advertisement

Editors Selection IGR 12-1

Comments

Comment by Tarek Shaarawy 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

Three-year results from Tube versus Trabeculectomy Study: are we going to change our surgical behavior?

The surgical management of difficult glaucomas may be slowly changing, for better we hope. Traditionally, glaucoma specialists tended to performing a second trabeculectomy in eyes that failed a first one. The same is true of pseudophakic eyes, especially when there is no significant conjunctival scarring. This has been based on the premise that placement of a glaucoma drainage device is bound to be less successful than a second trabeculectomy or a first trabeculectomy in pseudophakic eyes. Surgical management of glaucoma in such eyes took another turn with the publication of the three-year results from the Tube versus Trabeculectomy Study (TVT). The TVT investigators reported that, over three years of follow-up, placement of a 350-mm2 Baerveldt implant led to overall equivalent IOPs with a lower risk of failure in eyes with prior trabeculectomy or cataract surgery compared to eyes that had a trabeculectomy with Mitomycin C. The results are based on 158 patients from the original 212 patients. Ninety-three percent of follow-up visits were completed by three years.

There are some facts about the two groups that are worthy to be mentioned and kept in mind when interpreting the results. Most patients in the trabeculectomy group underwent limbus-based surgery. A higher concentration of MMC (0.4 mg/ml for four minutes) compared to our current standards was used. Interestingly, about one-third of the failure rate in the trabeculectomy group was related to hypotony. The surgical technique for trabeculectomy was not standardized among surgeons, a reflection of the multicenter nature of the study. The TVT investigators described failure as IOP > 21 mmHg or less than 20% IOP reduction, or IOP ≤ 5 mmHg, reoperation for glaucoma, or loss of light perception. Incidence of failure (and hypotony) was higher in the trabeculectomy group according to above criteria and also when failure was described as IOP ≤ 17 or ≤ 14 mmHg. It is notable that as more stringent IOP cutoff points were used the difference in failure rates between the two groups became progressively smaller and became non-significant for cutoff points of ≤ 17 mmHg (p = 0.057) and ≤ 14 mmHg (p = 0.12) and the absolute differences also became less and less significant clinically although the Baerveldt group still tended to do better. The latter criteria are more consistent with our current thinking about target pressures especially in a group of eyes with an average MD of −16 dB.

The average IOP after surgery was comparable in the two groups at one and three years, although there was a trend for better IOP control in the trabeculectomy group during the first year, likely a result of a higher rate of hypotony in these eyes. The IOP outcomes should be interpreted taking into account the fact that the trabeculectomy eyes were on fewer medications during the two years of follow-up, but the number of supplemental medications slowly increased over the three years of follow-up in this group. One of the interesting findings was the somewhat larger treatment difference between the two groups in eyes with prior cataract extraction. Only one-third of the pseudophakic eyes had prior clear-cornea cataract surgery although the trabeculectomy eyes were more likely to have had a prior clearcornea phacoemulsification (34% vs. 20%), which would have been expected to favor the trabeculectomy group.

The incidence of serious postoperative complications (22% in the tube group vs. 27% in the trabeculectomy group) was similar and the average reduction of visual acuity was also comparable in the two groups. Despite the more common prevalence of hypotony in the trabeculectomy group, incidence of loss of two lines of vision was similar between the two groups (34% loss of two lines vs. 31% in the tube group). Only one patient in the trabeculectomy group required bleb revision for hypotony.

What are we to make of these findings? Are the reported results for trabeculectomy consistent with the results for this group of patient in the literature? My answer is a tentative yes. Outcomes reported with regard to results of repeat trabeculectomy/MMC or trabeculectomy/ MMC in pseudophakic eyes are consistent with those reported in TVT with one caveat. Unpublished results from our center seem to indicate that outcomes of an initial trabeculectomy after prior clearcornea phacoemulsification are almost as good as those in phakic eyes. Generalizability of these surgical results should be a function of how concordant glaucoma surgeons think their perioperative techniques and trabeculectomy outcomes are with those of the TVT. Also, we need to remember that hypotony is compatible with good visual function and is not necessarily a bad outcome. In eyes in need of very low target pressures, it might even be the desired outcome. One more thing to remember, as the authors have rightfully mentioned, is that placement of a tube is almost always an option after a trabeculectomy, albeit at the cost of a possibly higher failure rate, while the reverse may not be true. Another issue that has to be borne in mind is that the baseline risk for failure was not quite uniform across the study sample. Eyes with a prior clear-cornea phacoemulsification are probably at lower risk of failure than eyes that have had a failed combined same-site superior phacotrabeculectomy with MMC. The fact that there were no treatment differences across diagnostic strata is most likely the result of low power when such interactions are explored. Also, some complications, such as endophthalmitis and other bleb-related issues are expected to be less frequent with a fornix-based approach.

How will I incorporate the results in my practice? As mentioned above, I tend to separate, in my mind, eyes with a prior temporal clear-cornea phacoemulsification from eyes with prior conjunctival surgery, be it cataract surgery or trabeculectomy. For the former, I will continue to do a fornix-based trabeculectomy with lower doses of MMC (0.2-0.28 mg/ml for one minute). Also, for patients with target pressures of ten or less, I will continue to use trabeculectomy if there is an adequate area of healthy conjunctiva in the superior quadrant. For the second category, I will seriously consider placing a 350-mm² Baerveldt implant as an alternative to trabeculectomy with Mitomycin C. I will discuss the risks and benefits of the tow types of surgeries with the patient and tailor my decision according to patient's wishes. I will place a Baerveldt implant in cases where a competent glaucoma surgeon performed the first trabeculectomy, which failed within a short period of time after surgery.

It is important that, as the authors have recommended, the results are not generalized to other glaucoma patients with different risk profiles for failure and to other glaucoma implants. Results have shown that outcomes after insertion of Ahmed Glaucoma Valve are not as good as those after trabeculectomy with MMC.

I would like to congratulate the TVT investigators for providing us with vital information regarding surgical management of difficult glaucoma patients. Results of a similar study in eyes without prior ocular surgery are pending but in the mean time, I expect that we'll see more frequent and earlier use of glaucoma drainage devices in eyes with a history of prior conjunctival surgery.

• ← Previous
• Next →