Aphakic Glaucoma After Congenital Cataract Surgery

Objective  To describe the largest series of patients who developed aphakic glaucomaafter lensectomy for congenital cataracts.

Methods  A retrospective review was performed of all patients seen by a pediatricglaucoma specialist between October 1, 1970, and November 30, 2002. Patientswith intraocular pressures greater than 25 mm Hg after lensectomy were studied.Patients with either conditions independently associated with glaucoma orany signs of glaucoma before lensectomy were excluded.

Results  We studied 170 eyes of 117 patients. The mean ± SDfollow-up period was 8.6 ± 7.6 years. The most common cataracttypes were complete (40.8%) and nuclear (22.5%). For the lensectomies, 103eyes underwent modern vitrectomy techniques, and 10 underwent older techniques.Lensectomies were performed in 80.6% of eyes before age 1 year. Onset of glaucomaafter lensectomy was by 1 year in 37.1% of eyes, by 6 years in 75.9%, andby 33 years in 100%. Of eyes that had gonioscopy, 93.9% had open angles. Glaucomasurgery was needed in 57.1% of eyes. The median final visual acuity was 20/400.

Conclusions  Most cases of aphakic glaucoma are of the open-angle type. Various riskfactors are suggested, and the prognosis is guarded. Lifelong follow-up isneeded to screen for glaucoma.

Glaucoma continues to be one of the most common complications of congenitalcataract surgery and ranges from 0% to 32%.^1-14 Theincidence of aphakic glaucoma also seems to increase with longer follow-up.^1-5 Inpatients who receive follow-up for longer than 5 years, the incidence of glaucomahas been reported to be as high as 41%.³

Although it was commonly believed that newer automated techniques woulddecrease the incidence of glaucoma,^1,3,6,15-18 theincidence after automated lensectomy and vitrectomy is still relatively highand unchanged^2,6,19 at0% to 32%,^{1,3,4,8,9,11,16,17} whichis similar to the incidence of 5.5% to 27.0% in patients who underwent olderneedling and aspiration procedures.^6-8,10,12 However,with the introduction of vitrectomy instruments for childhood cataract surgeryin 1975,⁹ angle-closure glaucoma with pupillaryblock after lensectomy has become less common.^{1-3,6,7,16-20} Now,most cases of glaucoma after pediatric lensectomy are of the later-onset open-angletype, accounting for 75.0% to 93.8%^5,19,20 ofglaucoma cases after lensectomy.

We describe the largest series in the literature of patients with aphakicglaucoma to better define the risk factors, clinical course, and prognosisof this disease.

A retrospective medical record review was performed for all patientsseen by a pediatric glaucoma specialist (D.S.W.) between October 1, 1970,and November 30, 2002. Patients were included if they had repeated intraocularpressures (IOPs) greater than 25 mm Hg after congenital cataract surgery.Goldmann or Perkins applanation tonometry was used. Cataracts were definedas congenital if they were identified in the first 6 months of life, weredominantly inherited, or were of the lamellar type. Patients were includedif they underwent lensectomy before age 20 years.

Inclusion criteria consisted of having at least 1 month of follow-upafter aphakic glaucoma was diagnosed. Patients were excluded if they had ahistory of trauma, intraocular neoplasm, radiation therapy, anterior uveitis,anterior segment dysgenesis, Stickler syndrome, Lowe syndrome, maternal rubellasyndrome, or trisomy 13. Patients with a history of corticosteroid use beforelensectomy or a history of maternal corticosteroid use, those with signs ofcongenital glaucoma before lensectomy, and those with retinal detachment orvitreous hemorrhage before lensectomy were also excluded.

Information included a complete history and eye examination. Also recordedwere the intraoperative and postoperative courses. Results are expressed asmean ± SD.

One hundred seventy eyes of 117 patients with aphakic glaucoma met theinclusion and exclusion criteria. There were 55 males and 62 females. Mean ± SDfollow-up time was 8.6 ± 7.6 years (range, 1 month to 28years). Patient ages at the last follow-up visit ranged from 5 months to 63years.

Fifty-three patients had bilateral aphakic glaucoma, 30 had bilaterallensectomy but aphakic glaucoma in only 1 eye, and 34 had unilateral lensectomywith subsequent aphakic glaucoma. No patient with unilateral lensectomy developedglaucoma in the other eye. In patients in whom cataract type was recorded(71 eyes), the most common type was the complete cataract (Table 1).

Thirty-nine (22.9%) of 170 eyes were operated on by D.S.W., and 131eyes (77.1%) were operated on by other surgeons. Modern vitrectomy instrumentswere used in 103 eyes (91.2%), whereas older techniques (ie, needling, aspiration,or linear extraction) were used in 10 eyes; the operative techniques usedin 57 eyes were unknown. Fifty-five (53.4%) of 103 eyes underwent lensectomywith primary posterior capsulotomy and anterior vitrectomy. One patient underwentextracapsular cataract extraction with a posterior chamber intraocular lensat age 1 month. The mean ± SD age at the time of lensectomywas 0.7 ± 1.4 years (range, 1 week to 10 years). One hundredthirty-seven (80.6%) of 170 eyes underwent surgery before the patient wasaged 1 year (Figure 1).

No intraoperative complications were recorded in any of the eyes. Eighty-five(50.0%) of 170 eyes had complications after lensectomy (Table 2).

Forty-eight (28.2%) of 170 eyes had a family history of congenital cataractsin first-degree relatives. Thirty-eight (79.2%) of these 48 eyes were from19 patients with bilateral congenital cataracts.

Associated ocular conditions were present in 58 (34.1%) of 170 eyes(Table 3). Associated systemic anomalieswere seen in 39 (33.3%) of 117 patients (Table4). The mean ± SD IOP before lensectomy was 15.2 ± 3.5mm Hg (range, 5-22 mm Hg) and after the development of aphakic glaucoma was34.0 ± 7.1 mm Hg (range, 25-50 mm Hg). The mean ± SDcorneal diameter before lensectomy was 9.9 ± 1.2 mm (range,7-13 mm). The mean ± SD axial length by ultrasonography (Ascan) before lensectomy was 15.6 ± 2.3 mm (range, 12-21 mm).The mean ± SD birth weight was 3.2 ± 0.8kg (range, 0.6-4.5 kg).

Aphakic glaucoma developed after lensectomy in 63 (37.1%) of 170 eyesby age 1 year and in 129 (75.9%) by age 6 years (Table 5). The mean ± SD interval that glaucomawas diagnosed after lensectomy was 4.0 ± 4.6 years.

Open-angle glaucoma was the most common type of aphakic glaucoma. Gonioscopyshowed open angles in 139 (93.9%) of the 148 eyes examined. In 74 (50.0%)of the 148 eyes, there was some degree of peripheral anterior synechiae. Someeyes also showed increased pigmentation and a glazed appearance of the trabecularmeshwork. Nine (6.1%) of 148 eyes had closed angles. Four eyes had cornealhaze, which precluded gonioscopy. The gonioscopy findings of 18 eyes werenot known.

After the development of glaucoma, the posterior pole could be seenin 169 eyes. Ninety-four (55.6%) of the 169 eyes showed good vascularizationof the optic disc, whereas 58 (34.3%) showed varying degrees of pallor. Afterthe development of aphakic glaucoma, the mean ± SD cup-discratio was 0.29 ± 0.13 (range, 0.2-0.99) (Table 6).

Seventy (41.2%) of 170 eyes were operated on for postoperative complications,with the most common operation being pupillary membrane removal in 39 (56%)of 70 eyes (Table 7). Surgical proceduresunrelated to postoperative complications were performed in 28 (16.5%) of the170 eyes. Strabismus surgery was performed in 22 (12.9%) of 170 eyes. In 8of these 22 eyes, strabismus was present before lensectomy. In the remaining14 eyes, it is unknown whether the onset of strabismus was before or aftercataract surgery. Secondary intraocular lens surgery was performed in 4 eyes,nasolacrimal duct probing in 1 eye, and refractive corneal surgery in 1 eye.

Topical glaucoma medications were applied to 156 (91.8%) of 170 eyesafter the onset of glaucoma (Table 8).The initial treatments in the other 14 eyes were either unknown or consistedof systemic medications or surgery. The most common single agent used wasa β-blocker in 84 (49.4%) of 170 eyes (Table9). Sixty (35.3%) of 170 eyes needed systemic carbonic anhydraseinhibitor therapy.

Glaucoma surgery was performed in 97 (57.1%) of 170 eyes (Table 10). Sixty-seven (69.1%) of 97 eyes underwent a single procedure,and 30 eyes (30.9%) needed multiple surgical procedures to control IOP. Goniotomiesor trabeculotomies were performed in 24 (24.7%) of 97 eyes, with a successrate of 16.0% (ie, IOP ≤21 mm Hg, with or without medications, and no needfor further surgery). Of the 61 (62.9%) of 97 eyes that underwent trabeculectomy,only 1 was without antimetabolites. Forty-three eyes used mitomycin, and 17eyes used fluorouracil. Fifteen (24.6%) of these 61 eyes had successful outcomes(ie, IOP ≤21 mm Hg, with or without medications, and no need for furthersurgery).

Thirty-four (35.1%) of 97 eyes had tube surgery (ie, 32 Ahmed valves[New World Medical, Inc, Rancho Cucamonga, Calif] and 2 Molteno valves [MoltenoOphthalmic Limited, Dunedin, New Zealand]), with a success rate of 44.1% (ie,IOP ≤21 mm Hg, with or without medications, and no need for further surgery).

Cyclodestructive procedures were performed in 21 (21.6%) of 97 eyesand included cyclocryotherapy, diode laser transscleral cyclophotocoagulation,and contact Nd:YAG laser cyclotherapy. The success rate for cyclodestructiveprocedures was 14.3% (ie, IOP ≤21 mm Hg, with or without medications, andno need for further surgery).

At the last follow-up visit, median and mean visual acuities were 20/400and 20/515, respectively. Visual acuities ranged from 20/25 to no light perception. Figure 2 shows that 10.6% of eyes had visualacuity of 20/25 to 20/40, 23.0% of 20/50 to 20/200, and 66.2% of 20/400 orworse at the last follow-up visit.

With modern lensectomy techniques, most cases of aphakic glaucoma areof the open-angle type. Various risk factors are associated with aphakic glaucoma,which has a guarded prognosis. Because glaucoma may not develop for yearsafter lensectomy, lifelong follow-up is needed.

In this study, most cases of aphakic glaucoma after lensectomy for congenitalcataracts were of the open-angle type (139 [93.9%] of 148 eyes). This is consistentwith the literature,^5,19,20 whichdescribes 75.0% to 93.8% of patients with open angles. Although most patientswith aphakic glaucoma have open angles, acquired angle changes can be seenin 96% of eyes and are characterized by circumferential forward positioningof the iris to the posterior or middle trabecular meshwork.¹⁹ Theresidual exposed trabecular meshwork is sometimes noted to have scatteredpigment deposits,^5,15,19,21 whitecrystalline deposits suggesting residual lens tissue caught in the trabecularmeshwork,¹⁹ or peripheral anterior synechiae.^2,3,19,22

Most cases of aphakic glaucoma develop years after lensectomy. The mean ± SDtime after lensectomy that glaucoma developed in our patients was 4.0 ± 4.6years (Table 5), compared with otherreports of 5.3 to 12.2 years.^{2,3,5,7,20,23-25} Therefore,buphthalmos, which usually occurs in the first 2 to 3 years of life, is generallynot present. The latest time after lensectomy that glaucoma developed in ourstudy was 33 years, compared with the literature’s citing of 45¹⁵ to 65²⁶ years. Becauseit is also difficult to predict which eyes will develop aphakic glaucoma,²⁵ lifetime surveillance for glaucoma after congenitalcataract surgery is necessary.^15,19

Even after 5 years of follow-up,²⁷ childrenwith congenital cataracts rarely, if ever, develop glaucoma if they do notundergo lensectomy.¹⁶ This finding is alsosupported by other researchers who have noted that no patient with bilateralaphakia developed unilateral glaucoma^3,25 andthat no patient with unilateral aphakia had bilateral glaucoma.^3,25 Ourstudy, however, had 30 patients with bilateral lensectomy but unilateral glaucoma.

The etiology of aphakic glaucoma is still unclear.^28,29 Undergoinglensectomy at a young age,^1,2,4,7,8,28 especiallyin the first year of life,^2,19,23 maybe a risk factor for the development of aphakic glaucoma. Only 1 study³ suggests that it may not be a risk factor. Most ofour patients with aphakic glaucoma (80.6%) underwent lensectomy in the firstyear of life (Figure 1). This is consistentwith the 77.0% to 92.9% reported in the literature.^2,4,19 Ithas been suggested that the immaturity of the developing infant’s angleleads to increased susceptibility to secondary surgical trauma.² Additionalcontributing factors seen in younger patients undergoing surgery include increasedpostoperative inflammation, associated congenital ocular anomalies, and increasedtechnical difficulties with their associated increased postoperative complications.^2,6,15,16,30 Despitethe increased risk of aphakic glaucoma with early surgery, it seems inadvisableto delay surgery solely for fear of aphakic glaucoma.² Theincreased risks of early lensectomy must be balanced against the need to decreasethe period of visual deprivation.¹

Corneal diameters less than 10 mm have also been associated with 88.5%to 94.0%^23,24 of patients withaphakic glaucoma and may be another risk factor.^2,7,23 Onestudy³ with only 8 eyes found no association.Our patients had a mean ± SD corneal diameter of 9.9 ± 1.2mm. Parks et al²³ noted that aphakic glaucomadeveloped in only 2.9% of patients with normal corneal diameters comparedwith 31.9% of eyes with small corneal diameters. A smaller cornea may reflectan abnormal anterior segment and subtle filtration angle defects, which increasethe risk of aphakic glaucoma.^19,24

In our study, one third of the eyes had associated ocular anomalies(Table 3). Other researchers^6,31 have reported that 50% of patientswith aphakic glaucoma have additional eye abnormalities. Microphthalmos^8,29 and persistent hyperplastic primaryvitreous (PHPV)²³ have been suggested as riskfactors for the development of glaucoma. In the 170 eyes in our study, microphthalmos(14.1%) and PHPV (5.3%) were 2 of the more common ocular abnormalities. Otherresearchers^5,6,8,14,17 havealso reported microphthalmos in 8.3% to 26.6% of eyes. Johnson and Keech⁴ did not believe that PHPV was a risk factor and notedthat patients with and without PHPV have a 32% incidence of aphakic glaucomaafter pediatric lensectomy. Wallace and Plager²⁴ pointedout that the underlying risk factor for these conditions may primarily bethe smaller cornea.

Our study is consistent with others in that an increased risk of aphakicglaucoma may be related to certain cataract types, such as complete,² nuclear,^2,15,23 andPHPV²³ (Table 1). Although 1 study³ suggested thatcataract type was not associated with subsequent glaucoma, that series of8 eyes may have been too small to accurately assess this. These congenitalcataract types often necessitate early surgery for better visual rehabilitation.Also, nuclear and PHPV cataracts are more commonly associated with smallercorneal diameters.^23,24

It has been suggested that in some patients cataracts and glaucoma maybe signs of a yet undescribed congenital ocular syndrome. This possibilityis supported by the bilateral lensectomy cases with atypical angle abnormalitiesin some but is less likely because of the late onset of the glaucoma and thenormal angle appearance in others.^15,25 Althoughsuch an ocular syndrome may not necessarily be inherited,^1,15 48(28.2%) of 170 eyes in our study involved a family history of congenital cataractsin first-degree relatives. Thirty-eight (79.2%) of these 48 eyes belongedto 19 patients with bilateral cataracts. All of these patients with bilateralcataracts developed bilateral glaucoma.

It has also been suggested that the aphakic state per se may increasethe risk of glaucoma. In our study, 169 of the 170 eyes were left aphakicat the time of initial cataract surgery. One theory as to why aphakia causesglaucoma was suggested by Paul Kaufman, MD, in response to Walton’s¹⁹ American Ophthalmological Society presentation. Takingout the lens during the first year of life without putting anything in itsplace may prevent normal meshwork development, which may require certain normalstructural interactions between the native lens, zonules, ciliary body, andtrabecular meshwork.¹⁹ At that same AmericanOphthalmological Society meeting, Albert W. Biglan, MD, observed that patientswith corneal diameters less than 10 mm are usually left aphakic.¹⁹ Theaphakic state may allow blockage of the angle by vitreous⁵ ormay allow vitreous factors to alter trabecular meshwork structure and maturation.^25,29

Conversely, it has been suggested that primary pseudophakia decreasesthe incidence of glaucoma.²⁵ Only 1 of the170 eyes in our study had primary lens implantation at the time of initialcataract surgery. Asrani et al²⁵ noted thatof 377 eyes with primary lens implantation for congenital cataracts, only1 eye developed glaucoma, with 3.9 years of follow-up, which is less thanthe mean time of aphakic glaucoma development of 5.3 to 12.2 years.^3,20,23-25 Unlikeour study, Asrani et al²⁵ excluded patientswith risk factors for aphakic glaucoma (ie, associated ocular anomalies andcorneal diameters <10 mm). Also, only 19.4% of patients underwent lensectomyat age 1 year or younger. Another study³² notedthat only 3 (6.7%) of 45 eyes developed glaucoma after pediatric lensectomyand primary lens implantation; however, this study also had a shorter follow-upperiod of 23 months and excluded certain high-risk patients (ie, patientswith corneal diameters <10 mm and patients <1 year). Perhaps if certainhigh-risk patients had not been excluded from these studies and if there werelonger follow-up times, the incidence of glaucoma in these studies would havebeen higher.

Retained degenerating lens proteins may be toxic to the trabecular meshwork^15,18,19 and may increasethe risk of aphakic glaucoma. Cytokines released by residual lens epithelialcells may also be responsible.¹⁶ In our study,15.0% of eyes had retained lens material. Other studies^6,19 havereported 41.6% to 78.0% of patients with residual cortex or lens material.Perhaps modern phacoemulsification techniques with complete cortical removalmay benefit patients with congenital cataracts.^19,33 Thecontinuous tear capsulorrhexis may also eliminate residual capsular tags towhich the iris can adhere and that may cause chronic inflammation.¹⁹

Poor pupillary dilation can make the surgery technically more difficult¹⁹ and has been associated with an increased risk ofglaucoma.^2,29 This may increasethe rate of complications and the chance for residual cortex.

A corticosteroid-induced mechanism seems unlikely because postoperativecorticosteroids are usually used for only 1 to 2 months.²⁵ Thisis inconsistent with the usual later onset of aphakic glaucoma. However, perhapsthe use of high-dose corticosteroids³⁴ in infancymay alter the still-developing microstructure of the trabecular meshwork.²⁹ None of our patients had a history of prolonged postoperativecorticosteroid use.

The increased postoperative inflammation seen in pediatric patientsdoes not seem to be a substantial risk factor. Only 2 of our patients werenoted to have severe postoperative inflammation. Although the occurrence ofchronic inflammation was not specifically assessed in the study by Walton,¹⁹ he noted that the anterior segments of these patientswere not characterized by cells, flare, and band keratopathy and, in thisway, were different from those of children with chronic anterior uveitis whodevelop glaucoma.¹⁹

A history of secondary surgery is seen more often in children who developaphakic glaucoma.^6,16 Seventy(41.2%) of 170 eyes in our study had surgery for postoperative complications,most commonly for pupillary membrane removal (Table 7). Other researchers^6,19 havesimilarly reported that 58.6% to 75.0% of eyes with glaucoma undergo 1 ormore secondary membrane procedures. Retained lens protein and repeated surgeryfor secondary membrane may cause increased postoperative inflammation andtrabecular meshwork damage.^6,35

Once glaucoma develops, medical treatment in children may be difficult(Table 8 and Table 9). Miotics may reduce visual acuity when dense capsular opacificationsare present. In one study,¹⁵ 2 (11.1%) of 18patients experienced retinal detachments related to miotic treatment. Epinephrine-inducedmaculopathy is also a concern in these patients with aphakia and may not berecognized if vision is reduced from other causes.^3,15 Asraniet al²⁵ noted that 63.6% of patients were treatedwith medications alone.

Glaucoma surgery was performed in 97 (57.1%) of 170 eyes (Table 10) compared with 36.4% in the literature.⁵ Successrates for all types of glaucoma surgery (primary and secondary) are guardedand range from 14.3% to 44.1% (ie, IOP ≤21 mm Hg, with or without glaucomamedications, and no need for further surgery). Sixty-seven (69.1%) of the97 eyes underwent a single procedure. Asrani and Wilensky⁵ notedthat 78.6% of patients achieved successful IOP control with 1 operation withoutusing glaucoma medications.

Systemic abnormalities were seen in 33.3% of our patients. As in ourstudy (Table 4), Peyman et al³³ noted that developmental delay (4 of 25 patients)and cardiac abnormalities (5 of 25 patients) were most common.

The visual prognosis is also guarded. In our study, visual acuity atthe last follow-up visit was 20/400 or worse in 66.2% of eyes (Figure 2). The literature^5,15,18 notesthat 11.4% to 100% of patients achieve a final visual acuity of less than20/200 or 20/400 in the affected eye. The mean visual acuity in our patientswas 20/515, which is similar to the 20/510 in the study by Simon et al.³ Once aphakic glaucoma was diagnosed, Asrani and Wilensky⁵ noted stable vision in 68.8% of patients. They alsofound that poor vision in most patients was partly due to delay in diagnosisand treatment. Poor visual acuity in patients with aphakic glaucoma is attributableto secondary membranes, nystagmus, strabismic or deprivational amblyopia,and glaucomatous optic nerve damage.^1,6

Ophthalmoscopy of eyes in our study after the development of aphakicglaucoma revealed a mean ± SD cup-disc ratio of 0.29 ± 0.13(Table 6). Simon et al³ reporteda mean cup-disc ratio of 0.54 (range, 0.15-0.80) in glaucomatous eyes.

Despite the introduction of automated surgical techniques in the 1970s,the incidence of aphakic glaucoma remains high. Open-angle glaucoma is nowthe most common type of glaucoma after congenital cataract surgery, and itoccurs at a mean ± SD duration of 4.0 ± 4.6years after lensectomy. Various risk factors have been suggested. Treatmentmay be difficult, with 57.1% of patients needing glaucoma surgery. Prognosisis guarded, with about two thirds of our patients having a final visual acuityof 20/400 or worse. Close lifelong follow-up is needed in these patients todetect this condition in its earliest stages and prevent loss of vision.

Correspondence: Teresa C. Chen, MD, GlaucomaService, Department of Ophthalmology, Harvard Medical School and MassachusettsEye and Ear Infirmary, 243 Charles St, Boston, MA 02114 (teresa_chen@meei.harvard.edu).

Submitted for Publication: October 10, 2003;final revision received April 12, 2004; accepted August 12, 2004.

Financial Disclosure: None.

Funding/Support: This study was supported inpart by the Miles Pediatric Research Fund, Boston, Mass.

Previous Presentation: This study was presentedin part at the Annual Meeting of the Association for Research in Vision andOphthalmology; May 4, 2003; Fort Lauderdale, Fla.