The Effect of Prior Trabeculectomy on Refractive Outcomes of Cataract Surgery


To examine surgical and refractive outcomes of phacoemulsification with intraocular lens (IOL) implant in eyes with prior trabeculectomy.


Retrospective observational case-control study.


The study compared eyes that underwent phacoemulsification with IOL implant at least 3 months post-trabeculectomy (n = 77) with eyes with either medically controlled glaucoma (n = 43) or no glaucoma (n = 50) at an academic institution. The main outcome measure was the difference between the expected and the actual postoperative refraction.


Mean intraocular pressure (IOP) increased in trabeculectomy eyes from 8.7 ± 4.2 mm Hg to 10.7 ± 4.0 mm Hg ( P < .0001), whereas it decreased in glaucoma control and normal control groups by 2.0 mm Hg ( P = .003) and 2.1 mm Hg ( P < .00001), respectively, with concurrent decrease in drops in the glaucoma control group (0.76 to 0.23, P < .0001). The difference from expected refractive outcome was −0.36 (more myopic) in trabeculectomy eyes compared with +0.23 (more hyperopic) in nonglaucoma controls and +0.40 in glaucoma controls ( P < .0001). The correlation between change in IOP vs extent of refractive surprise was statistically significant ( P = .01, r = −0.20). Final visual acuity was not affected by the difference in refractive error.


The refractive surprise correlated to IOP change, with 2 mm Hg rise resulting in a −0.36 diopter shift between predicted and actual refraction. After cataract extraction, IOP decreased in controls and fewer drops were required, but IOP increased in the study group. Factors affecting refractive surprise in cataract surgery after trabeculectomy, especially IOP change and axial length, require further investigation.

Glaucoma is the second-leading cause of blindness in the world. Primary open-angle glaucoma affected about 66 million people worldwide in the year 2000 and this figure is expected to increase to 79.6 million by the year 2020, disproportionately affecting women and Asians. The gold standard of initial surgical treatment in patients with uncontrolled glaucoma by medical and laser therapy remains, in most cases, trabeculectomy, which has been modified in a number of ways since it was first introduced by Cairns and Watson in 1968. A characteristic bleb is created by making a fistula between the anterior chamber and the subconjunctival space. Following trabeculectomy, patients are at a higher risk of developing cataracts. Management of visually significant cataracts in patients with prior trabeculectomy represents a clinical challenge. In addition to the risk that cataract surgery may compromise the success of an earlier trabeculectomy and result in loss of intraocular pressure (IOP) control, changes in axial length and corneal astigmatism following trabeculectomy have been reported, resulting from variances in IOP. The purpose of this study is to evaluate refraction outcomes in patients who have had cataract extraction with intraocular lens implant (IOL) after having previously undergone trabeculectomy to better aid in the process of selecting an appropriate intraocular lens preoperatively.


This is a retrospective observational case-control study. Approval of this study was granted by the Indiana University Institutional Review Board for retrospective review of existing patient data. Medical records of patients who have undergone phacoemulsification with posterior chamber intraocular lens implantation at least 3 months after prior trabeculectomy by a single surgeon (L.B.C.) in an academic medical center between September 1, 1997 and November 30, 2008 were reviewed. All cataract surgeries were uncomplicated phacoemulsification performed by the same surgeon who performed the trabeculectomy. All trabeculectomies were performed with limbal-based conjunctival flap with either intraoperative 5-fluorouracil (50 mg/mL for 5 minutes) or intraoperative mitomycin C (0.2 mg/mL for 2 minutes). Exclusion criteria included surgical complications either at trabeculectomy or phacoemulsification surgery, revision of bleb, sclerotomy or choroidal drainage performed at time of phacoemulsification, concomitant vitrectomy procedures, follow-up shorter than 6 months post-phacoemulsification, sulcus or anterior chamber intraocular lens placement, and combined procedures such as phacoemulsification/trabeculectomy or phacoemulsification/canaloplasty. Only patients receiving 3-piece silicone lenses (SI40NB or CLRFLXC; Abbot Medical Optics, Abbott Park, Ilinois, USA) were included because of their similar design and A-constant.

The study group included 77 eyes of 65 patients who underwent phacoemulsification with intraocular lens implant after trabeculectomy. Two control groups were used based on whether a diagnosis of glaucoma existed or not: 43 eyes of 36 patients with medically controlled glaucoma (no prior trabeculectomy) who underwent phacoemulsification with intraocular lens implant only (glaucoma control group); and 50 eyes of 36 patients who did not have glaucoma and underwent phacoemulsification with intraocular lens implant (normal control group). The preoperative axial length was measured by contact or immersion A-scan ultrasonography.

The data collected included time between trabeculectomy and phacoemulsification and age at phacoemulsification. Preoperative visual acuity, manifest refraction, axial length, and IOP were collected from the last visit prior to phacoemulsification. Corresponding postoperative values were taken at the first visit after inflammation had resolved and refraction was possible. Charts with missing data were excluded in all 3 groups (n = 55). Refractions were converted to spherical equivalents (spherical power + ½ cylinder power) for comparison purposes. Postoperative refractions were compared with the expected refractive outcomes as determined by standard formulas SRK II/SRK-T. Visual acuities were converted to logMAR equivalent for comparison. Statistical analysis was performed with SPSS, ver. 19 (IBM, Somers, New York, USA). The main outcome measure was the difference between the expected spherical equivalent refractive outcome and the actual postoperative spherical equivalent refraction among the study and control groups.


Patient characteristics are summarized in Table 1 . There was no statistically significant difference in patient demographics between the study and control groups. The axial length of eyes with prior trabeculectomy tended to be shorter than eyes without prior trabeculectomy. Of the 65 patients in the study group, 53 did not have a trabeculectomy in the fellow, nonstudy eye. Among these 53 patients, axial length data were available on the fellow eye in 37. Compared with the axial length of the fellow eye that did not have prior trabeculectomy, the study eyes were significantly shorter (23.8 ± 1.2 mm vs 24.1 ± 1.4 mm, P = .0002, paired t test). Among the study subjects, there was a mild correlation between difference in axial lengths between fellow eyes and difference in IOP between fellow eyes (Pearson r = 0.39, P = .018). In contrast, among the 29 patients in the control group without glaucoma who had axial length measurements in fellow eyes (excluding second eyes that underwent cataract surgery), the difference between cataract and fellow eyes was not statistically significant (23.8 ± 1.7 mm vs 23.8 ± 1.8 mm, P = 0.51, paired t test). Similarly, among the 29 glaucoma control patients with axial length data in both eyes (excluding second eyes that underwent cataract extraction), the difference between fellow eyes was not statistically significant (23.7 ± 1.4 mm vs 23.7 ± 1.5 mm, P = .80, paired t test).

Table 1

Baseline Patient Characteristics for Control and Study Groups: Patients Without Glaucoma (Normal Control Group), Patients With Medically Controlled Glaucoma (Glaucoma Control Group), and Patients who Have Previously Undergone Trabeculectomy (Study Group)

Group P Value
Normal Control Glaucoma Control Study
Number of eyes/patients 50/36 43/36 77/65
Right eye : left eye 25:25 23:20 32/45 .40 a
Age at surgery (y ± SD) 71.3 ± 12.7 75.1 ± 10.5 72.6 ± 12.1 .29 b
Age range (y) 30-92 52-94 38-91
Sex (M:F) 24:26 16:27 25:52 .21 a
Glaucoma diagnosis
Primary open-angle 0 31 58 .70 c
Angle-closure 0 7 8 .39 c
Normal tension 0 1 7 .26 c
Pigmentary 0 1 2 >.999 c
Pseudoexfoliative 0 0 1 >.999 c
Secondary 0 3 1 .13 c
None 50 0 0

a χ 2 test (df = 2).

b 1-way ANOVA.

c Fisher exact test for glaucoma control and study groups.

All patients in study and control groups had significant vision improvement after phacoemulsification with intraocular lens implant. Patients in the study group took an average of 32 more days to stabilize and achieve their final refraction compared with those in the control groups ( P = .005, Mann-Whitney test). They also had a greater absolute refractive difference and variability in postoperative refraction. However, final visual acuity was not affected by the difference in refractive error.

Mean IOP increased in trabeculectomy eyes from 8.7 ± 4.2 mm Hg to 10.7 ± 4.0 mm Hg ( P < .0001, paired t test), whereas mean IOP decreased in normal control and glaucoma control groups by 2.1 mm Hg ( P < .00001) and 2.0 mm Hg ( P = .003), respectively, with a concurrent decrease in required drops in the glaucoma control group (0.76 to 0.23, P < .0001, Wilcoxon matched pair test). Additionally, mean logMAR visual acuity improved significantly in all groups after phacoemulsification and was not statistically different among the groups ( Table 2 ).

Table 2

Summary of Results Before (Preoperative) and After (Postoperative) Cataract Extraction With Intraocular Lens Implantation in the 3 Groups: Eyes Without Glaucoma (Normal Control), Eyes With Medically Controlled Glaucoma (Glaucoma Control), and Eyes With Prior Trabeculectomy (Study)

Group Preoperative Mean ± SD Postoperative Mean ± SD P Value a
IOP (mm Hg) Normal control 15.3 ± 2.4 13.3 ± 2.7 .00001
Glaucoma control 16.3 ± 3.9 14.3 ± 3.3 .003
Study 8.7 ± 4.2 10.7 ± 4.0 .0001
Visual acuity (logMAR) Normal control 0.42 ± 0.37 0.19 ± 0.24 <.0001
Glaucoma control 0.52 ± 0.41 0.24 ± 0.33 <.0001
Study 0.65 ± 0.52 0.23 ± 0.23 <.0001
Refraction: predicted/actual (spherical equivalent) Normal control −0.37 ± 0.44 −0.14 ± 0.94 .035
Glaucoma control −0.37 ± 0.29 0.03 ± 1.08 .047
Study −0.34 ± 0.39 −0.70 ± 1.59 .0004
Number of medications Glaucoma control 0.76 ± 0.58 0.23 ± 0.43 <.0001
Study 0.12 ± 0.40 0.14 ± 0.39 .56

a Paired t test for IOP; Wilcoxon matched-pairs signed ranks test for all others.

Eyes undergoing cataract extraction after prior trabeculectomy had significantly greater refractive surprise than those in the control groups. The difference from expected refractive outcome was −0.36 (more myopic) in trabeculectomy eyes compared with 0.23 (more hyperopic) in nonglaucoma control eyes and 0.40 in glaucoma control eyes ( P < .0001, Kruskal-Wallis test) ( Table 2 ). The postoperative refractive difference was significantly different between the study group and each of the other 2 groups ( P = .004, 1-way ANOVA with Tukey post test). For all eyes, the refractive difference correlated to intraocular pressure change ( P = .01, r = −0.20). Excluding 1 patient’s highly anomalous point (+9 diopter surprise), findings are even more significant ( P < .0001, r = −0.33, r 2 = 0.11) ( Figure 1 ). In addition, the most myopic refractive surprise occurred in eyes that had postoperative increase in IOP, particularly those with preoperative IOPs below 8 mm Hg ( Figures 2 and 3 ).

Jan 9, 2017 | Posted by in OPHTHALMOLOGY | Comments Off on The Effect of Prior Trabeculectomy on Refractive Outcomes of Cataract Surgery

Full access? Get Clinical Tree

Get Clinical Tree app for offline access