Ahmed Glaucoma Valve Implantation in African American and White Patients

Objective  To evaluate the results of Ahmed Glaucoma Valve implantation in African American and white patients.

Methods  In this retrospective, comparative case-control study, we reviewed 86 eyes of 86 patients, comparing the surgical outcomes in white patients (n = 43) with matched African American patients (n = 43). Success was defined as an intraocular pressure (IOP) between 6 mm Hg and 21 mm Hg with or without glaucoma medicines, without further glaucoma surgery, and without loss of light perception (definition 1) and an IOP between 6 mm Hg and 21 mm Hg and achievement of a 20% reduction in IOP from the preoperative level (definition 2).

Results  The mean follow-up was 2.3 years for white patients and 2.5 years for African American patients (P = .50). At the last follow-up, the mean ± SD IOP was 15.3 ± 3.3 mm Hg and 15.3 ± 3.5 mm Hg (P = .77) in white and African American patients, respectively. Life table analysis showed a significantly lower success rate for African American patients compared with white patients by both definition 1 (P = .03) and definition 2 (P = .006). Cox proportional hazards regression analysis detected African American race as a risk factor for surgical failure by both definitions. Visual outcomes and complications were comparable between the 2 groups.

Conclusion  African American patients have a greater risk of surgical failure after Ahmed Glaucoma Valve implantation compared with white patients.

Glaucoma is one of the leading causes of blindness that may disproportionately affect certain racial or ethnic groups, including African Americans.^1-4 The Baltimore Eye Survey and other studies have estimated that the prevalence of primary open-angle glaucoma (POAG) may be 6 times more common in black persons than white persons in the same age group, and glaucoma develops on average 10 years earlier in black individuals.^5-11 Previous clinical studies have also found that glaucoma in black persons is more advanced at the time of diagnosis,^5-9 progresses more rapidly,^6,8,9,12-14 and is less responsive to glaucoma medical therapy than in white individuals.^15-17 Possible explanations for these findings include differences in corneal thickness,^18,19 iris pigmentation,^20,21 anatomical differences in the optic disc,^22-24 and environmental or socioeconomic factors.^2-4

African American patients with glaucoma are reportedly more likely to require surgical interventions than white patients (2.95 vs 1.38 glaucoma surgical procedures per 1000 person-years).²⁵ Early studies^26-31 have supported the strong clinical impression that black patients are at high risk for failure of a trabeculectomy. It is presumed that there is increased fibrovascular proliferation and more vigorous wound healing in African Americans.²⁷ Antifibrotic agents, such as mitomycin C, have been used to improve the success rate of trabeculectomy in blacks.^28,29 However, Morris et al²⁷ and Scott et al³⁰ reported that black race is still a strong risk factor for surgical failure of primary glaucoma surgery with adjunctive mitomycin C.

Glaucoma drainage implants are indicated when other surgical treatments have a poor prognosis for success, prior conventional surgery has failed, or significant conjunctival scarring precludes filtration surgery. A noncomparative retrospective case series described results after single-plate Molteno implantation in black patients,³² with no control group for comparison. Previous studies have not described whether black patients have a poor prognosis for clinical success with glaucoma drainage implants compared with white patients. Our purpose was to compare the outcomes of glaucoma drainage implant surgery using the Ahmed Glaucoma Valve in African American and white patients.

This was a single-center, retrospective, comparative (case-control) study. The study was approved by the institutional review board of the University of Tennessee Health Science Center, and consent was obtained before surgery in all patients. We reviewed the consecutive records of self-identified white and African American patients who were treated with the Ahmed Glaucoma Valve. All implantations were performed by 1 surgeon (P.A.N.) for increased intraocular pressure (IOP) that was not responsive to medical therapy, laser treatment, or previous glaucoma surgery.

Exclusion criteria included patient age younger than 18 years, elevated IOP associated with silicone oil, neovascular glaucoma, treatment with other types of drainage implants besides the Ahmed Glaucoma Valve (model S2), previous cyclodestructive treatment, and previous glaucoma drainage device implantation. Patients with less than 6 months of follow-up were excluded from the analysis to allow evaluation of the results after the immediate postoperative period. Cases were matched to controls by age and sex, with both cases and controls chosen from the same period. In patients with bilateral implantation, data analysis was performed for the first eye that had undergone surgery.

A total of 200 consecutive medical records from a 6-year period were reviewed. Of these medical records, 98 were excluded according to the predetermined exclusion criteria. Nine eyes with less than 6 months of follow-up were excluded from the analysis. In 7 patients with bilateral implantation, data analysis was performed for the first eye treated with the Ahmed Glaucoma Valve. Thus, a total of 86 eyes of 86 patients were included in the analysis (43 eyes in white patients and 43 eyes in African American patients).

Preoperative information included sex, ethnicity, age, history of laser and surgical treatments, glaucoma diagnosis, lens status, glaucoma medications, IOP measured by Goldmann applanation tonometry, and visual acuity. Postoperative data regarding IOP, number of glaucoma medicines, visual acuity, and complications were obtained at 1 day; 1 week; 2 weeks; 1, 2, 3, 6, 9, and 12 months; and every 6 months thereafter. After the first month, visits within 2 weeks of the scheduled time point were included in the analysis. Results of the most recent examination were for analysis of IOP, visual acuity, and success at the last follow-up examination.

Criteria for success were defined before reviewing the data. Surgical success was characterized by the following 2 definitions. Definition 1 was an IOP of 6 mm Hg or greater and 21 mm Hg or less with or without the use of additional glaucoma medicines, without further glaucoma surgery including cyclophotocoagulation or devastating complications that required removal of the implant, and without loss of light perception. Because of the possibility of modest preoperative IOP with a low postoperative target IOP, a second definition of success was included with more stringent criteria. Definition 2 required that the criteria for definition 1 be met and also a 20% reduction in IOP from the preoperative level. The definition of hypotony was an IOP of 5 mm Hg or less on 2 consecutive visits. A hypertensive phase was defined as an IOP greater than 21 mm Hg during the first 3 months after surgery.

The implant used was the Ahmed Glaucoma Valve (New World Medical Inc, Rancho Cucamonga, Calif), model S2 (184-mm² surface area). A fornix-based conjunctival flap was created between the superior and lateral or the superior and medial rectus muscles. The tube of the Ahmed Glaucoma Valve was irrigated with balanced saline solution to prime the valve. The plate was secured 8 to 9 mm posterior to the surgical limbus with interrupted 9-0 nylon sutures. The tube was trimmed to extend approximately 3 mm beyond the surgical limbus with the bevel facing up and was inserted in the anterior chamber through a 23-gauge needle tract. The tube was anchored to the sclera using a 9-0 nylon suture and covered with a rectangular piece of pericardium patch graft (IOP Inc, Costa Mesa, Calif, and New World Medical Inc). The conjunctiva was closed with 9-0 polyglactin sutures, and subconjunctival injection of an antibiotic and steroid was administered away from the surgical site.

Demographic, preoperative, and postoperative data were compared between the white and African American patients. Mann-Whitney U tests were used to compare continuous parameters between the 2 groups. The Wilcoxon signed rank test was used to compare continuous parameters within groups. The χ² and Fisher exact tests were used for categorical variables. The cumulative probability of success was analyzed by the Kaplan-Meier life table method. The log-rank test was used to compare the success rates calculated by life table analysis. Cox proportional hazards regression models were performed to assess the relationship between survival outcomes and multiple predictors, including age, ethnicity, glaucoma diagnosis, lens status, preoperative IOP, glaucoma medications, and preoperative number of surgical procedures and laser therapies. P<.05 was considered statistically significant.

A total of 86 eyes of 86 patients were analyzed in the study, including 43 eyes in African American patients compared with 43 eyes in control white patients. In 7 patients with bilateral implantation, medical record analysis was performed for the first eye that had undergone surgery. Nine eyes with less than 6 months of follow-up were excluded from the analysis. To evaluate for potential bias, the 9 eyes (in 5 African American and 4 white patients) with less than 6 months of follow-up were further analyzed. These 9 eyes had no failures by our criteria and were found to have no statistically significant differences for age, sex, diagnosis, lens status, history of surgery and laser treatment, preoperative IOP, medications, and visual acuity when compared with the 86 eyes that were included in the study.

Demographic and preoperative data are given in Table 1. The 2 groups were statistically comparable with respect to age (P = .87), sex (P>.99), diagnosis (P = .27), lens status (P = .06), history of surgery (P>.99) and laser treatment (P=.38), preoperative IOP (P = .64), medications (P = .76), and visual acuity (P = .07). The mean ± SD follow-up period was 2.3 ± 1.4 years (range, 0.5-5.6 years) in white patients and 2.5 ± 1.7 years (range, 0.5-6.8 years) in African American patients (P = .50).

As indicated in Table 2, the mean ± SD preoperative IOP was 30.8 ± 7.4 mm Hg (median, 29 mm Hg) with a mean ± SD of 3.8 ± 1.0 (median, 4) glaucoma medications for white patients and 33.6 ± 12.6 mm Hg (median, 30 mm Hg) with a mean ±SD of 3.9 ± 0.6 (median, 4) glaucoma medications in African American patients (P = .64 and .76, respectively). The mean ± SD IOP at last follow-up was 15.3 ± 3.3 mm Hg in white patients and 15.3 ± 3.5 mm Hg in African American patients (P = .77). The mean ± SD number of glaucoma medications at most recent follow-up was 1.4 ± 1.2 (median, 1) in white patients and 1.9 ± 1.5 (median, 2) in African American patients (P = .14). Postoperatively the mean IOP and mean number of medications decreased significantly in both groups (P<.001). At the most recent visit, the visual acuity was improved or within 1 Snellen line in 34 white patients (79%) and 30 African American patients (70%) (P = .32). The success at the last follow-up examination was 98% (42 of 43 patients) in white patients and 84% (36 of 43) in African American patients by definition 1 (P = .06) and 98% (42 of 43 patients) in white patients and 74% (32 of 43) in black patients by definition 2 (P = .003).

According to success definition 1, 1 white patient in whom surgery had failed had additional cyclophotocoagulation; however, the IOP at the last follow-up was controlled at 21 mm Hg or less. Of the 7 eyes considered surgical failures in African American patients, 2 eyes had additional cyclophotocoagulation with the IOP at the last follow-up controlled at 21 mm Hg or less, 4 eyes had an IOP higher than 21 mm Hg (1 requiring additional Ahmed Glaucoma Valve implantation), and 1 eye had persistent hypotony. According to success definition 2, surgery was considered to have failed in 4 additional African American patients because the IOP was not reduced by 20% postoperatively. No patient had loss of light perception.

Figure 1 shows the Kaplan-Meier survival analysis for success definition 1 for the 2 groups. The success for white patients and African American patients was 100% and 91% at 1 year and 96% and 79% at 3 years, respectively. The difference in the survival curves between the 2 groups was statistically significant (P = .03). The number of patients who completed the follow-up period and were not classified as having undergone failed surgery in the analysis of success in white and African American patients, respectively, was 43 and 43 at 6 months, 34 and 35 at 1 year, 23 and 24 at 2 years, 15 and 16 at 3 years, and 6 and 6 at 4 years. In the analysis of success definition 2 (Figure 2), the cumulative success for white and African American patients was 100% and 91% at 1 year, 96% and 75% at 3 years, and 96% and 55% at 4 years, respectively. There was a statistically significant difference between the 2 groups (P = .006).

A Cox proportional hazards regression model was used to evaluate risk factors for surgical failure (Table 3). Factors analyzed included sex, ethnicity, age, glaucoma diagnosis, lens status, preoperative IOP, number of preoperative glaucoma medications, and number of preoperative laser or surgical treatments. African American race was detected as a risk factor for surgical failure by both definition 1 (risk ratio = 10.2; P = .04) and definition 2 (risk ratio = 14.4; P = .02), whereas other factors were not statistically significant risk factors for failure.

Figure 3 depicts the mean preoperative IOP and the IOP at each of the postoperative time points. The mean ± SD preoperative IOP of 30.8 ± 7.4 mm Hg in white patients and 33.6 ± 12.6 mm Hg in African American patients (P = .64) decreased to 8.8 ± 5.2 mm Hg and 8.5 ± 4.3 mm Hg at 1 day (P = .87), 17.7 ± 6.9 mm Hg and 20.4 ± 9.5 mm Hg at 1 month (P = .26), 16.2 ± 3.6 mm Hg and 15.8 ± 7.0 mm Hg at 1 year (P = .25), 15.5 ± 3.8 mm Hg and 14.2 ± 3.9 mm Hg at 3 years (P = .43), and 11.3 ± 1.5 mm Hg and 15.4 ± 4.6 mm Hg at 5 years (P = .23) in white and African American patients, respectively. The mean IOP differences between the 2 groups were not statistically significant throughout the follow-up period. After the initial reduction in IOP at postoperative day 1, the mean IOP gradually increased to a peak at 1 month, then gradually decreased and stabilized after 1 month. The number of patients who had a hypertensive phase was comparable between 2 groups. However, the peak IOP at 1 month postoperatively was significantly higher than the IOP at 1 year after surgery in African American patients (P = .02) but not in white patients (P = .46).

Figure 4 shows the mean number of glaucoma medications preoperatively and at each of the postoperative time points. The mean ± SD preoperative number of medications of 3.8 ± 1.0 and 3.9 ± 0.6 (P = .76) decreased to 0 and 0 at 1 day (P > .99), 0.1 ± 0.5 and 0.3 ± 0.9 at 1 month (P = .22), 1.3 ± 1.2 and 1.8 ± 1.5 at 1 year (P = .15), 1.6 ± 1.2 and 1.4 ± 1.3 at 3 years (P = .52), and 2.3 ± 0.6 and 2.0 ± 1.6 at 5 years (P = .76) in white and African American patients, respectively. Compared with African American patients, white patients were taking significantly fewer medications at 3 months after surgery (P = .05); however, this difference was not observed at subsequent time points.

Complications occurred in 14 (33%) of 43 eyes in white patients and 21 (49%) of 43 eyes in African American patients after Ahmed Glaucoma Valve surgery (Table 4). No statistically significant differences were detected between white and African American patients (P = .12). Five white patients (12%) and 6 African American patients (14%) developed choroidal effusion or hemorrhage, with 3 requiring surgical drainage. Three white patients and 6 African American patients had corneal complications. Obstruction of the tube occurred in 2 white patients and 5 African American patients and was treated by needling or Nd:YAG laser; however, 3 patients needed surgical repositioning of the tube. A Tenon cyst was observed in 3 white patients (7%) and 7 African American patients (16%). All other complications (tube-cornea touch, retinal detachment, iritis, migration of the plate, malignant glaucoma, pupillary block, erosion of the tube, and hyphema) occurred infrequently. Implants were not removed in any of the patients.

Glaucoma is a leading cause of blindness among African Americans and is more prevalent in African Americans compared with the general US population.^1-4 Black individuals appear to respond differently than white persons to medical, laser, and surgical treatment with trabeculectomy with mitomycin C.^15-17,26-31 However, African American race has not been evaluated as a risk factor for failure of glaucoma drainage device implantation. In our study, we found that success was lower over time in African American compared with white patients. We also found that African American race was a risk factor for failure of glaucoma drainage implant surgery.

In our study, the success at the last follow-up visit in white and African American patients was significantly different. In a noncomparative study, Freedman and Rubin³² reported that surgical success (IOP ≤21 mm Hg) was achieved in 59 (72%) of 82 black patients treated with the single-plate Molteno implant, with a mean follow-up of 30 months. By using their definition in our African American patients, success was achieved in 86% of patients, with a mean follow-up of 31 months. When only POAG cases were compared between our study and that by Freedman and Rubin, surgical success was achieved in 23 (73%) of the 31 patients with POAG, with a mean follow-up of 19 months, in the study by Freedman and Rubin and in 23 (89%) of 26 patients with POAG, with a mean of 30 months of follow-up, in our study. It is not possible to directly compare these studies because of differences in implant types, patient populations, and other factors, and there was no control group for comparison in the study by Freedman and Rubin.

Life table analysis success rates in our study were comparable with those of previous studies. Using a definition of success identical to our definition 1, Huang et al³³ reported 87% success at 1 year and 75% success at 2 years after Ahmed Glaucoma Valve implantation in 159 eyes of 144 patients, including 22 black, 89 white, 32 Hispanic, and 8 Asian patients. Tsai et al³⁴ found 83% success at 6 months of follow-up in 40 white patients and 8 black patients, with success defined as the last 2 IOP values of more than 5 mm Hg and 21 mm Hg or less. Topouzis et al³⁵ defined surgical failure as an IOP greater than 21 mm Hg and less than 6 mm Hg at the last 2 visits or devastating complications (retinal detachment, malignant glaucoma, endophthalmitis, or phthisis bulbi) and found success rates of 87% at 1 year, 82% at 2 years, and 76% at 3 and 4 years in 60 patients, including 37 white, 12 Hispanic, 8 black, and 3 Asian patients, with a mean follow-up of 31 months. When we applied the same criteria in our patients, the success rate was 94% at 1 year, 93% at 2 years, and 90% at 3 and 4 years.

We used Cox regression analysis to determine risk ratios for various preoperative factors, including age, sex, ethnicity, glaucoma diagnosis, lens status, preoperative IOP, number of preoperative medications, and preoperative laser and surgical therapies. African American race was detected as a risk factor for surgical failure by success definition 1 (risk ratio = 10.2; P = .04) and success definition 2 (risk ratio = 14.4; P = .02). When corneal implantations were excluded from the definition of failure, Topouzis et al³⁵ found increased odds of failure in patients from minority backgrounds compared with white patients (odds ratio = 9.61; 95% CI, 1.44-64.15) in 60 patients treated with the Ahmed Glaucoma Valve. However, their study included black, Hispanic, and Asian patients in the minority group (8 black, 12 Hispanic, and 3 Asian patients). Because of our exclusion criteria, we were unable to determine whether other variables, such as neovascular glaucoma and silicone oil, were associated with a risk of failure after treatment with the Ahmed Glaucoma Valve.

In this study, the mean IOP was 15.3 mm Hg and the mean number of glaucoma medications was 1.6 at the last follow-up examination in all patients, with a mean follow-up of 29 months. In results reported by Huang et al,³³ the mean IOP was 15.1 mm Hg and the mean number of glaucoma medications was 1.1 at the last follow-up examination, with a shorter mean follow-up of 13 months. In 156 eyes of 139 patients (98 white, 14 Hispanic, 14 Asian, and 13 black patients) with a mean follow-up of 15 months, the mean IOP was 16.5 mm Hg and the mean number of medications was 1.4,³⁶ similar to the findings in our study.

The mean IOP during the follow-up period in our study was similar in white and African Americans patients. Although the white patients required fewer glaucoma medications compared with African American patients at up to 2 years of follow-up, only 1 time point was statistically significantly different (P = .05 at 3 months). Nouri-Mahdavi and Caprioli³⁶ reported that a hypertensive phase was observed in 56% of patients who had undergone Ahmed Glaucoma Valve implantation. Ayyala et al³⁷ noted that 84% of patients experienced a hypertensive phase after Ahmed Glaucoma Valve implantation. By using the same definition for hypertensive phase, 11 white patients (26%) and 15 African American patients (35%) developed a hypertensive phase in our study, which was not statistically significantly different (P = .35).

The number of complications experienced during the follow-up period by African American patients was slightly greater compared with white patients in our study, but this difference was not statistically significant (P = .12). Freedman and Rubin³² noted that the development of a Tenon cyst over the acrylic reservoir of a single-plate Molteno device was a frequent complication, which developed in 14 (88%) of 16 black patients without tenonectomy and 2 (3%) of 64 black patients with tenonectomy. Huang et al³³ observed a Tenon cyst in 10% of their mixed-ethnicity group. In our study, we did not excise a Tenon capsule during surgery and observed a Tenon cyst in 3 white patients (7%) and 7 African American patients (16%) (P = .31). Complications observed in our study were comparable to those in other studies,^32-36 and no statistically significant difference was seen when comparing white and African American patients.

The nonrandomized, retrospective design of this study has limitations. Also, our study did not attempt to find the underlying cause of the differences in success observed in African American and white patients. However, our comparative study was able to identify significant differences in success rates over time in African American and white patients. The significant treatment outcome difference found in this study indicates that African American race is a risk factor for failure of glaucoma drainage implant surgery.

Correspondence: Peter A. Netland, MD, PhD, Hamilton Eye Institute, University of Tennessee Health Science Center, 930 Madison Ave, Suite 100, Memphis, TN 38163 (mesmith@utmem.edu).

Submitted for Publication: August 18, 2005; final revision received November 23, 2005; accepted December 12, 2005.

Financial Disclosure: None.