Efficacy of Combined Retisert and Ahmed Glaucoma Valve for Medically Uncontrolled Uveitis and Glaucoma
We read with interest the article by Hennein et al which demonstrated a lower mean intraocular pressure (IOP) and the number of glaucoma drops used at 1 year postoperatively in patients with uncontrolled uveitis who received both Retisert and Ahmed glaucoma valve (AGV) compared with those with different etiologies of medically uncontrolled glaucoma that received AGV only. The authors concluded that combined Retisert-AGV may improve surgical outcome.1 We would like to highlight some aspects which, in our opinion, merit further discussion and considerations.
Retisert provides sustained and localized corticosteroid release into the vitreous cavity. A more direct approach would be pars-plana tube placement instead of anterior chamber placement given the closer proximity between the pars-plana tube and the pars-plana Retisert, this technique demonstrated good uveitis and IOP control without significant adverse effects.2 Hennein et al reported a difference in outcomes at 1 year only, whereas figures 1 and 2 in their article suggest there are no differences between any of the time points before 1 year.1 Hypertensive phase is most commonly seen in the early postoperative course after AGV,3 if Retisert does indeed modulate the endplate encapsulation with its antifibrotic and anti-inflammatory abilities, we would expect an early and sustainable effects. Although there is no single definition of failure in glaucoma surgery, cumulative probability of failure from a Kaplan-Meier analysis and the log-rank test are useful in determining whether the addition of Retisert has an effect on AGV survival before we can conclude Retisert can improve surgical outcomes. Cumulative success or failure probability is an important primary outcome, but was not reported in the current study.4
The study eyes had a range of time lag (0 to 18 mo) between Retisert and AGV insertions, the authors attempted to justify this time lag difference is not a confounding factor with a multivariate regression analysis which failed to demonstrate that it had an effect on the outcomes at 1 year. Most of the study patients had AGV after Retisert, we believe the observed reduction in IOP and glaucoma drops at 1 year may reflect the better controlled uveitis activity with Retisert and the IOP lowering effect of AGV which these patients later required, thus may not reflect the corticosteroid endplate modulation hypothesis.5 In the control group, they consist of different etiologies of glaucoma. The authors attempted to justify their method by demonstrating that there was no difference at the 1 year IOP between the uveitic eyes and primary open angle glaucoma eyes in the control group. However, patterns of IOP reduction between the uveitic glaucoma and primary open angle glaucoma are different after AGV implantation.6 Furthermore, failure to detect a statistical difference by these post hoc analysis does not prove such a difference does not fundamentally exist. The small sample size in the study group and the unbalanced etiologies in the control eyes may have inadequate power to detect this difference in a retrospective study. The confounding effects in the study and control group may have affected the results.
Combined Retisert-AGV has not been shown to result in serious adverse effect.7 However, cataract formation has been reported as one of the complications of Retisert.8 We believe that Retisert may not be contraindicated in eyes with uncontrolled uveitis and glaucoma if implanted at the same time as AGV, but this combination surgery should not be used indiscriminately in patients with uncontrolled glaucoma especially in pediatric cases because of the risk of cataract formation,8,9 as the additional benefits, if any, still remains unanswered.