The AVB Study is an international, multicenter trial that randomly assigned 238 patients with refractory or high-risk glaucoma to implantation of one of two glaucoma drainage devices (Ahmed-FP7 Valve, New World Medical or Baerveldt-350 Implant, Abbott Medical Optics). After 2 years, both devices were effective in lowering IOP and the need for glaucoma medications.
The Baerveldt group had a significantly lower failure rate than the Ahmed group (33% versus 48%) and required fewer glaucoma medications (mean 1 versus 1.8), but required a greater number of interventions. Vision deteriorated from baseline on average in both groups, and both devices had high rates of complications and interventions, although most were transient or minor.
"Although these two glaucoma drainage devices are used frequently, all comparisons to date have been retrospective, nonrandomized, and compared different device models and patient populations using different criteria for success," said Christakis, a fourth-year medical student at Yale University, New Haven, CT, and study director and lead author of the AVB Study.
"This has made the results difficult to interpret," he said. "We hope that our results, as well as those of a concurrent multicenter randomized trial (The Ahmed Baerveldt Comparison Study), will provide insight into the relative efficacy of these devices."
The AVB Study is under the leadership of Ike Ahmed, MD, assistant professor, Department of Ophthalmology and Vision Sciences, University of Toronto, Ontario, Canada. James Tsai, MD, chairman, Department of Ophthalmology and Visual Science and the Robert R. Young Professor of Ophthalmology and Visual Science, Yale School of Medicine, is a consultant and co-author of the AVB Study and Christakis' thesis advisor at Yale.
About the AVB Study
Seven clinical centers and 10 surgeons are participating in the AVB Study. A standardized surgical technique was employed that included ligating the Baerveldt implant with a releasable suture. The option to place proximal fenestrations was made at the discretion of the surgeon in patients with advanced glaucoma requiring immediate IOP reduction. No additional procedures were performed at the time of implantation, and the rate of significant intraoperative complications was 4% in both groups.
At baseline, both groups were similar in demographic and ocular characteristics, although there were more women in the Baerveldt group. Mean patient age was 66 years, mean IOP was 31 mm Hg, mean glaucoma medications was 3, and median Snellen acuity was 20/100. About half of the eyes had open-angle glaucoma, 20% had neovascular glaucoma, and 10% had uveitic glaucoma; 40% of eyes had a previous failed trabeculectomy and 25% had undergone laser trabeculoplasty.
Complete success was defined as an IOP of 5 to 18 mm Hg with a reduction of at least 20% from baseline at all visits beginning at 3 months, with no significant loss of vision, significant complications, or additional glaucoma procedures required. Success was qualified if glaucoma medications were required in order to achieve no two consecutive visits after 3 months in which the IOP criteria were not met. Progression to no light perception vision, devastating complications, and additional glaucoma surgery were considered failures. High IOP was the most common cause of failure in both groups, accounting for about two-thirds of failures in the Ahmed group and half of failures in the Baerveldt group.
In the early postoperative period, the Ahmed group had lower IOP and required fewer medications than the Baerveldt group, which can be attributed to the Baerveldt tubes being ligated. By 2 months, the Baerveldt group had a lower IOP and required fewer medications, a trend that continued through 2 years. Although the mean IOP-lowering effect was similar in both groups at the 2-year visit (48% Ahmed, 54% Baerveldt), the Baerveldt implant reduced the need for glaucoma medications more than the Ahmed valve (69% versus 43%). Visual acuity data showed a similar worsening in the two treatment groups, corresponding to a decrease in median Snellen visual acuity from 20/100 at baseline to 20/200 at 2 years.
The overall complication rate was lower in the Ahmed group, but not significantly different from in the Baerveldt group (49% versus 61%). Devastating hypotony-related complications occurred only in Baerveldt eyes (7% versus 0%), resulting in poor visual outcomes for these patients. The Baerveldt group also had a significantly higher rate of corneal edema (12% versus 5%), while bleb encapsulation was significantly more common in the Ahmed group (11% versus 3%). Interventions were performed significantly more often in eyes receiving the Baerveldt implant than the Ahmed valve (47% versus 32%), driven by a greater need for paracentesis to control IOP spikes in the early postoperative period (14% versus 4%) and also by a higher number of tube revisions to address tube obstruction (12% versus 6%).
"Follow-up in the AVB Study will continue until all patients reach 5 years, but the interim data show that each device has advantages and disadvantages," Dr. Ahmed said. "Consequently, selecting a device should be individualized, taking into account the profiles of each device, the surgeon's experience, and the patient's treatment goals."
Published papers on the AVB Study include one reviewing the design, baseline patient characteristics, and intraoperative complications [Ophthalmology. 2011;118:2172-2179] and another reporting the 1-year results [Ophthalmology. 2011;118:180-189].