Primary pediatric keratoplasty: Indications,graft survival, and visual outcomeAli Al-Ghamdi, MD, Ali Al-Rajhi, MD, and Michael D. Wagoner, MD

BACKGROUND Penetrating keratoplasty in children has been documented to have a higher rate of graftfailure and a worse visual prognosis than adult keratoplasty.

METHODS We undertook a retrospective review of all cases of primary penetrating keratoplastyperformed in children 12 years of age or younger at the King Khaled Eye SpecialistHospital between January 1, 1990, and December 31, 2003.

RESULTS One hundred sixty-five primary penetrating keratoplasties were performed in 134 childrenduring the study interval. The surgical indications were congenital opacities in 130 eyes(78.8%), acquired, traumatic opacities in 18 eyes (10.9%), and acquired, nontraumaticopacities in 17 eyes (10.3%). Among congenital opacities, 35 cases were caused bycongenital hereditary endothelial dystrophy (CHED). The median follow-up for 73 grafts(44.2%) that remained clear was 50 months (range, 12-50 months), whereas the medianfollow-up for 92 grafts (55.8%) that failed was 6 months (range, 1-54 months). Kaplan-Meier graft survival was significantly higher at all postoperative intervals in eyes withCHED than for other surgical indications ( p � 0.001). Eyes with CHED were signifi-cantly more likely to achieve ambulatory vision or vision �20/200 than eyes with otherindications ( p � 0.001).

CONCLUSIONS Pediatric keratoplasty was associated with an excellent prognosis for graft survival in eyeswith CHED and a fair prognosis for graft survival in eyes with non-CHED congenitalopacities and acquired opacities. The best visual prognosis was obtained in eyes withCHED and the worst prognosis was for non-CHED congenital opacities. ( J AAPOS2007;11:41-47)

P enetrating keratoplasty in children is a high-riskprocedure. The technical complexity of the proce-dure the difficulties in patient follow-up and man-

agement, the frequent presence of other ocular pathology,the inability of young children to verbalize subjectivesymptoms, and the occurrence of serious postoperativecomplications result in a relatively high rate of graft fail-ure.1-5 When graft clarity is maintained, the visual out-come is often disappointing because of the amblyogenicpotential of corneal opacities in pediatric patients,6,7 aswell as other associated ocular abnormalities.1-5 Despitethese obstacles, several large studies have reported success-ful outcomes with respect to graft survival and visual im-provement in a reasonable percentage of cases.1-5 Datafrom these reports have helped provide realistic expecta-tions for clinicians and parents considering surgery for achild with a corneal opacity.

Author affiliations: Department of Ophthalmology, King Khaled Eye Specialist Hospital,Riyadh, Kingdom of Saudi Arabia

Submitted February 2, 2006.Revision accepted September 4, 2006.Reprint requests: Michael D. Wagoner, MD, Medical Director, King Khaled Eye

Specialist Hospital, PO Box 7191, Riyadh 11462, Saudi Arabia (email: mwagoner@kkesh.med.sa).

Copyright © 2007 by the American Association for Pediatric Ophthalmology andStrabismus.

1091-8531/2007/$35.00 � 0doi:10.1016/j.jaapos.2006.09.012

Journal of AAPOS

We report the results of one of the largest series of pene-trating keratoplasty performed in children 12 years of age oryounger at a single center and provide additional evidence ofthe potential benefits of surgery in the restoration of cornealclarity and visual function in young children. In addition toits large size, the single center location also provides thisstudy with the advantage of a composition of cases in whichfairly consistent surgical techniques and management strate-gies were used.

Subjects and MethodsAfter obtaining approval from the Institutional Review Board,the medical records of every child 12 years of age or younger whounderwent primary penetrating keratoplasty at the King KhaledEye Specialist Hospital (KKESH) between January 1, 1990, andDecember 31, 2003, were retrospectively reviewed. Patients whohad primary penetrating keratoplasty at another institution andsubsequent repeat keratoplasty at KKESH were excluded fromthe study. Repeat keratoplasty for failed primary cases performedat KKESH were not included in the statistical analysis.

Penetrating keratoplasty was performed by 41 different mem-bers of the KKESH Anterior Segment Division during the 13-year study period. In most cases, the donor trephination size was0.5 mm greater than that of the recipient, with a donor-recipientdisparity of 0.25 mm in the remainder. Interrupted suture tech-nique was used in all cases, with 16 interrupted 10-0 nylon

sutures in most cases and 24 interrupted 10-0 nylon sutures in the

41

Volume 11 Number 1 February 200742 Al-Ghamdi, Al-Rajhi, and Wagoner

remainder. Suture removal was complete in most cases by 6months. All procedures were performed on inpatients, and everypatient remained in the hospital for at least 3 postoperative days.After discharge, the postoperative examination schedule for mostpatients was every 2 weeks for the first 3 postoperative months,monthly for the next 3 months, and every other month for thenext 6 months. All patients were treated with topical corticoste-roids for a minimum of 6 months. Most patients were treatedwith prophylactic topical antibiotics for a minimum of 2 monthspostoperatively, but subsequent use of antibiotic prophylaxis var-ied widely between the participating surgeons. The preferredprophylactic antibiotic was combination neomycin-polymyxinbefore 1999, after which ofloxacin was used almost exclusively.Preoperative, intraoperative, and postoperative glaucoma surgi-cal procedures were performed by members of the GlaucomaDivision. Postoperative refractive correction and amblyopiamanagement was performed by members of the Pediatric Oph-thalmology and Strabismus Division.

We used the scheme devised by Stulting and associates1 toclassify the indications for pediatric keratoplasty into “congenitalopacities,” “acquired traumatic opacities,” and “acquired non-traumatic opacities.” In recognition of previously reported dif-ferences in prognosis between congenital hereditary endothelialdystrophy (CHED) and other congenital opacities,8-12 we ana-lyzed graft survival and visual outcome data separately for thesecategories. Data extracted from the charts included age, sex,penetrating keratoplasty procedures (unilateral vs bilateral), sur-gical indication (congenital opacities, acquired nontraumaticopacities, or acquired traumatic opacities), associated surgicalprocedures ( previous, concomitant, or subsequent to kerato-plasty), major postoperative complications, graft clarity, and vi-sual acuity.

Glaucoma escalation was defined as the postoperative onset ofelevated intraocular pressure requiring one or more medicationsin an eye that did not previously require treatment or the needfor additional medications or surgery in an eye previously beingtreated. A persistent epithelial defect was defined as a postoper-ative epithelial defect or corneal erosion occurring anytime in thepostoperative course that lasted for more than 14 days.

Graft clarity was based on findings at the most recent examina-tion. For the grafts that remained clear, the duration of follow-upwas defined as the interval between surgery and the most recentexamination. For grafts that failed, the duration of follow-up wasdefined as the interval between surgery and the date that irre-versible loss of central graft clarity was first documented. Micro-bial keratitis was defined as a new-onset stromal infiltrate thatwas either culture-positive or had a clinical course that wasstrongly suggestive of a microbial etiology. Graft failure wasstrictly defined as irreversible loss of central graft clarity from anycause, irrespective of the level of visual acuity.

Visual acuity was defined as the best visual acuity measured atthe most recent examination of the primary graft. In cases inwhich verbal or cognitive skills of the patient were insufficient toobtain precise determination of visual acuity, the ability to fixateon and follow visual targets was recorded. Ambulatory vision was

defined as vision of counting fingers at 3 feet or better.

All data were entered onto a spreadsheet and analyzed usingtwo commercial statistical software packages, SPSS 1998 (Chi-cago, IL) and EPINFO 1994 6.0 (Santa Clara, CA). Chi-square,Fisher’s exact, and t-tests were used for analysis of variance.Nominal p-values were used for all comparisons, and the term“significance” was accepted if the p-value was �0.05. Graft sur-vival curves were produced using the standard Kaplan-Meiermethod and the life table.13 Data were entered and analyzedseparately for each eye for patients undergoing bilateral kerato-plasty.

ResultsOne hundred sixty-five primary penetrating keratoplastyperformed in 134 children (103 unilateral, 31 bilateral)met the inclusion criteria. There were 35 cases performedwhen the child was younger than 6 months of age, 57 caseswhen between 6 months and 2 years, and 73 cases whenolder than 2 years.

IndicationsCongenital opacities accounted for 130 cases (78.8%),including 35 cases caused by CHED and 95 cases thatwere unrelated to CHED. Non-CHED congenital opac-ities included congenital glaucoma (49 cases), Peter’sanomaly (28 cases), sclerocornea (6 cases), and others (12cases). Preoperatively, adequate intraocular pressure con-trolled had been achieved in 31 eyes (63.2%) with con-genital glaucoma. Concomitant glaucoma surgical proce-dures were performed at the same time as penetratingkeratoplasty in the remaining 18 eyes (36.8%). Acquiredopacities accounted for only 35 cases (21.2%). Amongthese, there was almost equal distribution of nontraumatic(18 cases) and traumatic (17 cases) etiologies. Microbialkeratitis was responsible for all nontraumatic cases, whilepenetrating injuries and blunt trauma were almost equallyresponsible for traumatic cases.

For non-CHED congenital opacities, unilateral surgerywas performed more frequently than bilateral surgery (59children vs 18 children). For CHED, bilateral surgery wasperformed more frequently (13 children vs 9 children). Allchildren with acquired opacities had unilateral surgery.

Graft SurvivalAt the most recent examination, 73 primary penetratingkeratoplasty (44.2%) remained clear. The median follow-upfor all grafts that remained clear was 50 months (range,12-120 months). Ninety-two grafts (55.8%) failed. Themedian follow-up to graft failure was 6 months (range,1-54 months). Sixty grafts (65.2%) failed in the first 12months and 81 (88.0%) failed in the first 24 months (Ta-ble 1). There was no correlation between corneal donorage, death-to-preservation time, media storage time, ordonor endothelial cell counts and graft survival.

Overall graft survival was also significantly greater in eyeswith CHED (85.7%) than in eyes with acquired, traumatic

opacities (41.2%; p � 0.001), non-CHED congenital opaci-

Journal of AAPOS

Volume 11 Number 1 February 2007 Al-Ghamdi, Al-Rajhi, and Wagoner 43

ties (32.6%; p � 0.001), and acquired, nontraumatic opac-ities (27.8%; p � 0.001). Among non-CHED congenitalopacities, overall graft survival was 40.8% for congenitalglaucoma, 28.6% for Peter’s anomaly, 0% for sclerocor-nea, and 25.0% for other conditions. Graft survival wassimilar in bilateral and unilateral cases in eyes with con-genital opacities.

The Kaplan-Meier probability of graft survival at 1, 2, 3, 4,and 5 years for each surgical indication is summarized in

Table 1. Annual graft failure rates after primary pediatric keratoplasty

IndicationPrimarygrafts

Failedgrafts

Yearn (%

Congenital opacity† 95 64 42CHED 35 5 1Acquired, nontraumatic opacity 18 13 9Acquired, traumatic opacity 17 10 8All cases 165 92 60

*Percentage of total failed grafts for the surgical indication.†Excludes cases of CHED (congenital hereditary endothelial dystrophy).

Table 2. Graft survival after pediatric keratoplasty vs surgical indication

Indication Congenital opacity*

Number 95Graft survival %

Overall 32.6Unilateral 33.9Bilateral 30.6

Kaplan-Meier % (95% CI)1 year 55.8 (45.2-65.1) 972 year 39.9 (29.8-49.7) 903 year 37.4 (27.5-47.2) 834 year 33.2 (23.6-43.1) 835 year 28.9 (19.6-38.8) 83

Follow-up (months)Clear grafts§

Median 72Range 12-120

Failed grafts¶Median 6Range 1-54

*Excludes cases of CHED (congenital hereditary endothelial dystrophy).†N/A � not applicable; there were no bilateral cases performed in this catego‡N/A � not applicable; there were no eyes with 4-year or 5-year follow-up.§Months between surgery and most recent examination.¶Months between surgery and irreversible graft failure.

Table 3. Overall graft survival after primary pediatric keratoplasty vs age

CongenitalOpacity,* n

Graftsurvival (%) CHED, n

Graftsurvival (%)

Age0.0 to 0.5 32 21.9 1 0� 0.5 to 2.0 44 36.4 9 88.9� 2.0 19 42.1 25 88.0

*Excludes cases of CHED (congenital hereditary endothelial dystrophy).

Table 2. Graft survival was significantly higher at all post-

Journal of AAPOS

operative intervals in eyes with CHED than in eyes withother surgical indications ( p � 0.001).

The relationship between graft survival and recipientage at the time of surgery is summarized in Table 3. Fornon-CHED congenital opacities, there was a higher per-centage of graft survival if the surgery was performed afterthe age of 6 months (38.1% vs 21.9%), but this differencewas not statistically significant ( p � 0.16). Although therewas a higher rate of graft survival with increasing patient

Year 2,n (%)*

Year 3,n (%)*

Year 4,n (%)*

Year 5,n (%)*

14 21.9 2 3.1 3 4.7 3 4.72 40.0 2 40.0 0 0 0 04 30.8 0 0 0 0 0 01 10.0 0 0 1 10.0 0 0

21 22.8 4 4.3 4 4.3 3 3.3

DAcquired,

nontraumatic opacityAcquired,

traumatic opacity

18 17

.7 27.8 41.2

.9 27.8 41.2

.6 N/A† N/A†

4-99.5) 50.0 (25.9-70.1) 52.9 (27.6-73.0)3-97.0) 27.8 (10.1-48.9) 39.7 (13.3-65.5)8-92.9) 27.8 (10.1-48.9) 39.7 (13.3-65.5)8-92.9) 27.8 (10.1-48.9) N/A‡8-92.9) 27.8 (10.1-48.9) N/A‡

38 1402 36-60 12-46

6 60 1-18 6-46

Acquired,aumatic opacity, n

Graftsurvival (%)

Acquired,traumatic opacity, n

Graftsurvival (%)

2 0 – –4 0 – –

12 41.7 17 41.2

1,)*

65.620.069.280.065.2

CHE

35

858884

.1 (81.

.9 (74.

.6 (64.

.6 (64.

.6 (64.

5112-1

186-3

ry.

nontr

age in all surgical groups, this was not statistically signif-

Volume 11 Number 1 February 200744 Al-Ghamdi, Al-Rajhi, and Wagoner

icant as a continuous variable for non-CHED congenitalopacities ( p � 0.69; odds ratio [OR] � 1.04; 95% confi-dence interval [95% CI] � 0.87–1.23), CHED ( p � 0.61;OR � 1.10; 95% CI � 0.77–1.56), acquired nontraumaticopacities ( p � 0.47; OR � 1.10; 95% CI � 0.85–1.43), oracquired traumatic opacities ( p � 0.08; OR � 1.27; 95%CI � 0.97–1.67).

The relationship between additional surgical proceduresand postoperative complications is summarized in Table 4.Eyes in which additional surgical procedures were per-formed had a statistically significant greater rate of graftfailure than those in which other procedures were notperformed (38.5% vs 51.3%; p � 0.02). Glaucoma surgicalprocedures were overwhelmingly the most common othersurgical procedure performed, with 68 eyes undergoingone or more glaucoma procedures before penetrating ker-atoplasty, 25 at the same time, and 37 afterward. Graftsurvival was similar in eyes that had glaucoma surgerybefore and after penetrating keratoplasty. Glaucoma sur-gery performed at the same time as penetrating kerato-plasty was associated with reduced graft survival (28.0%),but this was not quite statistically significant ( p � 0.08).Eyes in which surgical control of glaucoma was achievedpreoperatively had a greater rate of graft survival thanthose that required glaucoma intervention at the time ofpenetrating keratoplasty or afterward, although these dif-ferences were not statistically significant ( p � 0.33, 0.68,

Table 4. Overall graft survival after primary pediatric keratoplasty vssurgical interventions and complications

All casesn

Graft survival(%) p value

All cases 165 44.2Surgical Interventions*

PreviousGlaucoma 68 41.2 0.52Cataract 5 40.0 1.0

SimultaneousGlaucoma 25 28.0 0.08Cataract 26 19.2 0.005

SubsequentGlaucoma 37 35.1 0.26Cataract 14 50.0 0.78Vitreoretinal 10 10.0 0.04Any of above 91 38.5 0.02

Complications†Glaucoma‡ 55 32.7 0.046Microbial keratitis 44 15.9 �0.001PED 21 23.8 0.06Vitreoretinal 16 18.8 0.04Rejection 14 28.6 0.27Endophthalmitis 7 0 0.02Trauma 6 16.7 0.23Any of above 86 27.9 �0.001

PED: persistent epithelial defect.*Some eyes had more than one surgical intervention.

†Some eyes experienced more than 1 complication.‡Worsening of pre-existing glaucoma or new onset of postoperative glaucoma.

respectively).

Cataract procedures were the second most commonadditional surgical procedure, with 5 cases being per-formed before penetrating keratoplasty, 26 at the sametime, and 10 subsequent afterward. Graft survival wassignificantly reduced in cases where cataract surgery wasperformed at the same time as penetrating keratoplasty( p � 0.005), but not when it was performed prior to orafterward. Ten cases of vitreoretinal surgery that wereperformed after penetrating keratoplasty were associatedwith significantly reduced graft survival ( p � 0.02).

Eyes in which major complications occurred had a sta-tistically significant higher rate of graft failure than thosein which complications did not occur (27.9% vs 62.0%;p � 0.001). Complications that were significantly associ-ated with an increased risk of graft failure were glaucoma( p � 0.046), microbial keratitis ( p � 0.001), vitreoretinalcomplications ( p � 0.04), and endophthalmitis ( p �0.02).The most common postoperative complication wasglaucoma escalation, followed by microbial keratitis.Among 44 cases of microbial keratitis, none occurred inthe first postoperative month, 5 (11.4%) occurred in thesecond postoperative month, and 39 (88.6%) occurredlater. Prophylactic antibiotics were being used in all 5 casesthat occurred before 60 days. Streptococcus pneumonia wasresponsible for 4 cases and Staphylococcus aureus for 1 case.Prophylactic antibiotics were not being used in any casesthat developed after 60 days. Eleven (33.3%) cases ofmicrobial keratitis after 60 days were suture-related.

Visual OutcomeThe final postoperative visual acuities obtained for each sur-gical indication are summarized in Table 5. Best-correctedvisual acuity was available in all but 6 children with am-bulatory vision. A final visual acuity of “fixes and follows”was recorded in 3 children with CHED, 2 children withnon-CHED congenital opacities, and 1 child with an ac-quired traumatic opacity. Overall, 60 eyes (36.4%)achieved ambulatory vision, compared with only 31(18.8%) preoperatively. Vision better than 20/200 wasobtained in 32 eyes (19.4%), compared with only 3 (1.8%)preoperatively. Only 8 eyes (4.8%) achieved final vision of20/40 or better.

Eyes with CHED were significantly more likely toachieve ambulatory vision than those with non-CHEDcongenital opacities CHED (94.2% vs 12.6%; p � 0.001),those with acquired, traumatic opacities (35.3%; p �0.001), and those with acquired, nontraumatic opacities(50.0%; p � 0.001). Eyes with CHED also were signifi-cantly more likely to achieve vision better than 20/200than those with non-CHED congenital opacities (48.6%vs 9.5%; p � 0.001) and acquired, nontraumatic opacities(11.1%; p � 0.007), but not those with acquired, traumaticopacities (23.5%; p � 0.07). If a presumption is made thatthe 6 eyes with visual acuity of fixes and follows werebetter than 20/200, the percentage of eyes achieving thislevel of vision increases slightly to 57.1% in eyes with

CHED, 11.6% in eyes with non-CHED congenital opac-

Journal of AAPOS

Volume 11 Number 1 February 2007 Al-Ghamdi, Al-Rajhi, and Wagoner 45

ities, and 16.7% for acquired, nontraumatic opacities, withno change in the statistically significant better outcome ineyes with CHED. No surgical group was significantlymore likely to obtain vision of 20/40 or better.

The likelihood of obtaining ambulatory vision was almostthe same in bilaterally performed cases as in unilaterallyperformed cases. In eyes with non-CHED congenital opac-ities, ambulatory vision was obtained in 7 unilateral cases(11.9%) and 5 bilateral cases (13.9%). In eyes with CHED,ambulatory vision was obtained in 9 unilateral cases (100%)and 24 bilateral cases (92.3%).

DiscussionIndications for pediatric keratoplasty vary considerablyamong several major published series because of the dif-ferent patient populations.1-5 Congenital opacities (alltypes) were the most common indication for keratoplastyin our series, accounting for nearly 80% of cases. Thiscontrasts sharply with a large series from India, where con-genital opacities accounted for only 30.1% of cases and themajority of cases were performed for corneal opacities ac-quired after microbial keratitis.4 Our experience differedfrom that reported in four large North American series inwhich congenital opacities accounted for 48.6% of cases.1-4

This difference is probably related to the higher preva-lence of consanguinity and genetically related congenitalopacities in the Saudi patient population. Congenital glau-coma and CHED, which have a presumptive and con-firmed genetic etiology, respectively, accounted for nearlytwo-thirds of our congenital opacities. In contrast, these 2indications accounted for 21.1% of cases in the NorthAmerican studies,1-4 where Peter’s anomaly accounted fornearly 50% of all keratoplasties for congenital opacities.1-4

Evaluation of graft survival of pediatric keratoplasty as asingle statistic and comparing this result between variouspublished reports is flawed in many respects. Althoughcompensation can be made for differences in postoperativefollow-up with the use of Kaplan-Meier survival curves,13

differences in patient populations, as well as the upper agelimit for patient inclusion invalidate interseries compari-sons. For example, the cumulative survival curve will begreatly influenced by the relative proportion of patients

Table 5. Final visual acuity after primary pediatric keratoplasty vs surgic

AcuityCongenital opacity,* CHED, Acqu

n (%)† n (%)†

�20/40 3 (3.2) 3 (8.6)20/50-20/160 6 (9.5) 14 (48.6)Ambulatory‡ 3 (12.6)§ 16 (94.3)§Nonambulatory¶ 83 2Total 95 35

*Excludes cases of CHED (congenital hereditary endothelial dystrophy).†Cumulative % of eyes achieving this level of vision or better.‡20/200 to 20/800, counting fingers, or fixes and follows.§Visual acuity of fixes and follows was recorded in 3 eyes with CHED, 2 eyes¶Hand motions, light perception, no light perception.

that are derived from each of the major surgical indication

Journal of AAPOS

groups, with studies consisting of a relatively higher per-centage of congenital opacities generally having poorersurvival statistics. Even interseries comparison of graftsurvival for congenital opacities is subject to invalidationby the mix of patients, with series including a substantialrepresentation of eyes with CHED doing much betterthan a series with heavy representation of eyes with con-genital glaucoma, sclerocornea, or aniridia. For series withan upper age limit of 18, inclusion of substantial numbersof eyes with keratoconus, which often is classified as acongenital opacity in the statistical analysis, will facti-tiously improve the apparent prognosis for this category.For these reasons, we decided to evaluate graft survivalwithin our patient population by strictly adhering to thediagnostic groups originally described by Stulting and as-sociates,1 but with the separation of eyes with CHED as afourth surgical subgroup, and placement of a maximumage of 12 for inclusion in the study.

As with previous reports, most cases of graft failureoccur relatively early in the clinical course for non-CHEDcongenital opacities and acquired opacities.1,3 Stulting andassociates1 reported most failures in the first 12 months,whereas Erlich and associates3 documented most failuresin the first 6 months. In our series, 65.2% of graft failureshad occurred by the end of the first year and 88.0% hadoccurred by the end of the second year. In eyes withcongenital opacities not caused by CHED, there was asharp decrease in graft survival to at 1 year, a more modestdrop during the second year, and a very modest declineover the next 3 years. The survival curves for acquiredtraumatic and nontraumatic opacities were similar. Inas-much as most cases of graft failure occur during the first 2years, we recommend intensive observation and follow-upduring this interval, with examinations as frequently asevery 2 weeks during the first 3 months, every 4 weeks forthe remainder of the first year, and at least every 6-8 weeksduring the second year. As with previous reports fromother institutions9,11-13 and from our facility,10 the prog-nosis for overall graft survival in eyes with CHED wasexcellent.

Although previous investigators found a statistically sig-nificant correlation between increasing age at the time of

ations

ontraumatic opacity, Acquired, traumatic opacity, Total,n (%)† n (%)† n (%)†

0 (0) 2 (11.7) 8 (4.8)2 (11.1) 2 (23.5) 24 (19.4)7 (50.0)§ 2 (35.3) 28 (36.4)

9 11 10518 17 165

ngenital opacities, and 1 eye with acquired, nontraumatic opacity.

al indic

ired, n

with co

surgery and graft survival,2,5 the present study did not

Volume 11 Number 1 February 200746 Al-Ghamdi, Al-Rajhi, and Wagoner

confirm these findings for any of the 4 categories of sur-gical indications. In eyes with non-CHED congenitalopacities, the percentage of eyes with graft survival didimprove with increasing age, especially for children olderthan 6 months, but this factor was not statistically signif-icant. The majority of cases that were performed forCHED and acquired opacities were in children 2 years ofage or older, and analysis of increasing age as a continuousvariable did not demonstrate increased graft survival inolder children.

In the present study, the need to perform additionalsurgical procedures and the postoperative development ofmicrobial keratitis, endophthalmitis, or vitreoretinal pa-thology were statistically significant risk factors for graftfailure. Single procedures that were associated with anincreased risk of graft survival were concomitant lensec-tomy and subsequent vitreoretinal procedures. The reasonthat lensectomy increases the risk of graft failure is notclear, but a similar association was reported by Stultingand associates.1 Graft failure after vitreoretinal surgery canbe attributed not only to the trauma of the surgical inter-vention, but the poor prognosis for graft survival that waspresent in eyes that developed retinal detachments orendophthalmitis.

The occurrence of microbial keratitis in slightly morethan one-quarter of cases is worrisome, but recommenda-tions that may lead to better future results can be made onthe basis of our data. More than one-third of the cases wereattributed to the prophylactic antibiotic-resistant Gram-positive keratitis (5 cases) or the development of a suture-abscess after discontinuation of prophylactic antibiotics (11cases). Accordingly we recommend that suture removal beperformed as soon as wound healing is complete in thesehigh-risk cases and the use of prophylactic antibiotics thatare effective against Streptococcus and Staphylococcusspecies until all sutures have been removed.

Visual outcome after pediatric keratoplasty is not alwayscommensurate with that which would be expected fromobserved graft survival rates, especially in eyes with con-genital opacities. Poor vision in the setting of clear graft isusually attributed to development of amblyopia due tovisual deprivation before the establishment of a clear op-tical pathway,6,7 but the presence of coexisting ocularmorbidity, especially glaucoma,14,15 may also adverselyaffect the final result. The results of our study and those ofothers,1-4 demonstrate that many patients achieve im-provement in visual acuity compared with preoperativelevels, especially with respect to obtaining ambulatory vi-sion, thereby justifying the considerable investment intime and effort to perform pediatric keratoplasty. Unfor-tunately, very few eyes achieved a final visual acuity of20/40 or better, a finding consistent with several serieswhere less than 10% of eyes achieved this level of visualacuity.1-4

Congenital opacities caused by CHED and those causedby other etiologies represented 2 distinctly separate cate-

gories with respect to visual prognosis. More than 90% of

eyes with CHED achieved ambulatory vision after kera-toplasty, a finding that is consistent with previous obser-vations.9-12 In addition, approximately half of eyes ob-tained vision that was better than 20/200, although, likethe other surgical categories, very few achieved final visionof 20/40 or better. The excellent outcome with respect toambulatory vision correlates well with the success rate inmaintaining clear grafts, while the lower percentage ofpatients at progressively higher levels of vision is commen-surate with the relative degree of success with amblyopiamanagement.

In conclusion, the recommendation to perform pedi-atric keratoplasty should be made cautiously but shouldnot be withheld when the alternative is certain, lifelongdisability. The functional and social benefits of restor-ing useful vision to an eye otherwise doomed to alifetime of blindness cannot be overstated. The resultsof our study and as well as those of others suggest thatmany young children achieve improvement in visualacuity compared with preoperative levels, thereby justi-fying the considerable investment in time and effort toperform these procedures. When pediatric keratoplastyis performed, participation of the ophthalmologist,child and parents as a team in meticulous follow-up careand amblyopia management is essential to maintain aclear graft and optimize the chances of achieving asatisfactory visual outcome.

AcknowledgmentsWe thank the contributions of the King Khaled Eye Specialist HospitalCornea Transplant Study Group to this study. Physician members:Klaus Teichmann, MD, and Abdul-Elah Al-Towerki, MD. Datacoordination staff: Ritzchie O. Burog, BSSA, Barbara Elias, CEBT,El-Sayed Gonnah, CEBT, Jamila Al-Shahrani, BSc, and Abid Al-Talhi, CEBT. Biostatisticians: Yaser Shaheen, BSSA, Abdulaziz A.bin Saeed, PhD, M. Bridget Zimmerman, PhD.

References1. Stulting RD, Sumers KD, Cavanagh HD, Waring GO 3rd, Gam-

mon JA. Penetrating keratoplasty in children. Ophthalmology 1984;91:1222-30.

2. Cowden JW. Penetrating keratoplasty in infants and children. Oph-thalmology 1990;97:324-9.

3. Erlich CM, Rootman DS, Morin JD. Corneal transplantation in infants,children and young adults: experience of the Toronto Hospital for SickChildren, 1979-88. Can J Ophthalmol 1991;26:206-10.

4. Dana MR, Moyes AL, Gomes JA, et al. The indications for andoutcome in pediatric keratoplasty: a multicenter study. Ophthalmol-ogy 1995;102:1129-38.

5. Aasuri MK, Grag P, Gokhle N, Gupta S. Penetrating keratoplasty inchildren. Cornea 2000;19:140-4.

6. Hubel DH. Exploration of the visual cortex 1955-78. Nature 1982;299:515-24.

7. Wiesel TN. Postnatal development of the visual cortex and theinfluence of environment. Nature 1982;299:583-91.

8. Kirkness CM, McCartney A, Rice NS, Garmer A, Steele AD. Congen-ital hereditary corneal edema of Maumenee: its clinical features, man-agement, and pathology. Br J Ophthalmol 1987;71:130-44.

9. Sajjadi H, Javadi MA, Hemmati R, Mirdehghan A, Parvin M, Nassiri N.

Results of penetrating keratoplasty in CHED. Cornea 1995;14:18-25.

Journal of AAPOS

Volume 11 Number 1 February 2007 Al-Ghamdi, Al-Rajhi, and Wagoner 47

10. Al-Rajhi AA, Wagoner MD. Penetrating keratoplasty in congenitalhereditary endothelial dystrophy. Ophthalmology 1997;104:956-61.

11. Schaumberg DA, Moyes AL, Gomes JAP, Dana MR. Corneal trans-plantation in young children with congenital hereditary endothelialdystrophy. Am J Ophthalmol 1999;127:373-8.

12. Javadi MA, Baradaran-Rafii AR, Zamani M, et al. Penetrating kera-toplasty in young children with congenital hereditary endothelial

dystrophy. Cornea 2003;22:420-3.

Journal of AAPOS

13. Kaplan EL, Meier RF. Nonparemetric estimation from incompleteobservations. J Am Stat Assoc 1958;53:457-81.

14. Al-Torbak AA. Outcome of combined Ahmed glaucoma valve im-plant and penetrating keratoplasty in refractory congenital glaucomawith corneal opacity. Cornea 2004;23:554-9.

15. Huang SC, Soong HK, Brenz RM, Meyer RF, Sugar A. Problemsassociated with penetrating keratoplasty for corneal edema in con-

genital glaucoma. Ophthalmic Surg 1989;20:399-402.

An Eye on the Arts – The Arts on the Eye

And old men tell a half-remembered taleOf horror done where a dark ledge splits the seaAnd a double surf beats on the gráy shóres:How a king’s new woman, sickWith hatred for the queen he had imprisoned,Ripped out his two sons’ eyes with her bloody

handsWhile grinning Arês watched the shuttle plungeFour times: four blind wounds crying for revenge,—Sophocles, The Oedipus Cycle, Harcourt, 1977Translated by Dudley Fitts and Robert Fitzgerald