advertisement

Editors Selection IGR 18-1

Comments

Comment by Tina Wong on:

24877 Three-year follow-up of the tube versus trabeculectomy study, Gedde SJ; Schiffman JC; Feuer WJ et al., American Journal of Ophthalmology, 2009; 148: 670-684

Find related abstracts

The TVT trial was one the best-conceived and well-written studies published on this subject. Although this study found a higher success rate at three years in the tube shunt group, there are at least three significant issues to consider before burying trabeculectomy in favor of shunts for your individual patient.

1. The study sample is too heterogeneous and divergent in terms of prior ocular surgery, with subjects ranging from having had simple phacoemulsification (the largest single group in this study) to having had prior MMC trabeculectomy. Although both are technically 'ocular surgery', these are very different surgeries and their impact on the success of glaucoma surgery is likely, or at the very least possibly, different. In essence, the TVT is four studies in one with four different study populations. It is unclear whether the study was originally designed in a predetermined 'stratified sampling' design (if so, certainly would need a lot large sample size), or whether this is 'poststratification'. Is it possible that such a divergent sample was combined into one large group to enhance enrollment/numbers? Regardless, as a result,it makes it quite difficult to generalize the overall results, with enough statistical power, to an individual strata or furthermore to an individual patient. Although one could compare the results within each stratum and between strata, there are insufficient numbers (underpowered) within each of the four strata to do so. So we are left wondering about whether to choose a tube or a trabeculectomy for one patient with a history of prior phaco, and for another patient with prior trabeculectomy. We simply do not have large enough numbers to provide sufficient evidence of superiority for these individual groups, and to combine them in one large group, as the TVT trial did, makes for a very divergent group.

2. Looking at the Kaplan Meier curves, most of the failures in the trabeculectomy group were early on (i.e., around six months). After this point, the survival curves of both groups appear to be quite similar. Why the early failures in the trab group? Here's one explanation. One of the things that stands out in the early postop complications are the high number of early bleb leaks (10%) in the trab group, which seems to be higher than other recent large studies (5% in CIGTS¹ and 6.5% in AGIS²). As early wound leaks have been associated with bleb failure,³ could this be why we see the significant number of early failures in the trab group resulting in the tube-shunts being found to be overall superior? Perhaps for those surgeons who don't have such a high leak rate, or with alternative/ water-tight conjunctival closures, the success rate would be higher and thus no different than tube-shunts?

3. Even if we were certain these results could be generalized, and if we ignore the possible issue of trab technique/closure and failure, the fact is that when we look at qualified success rates over three years, there was no difference. That looks pretty good for trabeculectomy when we consider the most worrisome issue with otherwise uncomplicated tube-shunts is long-term corneal endothelial loss, which appears to be progressive (~20% at 24 months) and could be time-bomb for many of our patients.⁴ Unfortunately, it would take many years - likely ten years - to assess this complication and differences between the groups.

No doubt, the TVT study has shed more light on selection of glaucoma surgery in patients with prior ocular surgery (namely cataract extraction and/or trabeculectomy). Unfortunately the study sample is non-uniform and thus may not be generalized, and individual subgroups are too small to provide conclusive evidence of superiority within each strata.

Until further evidence, it still boils down to individualizing approach to our given patient

Although there is an argument based on this study to shift towards tube-shunts over trabeculectomy in patients with prior surgery, I think if you do a good MMC trabeculectomy, minimize early wound leaks and use postoperative adjuncts like needling when needed, IOP results are just as good as a tube-shunt with less cost and without the long-term risk of corneal decompensation. Until further evidence, it still boils down to individualizing approach to our given patient.

• ← Previous
• Next →