The Tube-versus-Trabeculectomy (TVT) study is a randomized, multicentre trial evaluating treatment options for eyes with previous failed glaucoma surgery. The three-year results, presented at the 2008 meeting of the American Academy of Ophthalmology,¹ demonstrate certain differences between the two surgical options but nevertheless highlight the many shortcomings of both. Although the outcomes of the tube surgery were marginally better, around 35% of subjects in both groups lost more than two lines of Snellen visual acuity (VA).

The European Glaucoma Society (EGS) guidelines suggest that non-penetrating glaucoma surgery (NPGS) should be reserved for eyes that are at low risk of failure, where the target intraocular pressure (IOP) is set in the high teens, rather than for treating those eyes with high-risk characteristics (including previous failed surgery). It is true that bleb-independent NPGS procedures, such as viscocanalostomy and canaloplasty, are not indicated in high-risk eyes; this is because Schlemm's Canal is usually irreversibly closed at the site of the failed trabeculectomy, and, in our experience, Nd:YAG laser goniopuncture (LGP) tends not to be effective in the absence of a filtering bleb. Nevertheless, deep sclerectomy (DS), a bleb-dependent variant of NPGS, may be a more appropriate and effective treatment for high-risk eyes than either tube or trabeculectomy. This is especially true when the IOP-lowering efficacy of DS is increased by subconjunctival mitomycin-C (MMC) application, postoperative 5-fluorouracil (5-FU) injections, needle revision and LGP.

Why consider deep sclerectomy?

Rebolleda and co-workers, in a prospective study of DS-MMC on 20 eyes with failed trabeculectomy,² reported that all patients achieved an IOP of less than 21 mmHg, with or without medication; at one year, mean IOP had dropped from 25.8±7.3 mmHg to 14.6±3.2 mmHg. LGP was performed in 40% of eyes. Crucially, no complications were observed, and only one eye with very advanced visual field damage had a VA decrease of more than two lines.

In my practice, DS with MMC has been the procedure of choice for most open angle glaucoma cases for over eight years; since 2003, we have been performing DS for a wide variety of high-risk situations, including African race, uveitis and previous failed glaucoma surgery. The major benefit of DS is the very low incidence of postoperative complications and the lack of postoperative anterior chamber inflammation.

Putting it to the test in the real world

I conducted a study to evaluate the efficacy of DS in high-risk eyes. The study was a single surgeon series with intent-to-treat analysis: a retrospective database search (August 2001–January 2006) was performed, and 57 eyes of 57 consecutive patients treated with DS after previously failed glaucoma surgery were identified. The patients had a mean age of 71.0±16.0 years and were using 2.2±1.0 medications preoperatively. Most patients were diagnosed with primary-open angle glaucoma (n=50); five had uveitic glaucoma and the final two patients had other forms of glaucoma.

Figure 1: Surgical technique for deep sclerectomy with MMC in eyes with previous failed glaucoma surgery

Preoperative conjunctival mobility testing, to identify an area of superior conjunctiva free of adhesions and gonioscopy, was performed. The surgical technique is shown in Figure 1. MMC 0.2 mg/ml was applied for three minutes before scleral dissection in all cases except one, in which I discovered intraoperatively that the conjunctiva was very fragile and there were hardly any subconjunctival tissues, making the use of MMC hazardous. DS with MMC was performed in 49 eyes; combined phacoemulsification with DS and MMC was performed in seven eyes, and DS alone was performed in the final eye.

A number of these surgeries involved a significant learning curve due to their difficulty. It was often necessary to perform the surgery in nasal quadrants, and access was limited. Problems encountered during surgery included excessive bleeding (perforators) in eight eyes (14%), superficial dissection of deep scleral flap with limited aqueous flow from the trabeculo-descemet's membrane (TDM) (eight eyes; 14%), fibrosis of juxtacanalicular meshwork and Schlemm's canal (seven eyes; 12%), iris synechiae at site of dissection (seven eyes; 12%) and deep dissection exposing the choroid (five eyes; 7%).