The 3-year results of the Tube Versus Trabeculectomy (TVT) Study have concluded that tube shunt surgery had higher success compared to trabeculectomy with mitomycin C, and the incidence of complications was greater in the trabeculectomy group. The accompanying editorial states that “success rates in the tube shunt group must be tempered by jaded cynicism.” We would like to explore the possibility that the study design and the reporting of the results are biased in favor of tube implant.
At the end of 3 years, 62% in the tube group and 58% in the trabeculectomy group achieved intraocular pressure (IOP) less than 14 mm Hg ( P = .76, χ 2 test). In an intent-to-treat analysis, the final IOP was 13.0 ± 4.8 mm Hg in the tube group and 13.1 ± 6.5 mm Hg in the trabeculectomy group ( P = .91, Student t test). If the primary outcome measure was IOP control, then these results indicate at least equality between the groups.
The conclusion that the failure rate was greater in the trabeculectomy group seems to be because of definitions of failure. The greater failure in the trabeculectomy group is attributable to greater number of repeat surgeries in the trabeculectomy group than in the tube group. It looks like the threshold for second surgery was low in the trabeculectomy group as compared to the tube group. The authors do concede that additional surgery in eyes that failed a tube shunt surgery was complex and involved placement of a second tube or cyclodestruction. This would result in more medications in the tube group for control of IOP (resulting in greater qualified success). The authors say that they have explored the possibility that the surgeons may have had a higher threshold to perform additional surgery in the tube group than in the trabeculectomy group. They claim that there was no selection bias for additional glaucoma surgery, as the mean IOP before repeat surgery was similar in both groups. We would like to suggest that the number of eyes with qualified success in the 2 groups is a correct indicator for this bias. Qualified success (which indicates higher medication use to keep the eyes in the successful group) is not formally compared in Table 3. The qualified success in the tube shunt group (54%) is double that in the trabeculectomy group (26%). This difference (28.0%; 95% CI: 13.8, 42.2) is highly significant ( P = 0.0004, χ 2 test). This significant difference in qualified success between the groups would indicate a higher threshold for second surgery in the tube group.
Though the number of overall complications were more in the trabeculectomy group, the serious complications associated with reoperation and/or vision loss during the 3 years of follow-up were similar in both groups (Table 8).
We tend to agree with the editorial that “clinicians may have different views interpreting the rich body of data in the TVT trial,” and feel that the final verdict on the supremacy of tube shunt over trabeculectomy is still open.