Purpose
To identify the risk factors associated with failure of tube shunt surgery.
Design
Pooled analysis of 3 prospective multicenter, randomized clinical trials.
Methods
A total of 621 patients with medically uncontrolled glaucoma were enrolled, including 276 from the Ahmed Baerveldt Comparison Study, 238 from the Ahmed Versus Baerveldt Study, and 107 from the tube group of the Tube Versus Trabeculectomy Study. Patients were randomized to treatment with an Ahmed glaucoma valve (model FP7) or Baerveldt glaucoma implant (model 101-350). The associations between baseline risk factors and tube shunt failure were assessed using a Cox proportional hazards regression model. The primary outcome measure was the rate of surgical failure defined as intraocular pressure (IOP) > 21 mmHg or reduced < 20% from baseline, IOP ≤ 5 mmHg, loss of light perception vision, reoperation for glaucoma, or removal of implant.
Results
The cumulative probability of failure after tube shunt surgery was 38.3% after 5 years. In multivariable analyses, baseline factors that predicted tube shunt failure included preoperative IOP (≤ 21 mmHg compared to IOP > 21 and ≤ 25 mmHg; HR, 2.34; 95% CI, 1.52-3.61; P < .001), neovascular glaucoma (HR, 1.79; 95% CI, 1.28-2.52; P = .001), randomized treatment (for Ahmed glaucoma valve; HR, 1.36; 95% CI, 1.04-1.78; P = .025), and age (for 10 year decrease in age; HR, 1.19; 95% CI, 1.09–1.31; P < .001).
Conclusions
Lower preoperative IOP, neovascular glaucoma, Ahmed implantation, and younger age were predictors of tube shunt failure. This Study provides the largest prospectively collected dataset on tube shunt surgery.
INTRODUCTION
P ractice patterns in the surgical management of glaucoma have changed in recent decades. Tube shunt surgery has traditionally been indicated for reduction of intraocular pressure (IOP) in patients who have failed trabeculectomy or have secondary glaucoma at high risk of filtration failure, such as neovascular and uveitic glaucoma. However, growing experience with tube shunts has prompted many surgeons to use these devices in patients with less refractory glaucoma. Medicare claims data show a 410% increase in tube shunt implantation and a 72% decrease in trabeculectomy procedures between 1994 and 2012. Recent surveys of the American Glaucoma Society membership have shown a steady rise in the proportion of surgeons who are selecting tube shunts as an alternative to trabeculectomy in a variety of clinical settings.
Recent multicenter, randomized clinical trials have evaluated the safety and efficacy of tube shunts. The Ahmed Baerveldt Comparison (ABC) Study and Ahmed Versus Baerveldt (AVB) Study enrolled patients with refractory glaucoma requiring tube shunt surgery, and they were randomized to placement of an Ahmed glaucoma valve (AGV; New World Medical) or Baerveldt glaucoma implant (BGI; Abbott Medical Optics). , The Tube Versus Trabeculectomy (TVT) Study recruited patients with medically uncontrolled glaucoma who had prior cataract extraction with intraocular lens implantation and/or failed trabeculectomy, and randomly assigned them to treatment with a BGI or trabeculectomy with mitomycin C. Data throughout 5 years of follow-up are available from these clinical trials.
Several risk factors for failure of trabeculectomy have been identified and include younger age, black or Hispanic race, prior conjunctival surgery, prior laser trabeculoplasty, higher preoperative IOP, and secondary glaucoma such as uveitic, neovascular, and iridocorneal endothelial syndrome-associated glaucomas. These patient characteristics are largely thought to result in increased scarring after filtration surgery. There is limited information in the medical literature regarding risk factors associated with failure of tube shunt surgery. Case series have suggested that younger age, male gender, African descent, lower preoperative IOP, and neovascular glaucoma are predictors of tube shunt failure. The current Study aimed to better elucidate the risk factors for failure of tube shunt surgery by utilizing pooled data from the ABC, AVB, and TVTStudies.
METHODS
Data were obtained on a per patient level and pooled from 3 independent prospective randomized clinical trials, including the ABC Study, AVB Study, and the tube group of the TVT Study. Patients were recruited from multiple clinical centers in each trial (see Appendix). Written informed consent was obtained from all subjects for both treatment and research. The Institutional Review Board at each clinical center approved the study protocol before initiating recruitment. Each study conformed to the tenets of the Declaration of Helsinki. The studies were registered on ClinicalTrials.gov (ABC Study: NCT00376363, AVB Study: NCT00940823, and TVT Study: NCT00306852). The design and methods of the ABC Study, AVB Study, and TVT Study, were previously described in detail, and are summarized below.
ELIGIBILITY CRITERIA
All patients in this study were aged ≥18 years and had uncontrolled glaucoma on maximum tolerated medical therapy. The TVT Study included patients who had undergone prior cataract extraction with intraocular lens implantation and/or failed trabeculectomy. The ABC and AVB Studies included patients at high risk of trabeculectomy failure because of previous ocular surgery or secondary glaucoma, such as neovascular, uveitic, or iridocorneal endothelial syndrome-associated glaucoma. In contrast, the TVT Study excluded eyes with superior conjunctival scarring precluding trabeculectomy, aphakia, active iris neovascularization, active proliferative retinopathy, chronic or recurrent uveitis, or iridocorneal endothelial syndrome because of the increased risk of filtration failure. Other exclusion criteria for all patients included a prior cyclodestructive procedure, no light perception vision, unwilling or unable to provide informed consent or otherwise fulfill the requirements of the study, or need for additional procedures at the time of glaucoma surgery (eg, cataract surgery, penetrating keratoplasty, or retinal surgery).
RANDOMIZATION
Patients in the ABC and AVB Studies were randomly assigned to treatment with an AGV (model FP7) or BGI (model 101-350). Patients recruited in the TVT Study were randomized to a BGI (model 101-350) or trabeculectomy with mitomycin C (0.4 mg/mL for 4 minutes), and only patients in the tube group were included in the present study. The ABC and TVT Studies used a permuted variable block randomization design with stratification. The AVB Study: used simple randomization without stratification. Neither the patient nor the clinician was masked to the randomization assignment.
SURGICAL TREATMENT
All implants were placed in the superotemporal quadrant, with insertion of the tube into the anterior chamber. A limbus-based or fornix-based conjunctival flap was used depending on the surgeon’s preference. The end plate was sutured to sclera 8 to 10 mm posterior to the limbus. The AGV was primed with balanced salt solution. The BGI was completely occluded to temporarily restrict aqueous flow; tube fenestration and the method of tube occlusion was left to the discretion of the surgeon. A patch graft was used to cover the limbal portion of the tube. The conjunctival flap was closed with sutures.
PATIENT VISITS
Baseline ocular and demographic characteristics were collected preoperatively. Enrolled patients had follow-up visits scheduled at 1 day, 1 week, 1 month, 3 months, 6 months, 12 months, 18 months, 2 years, 3 years, 4 years, and 5 years. Best-corrected Snellen visual acuity and IOP were assessed at each follow-up visit.
OUTCOME MEASURES
The primary outcome measure in this pooled analysis was the cumulative probability of failure. Surgical failure was defined as IOP > 21 mmHg or reduced < 20% from baseline on 2 consecutive visits after 3 months, IOP ≤ 5 mmHg on 2 consecutive visits after 3 months, loss of light perception vision, reoperation for glaucoma, or removal of implant. Reoperation for glaucoma was defined as additional glaucoma surgery requiring a return to the operating room. Cyclodestruction was also considered a reoperation, irrespective of whether it was performed in the clinic or operating room.
STATISTICAL ANALYSIS
The time to failure was defined as time from surgical treatment to reoperation for glaucoma, loss of light perception vision, implant removal, or the first of 2 consecutive study visits after 3 months, in which the patient had persistent hypotony (IOP ≤ 5 mmHg) or inadequately reduced IOP (IOP > 21 mmHg or reduced < 20% below baseline). Risk factors for failure were evaluated with the Kaplan-Meier survival log-rank test and Cox proportional hazard regression analysis. Multivariable analysis was performed using forward stepwise entry of risk factors into a multiple Cox proportional hazard regression model. P < .05 was considered statistically significant.
RESULTS
BASELINE CHARACTERISTICS
The study had a total of 621 patients, including 276 patients from the ABC Study, 238 patients from the AVB Study, and 107 patients from the tube group of the TVT Study. Baseline demographic and ocular characteristics are shown in Table 1 . The age (mean ± SD) of the overall group was 65.8 ± 14.3 years, and 329 patients (53%) were female. The preoperative IOP (mean ± SD) of the entire study group was 30.4 ± 10.8 mmHg on 3.3 ± 1.1 glaucoma medications. The most common diagnosis was primary open-angle glaucoma in 318 patients (51%), and 424 patients (68%) were pseudophakic at baseline. Patients were older and had lower preoperative IOP in the TVT Study compared with the ABC and AVB Studies. A significant proportion of patients had a diagnosis of neovascular glaucoma in the ABC Study (29%) and AVB Study (21%). No patients in the TVT Study had a diagnosis of NVG, uveitic glaucoma, or aphakic glaucoma.
| Overall Group (n = 621) | ABC Study | AVB Study | TVT Study | |||
|---|---|---|---|---|---|---|
| Ahmed Group (n = 143) | Baerveld tGroup (n = 133) | Ahmed Group (n = 124) | Baerveldt Group (n = 114) | Tube Group (n = 107) | ||
| Age (years), mean ± SD | 65.8 ± 14.3 | 65.4 ± 12.8 | 62.2 ± 14.2 | 65 ± 17 | 67 ± 15 | 70.9 ± 11.0 |
| Race, n (%) White Black Hispanic Other | 353 (57) 136 (22) 57 (9) 72 (12) | 66 (46) 43 (30) 12 (8) 22 (16) | 68 (51) 25 (19) 21 (16) 19 (14) | 91 (73) 15 (12) 5 (4) 13 (11) | 76 (69) 13 (11) 7 (6) 15 (13) | 52 (49) 40 (37) 12 (11) 3 (3) |
| Sex, n (%) Male Female | 289 (47) 329 (53) | 73 (51) 70 (49) | 70 (52) 63 (48) | 62 (52) 59 (48) | 41 (36) 73 (64) | 43 (40) 64 (60) |
| IOP (mmHg), mean ± SD | 30.4 ± 10.8 | 31.2 ± 11.2 | 31.8 ± 12.5 | 31.1 ± 10.5 | 31.7 ± 11.1 | 25.1 ± 5.3 |
| Glaucoma medications (n), mean ± SD | 3.3 ± 1.1 | 3.4 ± 1.1 | 3.5 ± 1.1 | 3.1 ± 1.0 | 3.1 ± 1.1 | 3.2 ± 1.1 |
| Snellen VA Median Range | 20/70 20/15-LP | 20/80 20/15-LP | 20/70 20/15-LP | 20/100 20/20-LP | 20/100 20/20-LP | 20/30 20/17-HM |
| Study eye, n (%) Right Left | 319 (51) 293 (47) | 83 (58) 60 (42) | 65 (49) 59 (48) | 65 (52) 59 (48) | 63 (55) 51 (45) | 43 (40) 64 (60) |
| Diagnosis, n (%) POAG CACG NVG Uveitic Other | 318 (51) 39 (6) 130 (42) 41 (7) 93 (15) | 58 (41) 10 (7) 41 (29) 11 (8) 23 (16) | 53 (40) 8 (6) 39 (29) 7 (5) 26 (20) | 64 (52) 7 (6) 28 (23) 10 (8) 15 (12) | 55 (48) 7 (6) 22 (19) 13 (11) 17 (15) | 88 (82) 7 (7) 0 0 12 (11) |
| Lens status, n (%) Phakic Pseudophakic Aphakic | 173 (28) 424 (68) 20 (3) | 47 (33) 95 (66) 1 (1) | 44 (33) 79 (59) 6 (5) | 29 (23) 87 (70) 8 (7) | 29 (25) 80 (70) 5 (4) | 24 (22) 83 (78) 0 |
RETENTION
The median follow-up time of the overall study group was 60 months (range 1-71). A total of 55 patients (9%) died within 5 years of enrollment. An additional 142 patients (23%) missed their 5-year study visit.
SURGICAL FAILURE
Kaplan-Meier survival analysis was used to determine the rate of failure of tube shunt surgery, and the results are shown in Figure 1 . The cumulative probability of failure of the overall study group was 38.3% at 5 years. Table 2 presents the reasons for surgical failure. Inadequate IOP reduction (ie, IOP > 21 mmHg or reduced < 20% from baseline) was the most common cause of failure, accounting for 135 failures (53%). Reoperation for glaucoma accounted for 68 failures (27%) in the overall study group.
