Vol 31 No 1 Original Article PDF

Risk of visual loss in advanced glaucoma after trabeculectomy and combined cataract surgery

Raquel M. Joaquin-Quino, MD , Patricia M. Khu, MD, MSc , Nilo Vincent Dg. Florcruz, M.D. , Jesus N. Sarol, PhD

PATIENTS with advanced glaucoma are at substantial risk of blindness arising from such causes as elevated intraocular pressure (IOP), progression of field loss, and loss of central fixation. Eyes with more advanced damage deteriorate 11 times faster, and progression of field loss may occur in spite of seemingly good control of glaucoma. The need for much lower IOP and individualized care in these eyes cannot be overemphasized. Bringing down IOP in advanced glaucoma usually requires surgery, generally trabeculectomy, the most commonly performed. Trabeculectomy was successful in preventing blindness with a 54% probability after 10 years, 42% after 15. Risks of filtration surgery in advanced glaucoma include endophthalmitis, increased rate of cataract surgery, progression of glaucomatous damage leading to reduced visual acuity or loss of central fixation, hypotony, and aqueous misdirection. Cataract contributes to both acuity and field loss. In the Advanced Glaucoma Intervention Study, it contributed more to acuity loss than to field loss. For eyes with significant cataract and further need of IOP lowering, a combined trabeculectomy and phacoemulsification done either on the same or different sites showed better longterm IOP control than cataract surgery alone. Removal of the cataract was also found to reduce visual-field deficits significantly. In advanced glaucoma with visual-field deficits encroaching central fixation, the risk of “wipeout” is a real concern. The existence of wipeout, defined as loss of central visual field in the absence of other explanation as a cause of postoperative loss of visual acuity, has been debated. Considered rare, it has been shown in several studies to have an incidence of less than 2%, 6, 7, 8, 9 with one study reporting as high as 14%. Its occurrence has made some ophthalmic surgeons reluctant to perform intraocular surgery in eyes with advanced glaucoma; however, its rarity should be a reason for not deferring surgery. Thus, this study evaluated the risks of postoperative visual loss in patients with advanced glaucoma who underwent trabeculectomy or combined surgery. Specifically, we determined and compared the incidence of visual loss after trabeculectomy and combined surgery in patients with advanced glaucoma and the factors affecting the decrease in visual acuity postoperatively.

We reviewed the charts of patients 18 years or older seen at the Glaucoma Clinic of the University of the Philippines–Philippine General Hospital from 2002 to 2004 who had adv anced gl aucoma and under went trabeculectomy or combined cataract surgery [phacoemulsification with intraocular-lens (IOL) implantation and trabeculectomy or extracapsular cataract extraction (ECCE) with intraocular-lens implantation and trabeculectomy]. Eyes were defined as having advanced glaucoma if absolute field defects occurred within 10 degrees of fixation or the cup-disc ratio is greater than 0.8 associated with markedly excavated or atrophic disc as confirmed by optic-disc evaluation in the affected eye. Included were eyes with best-corrected visual acuity (BCVA) of counting fingers (CF) at 1 foot or greater in the affected eye, no previous intraocular surgery, and a follow-up of at least 2 months. Eyes with macular problems, diabetic retinopathy, and severe myopic retinal degeneration were excluded.
Records were reviewed for the following parameters: age, gender, presence of systemic disease, family history of glaucoma, type of glaucoma, duration of glaucoma from first diagnosis, and presence of cataract. Visual acuity and IOP were recorded from the visit before surgery and postoperatively at 1 week, 1 month, and 2 months, including the use of glaucoma medication at 1 month postoperatively. Complications such as vitreous loss, shallow anterior chamber, overfiltration, and hypotony were also recorded. In the trabeculectomy group, all surgeries were done under local anesthesia. A superior corneal bridle suture was placed for a wide exposure of the superior conjunctiva. Limbal-based conjunctival flap was made, followed by the creation of a half-thickness scleral flap. In all surgeries, 0.2-0.4 mg/ml of mitomycin-C on cotton swabs was applied for 2 to 4 minutes under the scleral flap, followed by copius irrigation. Paracentesis was placed temporally, and sclerectomy was done with the use of a Kelly-Descemet punch. A peripheral iridectomy was performed before scleral-flap closure with nylon 10-0 sutures. Adequate closure showed slow oozing of aqueous humor. Watertight conjunctival closure was achieved at the conclusion of the surgery. Topical steroid, antibiotic, and atropine 1% eye drops were given postoperatively for an average of three months, the doses gradually tapered. In combined cataract and trabeculectomy done under local anesthesia, a limbalbased conjunctival flap and a triangular scleral flap were created followed by mitomycin-C application prior to phacoemulsification on the temporal side and IOL insertion. Sclerectomy was subsequently performed, fol lowed by iridectomy, closure of the scleral and conjunctival flaps. Those who underwent ECCE had a fornix-based conjunctival flap, followed by creation of the scleral flap, and application of mitomycin-C before the completion of the ECCE incision. The lens was delivered and IOL inserted before the wound was closed. Sclerectomy, iridectomy, and closure of the scleral flap and the conjunctiva were subsequently done. After surgery, topical steroid and antibiotic eye drops were administered.

Outcome measures included visual acuity (VA), IOP, and visual-field indices of the Humphrey perimetry such as mean defect (MD) and corrected pattern standard deviation (CPSD). The VA was measured at each visit using the standard Snellen chart; it was modeled as the logarithm of the reciprocal of the minimum angle of resolution (logMar). Eyes that lost 2 or more Snellen lines or worsened one category [categories were CF, Hand Movement (HM), Light Perception (LP), No Light Perception (NLP)] 2 months after the surgery were identified. Central vision is lost (wipeout) when BCVA is less than 20/200 or when the result of the visual-field examination confirmed the existence of a central scotoma.
It is characterized as a sudden, permanent, and unexplained visual loss within 2 months postoperatively. Return to the preoperative level at anytime after 2 months of follow-up was considered not to have lost central vision from the operation. Mean IOP, measured with the Goldmann applanation tonometer, was calculated for each period. IOP maintained at less than 14 mm Hg with or without medication was considered successful control of IOP. Visual-field results (MD and CPSD), obtained with a Humphrey Field Analyzer (Zeiss Corporation, San Leandro, CA, USA), were compared preoperatively and between 1 and 6 months postoperatively. Since not all eyes had pre- and postoperative visual fields, a subgroup analysis was done only on eyes with both fields. Data collected were entered into MS Excel (Microsoft Corporation, Redmond, WA, USA) worksheet and subjected to statistical analysis using Statistical Analysis System (SAS) version 6.12 (SAS Institute, Cary, NC, USA). ANOVA was used to determine differences between groups. Logistic regression was performed to determine the factors predictive of decrease in VA postoperatively. The p value was considered significant if less than or equal to 0.05.

A total of 98 eyes of 92 patients were included in the study. Over half were female (56%). Only 10% had family history of glaucoma. Among those with systemic illnesses, 17 had hypertension and diabetes. Most had chronic angle-closure glaucoma (CACG) (52%), followed by open-angle glaucoma (OAG) (43%). Twenty patients had clear lens, the rest had mild to moderate cataracts (Table 1). The mean age at the time of surgery was 54.4 ±12.1 years in the trabeculectomy group and 66.5 ± 9.8 in the combined group; patients were older in the combined group (p < 0.001) , and had wor se preoper a t i ve VA (p < 0.001)(Table 1). Twenty-six percent (25/98) had VA worse than 20/200 with no visual-field analysis done. Among those with VA better than 20/200 (73/98), 46% (34/73) had visual-field analysis, 5 in the combined group and 29 in the trabeculectomy group. Sixty-five percent (23/34) had fixation-splitting defect, described as an absolute defect extending to the horizontal meridian and completely crossing the vertical midline. The others had central 5- 10 tunnel remaining, and 1 had temporal island only. Visual-field indices for the two groups were not significantly different. The mean preoperative IOP was also similar.

Postoperative VA within 2 months did not differ significantly from baseline in the combined group. In the trabeculectomy group, similar findings were noted until the first month; by the second month, vision improved by 1 l ine ba sed on the mean logMa r, whi ch wa s a l so significantly different from the combined group (p = 0.01). IOP was lowest at week 1 for both groups, which gradually increased at 2 months. By 2 months, the mean IOP in the combined group was higher than in the trabeculectomy group (p = 0.04). ANOVA, however, showed no statistically significant difference in IOP at 1 week, 1 month, and 2 months postoperatively in each group. Only 14 patients in the trabeculectomy group and 1 in the combined group had visual fields tested at 6 months postsurgery. Subgroup analysis of these eyes comparing visual-field defects pre- and postsurgery showed no statistical change.
There were 8 in the combined and 15 in the trabeculectomy group with decrease in VA within 2 months after surgery (Table 2). Most were females (61%). Fifty-two percent (52%) had CACG and 48% OAG. Baseline logMar VA for the trabeculectomy group was better (p = 0.01). Vision worsened 2 months postoperatively up to a logMar mean of –1.67 in the combined group (p = 0.02). The mean preoperative IOPs in both groups were not statistically different. Mean IOP at 1 week and 2 months postoperatively was higher in the combined group (p = 0.004 and 0.058 respectively). Twelve patients had preoperative visual-field analyses; 3 had central island and the rest had split fixation. Preoperative visual-field indices for both groups were similar. Only eyes in the trabeculectomy group had available visual-field analysis postoperatively. Twenty-three eyes had decrease in VA that improved, but 14 did not return to their preoperative level. Visual recovery took place at 2 (5/13), 3 (4/13), and 6 (4/13) months postoperatively. Factors predictive of decrease in VA postsurgery are shown in Table 3. Surgical complications (p = 0.045) and increase in postoperative IOP at 2 months (p = 0.05) were predictive. Older age and better logMar baseline VA were not statistically significant. The type of glaucoma, type of surgery, and preoperative IOP were not predictive.
Irreversible loss of central vision developed in 3 patients with 0.9 to 1.0 cup–disc ratio; 1 in the trabeculectomy group and 2 in the combined group (1 phacoemulsification and 1 ECCE). Other causes of visual loss were excluded based on the postoperative ophthalmologic ev a lua t ion. Al l had CACG, and the surger ies were uncomplicated. Mean preoperative logMar VA of both groups were similar. At 2 months postoperatively, all three eyes worsened. No visual recovery was observed even after 2 months of follow-up in both groups (Table 4). The preoperative IOP in the trabeculectomy group was higher than in the combined group, but postoperative IOPs in both groups were less than 14 mm Hg.

Eyes with severe glaucomatous damage have meager remaining viable optic-nerve tissue that a dramatic decrease in visual acuity may result if any damage occurs in the remaining nerve tissue.Trabeculectomy or a combined procedure is warranted for those with elevated IOP to achieve IOP control and preserve visual acuity. However, not all surgeries are without risks, one of which is the phenomenon of “wipeout” or “snuffout,” which may occur after an apparently uncomplicated glaucoma surgery. The existence of wipeout has been questioned and its incidence debated in several reports. In this study, the incidence of permanent postoperative visual loss in advanced glaucoma after trabeculectomy with and without cataract surgery was 3% (3/98), which is comparable to those reported in other studies. Most studies supported the existence of “wipeout” after trabeculectomy in advanced glaucoma, but this incidence is relatively low, generally less than 2%. In the study by Kolker, which reported the highest incidence of 14% (3/22), macular edema was considered an identifiable and reversible cause, violating the definition of wipeout. Theories postulated to explain “wipeout” include the following: (1) sudden visual-acuity loss secondary to opticnerve hemorrhage brought about by sudden decompression of the eye or sudden lowering of the IOP intraoperatively, (2) decreased perfusion pressure due to transient systemic intraoperative hypotension, and (3) toxic or direct injury to the optic nerve from retrobulbar injection. The low incidence in our study could be attributed to advances and improvements in microsurgical techniques such as avoidance of sudden decompression of the globe intraoperatively and knowledge in the management of advanced glaucoma. The occurrence of wipeout in 3 cases despite lowered IOP preoperatively could be due to further compromise of the few remaining nerve fibers regardless of the IOP level. Another possible cause is the occurrence of IOP spikes seen in 1 of the eyes within a month postsurgery. The incidence of postoperative visual loss that recovered was more common in our study (20%). Causes identified were: refractive changes, lens opacification, keratopathy, maculopathy, and slow recovery of visual function in eyes with advanced glaucoma damage. Other studies have indicated that the return of maximal visual acuity after trabeculectomy or combined surgery may take 6 months or more. Delayed clearing of an undiagnosed macular edema after cataract surgery is also possible. Factors predictive of decrease in postoperative visual a cui t y were deter mined. The presence of surgical complications had 5 times the likelihood of decrease in VA postoperatively (Table 3). In the combined group, vitreous loss was the most common complication encountered that led to increased inflammation and subsequent cystoid macular edema. There was increased chance of failure of the filtering surgery from the inf l amma t ion and the subsequent inc rea se in IOP postsurgery. In the trabeculectomy group, increased risk of cataract formation or worsening of the cataract, also noted in other studies, may result from persistent shallow anterior chamber postsurgery. Increase in postoperative IOP at 2 months, so called IOP spikes, had 1.2 odds of causing decrease in VA. Although the odds were low, this postoperative increase in IOP in advanced glaucoma with marginal nerve fibers left could be significant. Better logMar baseline VA appeared to be predictive of postoperative VA but was not statistically significant. Most of the cases in this series had tunnel vision of less than 5 degrees with good VA. The risks of split fixation, cataract formation, and postoperative maculopathy in advanced glaucoma are all possible. It is believed by some investigators that macular fixation has already been lost in eyes with preoperative vision of 20/200 or less, and that those with residual central island, defined as visual defects extending to within 5 of fixation, are more vulnerable to minor opacities and postoperative maculopathy than those whose fixation has split. Our study favored the view held by Lichter and O’Connell that wipeout did not occur only in patients with split fixation. Thus, the VA measurement in this study may not have adequately reflected the true state of the patient’s central acuity. Older age is another factor predictive of decreased VA postoperatively; it was not significant in this study in contrast to other studies where it was strongly correlated with progression to blindness. Eyes in the combined group had worse visual acuity than those in the trabeculectomy group pre- and postoperatively. Patients in the combined group were older and likely to have denser lens opacities and other subclinical retinal problems.

Other studies reported strong association between older age and late glaucoma presentation, with the risk increasing linearly with age over 40 years. Our study had several limitations. The relatively small sample size for the combined group may make comparisons with the trabeculectomy group biased. The lack of pre- and postoperative visual-field tests in many patients in this study resulted in dependence on findings based solely on acuity loss rather than on acuity and field loss. Nevertheless, the finding of atrophic disc with cupping to disc margin seen in advanced glaucoma was noted in all cases. It most likely corresponded to a remaining central island on visual field. In conclusion, we found a 3% incidence of “wipeout” among patients with advanced glaucoma who have VA greater than 20/200 and underwent trabeculectomy or combined surgery. Postoperative decrease in visual acuity with an incidence of 20% in this series may be explained by several factors including the presence of intraoperative and postoperative complications and uncontrolled IOP postoperatively. Future studies with larger sample size and longer follow-up can determine the reversibility of the decrease in visual acuity in advanced glaucoma.

1. Wilson R, Walker AM, Dueker DK, Crick RP. Risk factors for rate of progression of glaucomatous visual-field loss. Arch Ophthalmol 1982; 100: 737-741.
2. Kolker AE. Visual prognosis in advanced glaucoma: a comparison of medical and surgical therapy for retention of vision in101 eyes with advanced glaucoma. Tr Am Ophth Soc 1977; 75: 539-555.
3. Parc CE, Johnson DH, Oliver JE, et al. The long-term outcome of glaucoma filtration surgery. Am J Ophthalmol 2001; 132: 27-35.
4. AGIS Investigators. The Advanced Glaucoma Intervention Study (6). Effect of cataract on visual field and visual acuity. Arch Ophthalmol 2000; 118: 1639-1652.
5. Friedman DS, Jampel HD, Lubomski LH, et al. Surgical strategies for coexisting glaucoma and cataract. Ophthalmology 2002; 109: 1902-1913.
6. Costa VP, Smith M, Spaeth GL, et al. Loss of visual acuity after trabeculectomy. Ophthalmology 1993; 100: 599-612.
7. O’Connell EJ, Karseras AG. Intraocular surgery in advanced glaucoma. Br J Ophthalmol 1976; 60: 124-131.
8. Martinez JA, Brown RH, Lynch MG, Caplan MB. Risk of postoperative visual loss in advanced glaucoma. Am J Ophthalmol 1993; 115: 332-337.
9. Lichter PR, Ravin JG. Risks of sudden visual loss after glaucoma surgery. Am J Ophthalmol 1974;78: 1009-1013.
10. AGIS Investigators. The Advanced Glaucoma Intervention Study (7). The relationship between control of intraocular pressure and visual-field deterioration. Am J Ophthalmol 2000; 130: 429-440.
11. Ritch R, Shields MB, Krupin T. Complications of glaucoma filtering surgery. The Glaucomas, second ed. Missouri: Mosby Year Book, Inc. 1996; v 3, chap. 84: 1703- 1736.
12. Heijl A, Leske C, Bengtsson B, et al. Reduction of intraocular pressure and glaucoma progression. Arch Ophthalmol 2002; 120: 1268-1279.
13. Thiel HJ, Denk PO, Knorr M. Are filtering interventions in glaucoma patients with extensive visual-field defects associated with a higher functional risk? Ophthalmologe 2000; 97: 336-341.
14. AGIS Investigators. The Advanced Glaucoma Intervention Study (8). Risk of cataract formation after trabeculectomy. Arch Ophthalmol 2001;119: 1771-1780.
15. Hattenhauer MG, Johnson DH, Ing HH, et al. The probability of blindness from open-angle glaucoma. Ophthalmology 1998; 105: 2099-2104.
16. Fraser S, Bunce C, Wormald R. Retrospective analysis of risk factors for late presentation of chronic glaucoma. Br J Ophthalmol 1999; 83: 24-28.
17. Cantor L, Berlin MS, Hodapp EA, et al. Clinical Evaluation: Basic and Clinical Science Course, California: American Academy of Ophthalmology, 1999; section 10, chap.6: 36-47.