RESULTS OF PENETRATING KERATOPLASTY IN SYPHILITIC INTERSTITIAL

GOEGEBUER A.*, ***, AJAY LOHIYA**, ***, CLAERHOUT I.*, KESTELYN PH.*

ABSTRACT ratite interstitielle (KI) syphilitique avec ceux dé- crits dans la litérature. Purpose: To compare the results in our patient se- Méthode: Analyse rétrospective des cas, avec acui- ries after penetrating keratoplasty (PKP) for syphi- té visuelle, clarté du greffon, épisodes de rejet, pres- litic interstitial keratitis (IK) with those described in sion intraoculaire et comptage postopératoire des the literature. cellules endothéliales. Methods: Retrospective case series in which visual Résultats: L’acuité visuelle s’est améliorée dans tous acuity (VA), graft clarity,rejection episodes, intraocu- les cas. Il n’y avait pas de signes de déhiscence du lar pressure and endothelial cell density (ECD) were greffon, ni d’ occurrence d’une membrane rétrocor- examined postoperatively. néenne. L’ postopératoire n’était pas Results: Postoperative VA improved in all cases. plus prononcée chez nos patients avec une kératite There was no evidence of wound dehiscense or oc- interstitielle comparée aux patients opérés pour currence of retrocorneal membrane formation in any d’autres raisons. Un déclin normal du comptage des case. Postoperative inflammation was not more se- cellules endothéliales prouvait qu’il n’y avait pas d’in- vere in our patients with syphilitic IK compared to flammation infraclinique. patients undergoing PKP for other reasons. A nor- Conclusion: Une kératoplastie transfixiante pour KI mal decline in ECD proved that there was no sub- syphilitique a un bon pronostic en ce qui concerne clinical inflammation as well. la clarté du greffon. Tous les patients bénéficiaient Conclusion: PKP for syphilitic IK has a good prog- d’ une amélioration de l’acuité visuelle, bien que par- nosis in our case series as far as graft survival is con- fois limitée. Nous rencontrions moins de complica- cerned. Improvement in VA was present in all cas- tions postopératives chez nos patients que dans les es, though sometimes limited. In our case series, we séries décrites dans la litérature ancienne, ce qui est experienced less postoperative complications than probablement du a l’évolution des techniques mi- described in the older literature, which is probably crochirurgicales. due to better microsurgical techniques used nowa- days. KEY-WORDS Penetrating keratoplasty, syphilitic interstitial RÉSUMÉ keratitis, postoperative inflammation, But: Comparer les résultats obtenus chez nos pa- endothelial cell density tients après une kératoplastie transfixiante pour ké- MOTS-CLES Kératoplastie transfixiante, kératite zzzzzz interstitielle syphilitique, inflammation * Department of Ophthalmology, Ghent University postopératoire, comptage des cellules Hospital, Belgium endothéliales. ** Department of Ophthalmology, LTMMC & LTMGH, Sion, Mumbai, India *** Both authors contributed equally

received: 16.07.03 accepted: 16.09.03

Bull. Soc. belge Ophtalmol., 290, 35-39, 2003. 35 INTRODUCTION membrane (5). We compared our results after PKP for syphilitic IK with those described in Although has become rare, it still is an literature. existing cause of ocular disease. Syphilis is a multi-system, multi-symptom dis- order, caused by a spirochete, Treponema pal- MATERIALS AND lidum, that occurs primarily through sexual trans- METHODS mission, but can also be spread through blood transfusion or direct contact with an infected We retrospectively studied six cases of pene- lesion. Transplacental transmission leads to con- trating keratoplasty, performed on five eyes of genital syphilis. Clinically, if untreated, acquired four patients with syphilitic interstitial keratitis syphilis may follow three classic stages: pri- (one patient had bilateral grafts, and had a re- mary, secondary and tertiary syphilis. In cases graft in one eye) and reviewed the literature of congenital syphilis, patients may manifest concerning this subject. Hutchinson’s triad: interstitial keratitis, deaf- The diagnosis of syphilitic interstitial keratitis ness and teeth malformation. was based on the characteristic slit lamp find- In acquired syphilis, various inflammatory ocu- ingsofinterstitalkeratitis,combinedwithaposi- lar complications may occur in the secondary tive TPHA serology. or third stages: episcleritis, uveitis, glaucoma, The study group consisted of one male and three neuroretinitis, chorioretinitis, ischemic vascu- females aged between 39 and 71 years, with lopathy, infectious optic neuropathy, optic a mean age of 60 years. The average follow-up atrophy and pupil abnormalities. period was 42 months, with a range between Interstitial keratitis (IK) is a typical manifesta- 31 and 50 months (median: 40 months). tion of congenital syphilis (4), but may rarely Donor corneal buttons preserved in organ cul- occur in acquired syphilis. The active keratitis ture at 31°C, with an endothelial cell density begins with a corneal infiltrate, uveitis and of at least 2200 cells/mm2, were used for graft- Descemet’s membrane alterations. This ing. In all cases, the graft size was 7.25 mm progresses to heavy superficial and deep vas- for donor and 7.00 mm for recipient . cularisation, leading to the so called ’salmon Double continuous (10-0) non absorbable su- patches’. Clearing is associated with corneal tures were used in all patients. In two cases, thinning and ghost vessel formation. Chronic keratoplasty was combined with lens extrac- ocular signs of congenital syphilis include stro- tion and IOL implantation. Peripheral iridecto- mal scarring, ghost vessels and residual deep my was not done in any of the cases. vascularisation, cornea guttata, scrolls in De- scemet’s membrane, glaucoma, band keratop- Postoperatively, steroid eye drops were admin- athy, corneal thinning, lipid keratopathy and istered with tapering dosage for six months. Salzmann’s degeneration. Other late sequelae Systemic steroids were not used. of congenital syphilis include cataract, pigmen- All patients were followed-up weekly for the first tary retinopathy (salt and pepper fundus) and 3 weeks, then monthly for 6 months and then optic atrophy (7). 6-monthly for 3 years. On each follow-up visit, In cases of interstitial keratitis, therapeutic de- following examinations were done: visual acui- cision making is based on syphilitic disease ac- ty testing on Snellen’s chart, slit lamp biomi- tivity. Distinguishing between active and resi- croscopy, intraocular pressure measuring us- dual corneal disease is crucial when deciding ing applanation tonometry and fundus exami- to start a possible penicillin treatment (3). Pen- nation. etrating keratoplasty (PKP) is usefull in cases Endothelial cell counts were done 6 months af- of severe corneal opacification in quiet inter- ter surgery and then yearly thereafter with spec- stitial keratitis (8). Several problems after PKP ular microscope and compared to data obtained for syphilitic keratitis are reported in the liter- from a control group. The data from the con- ature: higher incidence of serious postopera- trol group were used to generate the expected tive uveitis (1) and glaucoma (2, 8), wound de- loss of ECD over the different time intervals in- hiscence (8) and rarely retained Descemet’s volved.

36 RESULTS in the control group are given as mean ± the standard deviation. Mean loss of ECD and stan- Patient characteristics at dard deviation were used to construct 90% ref- presentation: erence ranges at each time interval in the con- trol population (mean ± 1.64 times the stan- Preoperative slit lamp examination showed the dard deviation). A decline in ECD outside this presence of leucomatous corneae with ghost 90% reference range was considered signifi- vessels in all cases. One patient had miosis and cantly stronger than the expected decline. For . Another patient had exten- instance from transplantation to 6 months post- sive deep corneal vascularisation in both eyes, operatively the endothelial cell loss was 17% but also signs of an old uveitis with posterior with a standard deviation of 13. So a decline synechiae and a pigmentary retinopathy. over the six month period after the transplan- VDRL serology was positive in one patient, TPHA tation of more than 38% (17+1.64×13 gives was positive in all cases. 38%) was considered significantly stronger than expected. Similar reference ranges were con- Postoperative findings: structed for all time periods involved. Postoperative decline in endothelial cell densi- Three grafts of three patients remained clear on ty in our patient group was not significantly all follow-up visits. The fourth patient devel- stronger than that observed in the control group, oped an irreversible allograft rejection of the except for 1 patient (who had undergone 2 re- graft in her right eye nine months postopera- versible allograft rejection episodes). tively. She underwent a regraft in that eye one year later. A PKP was also performed in the left eye eleven months after the regraft in the right DISCUSSION eye. Both grafts underwent two reversible al- In this study we have examined the outcome lograft rejection episodes during the first post- of PKP in syphilitic interstitial keratitis. IK has operative year. On the last follow up visit, 50 classically been defined as a non suppurative months and 39 months postop respectively, infiltration of the corneal stroma, with stromal both grafts were clear. vascularisation. Years after active inflamma- In all eyes, the postoperative inflammatory re- tion, only ghost vessels and corneal scarring re- action did not exceed that observed after PKP main. Severe IK may leave prominent vessels, for other corneal pathologies. a dense scar and corneal thinning (4). Our pa- Postoperative best corrected visual acuity im- tients were diagnosed with syphilitic IK based proved remarkably in all cases. However, only on typical slit lamp findings and positive sero- one eye reached a BCVA of 10/10 and three logy for syphilis. All eyes had healed inter- eyes had a BCVA of less than 5/10 (fig. 1). stitial keratitis with leucomatous corneae with There was no evidence of wound dehiscence ghost vessels and in one case extensive deep or occurence of retrocorneal membrane forma- vascularisation. Signs of old uveitis were present tion in any of our cases. Postoperative IOP was in both eyes of 1 patient. normal in all eyes at all visits. Serologic testing consisted in VDRL- and TPHA- Endothelial cell density sceening. VDRL (venereal disease research lab- (ECD): oratory) is a non-specific test for Treponema pallidum used to assess disease activity and We compared the evolution of the endothelial provides an indication of response to treatment, cell density in our patients with data from a turning negative after succesful treatment. TPHA- control group. The control group consisted of testing (treponema pallidum hemagglutination 167 patients, who had had PKP for other cor- test) is a treponema-specific test, remaining neal pathologies, in which follow-up was un- positive for many years to a lifetime in syphilis eventful and in which no additional surgical patients and is used to confirm prior exposure, procedures were performed that could affect the whether adequately treated or not (6). This ex- ECD (table 1). The percentage loss relative to plains why all our patients had positive TPHA the initial density of the graft is also given. Data and only 1 had positive VDRL.

37 Figure 1. Evolution of pre- and postoperative best corrected visual acuity.

Table 1. Evolution of endothelial cell density (cells/mm2)

CASE age/sex Preoperative 6m postop 1y postop 2y postop 3y postop remarks ECD (%loss) ( % loss) (%loss) ( % loss) 1 77y, F 3050 2000 1720 1720 (34%) (44%) (44%) 2 65y, F 2950 2200 2200 2200 (25%) (25%) (25%) 3 74y, M 2360 2000 1480 1200 1200 (15%) (37%) (49%) (49%) 4 45y, F 2625 1600 1200 1080 regraft, 2 reversible rejection OD (39%) (54%) (59%) episodes at5&12months 4 45y, F 2200 Oedema 680 680 2 reversible rejection episo- OS (69%) (69%) desat2&9months Control (n=167) 2575 ± 283 2155 ± 411 1978 ± 469 1723 ± 519 1514 ± 493 (17% ± 13) (25% ± 15) (34% ± 18) (41% ± 17) Decline outside 90% >38% >50% >63% s69% reference range

Penetrating keratoplasty is a therapeutic op- state that the violent inflammation is frequent- tion for treating vision-threatening corneal opaci- ly associated with the formation of a retrocor- fication, due to syphilitic IK. However, several neal membrane and graft opacification. More- problems after PKP for syphilitic IK are report- over, they observed a weaker anti-inflammato- ed in literature. Goldman et al. (1) observed a ry response to massive steroid (and ) more severe postoperative inflammatory reac- administration. Rabb and Fine (8) reported a tion after PKP for IK in congenital syphilis, com- higher incidence of ’’homograft reactions’’ (=re- pared to cases of PKP for other causes. They jection episodes) in their series of 61 PKP’s for

38 syphilitic IK, than observed in keratoplasty for sible hypothesis could be the poor optical qual- other types of corneal disease. In their series, ity of the surface of the graft, due to limbal they found wound dehiscence following suture stemcell deficiency as a late sequel of IK, a removal and partial over-riding of the graft to hitherto unreported complication. In conclu- be the most common complications. They sug- sion, the results of our study show that prog- gest that wound healing is retarded due to thin- nosis of PKP for syphilitic interstitial keratitis ness of the recipient cornea. As glaucoma is is the same as for PKP for other corneal pa- more frequent in patients with syphilitic IK (2), thologies, regarding graft clarity, visual acuity there is also an increased risk for postopera- and ECD tive glaucoma after PKP for syphilitic IK (8). A rare complication of PKP for syphilitic IK is REFERENCES splitting of Descemet’s membrane with reten- tion of the posterior part at the time of surgery (1) GOLDMAN J.N., KENNETH M.D. and GIRARD − (5). F. Causes of opacity after corneal graft in in- In our patient series we did not experience the terstitial keratitis. Bull Mem Soc Fr Ophtalmol 1967; 80:126-131. major postoperative complications, described (2) GRANT W.M. − Late glaucoma after interstitial in the older literature: neither wound dehis- keratitis. Am J Ophthalmol 1975; 79:87-91. cence, retrocorneal membrane formation nor re- (3) HARISPRASAD S.M., MOON S.J., ALLEN R.C. taining of Descemet’s membrane. These good and WILHELMUS K.R. − Keratopathy from con- results are probably due to better microsurgi- genital syphilis. Cornea 2002; 21:608-609. cal techniques used nowadays, since the litera- (4) KNOX CM and HOLSCLAW DS. Interstitial ke- ture on these complications mainly dates back ratitis. Int Ophthalmol Clin 1998; 38:183-195. to the sixties and seventies. There were no cas- (5) KURZ G.H. and D’AMICO R.A. − Retained es of postoperative glaucoma. Descemet’s membrane after keratoplasty for old All grafts were clinically clear at the end of the interstitial keratitis. Am J Ophthalmol 1973; follow-up period. Postoperative inflammation 76:51-53 did not seem to exceed that after PKP for other (6) LARSON S.A., STEINER B.M. and RUDOLPH − reasons. However, one patient was subject to A.H. Laboratory diagnosis and interpretation of tests for syphilis. Clin Micriobiol Rev 1995; several allograft rejection episodes: an irrever- 8:1-21 sible rejection in one eye and two reversible re- (7) PROBST L.E., WILKINSON J. and NICHOLS B.D. jection episodes in both the regraft in that eye − Diagnosis of congenital syphilis in adults pres- and the primary graft in the other eye. This pa- enting with interstitial keratitis. Can J Ophthal- tient however, also had a clear history of ante- mol 1994; 29:77-80. rior uveitis with synechiae, seen on preopera- (8) RABB M.F. and FINE M. − Penetrating kerato- tive biomicroscopy. Moreover, on close exami- plasty in interstitial keratitis. Am J Ophthalmol nation, ghost vessels were not always com- 1969; 67:907-917 pletely empty and may have constituted an ad- ditional risk factor for allograft rejection. Acknowledgements: Ilse Claerhout is a research as- Endothelial cell density did not show a great- sistant for the Flemish Fund of Scientific Research- er-than-expected decline when compared to a Flanders (FWO-Vlaanderen)1 control group. This proves that these patients are probably not subject to subclinical inflam- mation postoperatively, since this would cer- zzzzzz tainly affect their ECD even in the absence of Corresponding author: clinical rejection episodes. Ilse Claerhout Visual acuity improved remarkably in all cas- Ghent University Hospital es. However, final visual acuity was 10/10 in De Pintelaan 185 only one eye and was less than 5/10 in 3 eyes. 9000 Ghent Belgium One possible reason for these moderate visual tel.: +32-9-240.43.88 results is some degree of amblyopia, due to the fax: +32-9-240.38.38 occurrence of IK at an early age. Another pos- [email protected]

39