode of treated angle closure, may result in chronic IOP elevation that is treated in the long term similarly to POAG, with the exception that is often appropriate to perform lens extraction earlier in an eye with chronic angle closure and chronic IOP elevation (Fig. 34.6). If medical treatment fails to control the IOP, lens extraction alone may be required to reduce the IOP. In eyes with significant glaucomatous optic neuropathy and chronic synechial angle closure, lens extraction may be beneficial in improving pressure control, but ultimately filtration surgery may also be required. EAGLE (Effectiveness in Angle-closure Glaucoma, of Lens Extraction) is a multicenter randomized surgical trial designed to establish whether lens extraction for newly diagnosed primary angle closure glaucoma results in better patient reported health, vision, lower IOP, and other outcomes compared with standard treatment and should help guide management of these patients.

Fig. 34.6 In this patient with chronic angle closure, gonisocopy shows uneven irido-trabecular contact obscuring the view of the trabecular meshwork and most of the angle structures, as well as some pigment or smudging on the visible angle. The smudging indicates that the areas that are open on gonisocopy in the clinic are likely also to be subject to iris contact and occlusion intermittently.

Surgical indications in secondary glaucoma

Secondary glaucomas are often asymmetrical or unilateral and cause vision loss more commonly than primary glaucomas. When first examining a patient with glaucoma look for signs of the common secondary glaucomas. Exfoliation syndrome (pseudoexfoliation) (Fig. 34.7) and pigmentary glaucoma may be missed if the pupillary area of the iris, drainage angle, and anterior lens surface are not examined carefully. These usually present in less dramatic fashion than other secondary glaucomas and can be managed with the same medical regimens as POAG, though treatment failure is more common. Both respond well to trabeculectomy, though cataract surgery has been reported to cause a greater IOP reduction in exfoliation than in POAG. Pigmentary glaucoma patients are often young, male, and myopic and therefore at a higher risk of maculopathy should hypotony develop after surgery; take care to avoid trabeculectomy over-drainage at the time of scleral flap closure. Patients with exfoliation more commonly develop cataract and cataract surgery is more often complicated, though trabeculectomy success appears to be comparable with POAG.

Fig. 34.7 Exfoliation syndrome (pseudoexfoliation), showing the characteristic exfoliation material on the central anterior lens surface. The clear area, due to pupil rubbing against anterior lens surface is easily seen just peripheral to this on pupil dilation (A). On the right (B), pigment deposition is seen on the anterior iris surface. Iris pigment is lost from the pupillary area due to rubbing against the lens in exfoliation syndrome. The dispersed pigment can also be seen in the trabecular meshwork, though the amount of deposition is usually less marked than in pigment dispersion syndrome.

Be aware of the many known risk factors for failure when considering trabeculectomy in secondary glaucomas as such eyes often have multiple risk factors for failure and GDDs are more often indicated (Box 34.2). In single chamber eyes, anterior segment proliferative conditions (e.g. neovascularization, fibrous/epithelial ingrowth, ICE syndrome), prior anterior chamber or subconjunctival silicone oil, or severe ocular surface scarring or thinning (scleral buckles, scleritis, chemical burns) greatly diminish prospects for successful trabeculectomy. In other conditions such as uveitic glaucoma, trabeculectomy may be successful if inflammation is well controlled, sometimes aggressively.

In eyes that do not have navigating vision (e.g. neovascular glaucoma after central retinal vein occlusion) control of IOP may be necessary to improve patient comfort. In that circumstance, trans-scleral diode laser cyclophotocoagulation (cyclodiode) may be more appropriate than more aggressive surgical intervention.

Surgical indications in pediatric glaucoma

Developmental glaucomas, such as primary congenital glaucoma, glaucoma secondary to aniridia, anterior segment dysgeneses or after congenital cataract surgery, require management by a surgeon specialized in the care of pediatric glaucoma. Trabeculectomy success is lower, hypotony rates higher, and complications such as corneal edema after GDDs more common. Angle surgery (goniotomy/trabeculotomy) is often the first procedure for primary congenital glaucoma. If this fails to control the IOP or the child is older at presentation then trabeculectomy or a GDD may be indicated.