A, Criterion A (intraocular pressure ≥21 mm Hg, P = .02). B, Criterion B (intraocular pressure ≥18 mm Hg, P = .04). C, Criterion C (intraocular pressure ≥15 mm Hg, P = .10).
eTable. Univariable Analysis to Identify Prognostic Factors for Failure of Trabeculectomy
Takihara Y, Inatani M, Ogata-Iwao M, et al. Trabeculectomy for open-angle glaucoma in phakic eyes vs pseudophakic eyes after phacoemulsification: a prospective clinical cohort study. JAMA Ophthalmol. Published online November 14, 2013. doi:10.1001/jamaophthalmol.2013.5605.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Takihara Y, Inatani M, Ogata-Iwao M, et al. Trabeculectomy for Open-angle Glaucoma in Phakic Eyes vs in Pseudophakic Eyes After Phacoemulsification: A Prospective Clinical Cohort Study. JAMA Ophthalmol. 2014;132(1):69–76. doi:10.1001/jamaophthalmol.2013.5605
Whether pseudophakic eyes are resistant to trabeculectomy remains unknown.
To determine the effect of previous phacoemulsification on surgical success of trabeculectomy with mitomycin C for open-angle glaucoma.
Design, Setting, and Participants
Prospective clinical cohort study at Kumamoto University Hospital, Kumamoto, Japan, among patients 55 years or older having open-angle glaucoma with intraocular pressure (IOP) of 22 mm Hg or higher, including 39 phakic eyes (phakic group) and 25 pseudophakic eyes after phacoemulsification (pseudophakic group).
Trabeculectomy with mitomycin C was performed.
Main Outcomes and Measures
The primary outcome measure was the probability of success at 1 year after trabeculectomy. Surgical failure was defined as the following 3 IOP levels: 21 mm Hg or higher (criterion A), 18 mm Hg or higher (criterion B), and 15 mm Hg or higher (criterion C). Secondary outcome measures included IOP, the number of postoperative antiglaucoma medications, and the number of laser suture lysis procedures, as well as postoperative complications.
The probabilities of success at 1 year in the phakic vs pseudophakic groups were 95% vs 74% for criterion A (P = .02), 84% vs 62% for criterion B (P = .04), and 67% vs 53% for criterion C (P = .10). Only pseudophakia was significantly associated with outcome in the multivariable analysis for criterion A (relative risk, 9.37) and for criterion B (relative risk, 5.52) (P = .01 for both). Postoperative IOP in the pseudophakic group was significantly higher than that in the phakic group at 6 months (P = .03) and 9 months (P = .047) after trabeculectomy. No significant difference between groups was noted in postoperative complications or in the number of postoperative antiglaucoma medications or the number of laser suture lysis procedures.
Conclusions and Relevance
Among patients with open-angle glaucoma, trabeculectomy with mitomycin C in pseudophakic eyes after phacoemulsification for target IOP of less than 21 mm Hg or less than 18 mm Hg is less successful compared with that in phakic eyes. No significant difference between phakic and pseudophakic eyes was observed for secondary outcome measures other than IOP.
clinicaltrials.gov Identifier: University Hospital Medical Information Network Clinical Trials Registry of Japan UMIN000001196
Trabeculectomy is the most common filtrating surgical intervention for control of intraocular pressure (IOP) in patients with open-angle glaucoma (OAG).1 Reduction of IOP depends on aqueous humor filtration through a surgically created scleral flap and subsequent bleb formation in the conjunctiva. Quiz Ref IDFibroblast activation in the subconjunctiva and ocular inflammation may cause bleb scarring; younger patients and patients with uveitic glaucoma have higher risk for failure of trabeculectomy.2,3 Some studies report the attenuation of bleb scarring with intraoperative use of mitomycin C4,5 and postoperative topical corticosteroid eyedrops,6 which enhance IOP reduction after trabeculectomy.
In a previous retrospective study,7 higher risk for failure of trabeculectomy with mitomycin C among patients with OAG was demonstrated for pseudophakic eyes after phacoemulsification compared with phakic eyes. In OAG, the inflammatory cytokine concentration in the aqueous humor in pseudophakic eyes after phacoemulsification was higher than that in phakic eyes.8 However, no prospective studies to date have compared the outcome of trabeculectomy between phakic eyes and pseudophakic eyes after phacoemulsification. In this prospective study, the outcome of trabeculectomy with mitomycin C in patients having OAG was directly compared between phakic eyes and pseudophakic eyes after phacoemulsification.
This study was approved by the institutional review board of Kumamoto University Hospital, Kumamoto, Japan. The protocol adhered to the tenets of the Declaration of Helsinki. The study was registered with the University Hospital Medical Information Network Clinical Trials Registry of Japan (identifier UMIN000001196; date of access and registration, June 18, 2008). Written informed consent was obtained from all patients after detailed explanation of the procedures involved.
Quiz Ref IDThis prospective clinical cohort study compared the outcome of trabeculectomy with mitomycin C between phakic eyes (phakic group) and pseudophakic eyes after phacoemulsification (pseudophakic group) among patients having OAG. Patients were recruited between June 18, 2008, and December 31, 2010, at Kumamoto University Hospital using the following inclusion criteria: age 55 years or older, primary OAG or exfoliation glaucoma, no history of intraocular surgery other than phacoemulsification, phakic eyes or pseudophakic eyes previously treated with phacoemulsification, and eyes with IOP of 22 mm Hg or higher (despite administration of antiglaucoma medication) measured at least once within 3 months before trabeculectomy. The pseudophakic group included only eyes with previous phacoemulsification and posterior chamber intraocular lens implantation. Exclusion criteria were aphakic eyes, eyes with previous vitrectomy, eyes with a history of glaucoma surgery before trabeculectomy, and pseudophakic eyes previously treated with intracapsular or extracapsular cataract extraction. If both eyes in the same patient satisfied the inclusion criteria, the eye with the higher preoperative IOP was included in the study.
All trabeculectomy procedures were identically performed by 3 experienced surgeons (M.I., T.I., and H.T.) during the study period. A 5-mm conjunctival incision was made along the limbus to create a fornix-based conjunctival flap at the superior conjunctiva. If mobility in the conjunctiva of the pseudophakic eye was disturbed because of previous scleral incisions related to phacoemulsification, the mobile conjunctiva, which was separated from the immobile conjunctiva, was chosen for the conjunctival flap during trabeculectomy. A 4-mm-wide half-layer scleral flap was also created. Mitomycin C (0.4 mg/mL) was applied on and under the scleral flap, as well as under the conjunctiva, for 4 minutes, after which the eye was irrigated with physiological saline (200 mL). A deep limbal block was excised to create a fistula in the anterior chamber, after which peripheral iridectomy was performed. The scleral flap was closed using 3 sutures of 10-0 monofilament nylon. The conjunctival flap was also sutured with 10-0 monofilament nylon. All patients received similar postoperative topical medication with levofloxacin, 0.5%, for 1 month and betamethasone sodium phosphate, 0.1%, for 3 months.
Patient data were collected, including sex, age, visual field, type of glaucoma, preoperative IOP, postoperative IOP, best-corrected visual acuity (BCVA), the number of laser suture lysis procedures, and the number of postoperative antiglaucoma medications, as well as postoperative complications. Visual field testing was performed using static automated white-on-white threshold perimetry program 24-2 (Swedish interactive thresholding algorithm standard, model 750; Zeiss). If the eye had advanced visual field loss, perimetry program 10-2 was used. A logarithm of the reciprocal of the decimal BCVA was used to approximate the logarithm of the minimal angle of resolution (logMAR).
The initial study-related visit was scheduled 2 weeks after surgery; thereafter, visits were conducted 1, 3, 6, 9, and 12 months later. The IOP, BCVA, and number of postoperative antiglaucoma medications were measured before surgery and at all postoperative visits. Complications were also assessed at all postoperative visits. Laser suture lysis and bleb needling were completed within 2 months of surgery depending on postoperative IOP and bleb formation. The study protocol included the timing of antiglaucoma medication following trabeculectomy. After surgery, IOP of all eyes was measured without antiglaucoma medication. If IOP was 21 mm Hg or higher at the time of the study visit at 1 month after surgery or later, we performed laser suture lysis or bleb needling and remeasured IOP 1 month later. If IOP was 21 mm Hg or higher at 2 consecutive visits after completion of laser suture lysis and bleb needling, antiglaucoma medication was initiated to prevent glaucoma progression.
The primary outcome measure was the probability of success at 1 year after trabeculectomy based on IOP. The definition of success was based on a previous retrospective study7 of trabeculectomy in phakic vs pseudophakic eyes, with slight modifications. Surgical failure was defined as the following 3 IOP levels at 3 months or longer after surgery, which was confirmed 1 month later with or without antiglaucoma medication: 21 mm Hg or higher (criterion A), 18 mm Hg or higher (criterion B), and 15 mm Hg or higher (criterion C). The IOP levels that corresponded to criterion A, B, or C at less than 3 months after trabeculectomy were considered early postoperative IOP fluctuation rather than surgical failure. However, surgical failure was declared for all criteria at IOP levels of 26 mm Hg or higher despite completion of laser suture lysis and bleb needling. These patients were immediately treated with antiglaucoma medication or underwent reoperation to prevent glaucoma progression. Moreover, if loss of light perception was found at any study-related visit, surgical failure was declared for all criteria. Kaplan-Meier survival curves of surgical success were compared between the phakic and pseudophakic groups.
Secondary outcome measures were assessed. These included IOP, the number of postoperative antiglaucoma medications, and the number of laser suture lysis procedures, as well as postoperative complications.
Univariable comparisons between groups were performed using χ2 test, Fisher exact test, and Mann-Whitney nonparametric test. The probability of success was analyzed using Kaplan-Meier survival curves and log-rank test. P < .05 was considered statistically significant. Multivariable analysis was performed to determine prognostic factors for failure of trabeculectomy using Cox proportional hazards models.
In a previous retrospective study,7 the probabilities of success (postoperative IOP ≥21 mm Hg) were 98% in the phakic group and 79% in the pseudophakic group. We hypothesized that the ratio of the number of phakic eyes to pseudophakic eyes would be 2:1. Therefore, the minimum sample size was calculated as 62 (41 phakic eyes and 21 pseudophakic eyes) with a 1-sided significance level of .05 and a power of 0.8. Additional patients were recruited to allow for follow-up loss, and recruitment was terminated at 64 patients.
In total, 64 patients (39 phakic eyes and 25 pseudophakic eyes) were enrolled in the study. One patient in the phakic group and 3 patients in the pseudophakic group were lost to analysis before the 1-year follow-up visit. No significant difference between groups was noted in the number of patients who completed the 1-year follow-up visit (P = .29, Fisher exact test).
Table 1 summarizes the baseline characteristics of all patients. Patients in the pseudophakic group were significantly older than those in the phakic group (P = .005). No other statistically significant differences between groups were found in preoperative status. All patients in the pseudophakic group were treated with sclerocorneal incision phacoemulsification.
Among patients in the phakic group who completed the 1-year follow-up visit, surgical failure occurred in 2 of 38 (5%), 6 of 38 (16%), and 11 of 38 (29%) based on criteria A, B, and C, respectively. Among patients in the pseudophakic group who completed the 1-year follow-up visit, surgical failure occurred in 6 of 22 (27%), 9 of 22 (41%), and 11 of 22 (50%) based on criteria A, B, and C, respectively. Significantly higher failure rates were found in the pseudophakic group for criterion A (P = .04, Fisher exact test) and criterion B (P = .03, χ2 test). No significant difference between groups was found for criterion C (P = .10, χ2 test). Insufficient IOP reduction was the cause of failure in all patients in both groups for all 3 criteria. No surgical failure because of loss of light perception was observed. No reoperation to reduce IOP was necessary during the follow-up period.
In the phakic group, success without antiglaucoma medication (complete success) was achieved for 35 of 38 (92%), 32 of 38 (84%), and 27 of 38 (71%) patients based on criteria A, B, and C, respectively. In the pseudophakic group, complete success was achieved for 15 of 22 (68%), 12 of 22 (55%), and 11 of 22 (50%) patients based on criteria A, B, and C, respectively. The rate of complete success was also significantly higher in the phakic group for criterion A (P = .02, χ2 test) and criterion B (P = .01, χ2 test). No significant difference between groups was seen for criterion C (P = .10, χ2 test).
The results of the Kaplan-Meier survival curve analysis comparing the 2 groups for criteria A, B, and C are shown in the Figure. Significantly lower cumulative probability of success for criterion A (P = .02) and criterion B (P = .04) was observed for the pseudophakic group. However, no significant difference between groups was found for criterion C (P = .10). The probabilities of success at the 1-year follow-up visit in the phakic vs pseudophakic groups were 95% vs 74% for criterion A (P = .02), 84% vs 62% for criterion B (P = .04), and 67% vs 53% for criterion C (P = .10).
Intraocular pressure values at various follow-up time points were compared between groups (Table 2). Significantly higher postoperative IOP levels were observed in the pseudophakic group compared with the phakic group at 6 months (P = .03) and 9 months (P = .047) after trabeculectomy.
No significant difference between groups was found in the number of postoperative antiglaucoma medications needed after surgery (Table 2). More patients in the pseudophakic group compared with the phakic group required postoperative antiglaucoma medication at follow-up completion, although the difference was not significant (23% [5 of 22] vs 8% [3 of 38]; P = .10, χ2 test). Fewer mean (SD) postoperative laser suture lysis procedures were performed in the phakic group (1.7 [1.2] sutures) than in the pseudophakic group (2.0 [1.3] sutures), but the difference was not significant (P = .33). No significant difference in the mean (SD) logMAR BCVA at 1 year after surgery was seen between the phakic group (0.40 [0.62]) and the pseudophakic group (0.55 [0.87]) (P = .85).
No significant difference between groups was found in postoperative complications (Table 3). Cataract progression occurred in 7 eyes (18%) in the phakic group. Choroidal detachment was observed more frequently in the pseudophakic group than in the phakic group, although the difference was not significant (P = .06). Bleb leak, wound leak, and flat anterior chamber were more common in phakic eyes, but no significant difference was observed compared with pseudophakic eyes. Deterioration of decimal BCVA to at least 0.3 at 1 year after surgery occurred more frequently in phakic eyes, but no significant difference was observed compared with pseudophakic eyes.
Quiz Ref IDBaseline characteristics, including sex, age, pseudophakia, logMAR BCVA, type of glaucoma, preoperative IOP, and the number of postoperative antiglaucoma medications, were evaluated as possible predictors of surgical failure. In analyses using univariable Cox proportional hazards regression models, pseudophakia was a significant risk factor based on criterion A (relative risk [RR], 5.20; P = .03) and a borderline risk factor based on criterion B (RR, 2.75; P = .05) (eTable in the Supplement). Multivariable analysis identified only pseudophakia as significantly associated with surgical outcome for criterion A (RR, 9.37; P = .01) and criterion B (RR, 5.52; P = .01) (Table 4). Pseudophakia was not significantly associated with surgical outcome for criterion C (RR, 2.33; P = .10).
In this study, the rate of failure of trabeculectomy with mitomycin C among patients with OAG was higher in pseudophakic eyes after phacoemulsification than in phakic eyes. In addition, significantly lower cumulative probabilities of success for criterion A (P = .02) and criterion B (P = .04) were observed among patients in the pseudophakic group. After adjustment for other potential prognostic factors in the Cox proportional hazards regression models, the independent contribution of pseudophakia to the prediction of surgical failure was confirmed for criterion A (RR, 9.37) and criterion B (RR, 5.52) (P = .01 for both).
Numerous studies9-13 indicate that patients with glaucoma who undergo cataract surgery exhibit resistance to maintaining target IOP after trabeculectomy. Previous reports included pseudophakic and aphakic eyes after extracapsular or intracapsular cataract extraction. Cataract extraction may cause extensive scarring in the superior location of the conjunctiva, where trabeculectomy should be performed. Surgical scarring in the conjunctiva may contribute to a poor outcome of trabeculectomy in pseudophakic or aphakic eyes. Phacoemulsification causes less scar formation in the conjunctiva of patients with glaucoma. A recent retrospective study compared the surgical prognosis of trabeculectomy between 175 phakic eyes and 51 pseudophakic eyes after phacoemulsification among patients with OAG.7 In that study, the probabilities of success at 1 year after trabeculectomy in phakic eyes vs pseudophakic eyes were 97.8% vs 78.6% for IOP of less than 21 mm Hg, 92.9% vs 72.8% for IOP of less than 18 mm Hg, and 73.1% vs 53.1% for IOP of less than 15 mm Hg. In another recent study,3 prognostic factors for failure of trabeculectomy in uveitic glaucoma eyes were retrospectively analyzed. In addition to the association with granulomatous uveitis, previous cataract surgery was a prognostic factor for failure of trabeculectomy. In contrast, in a 20-year longitudinal study of surgical outcomes of patients who underwent trabeculectomy between January 1988 and December 1990, Landers et al2 showed that extracapsular cataract extraction was not significantly associated with qualified success of trabeculectomy. In another retrospective study, Supawavej et al14 reported that the probability of success of trabeculectomy for pseudophakic eyes was comparable to that for phakic eyes. The present study is unique in its prospective comparison between phakic eyes and pseudophakic eyes after phacoemulsification. In addition, the surgical procedure of trabeculectomy with mitomycin C was identical for all patients in this prospective study.
The pathomechanism for failure of trabeculectomy in pseudophakic eyes is not fully understood.15Quiz Ref IDAlterations in the nature of the aqueous humor may contribute to failure of trabeculectomy. Intraocular surgery causes ocular inflammation and breakdown of the blood-aqueous barrier, which may cause bleb failure after trabeculectomy.16 A high concentration of inflammatory cytokines has been shown in the aqueous humor of pseudophakic eyes with glaucoma.8 A prospective clinical study17 compared the concentration of monocyte chemoattractant protein 1, an inflammatory cytokine, in the aqueous humor of patients with no eye disease other than cataract before and after phacoemulsification. In all 21 patients, monocyte chemoattractant protein 1, which was collected 1 to 2 years after phacoemulsification, was upregulated in the aqueous humor. This cytokine recruits leukocytes in the anterior ocular segments, which may promote subconjunctival fibrosis and bleb scarring after trabeculectomy in pseudophakic eyes.
In this study, significant differences in failure rates, complete success rates, and Kaplan-Meier survival curves of the probability of success for criterion A and criterion B were observed between groups. For criterion C, no significant difference was found, although a poorer surgical outcome was observed in the pseudophakic group. The lack of statistical significance in the results for criterion C may be attributable to the few patients in the present study. A larger study may determine the true significance of this criterion.
While postoperative IOP values were significantly higher in the pseudophakic group at 6 months and 9 months after trabeculectomy, no significant difference was found at 12 months. During the follow-up period, antiglaucoma medication was administered for IOP reduction. More patients in the pseudophakic group than in the phakic group were treated with antiglaucoma medication at 12 months (23% [9 of 39] vs 8% [2 of 25]), but this trend was not significant. The difference between groups in IOP values at 12 months may have been influenced by postoperative administration of antiglaucoma medication, although the timing of such drug use after trabeculectomy was included in the study protocol.
In the analysis of postoperative complications, choroidal detachment was more often encountered in the pseudophakic group than in the phakic group, but the significance was borderline (P = .06). Jampel et al18 reported that older patients were more likely to experience choroidal detachment after trabeculectomy. In the present study, patients in the pseudophakic group were significantly older than those in the phakic group (P = .005). Therefore, the higher frequency herein of choroidal detachment in the pseudophakic group may have been associated with the fact that these patients were older. Despite a better surgical outcome, cataract progression occurred in 18% (7 of 39) of patients in the phakic group. Postoperative cataract development is frequently observed in phakic eyes among older patients after trabeculectomy.19,20 Future studies should include age-matched patients for more accurate comparison of these complications.
Deterioration of decimal BCVA to at least 0.3 at 1 year after surgery was more frequently observed in the phakic group than in the pseudophakic group (37% [14 of 38] vs 23% [5 of 22], P = .26). This higher incidence of visual disturbance may be associated with cataract progression. Moreover, phacoemulsification after trabeculectomy may have a detrimental effect on bleb function.21,22 In our study, phacoemulsification was not performed in eyes in which cataract progression was observed during the follow-up period. Future studies should include a longer follow-up period, in which phacoemulsification after trabeculectomy may be necessary in phakic eyes and in which the outcome of trabeculectomy in phakic eyes may be affected by phacoemulsification after trabeculectomy.
One major limitation of the present study was that patients in the pseudophakic group were significantly older than those in the phakic group. Although the patient recruitment criteria included age 55 years or older, no age matching was performed between groups. The findings of several clinical studies2,23,24 have suggested that trabeculectomy is more likely to fail in younger patients. In the present study, a better surgical outcome was observed among patients in the phakic group, who were younger than those in the pseudophakic group. However, pseudophakia was confirmed as a significant independent prognostic factor for surgical outcome described by criterion A and criterion B in the Cox proportional hazards regression models. This result suggests that the primary outcome in this study may not be significantly influenced by the age difference between groups. Another limitation was that the study did not evaluate whether clear corneal incision was effective for improving surgical success in pseudophakic eyes. While trabeculectomy was performed in the superior conjunctiva, which was separated from the previous corneoscleral incision in phacoemulsification, the possibility of a residual effect of the previous small conjunctival incision apart from that at the site of trabeculectomy cannot be excluded. A prior retrospective study14 of pseudophakic eyes treated with clear corneal incision showed that the success rate in the pseudophakic group was comparable to that in the phakic group. However, another study25 demonstrated that clear corneal incision did not contribute to an improved prognosis in trabeculectomy for pseudophakic eyes. In a previous retrospective study,7 the eyes that underwent trabeculectomy after clear corneal incision phacoemulsification had a lower probability of success than the phakic eyes. Because these retrospective studies included small samples, a larger-scale prospective study is necessary to determine the contribution of clear corneal incision phacoemulsification to prognosis after trabeculectomy for pseudophakic eyes.
In conclusion, among patients with OAG, pseudophakic eyes after phacoemulsification exhibit greater resistance than phakic eyes to maintaining target IOP of less than 21 mm Hg or less than 18 mm Hg after trabeculectomy with mitomycin C. A larger prospective study among patients with OAG is required to determine whether pseudophakic eyes are also more resistant than phakic eyes to preserving target IOP of less than 15 mm Hg.
Submitted for Publication: March 27, 2013; final revision received May 19, 2013; accepted June 8, 2013.
Corresponding Author: Masaru Inatani, MD, PhD, Department of Ophthalmology, Faculty of Medical Science, University of Fukui, 23-3 Shimoaizuki, Eiheiji, Yoshida, Fukui 910-1193, Japan (firstname.lastname@example.org).
Published Online: November 14, 2013. doi:10.1001/jamaophthalmol.2013.5605.
Author Contributions: Dr Inatani had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis
Study concept and design: Takihara, Inatani, Ogata-Iwao.
Acquisition of data: All authors.
Analysis and interpretation of data: All authors.
Drafting of the manuscript: Takihara, Inatani, Kawai, Tanihara.
Critical revision of the manuscript for important intellectual content: All authors.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology, Tokyo, Japan (Drs Inatani, Inoue, and Tanihara).
Role of the Sponsor: The Ministry of Education, Culture, Sports, Science, and Technology had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, and approval of the manuscript; and decision to submit the manuscript for publication.