Tan Y, Tsou P, Tan GS, Perera SA, Ho C, Wong TT, Aung T. Postoperative Complications After Glaucoma Surgery for Primary Angle-Closure Glaucoma vs Primary Open-Angle Glaucoma. Arch Ophthalmol. 2011;129(8):987-992. doi:10.1001/archophthalmol.2011.71
Author Affiliations: Singapore National Eye Centre and Singapore Eye Research Institute, Singapore.
Objectives To investigate the incidence of postoperative complications arising in the first year after trabeculectomy and combined phacotrabeculectomy in eyes with primary angle-closure glaucoma (PACG) vs those with primary open-angle glaucoma (POAG).
Methods This was a retrospective study of all glaucoma operations done at a Singapore hospital from January 9, 2001, to December 30, 2004. The types of glaucoma surgery included trabeculectomy and phacotrabeculectomy, all with mitomycin C or fluorouracil. The incidences of postoperative complications and reoperations were analyzed. For those who underwent bilateral or repeated operations, only the first operated eye of each subject was included.
Results A total of 446 subjects with PACG (112 who underwent trabeculectomy and 334 who underwent phacotrabeculectomy) and 816 subjects with POAG (208 who underwent trabeculectomy and 608 who underwent phacotrabeculectomy) were analyzed. Postoperative complications occurred in 65 of 1262 eyes (5.2%) overall, with 27 eyes (8.4%) in the trabeculectomy group (PACG: 8.0%; 95% confidence interval [CI], 4.3%-14.6%; POAG: 8.7%; 95% CI, 5.5%-13.3%; P >> .99) and 38 eyes (4.0%) in the phacotrabeculectomy group (PACG: 5.1%; 95% CI, 3.2%-8.0%; POAG: 3.5%; 95% CI, 2.3%-5.2%; P = .31). The rate of complications was significantly higher in the trabeculectomy group than the phacotrabeculectomy group overall (P = .003), but there was no significant difference between the POAG and PACG groups overall (POAG: 4.8%; 95% CI, 3.5%-6.5%; PACG: 5.8%; 95% CI, 4.0%-8.4%; P = .53). The commonest complication found was hypotony with overfiltration (23 cases [1.8%]), followed by bleb leak (11 cases [0.9%]). There was no significant difference in incidence of reoperations between POAG (2.7%; 95% CI, 1.8%-4.1%) and PACG (4.0%; 95% CI, 2.6%-6.3%) (P = .27).
Conclusion The incidences of postoperative complications and reoperations in the first year after glaucoma surgery were similar for PACG and POAG.
Primary angle-closure glaucoma (PACG) is a major form of glaucoma in East Asian persons,1,2 while primary open-angle glaucoma (POAG) is the predominant form of glaucoma among Caucasian and African persons.3,4 Recent glaucoma prevalence studies in southern India found that the prevalence of PACG in Indian persons is also high.5,6 It is estimated that PACG blinds more people than POAG in absolute terms, although the number of those with POAG worldwide is higher.7 By 2020, PACG will affect 20 million people, and 5.3 million will be blind.8
Laser peripheral iridotomy is the recommended initial treatment for PACG. It results in the relief of pupil block. In cases refractory to laser peripheral iridotomy and medical therapy, often as a consequence of angle closure associated with peripheral anterior synechiae, filtering surgery remains the only effective method to sufficiently lower the intraocular pressure (IOP) to a desirable level. The surgery often consists of trabeculectomy alone or combined phacotrabeculectomy. Both surgical methods have been reported to be associated with postoperative complications such as endophthalmitis, shallow anterior chamber and hypotony, aqueous misdirection syndrome, choroidal detachment, and vision loss.9- 11 However, as recently emphasized in the consensus meeting on angle-closure glaucoma organized by the World Glaucoma Association in 2006,12 few data have been reported on the incidence of complications after glaucoma surgery for PACG. There is a clinical impression that glaucoma surgery for PACG is associated with greater risk of complications such as hypotony and malignant glaucoma, but to our knowledge a comparison of postoperative complication rates in eyes with PACG and those with POAG has not been performed.
To address this, we aimed to investigate the incidence of postoperative complications arising in the first year after trabeculectomy and combined phacotrabeculectomy in eyes with PACG and to compare it with this incidence for POAG.
Data were obtained from an audit of all glaucoma operations done for PACG and POAG at the Singapore National Eye Centre during a 4-year period from January 9, 2001, to December 30, 2004. For this audit, the incidences of complications and reoperations within the first year after surgery were obtained independently by the center's audit team from a review of all of the records approximately 1 year after the surgery, and the audit team filled in a standardized audit sheet. The records of patients with postoperative complications (identified by the audit) were then retrospectively reviewed by 2 of us (Y.-L.T. and P.-F.T.) to confirm the complication(s) identified. Collected data included patient demographic characteristics, glaucoma diagnosis, details of surgery, type of antiscarring agent used, postoperative complications, and additional operations or procedures done. For those who underwent bilateral or repeated operations, we included only the first operated eye of each subject. Eyes that underwent surgery for acute primary angle closure were also excluded. Institutional review board approval was obtained for the study.
Glaucoma was defined as the presence of glaucomatous optic neuropathy (defined as loss of neuroretinal rim with a vertical cup-disc ratio >0.7) and associated visual field loss. A glaucomatous visual field defect was noted if the following were found: (1) glaucoma hemifield test results outside the reference range; (2) a cluster of 3 or more nonedge, contiguous points on the pattern deviation plot, not crossing the horizontal meridian with a probability of less than 5% being present in age-matched control subjects (1 of which is <1%); and (3) pattern standard deviation less than 0.05. These were repeatable on 2 separate occasions.13 We defined PACG as glaucoma in association with a closed angle (presence of ≥180° of angle in which the posterior trabecular meshwork was not visible on nonindentation gonioscopy) with or without peripheral anterior synechiae.
The surgical procedure and postoperative management were broadly similar, although multiple surgeons (8 surgeons) were involved in the operations and postoperative care. For trabeculectomy, a fornix-based conjunctival flap was created superonasally or superotemporally. A partial-thickness scleral flap was dissected to clear cornea. All cases received an application of either mitomycin C (0.2-0.4 mg/mL) or fluorouracil (25-50 mg/mL) for 1 to 3 minutes, followed by irrigation with 30 mL of balanced salt solution. A sclerostomy was then created in the scleral bed, and a peripheral iridectomy was performed. The scleral flap was sutured with 10-0 nylon, and the conjunctiva was then reapproximated with 10-0 nylon. For phacotrabeculectomy, similar steps as described for trabeculectomy were performed, but after irrigation of mitomycin C or fluorouracil, temporal clear corneal phacoemulsification with intraocular lens implantation was performed. Sclerostomy and subsequent surgical steps were then performed as described previously.
Postoperative treatment included a topical antibiotic and a topical steroid for 8 to 12 weeks. Patients were seen on the first postoperative day and thereafter depending on the clinical condition.
Statistical analysis was performed using SPSS statistical software version 16.0 (SPSS Inc, Chicago, Illinois). We analyzed the relationship between type of glaucoma and postoperative complication as well as the need for reoperation. We also stratified the analysis by type of operation performed. The Fisher exact test was used for comparison of categorical data. For quantitative variables, a t test was done to compare means. Statistical significance was defined as P < .05. Univariate logistic regression was used to delineate the relationship between individual risk factors and complications. A multivariate logistic model including all statistically significant risk factors was then performed to assess their independent effect.
A total of 446 subjects with PACG (112 who underwent trabeculectomy and 334 who underwent phacotrabeculectomy) and 816 subjects with POAG (208 who underwent trabeculectomy and 608 who underwent phacotrabeculectomy) completed 1 year of follow-up and were analyzed. The demographic features and the antiscarring agent used during surgery are summarized in Table 1. In both the trabeculectomy and phacotrabeculectomy groups, there were more men than women in the POAG group (P < .001) but more women than men in the PACG group (P < .001). With regard to the use of antiscarring agents, significantly more cases used mitomycin C than fluorouracil in all of the groups (P = .008).
In our series, we found that postoperative complications occurred in 65 eyes (5.2%) overall, with 27 eyes (8.4%) in the trabeculectomy group (PACG: 8.0%; 95% confidence interval [CI], 4.3%-14.6%; POAG: 8.7%; 95% CI, 5.5%-13.3%; P >> .99) and 38 eyes (4.0%) in the phacotrabeculectomy group (PACG: 5.1%; 95% CI, 3.2%-8.0%; POAG: 3.5%; 95% CI, 2.3%-5.2%; P = .31). The overall rate of complications was higher in the trabeculectomy group than in the phacotrabeculectomy group (P = .003). However, as a whole, there was no significant difference in the incidence of postoperative complications between the POAG and PACG groups (POAG: 4.8%; 95% CI, 3.5%-6.5%; PACG: 5.8%; 95% CI, 4.0%-8.4%; P = .53).
The types of postoperative complications are summarized in Table 2. The commonest complication found in our study was prolonged hypotony (defined as IOP <5 mm Hg) due to overfiltration, which was recorded in 23 eyes (1.8%). In the trabeculectomy group, overfiltration occurred in 4.5% (95% CI, 1.9%-10.0%) of eyes with PACG and 1.9% (95% CI, 0.8%-4.8%) of eyes with POAG (P = .32). In the phacotrabeculectomy group, the incidences were 2.1% (95% CI, 1.0%-4.3%) in eyes with PACG and 1.2% (95% CI, 0.6%-2.4%) in eyes with POAG (P = .42). Bleb leak was the second commonest complication found, with 11 reported cases (0.9%). This consisted of 7 cases (2.2%) among the trabeculectomy group (PACG: 0.9%; 95% CI, 0.2%-4.9%; POAG: 2.9%; 95% CI, 1.3%-6.2%; P = .44) and 4 cases (0.4%) in the phacotrabeculectomy group (PACG: 0.6%; 95% CI, 0.2%-2.2%; POAG: 0.3%; 95% CI, 0.1%-1.2%; P = .88). The higher incidence of bleb leak in the trabeculectomy group than in the phacotrabeculectomy group was found to be statistically significant (P = .008). Aqueous misdirection syndrome was found to be very rare in our series, occurring in only 1 eye with POAG following phacotrabeculectomy. With regard to infective complications, there were 4 cases of endophthalmitis (3 with POAG and 1 with PACG) and 2 cases of blebitis (both with POAG). One patient with POAG developed fluorouracil-related corneal epithelial toxic effects and another patient with PACG had bleb hemorrhage, both after phacotrabeculectomy.
Overall, there was no difference in complication rates between the mitomycin C and fluorouracil groups (mitomycin C: 5.9%; 95% CI, 4.5%-7.9%; fluorouracil: 4.0%; 95% CI, 2.7%-4.0%; P = .17). However, the mitomycin C group had more cases of overfiltration with hypotony than the fluorouracil group (mitomycin C: 2.6%; 95% CI, 1.7%-4.0%; fluorouracil: 0.8%; 95% CI, 0.3%-2.0%; P = .03).
Of the 65 eyes that encountered complications following surgery, 40 eyes (3.2%) required a second operation or procedure (Table 3). There were no significant differences comparing the overall incidence of reoperations between the glaucoma subtypes (POAG: 2.7%; 95% CI, 1.8%-4.1%; PACG: 4.0%; 95% CI, 2.6%-6.3%; P = .27). However, there was a statistically significantly higher incidence of reoperations in the trabeculectomy group than in the phacotrabeculectomy group (trabeculectomy: 5.0%; 95% CI, 3.1%-8.0%; phacotrabeculectomy: 2.6%; 95% CI, 1.7%-3.8%; P = .048). Twenty eyes (1.6%) with overfiltration subsequently underwent revision of trabeculectomy with anterior chamber re-formation. This consisted of 6 in the trabeculectomy group (PACG: 3.6%; 95% CI, 1.4%-8.8%; POAG: 1.0%; 95% CI, 0.3%-3.5%; P = .23) and 14 in the phacotrabeculectomy group (PACG: 2.1%; 95% CI, 1.0%-4.3%; POAG: 1.2%; 95% CI, 0.6%-2.4%; P = .38). Furthermore, another 10 eyes that developed bleb leak required conjunctival repair. Scleral buckle with trans pars plana vitrectomy was performed in 2 eyes that developed retinal detachment at postoperative weeks 5 and 7. Three of the 5 eyes that developed endophthalmitis underwent trans pars plana vitrectomy. One patient who developed severe endophthalmitis had to undergo evisceration. The 1 subject who developed aqueous misdirection syndrome was successfully treated with YAG capsulotomy and trans pars plana vitrectomy.
Logistic regression analysis was used to investigate risk factors for postoperative complications or need for reoperations. After adjusting for age and sex, combined phacotrabeculectomy was associated with lower risk of any complication (odds ratio [OR] = 0.467; P = .01), hypotony (OR = 0.356; P = .01), and endophthalmitis or blebitis (OR = 0.256; P = .03) compared with trabeculectomy. Combined phacotrabeculectomy was also associated with fewer reoperations (OR = 0.452; P = .04), while use of mitomycin C was associated with more reoperations than use of fluorouracil (OR = 2.373; P = .02). On the whole, PACG was not found to have any significant association with risk of complications or reoperations compared with POAG. On multivariate analysis, combined phacotrabeculectomy was associated with lower risk of any complication (OR = 0.482; P = .02), hypotony (OR = 0.373; P = .02), and endophthalmitis or blebitis (OR = 0.293; P = .04) compared with trabeculectomy. However, there was no significant association with reoperations. Use of mitomycin C as an adjuvant was still associated with higher risk of reoperations than use of fluorouracil (OR = 2.285; P = .03).
In this study, the overall incidences of postoperative complications in the first year after glaucoma filtering surgery for PACG and POAG were comparable. There was no significant difference between the PACG and POAG groups overall or between the trabeculectomy and phacotrabeculectomy groups. Many of the complications found were caused by specific problems usually unrelated to the underlying diagnoses. We found that trabeculectomy resulted in a significantly higher overall reported incidence of postoperative complications compared with phacotrabeculectomy (P = .003). However, statistical analyses between the 2 groups yielded no significant differences for individual complications except for bleb leak (P = .008). The reason for this is unknown. Interestingly, we found that there was a slightly higher incidence of overfiltration with hypotony in the trabeculectomy group than in the phacotrabeculectomy group, but this difference was not statistically significant (P = .21). We speculate that in cases of phacotrabeculectomy, viscoelastic material may have been left in the eye at the end of surgery either intentionally or unintentionally, thereby reducing the risk of hypotony. However, a prospective review of a larger group of patients would be required to confirm these findings.
The use of antiscarring agents fluorouracil and mitomycin C has been shown in clinical trials to increase the success rate of filtering operations.14- 17 Although effective in lowering IOP, their use has been associated with sight-threatening complications such as endophthalmitis and blebitis.18,19 We found a fairly low rate of endophthalmitis (0.3%) in our study. In several studies, the reported incidence of postoperative bleb-related infections ranged from 0.4% to 6.9%, with a composite incidence of 1.5% per year. This incidence is likely to increase with greater use of these agents intraoperatively.20 The overall incidence of 0.5% for bleb-related infections within the first year after surgery in our patients is lower than that reported from previous studies.21- 23 The incidence of infections (as logically expected) was also comparable between PACG and POAG and between trabeculectomy and phacotrabeculectomy. However, a previous study by Jampel et al21 suggested that combined phacotrabeculectomy may be protective against the development of infection. Generally, bleb-related infections occur several years after the initial surgery and, as stated earlier, there is a cumulative risk. This could explain why we did not find many bleb-related infections in our study with our short postoperative follow-up of 1 year. We also found that there was a higher incidence of overfiltration with hypotony and shallow anterior chamber in the mitomycin C group than in the fluorouracil group (P = .03). This is also not surprising as previous studies have shown that mitomycin C use has significant complications such as excessive hypotony with choroidal effusions and irreversible maculopathy.18,24,25
A few previous studies have reported surgical outcomes and complications after glaucoma surgery in eyes with PACG,26- 28 but no study to our knowledge has compared patients with PACG vs patients with POAG. In one series by Tsai et al,26 the long-term efficacy and safety of combined phacotrabeculectomy were compared with those of trabeculectomy alone for PACG. A total of 16 of 75 eyes (21.3%) in the phacotrabeculectomy group and 8 of 24 eyes (33.3%) in the trabeculectomy group experienced postoperative complications. This complication rate was far higher than our findings. Among these were 2 cases (2.7%) of shallow anterior chamber or hypotony in the phacotrabeculectomy group and 3 cases (12.5%) in the trabeculectomy group. Hyphema occurred in 6 eyes (8.0%) that underwent phacotrabeculectomy and 1 eye (4.2%) that underwent trabeculectomy. There were no cases of wound leak, blebitis, endophthalmitis, or aqueous misdirection syndrome. The higher incidence of shallow anterior chamber or hypotony in the trabeculectomy group as compared with the phacotrabeculectomy group was also seen in our study (9 [2.8%] and 14 [1.5%], respectively). Another study of trabeculectomy for acute primary angle closure also found much higher complication rates than our study.28 In particular, 19.6% had shallow anterior chamber and 12.5% had hyphema, while the corresponding figures for our study were 4.5% and 1.8%, respectively. The high complication rate of trabeculectomy for acute primary angle closure in that study is perhaps not surprising owing to the high risk of operating in these acutely inflamed eyes with markedly increased IOP. Hence, in our study, we excluded eyes with previous acute primary angle closure so as not to confound our results.
Our series is one of the largest analyzing Asian eyes, with a significant proportion of patients with PACG. The audit of complications in the records was standardized for all patients. The Singapore National Eye Centre has a 100% audit of glaucoma surgery complications, and we believe that these data accurately reflect postoperative complication rates after glaucoma surgery. However, any retrospective review has its own set of limitations. There were multiple surgeons involved in the operations and postoperative management of the patients. The choice of surgery may have been affected by the complexity of the patient's condition, and combined surgery may have been performed more often in less complicated cases. There may be underreporting of complications, although this was performed by an independent retrospective audit. Twenty percent of patients were lost to follow-up within the first year, and this could have decreased our reported complication rates. Some eyes may have had a bleb needling procedure with subconjunctival fluorouracil injections, the removal of conjunctival sutures, or suture lysis, all of which may have influenced the rate of complications, in particular the rates of bleb-related infections and overfiltration. However, these data were not available for analysis. Finally, we did not investigate the surgical outcomes such as IOP results; these will be examined as part of future work.
In conclusion, patients with PACG were not found to have any higher incidence of postoperative complications after either trabeculectomy or phacotrabeculectomy than patients with POAG. The rates of postoperative complications and reoperations were found to be low, and serious complications such as aqueous misdirection syndrome were rare in eyes with PACG undergoing glaucoma filtering surgery. Combined phacotrabeculectomy was associated with an overall lower risk of complications than trabeculectomy.
Correspondence: Tin Aung, MBBS, PhD, Singapore National Eye Centre, 11 Third Hospital Ave, Singapore 168751 (firstname.lastname@example.org).
Submitted for Publication: October 26, 2010; final revision received January 25, 2011; accepted February 20, 2011.
Published Online: April 11, 2011. doi:10.1001/archophthalmol.2011.71
Financial Disclosure: None reported.
Funding/Support: This work was supported by a grant from the National Medical Research Council, Singapore.