Managing concurrent uncontrolled uveitic glaucoma represents a surgical challenge. The surgical success rates in uveitic glaucoma varies markedly (50–100%) (5). There is a consensus that the surgical success rate of filtering surgery is lower for eyes with UG compared with primary open-angle glaucoma (POAG) (8).
Trabeculectomy is hailed as the surgical procedure of choice in uveitic glaucoma (9). Although use of antiproliferative agents is controversial, some studies suggest a benefit in terms of long-term IOP control (9). Stavrou et al. quote a 5-year success rate (IOP < 21 mmHg) as 78% in their uveitic glaucoma patients following trabeculectomy without antimetabolites, and with topical medication (10). With the use of antimetabolites (5 FU or MMC), Ceballos et al. reports a cumulative probability of complete or qualified success as 78% at 1 year and 62% at 2 years in uveitic glaucoma patients (11). However, in this group, 51.6% needed cataract surgery and 25% required repeat trabeculectomy. A retrospective study by Iwao et al. concluded that trabeculectomy with mitomycin c (MMC) in UG eyes is independently associated with a worse prognosis than the same procedure in POAG eyes after adjusting for all confounding factors (12).
Trabeculectomy with or without antimetabolites has well-documented complications such as hypotony maculopathy, aqueous leak, postoperative hyphema, worsening of intraocular inflammation, serous choroidal detachment, shallow or flat anterior chamber and bleb encapsulation (11, 12). Although the use of antimetabolites can improve the success rate of glaucoma surgery, it also increases the severity of complications (13).
In cases with high risk of trabeculectomy failure such as aphakic patients, juvenile idiopathic arthritis, previous trabeculectomy failure, retinal detachment surgery, active inflammation the surgical procedure of choice is a glaucoma drainage device (GDD). Studies have shown that all 3 common types of GDD (Ahmed, Baerveldt, and Molteno) are effective in lowering intraocular pressure and reducing the number of glaucoma medications (14–17). The most common complications after GDD in patients with uveitic glaucoma are: encapsulated bleb, transient hypotony, and hyphema, occlusion of the tube by inflammatory materials and corneal decompensation (8).
In an attempt to lower the incidence of all these complications following trabeculectomy or GDD, NPGS was developed.
The advantage of nonpenetrating procedures, such as deep sclerectomy (DS) and VC, is that there is no entry into the anterior chamber and the avoidance of an iridectomy may reduce intraocular inflammation and postoperative complications, which makes them of specific interest in uveitic glaucoma. In particular, VC, by attempting to restore an aqueous outflow pathway, is predominantly independent of external filtration and may improve clinical outcomes even in the presence of chronic inflammation (2). Furthermore, in NPGS, bleb manipulation is not required as it is bleb independent and subsequent anti-metabolite sub conjunctival injections are not needed. The conjunctiva has little or no scarring leaving room for future glaucoma surgeries if needed. However, one must be aware that this procedure is likely to be unsuccessful in UG patients with peripheral anterior synechiae (PAS) as the outcome depends on a clear trabecular Descemet membrane window (TDW).
Souissi et al. (18) reported a mean decrease of IOP by 52.9% in eight eyes that underwent DS without antimetabolites with mean follow-up of 42.2 months. Complete success was obtained in 50%, relative success was obtained in 37.5% and failure in 12.5%. Additionally, Dupas et al. (19) found a success rate of 89% at 12 months in the patients’ group that underwent trabeculectomy with MMC and a success rate of 88% in the group that underwent DS with MMC and with implant but with an increase number of postoperative adjustments in the DS group such as goniopuncture in 45% of the cases and needling in 15%. Similar results have been reported by Obeidan et al. (9) in patients that underwent DS with implant where a complete success rate was achieved in 84.6%, a qualified success was achieved in 7.7% and a complete failure in 7.7%.
Miscrocchi et al (2) reported a qualified success in 90.9% with a mean IOP reduction of 46.4% in all eyes that had VC and a mean final IOP of 18.1 mmHg. In our study, qualified success was achieved in 94% of patients at year 1, and 75% of patients at year 5 with a mean final IOP of 15.70. In the group of patients requiring further surgery, 50% of patients had previous surgeries. There was a mean drop in eye drops at 6 months of 0.25, at 1 year of 0.5 mean and at 5 years of 1.1 post-operatively.
The only complication encountered was TDW perforation which occurred in 2 patients. None of the eyes had endophthalmitis, flat anterior chamber, surgically induced cataract or a flare-up of their uveitis.
The overall outcome of VC in this study is very encouraging yielding an overall success rate of 9/10 (90%). The average reduction in IOP in this study was 49.6%.
To date there are no studies on PVC in uveitic glaucoma. Combined phaco-trabeculectomy has been reported to be less successful than trabeculectomy alone (20, 21) and most glaucoma surgeons advise the use of MMC in combined procedures (22). It has been suggested that the reason for a worse outcome following combined procedure is the added inflammation occurring from the cataract extraction giving rise to an increase in release of inflammatory mediators such as transforming growth factor-beta 1, which promote subconjunctival/episcleral scarring and failure of the filtering bleb (20, 21). By performing PVC, these aqueous derived vasoactive stimulators of fibroblast activation cross the TDW into Schlemm’s canal and probably leave the eye either through existing collector channels, or through the uveoscleral outflow and therefore do not come into contact with the episcleral/subconjunctival space fibroblasts as they would after trabeculectomy (23).
There are several limiting factors in this study. This is a retrospective case series with limited sample size and the included eyes were heterogeneous based on type of uveitis, type of procedure and lens status.
None of our patients were on immune-modulating medication and this may
have affected our success rate. Also, of relevance is that we only had idiopathic uveitis and Fuch’s heterechromic cyclitis (FHC) and hence a direct comparison with other studies (with different uveitic aetiology) is not possible. The limited sample size, heterogeneous subjects further reduces the possibility for analytical approach to the data. Also relevant is that the results are of surgery in a subset of patient with mild form of uveitis and cannot be easily generalized to other types of uveitis.