When confronted with the glaucoma patient requiring surgical intervention for filtration, the decision between selecting guarded filtration surgery (trabeculectomy) or an aqueous shunting device (tube shunt) is controversial. Trabeculectomy is the most common procedure selected, although the number performed is decreasing. In contrast, the number of tube shunts performed is increasing, but these remain much less frequent than trabeculectomy. Until the Tube Versus Trabeculectomy Study (TVT), there was little information comparing the 2 procedures. Previous publications have reported the design and 1-year and 3-year results. With the publication of the 2 articles describing the 5-year results concerning treatment outcomes and complications, there are now long-term, high-level data comparing these 2 procedures in glaucoma patients. This commentary discusses the possible implications of the TVT in clinical practice.

The TVT was a randomized, multicenter clinical trial comparing the safety and efficacy of tube shunt surgery (350 mm ² Baerveldt) with those of trabeculectomy with mitomycin C (MMC) 0.4 mg/mL for 4 minutes in 212 eyes of 212 glaucoma patients. Included were those with previous trabeculectomy failure, previous cataract surgery, or both. This is a heterogeneous group ranging from those eyes failing a previous trabeculectomy with MMC filter to eyes having undergone only clear cornea phacoemulsification, likely having a different expected response to filtration. However, the cases included are representative of those in which either procedure is potentially a reasonable option.

At 5 years, it was reported (1) that there was a higher success rate with tube shunt implantation than trabeculectomy, (2) that there was a higher rate of reoperation for glaucoma in the trabeculectomy group, and (3) that there was no difference in vision loss found between the 2 groups. Intraocular pressure (IOP) reduction and the number of supplemental IOP-lowering medications were not found to be different. There were a large number of early postoperative complications, more frequent in the trabeculectomy group, with most being transient and self-limited. Late postoperative complications, reoperation for complications, and cataract extractions were not different between the procedures.

The apparent improved outcome with tube shunt surgery compared with trabeculectomy must take into account several factors that may affect these findings. Among the most important considerations is the change in trabeculectomy technique since the TVT was designed and performed. Long-term studies typically suffer from this weakness as procedures evolve and improve, whereas the study protocol for the procedure does not. Lower doses and less duration as well as a larger area of MMC application are current practice to limit the risk of hypotony and bleb leaks.

The definition of failure in this study, including reoperation and hypotony, is noteworthy. Although most failures in the trabeculectomy group were attributed to inadequate IOP control, which potentially could be increased with less MMC, conversely, current use of MMC also may be expected to decrease failure resulting from hypotony (13 patients) that would have a substantial impact on success rates. Because failure included those requiring glaucoma reoperation, the authors point out that this may result in bias, because the next step after a failed trabeculectomy is tube shunt, and the next procedure after a tube shunt is likely a second tube or a cyclodestructive procedure, which are considered much less desirable. In this study, more reoperations occurred in the trabeculectomy group (n = 18) compared with the tube group (n = 8).

Although there were no statistically significant differences in IOP between groups, there actually was a 2.2-mm Hg difference in mean IOP reduction from baseline and a 1.8-mm Hg difference in mean IOP in those completing the 5-year study that favored trabeculectomy. Additionally, in evaluating the visual acuity of patients, the trabeculectomy group nearly reached statistically significant better Early Treatment Diabetic Retinopathy vision.

The 5-year TVT data were different than the findings reported in the 1- or 3-year end points. Specifically, the fewer medicines in the trabeculectomy group, although less than the number used in the tube group, was no longer statistically significant. Diplopia that was more common in the 1- and 3-year results in the tube group lost statistical significance in the 5-year data because of 2 cases in the trabeculectomy group, but this diplopia occurring several years after surgery likely is not related to the initial surgical intervention.

In evaluating the complications from both procedures, the interventions were overestimated because suture lysis, 5-fluorouracil injections, and removal of ripcords were included. Generally, these should be considered an expected part of the procedure, and although they occasionally result in complications, those would be identified separately. In the early complications, the difference in the 2 groups largely was related to wound leaks and hyphema in the trabeculectomy group, which mostly are self-limited. These did not increase the risk of failure and did not correlate with outcome. Late complications were similar in numbers in the 2 groups and, not surprisingly, were those most commonly associated with the procedures: cornea edema, diplopia, and tube erosion with the tube shunt and bleb dysesthesia, bleb leak, and blebitis with trabeculectomy. There were fewer reoperations for complications in the trabeculectomy group, but the major differences again were those expected with the procedure and its revision.

The most important clinical feature is the serious complications. Both of these procedures do have substantial risk that must be recognized. In this study, with the exception of corneal edema, slightly more common in the tube group, other serious complications and causes of decreased vision were relatively infrequent. Clinically, although the complications came from a variety of infrequent occurrences, approximately 20% of patients undergoing either procedure will have a serious complication within 5 years. The impact of this potential on these patients should be taken into consideration. It is very informative that in the TVT, there was no difference in the rate of visual loss between those in whom postoperative complications developed and those in whom they did not, either in presence or number, unlike some previous reports.

The authors should be congratulated for providing excellent-quality long-term descriptive and comparative information on the outcomes, including efficacy and complications of trabeculectomy and tube shunts, in this clinically challenging population. The information concerning IOP and visual outcomes is extremely valuable in decision making going forward, taking into consideration the possible implications of newer techniques, particularly for trabeculectomy, that could result in a more favorable profile. The TVT reflects the increasing use of tube shunts earlier in the therapeutic process and provides valuable information in this population. It demonstrates that both trabeculectomy and tube shunts have the potential to control IOP in glaucoma patients successfully, but that both have potential complications that are often specific to the procedure. Based on these data, increased indications for tube shunts, including all eyes with previous conjunctival incisions, such as limbal cataract extraction and trabeculectomy, seem reasonable.

The author’s conclusion is a key point. Although randomized, controlled, prospective trials are vital for scientific evaluation, the results should be a consideration in making individual patient treatment decisions, but other factors, such as surgeon experience and patient-specific factors, are extremely important.