Uneventful Phacoemulsification after Trabeculectomy in Pseudoexfoliation Glaucoma versus Primary Open-Angle Glaucoma

Conflicting results have been reported on phacoemulsification in the filtered eyes. Some studies have demonstrated that phacoemulsification following trabeculectomy results in increased intraocular pressure (IOP) [1, 2]. Meanwhile, others have shown no adverse effects on IOP control [3, 4].

Pseudoexfoliation (PEX) has an etiological association with cataract and frequent necessity for glaucoma surgery [5]. Cataract surgery alone has been shown to result in a greater reduction in IOP in eyes with PEX than those without PEX [6‒8]. In many previous studies that have compared the trabeculectomy results, pseudoexfoliation glaucoma (XFG) showed poorer long-term IOP control than primary open-angle glaucoma (POAG) [9‒11]. Multiple researches have examined the effect of cataract surgery on trabeculectomy function in mix glaucoma types [12‒14]. The purpose of this study was to compare the influence of phacoemulsification after trabeculectomy in XFG versus POAG and to identify the factors affecting the final success.

In this retrospective, comparative study, we reviewed the medical records of patients who had undergone trabeculectomy with 5-fluorouracil at the Health Sciences University, Haydarpaşa Numune Training and Research Hospital, Istanbul, Turkey, between 2010 and 2020. The study protocol adhered to the tenets of the Declaration of Helsinki and was accepted by the ethical committee of same hospital.

Two-hundred four eyes of 204 patients were enrolled in this study. The 102 eyes of 102 patients with XFG (52 eyes) and POAG (50 eyes) who had undergone uneventful clear-corneal phacoemulsification at least 6 months after trabeculectomy with 5-fluorouracil were included in the TRAB-PHACO group. For a matched comparison group, we selected 102 patients (52 eyes with XFG and 50 eyes with POAG) from a total of 167 patients who had trabeculectomy using the same surgical technique (with 5-fluorouracil) but did not have cataract surgery history (TRAB group). The patients were matched for glaucoma type, gender, age, IOP, and the time between the trabeculectomy and cataract surgery. “Matched follow-up date” in the TRAB group was date that corresponded to the date of phacoemulsification in the TRAB-PHACO group. Time period between trabeculectomy and phacoemulsification in the TRAB-PHACO group was defined as “matched follow-up period” in the TRAB group (shown in Fig. 1).

The diagnostic criteria of PXG were open-angle confirmed by gonioscopy, visible pseudoexfoliation material on the anterior segment structures with a dilated pupil, IOP >21 mm Hg before antiglaucomatous agents, glaucomatous optic nerve head changes (diffuse or focal thinning in the neuroretinal rim), and glaucomatous visual field defects with computerized visual field examination (SITA-standard algorithm, 24-2 test, Humphrey Visual Field Analyzer II; Carl Zeiss Meditec, Jena, Germany). The criteria for POAG included open-angle confirmed by gonioscopy, glaucomatous optic nerve head changes with a compatible visual field loss and IOP >21 mm Hg before antiglaucomatous agents without any secondary cause of glaucoma.

Patients who had previously undergone intraocular surgery other than one trabeculectomy, patients with IOP >21 mm Hg with or without medical treatment, patients had a follow-up less than 6 months, and who developed complications during phacoemulsification (capsular tear, vitreous loss, implanted IOL outside of the capsular bag) or during trabeculectomy surgery were excluded from the study. Eyes with minor complications such as transient corneal edema and inflammatory reaction following phacoemulsification and those with transient anterior chamber shallowing, hypotony, hyphema following trabeculectomy were included. These minor complications resolved without need for additional medication or intervention. If both eyes of a patient satisfied the inclusion criteria of the TRAB-PHACO or TRAB group, the first operated eye was chosen.

In both groups, trabeculectomy surgery was performed with 5-fluorouracil (5-FU) under limbus-based conjunctival flaps at the superior quadrant and rectangular scleral flaps of at least half thickness. Phacoemulsification was performed by two different, but experienced, phacoemulsification surgeons, and was always performed using standard phacoemulsification techniques with the Infiniti Vision System (Alcon Laboratories) with sub-Tenon anesthesia. The clear-corneal incision (2.8 mm) was in a different site from the filtering bleb. Foldable one-piece acrylic hydrophobic IOL (Eyecryl Plus ASHFY600; Biotech Vision Care Pvt. Ltd.) was implanted in the capsular bag. No antimetabolite injection or bleb needling was performed during cataract surgery. None of the patients received preoperative medication except glaucoma medications they already use. All patients used prednisolone acetate and moxifloxacin drops during the postoperative 1 month.

Patient data were recorded for age at phacoemulsification (and at the matched follow-up date in the TRAB group), gender, IOP before trabeculectomy, time between trabeculectomy and cataract surgery (and the matched follow-up period in the TRAB group) and follow-up time. Preoperative data included corrected-distance visual acuity (CDVA), the visual field (VF) tested by automatic computerized perimetry (24-2, SITA-standard, Humphrey Field Analyzer II; Carl Zeiss, Jena, Germany), IOP, and the number of glaucoma medications. Snellen chart was used for CDVA and then transformed into logarithm of the minimum angle of resolution (LogMAR). The average of two IOP measurements that had performed by Goldmann applanation tonometry was recorded. In the TRAB-PHACO group, follow-up information included IOP and the number of medications at 1, 3, 6, 12, 24 months and last visit, and also VF and CDVA at the last visit. In the first postoperative 24 h after phacoemulsification, IOP >25 mm Hg was defined as an IOP spike. In the TRAB group, follow-up data were collected after matched follow-up date; at 6, 12, 24 months, and last visit.

Surgical failure was defined as IOP >21 mm Hg at any time point during follow-up or IOP ≤21 mm Hg with additional antiglaucoma medications or additional glaucoma surgery. Eyes that required additional glaucoma surgery were excluded from any further CDVA, VF, IOP, and medication number analysis. Eyes that required additional medication were excluded from further IOP analysis. Final values before additional treatment were taken as last visit values.

Statistical analysis was performed using the SPSS for Windows (v.20.0, SPSS Inc.). The variables were investigated by using Kolmogorov-Simirnov/Shapiro-Wilk’s test to determine whether or not they are normally distributed. Groups were compared using the χ² test/Fisher’s exact test for categorical variables and independent T test/Mann-Whitney U test for continuous variables. The comparisons between preoperative and postoperative IOP, medication number, and CDVA were tested with paired Student’s t test/Wilcoxon signed-rank test. Survival analysis according to final success was performed by the Kaplan-Meier test. The distribution of survival rates for the groups was compared with the log-rank test. Univariate and multivariate logistic regression analyses were carried out in order to establish the risk factors for surgical failure. All predictors that had a p value <0.20 (cutoff) in the univariate analysis were included in the multivariate analysis. The values were given as mean ± standard deviation and also with range when required. p value <0.05 was considered statistically significant.

There was no statistically significant difference in the baseline patient characteristics between two glaucoma types both in the TRAB-PHACO and TRAB groups (Table 1). In the TRAB-PHACO group, the mean IOPs were significantly lower in the eyes with XFG than those with POAG only at 1-month and 3-month visits (p = 0.003 and p = 0.01, respectively). When compared with the baseline, in the TRAB-PHACO group, the mean IOP significantly increased at 24-month visit both in XFG and POAG (p = 0.02 and p = 0.004, respectively) (Table 2).

When compared with the preoperative, in the TRAB-PHACO group, the mean number of medication significantly increased only at the last visit (p = 0.001) in XFG. There was a statistically significant increase at 6-month visit (p = 0.02) in POAG and remained so throughout the rest of follow-up (Table 3).

Visual acuity improvement after phacoemulsification was statistically significant in both XFG and POAG groups (p ˂ 0.001). The mean preoperative and postoperative CDVA were statistically significantly worse in the eyes with XFG than those with POAG (p = 0.01 and p = 0.03, respectively). Change from baseline to last visit, demonstrated statistically significantly improvement for visual field mean deviation (MD) in the eyes with XFG (p = 0.001) and eyes with POAG (p = 0.003). The mean preoperative and postoperative MD were statistically significantly worse in the eyes with XFG than those with POAG (p = 0.006, both) (Table 4).

There was no statistically significance in failure rates between two glaucoma types at any follow-up visits in both TRAB-PHACO and TRAB groups (Table 5). The additional glaucoma surgery frequency was more common in XFG eyes without phacoemulsification. This difference was statistically significant between XFG and POAG in the TRAB group (p = 0.02).

Mean survival times for success were 23.2 ± 2.0 and 25.5 ± 2.1 months in the TRAB-PHACO and TRAB groups, respectively, in the eyes with XFG (shown in Fig. 2). Mean survival times for success were 26.7 ± 2.6 and 26.1 ± 1.9 months in the TRAB-PHACO and TRAB groups, respectively, in the eyes with POAG (shown in Fig. 3). The distribution of survival rates was not statistically different between XFG and POAG in both TRAB-PHACO and TRAB groups (log-rank test, p = 0.21 and p = 0.97, respectively).

Patient age, IOP before trabeculectomy, preoperative IOP and number of glaucoma medications, cataract surgery and glaucoma type (XFG/POAG) were investigated as potential risk factors. In addition, regression analysis was used to determine if time between trabeculectomy and phacoemulsification, CTR implantation and iris manipulation during cataract surgery, presence of an IOP spike at postoperative first day, influenced the failure in the TRAB-PHACO group. In 204 eyes with and without cataract surgery, preoperative IOP (p = 0.001) and cataract surgery (p = 0.18) were included in the multivariate analysis, only a higher baseline IOP at the study entry, was found to have influence on failure (p = 0.001; odds ratio = 1.17; 95% CI = 1.07–1.28). In the TRAB-PHACO group (102 eyes), age (p = 0.17), preoperative IOP (p = 0.03), and IOP spike (p = 0.01) were included in the multivariate analysis. The only significant factor related with failure was the presence of an IOP spike at postoperative first 24 h (p = 0.04; Odds ratio = 3.58; 95% CI = 1.04–12.36).

The influence of phacoemulsification on IOP control in patients with functioning filtering blebs has been previously investigated. In separate uncontrolled studies, statistically significant increases in IOP and/or number of glaucoma medications were reported following phacoemulsification in previously filtered eyes [15‒18]. However, none of these studies considered the time-dependent reduction in surgical success of trabeculectomy function [19]. Lindemann et al. [20] studied long-term results of trabeculectomy with 5-FU in eyes without cataract surgery and reported that the year 1 complete success (without medication) rate of 83% decreased to 64% at the end of year 3.

In a prospective case-control study, in 108 filtered eyes with open-angle glaucoma, it was found that uneventful phacoemulsification reduces the function of filtering bleb [21]. In this study, two groups had different ages and follow-up. Arimura et al. [22] reported that additional phacoemulsification deteriorates the outcome of Mitomycin C-augmented trabeculectomy in POAG and XFG, but the two groups were significantly different in terms of age and number of glaucoma medications at the study entry.

We thought that in order to reach valid and appropriate conclusions about the influence of phacoemulsification on filtering blebs, a control group (trabeculectomized eyes without cataract surgery) should be incorporated in the study design. Case and control groups should not statistically differ in terms of glaucoma type, gender, age, IOP, medication number, time interval between trabeculectomy and phacoemulsification, follow-up time.

In two retrospective, case-control studies, any adverse effect of temporal clear-corneal phacoemulsification on IOP control could not be demonstrated in eyes with filtering blebs [3, 23]. In Inal et al.’s prospective study, uncomplicated phacoemulsification was not found to have any additional unfavorable influence on IOP control in 30 eyes that had undergone trabeculectomy without antimetabolites [24].

Variations in patients’ characteristics, follow-up time, failure criteria, surgical technique, and methods of statistical analysis make comparisons among the previous studies difficult. In the present study, patients were individually matched according to the glaucoma type and the time interval between filtering surgery and study entry. There was no significant difference in terms of gender, age, IOP, number of glaucoma medications between groups with or without phacoemulsification and also between two glaucoma types. There were some differences between the previous studies and ours. We included eyes that had IOP <22 mm Hg with or without glaucoma medications. In our study, eyes with complicated cataract surgery were excluded because complications such as posterior capsular rupture and/or vitreous loss might lead to a possible unfavorable influence on the filtering bleb. Ehnrooth et al. [25] reported that the proportion of failures in the cataract surgery group was twice that in the no cataract surgery group in eyes with POAG or XFG, but six of 28 failured eyes had complicated cataract surgery and patients were significantly younger in the group without cataract surgery.

In the present study, all trabeculectomies were performed with 5-FU and therefore it would be appropriate to compare our results with the studies that included the eyes with a history of 5-FU augmented trabeculectomy [2, 13, 14]. Pylak et al. [2] included 50 eyes with POAG that had IOP <17 mm Hg without medication and demonstrated that phacoemulsification causes a significant elevation of IOP in the eyes after previous trabeculectomy with 5-FU. Swamynathan et al. [13] performed a retrospective analysis of 29 patients who had undergone antimetabolite-augmented trabeculectomy followed by phacoemulsification. This group was matched in terms of length of follow-up and type of antimetabolite used to a group that did not have cataract surgery. Failure was defined as an IOP >21 mm Hg or the need for additional surgery, they found no difference between the two groups. Husain et al. [14] included the eyes that had a previous trabeculectomy with either intraoperative 5-FU or placebo and reported that participants with open-angle glaucoma had failure significantly earlier than participants with primary closed-angle glaucoma.

It has also been shown that phacoemulsification may independently reduce IOP more in eyes with XFG than in those with POAG [26, 27]. In the previous report by American Academy of Ophthalmology [26], for POAG (9 studies, 461 patients) showed that phacoemulsification reduced IOP by 13% and glaucoma medications by 12%. For XFG (5 studies, 132 patients), showed phacoemulsification reduced IOP by 20% and glaucoma medications by 35%. The meta-analysis by Masis et al. [27] determined an IOP drop of −5.8 mm Hg and −2.7 mm Hg from baseline in the XFG and POAG eyes, respectively. In our study, IOP control was indeed slightly better in eyes with XFG in the short term (at 1- and 3-month visits). We found that, IOP control worsened earlier in the TRAB group compared to the TRAB-PHACO group, in eyes with XFG. No statistically significant increase in the number of medication was observed for a longer period of time after phacoemulsification in XFG compared to POAG. Although there was no significant difference between XFG and POAG in the TRAB-PHACO group, the additional surgery was significantly more common in the eyes with XFG than those with POAG in the TRAB group. To prevent the effect of IOP decrease after additional surgery or glaucoma medication, eyes that required additional surgery or medication were excluded from any further IOP analysis. These results can be explained as the IOP-lowering effect of the cataract surgery is greater in the filtered eyes with XFG than those with POAG.

The IOP-lowering effect of cataract surgery in the eyes with PEX, is assumed to be associated with the anterior chamber widening, reduced liberation of exfoliation material and pigment granules after surgery due to decrease in iridolenticular friction, aspiration of debris during surgery, partial removal of the exfoliation-producing anterior capsule, and improvement in trabecular outflow [28]. In the previous studies, which analyzed the long-term risk of glaucoma in PEX eyes, the number of newly diagnosed glaucoma cases was lower than expected at 6–7 years, following cataract extraction [29, 30]. In these studies, a possible explanation was that the removal of the biological lens and the central anterior capsule reduces the release of pseudoexfoliative material and pigment, in addition to wash-out of pseudoexfoliative material during the operation. In our study, it is possible that aspiration of debris from the trabecular meshwork during surgery, reduced release and production of exfoliation material after surgery, and consequent improvement in trabecular outflow were actually the reasons behind the favorable outcome of phaco after TRAB in PXG eyes. In some previous studies, it has been reported that the type of glaucoma (POAG/XFG) does not influence the outcomes of cataract surgery in the trabeculectomized eyes, but these studies had small sample size and did not describe the change in IOP, medication numbers and failure rates of the eyes with XFG [22, 24, 31].

A significant association between a higher IOP and medication number prior to cataract surgery and increased risk of bleb failure has been demonstrated in several studies [15, 21, 24]. Published studies indicate that the earlier cataract surgery is performed, the greater the risk of subsequent trabeculectomy failure [4, 14, 31]. Chen et al. [32] demonstrated that the younger age was associated with an increased risk of loss of IOP control in the filtered eyes after cataract surgery.

In our study, the only risk factor, associated with trabeculectomy failure after cataract surgery, was the presence of an IOP spike in the multivariate logistic regression analysis. Fifteen eyes of 102 eyes with phacoemulsification (15%) had an IOP spike at postoperative first 24 h and failure was observed in 10 of these eyes. Rebolleda et al. [18] reported that, on the first day after phacoemulsification, nine filtered eyes (18%) had an IOP spike of greater than 10 mm Hg above the IOP before phacoemulsification. The IOP spike did not have any significant effect on the failure rate. Chen et al. [32] reported that an IOP >25 mm Hg during the early postoperative period was significantly associated with loss of IOP control and additional glaucoma surgery need in trabeculectomized eyes.

This study has some limitations. The bleb morphology before or after phacoemulsification and the effect of cataract density on the results could not be analyzed because of our study’s retrospective nature. We excluded the complicated cataract surgeries. The patients had a variable follow-up. Future studies are necessary to overcome these limitations and give a clearer result about the long-term effects of phacoemulsification after trabeculectomy in XFG.

To our knowledge, this is the first study that has compared the effect of uneventful phacoemulsification in eyes with XFG versus POAG that have undergone a trabeculectomy with 5-FU. In our study, the IOP control progressively worsened in a time-dependent manner in the filtered eyes in both glaucoma types with or without cataract surgery. Uneventful clear-corneal phacoemulsification did not have any additional deleterious influence on that course. On the contrary, in XFG, phacoemulsification may delay time-related worsening in IOP control and may decrease the additional glaucoma surgery need in the filtered eyes. We suggest that cataract surgery may be performed in XFG eyes with well-functioning filtering blebs whenever necessary. An IOP spike after phacoemulsification may be a sign that the filtration is not working well and these patients should be followed carefully.

This study was approved by the Research Ethics Committee of University of Health Sciences, Haydarpasa Numune Education and Research Hospital (the approval reference number: HNEAH-KAK-KK-2020-6.) and was conducted in accordance with the tenets of the World Medical Association Declaration of Helsinki. Patient consent was not required as this study was based on publicly available data.

The authors have no conflicts of interest to declare.

The authors did not receive any funding.

Hatice Tekcan and Mehmet Serhat Mangan: study design, study performance, data collection and management, data analysis and interpretation, writing and review of the manuscript; Hatice Tekcan, Mehmet Serhat Mangan, and Serhat Imamoglu. Hatice Tekcan contributed to the manuscript as the first author, and as the corresponding author. All the authors have approved the final manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.