- Email: [email protected]
Outcomes of Corneal Transplantation for Irreversible Corneal Decompensation Secondary to Corneal Endotheliitis in Asian Eyes
Outcomes of Corneal Transplantation for Irreversible Corneal Decompensation Secondary to Corneal Endotheliitis in Asian Eyes MARCUS ANG, CHELVIN C.A. SNG, SOON-PHAIK CHEE, DONALD T.H. TAN, AND JODHBIR S. MEHTA PURPOSE:
To describe outcomes of corneal transplantation for irreversible corneal decompensation from corneal endotheliitis in Asian eyes. DESIGN: Retrospective, observational case series. METHODS: We reviewed consecutive patients with corneal endotheliitis (32 eyes of 31 subjects) who underwent keratoplasty (January 1, 2008-December 1, 2009). All eyes had preoperative aqueous polymerase chain reaction (PCR) analysis for viruses, including cytomegalovirus (CMV). CMV-positive patients were treated preoperatively with topical corticosteroids and antiCMV treatment (oral valganciclovir 900 mg twice daily, topical ganciclovir 0.15% 5 applications per day, for 6 weeks) with complete resolution of ocular inflammation, and quiescence for at least 6 months before corneal transplantation. Our main outcome measure was recurrence of endotheliitis within 1 year after corneal transplantation. RESULTS: Five eyes were CMV positive; the remaining 27 eyes were negative for all viruses on PCR analysis. CMV-positive patients had a higher rate of recurrence of endotheliitis within 1 year after corneal transplantation, compared with CMV-negative eyes (60% vs 7.4%, P [ .01). The CMV-positive eyes had recurrent endotheliitis at a median of 10 months (range 3-11 months) after corneal transplantation. After successful antiCMV treatment, all 5 CMV-positive eyes then continued to have clear grafts for a median duration of 21 months (range 13-44 months). CONCLUSION: Our study suggests that Asian patients with corneal endotheliitis may benefit from preoperative aqueous PCR analysis before corneal transplantation. Such patients were more likely to have a recurrence of endothelial inflammation if they were CMV positive preoperatively, despite successful anti-CMV treatment before surgery. (Am J Ophthalmol 2013;156: 260–266. Ó 2013 by Elsevier Inc. All rights reserved.) Accepted for publication Mar 14, 2013. From Singapore National Eye Centre (M.A., C.C.A.S., S.P.C., D.T.H.T., J.S.M.); Singapore Eye Research Institute (M.A., C.C.A.S., S.P.C., D.T.H.T., J.S.M.); Department of Ophthalmology, National University Health System (C.C.A.S., S.P.C., D.T.H.T., J.S.M.); and Duke-National University of Singapore Graduate Medical School (D.T.H.T., J.S.M.), Singapore. Inquiries to A/Professor Jodhbir S. Mehta, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751; e-mail: [email protected]
260
Ó
2013 BY
C
ORNEAL ENDOTHELIITIS IS AN ENTITY CHARACTER-
ized by localized corneal edema with the presence of mild iritis and keratic precipitates (KPs).1 Although corneal endotheliitis was once considered ‘‘autoimmune’’ or ‘‘idiopathic,’’ there have been increasing reports that viruses such as cytomegalovirus (CMV) and herpes simplex virus (HSV) may play a role as etiologic agents.2 Polymerase chain reaction (PCR) analysis of aqueous humor has been shown to be capable of detecting the virus3; however, real-time PCR and quantification of viral DNA may be useful in distinguishing an etiologic virus, as opposed to a ‘‘bystander’’ virus (the latter does not play a causative role in inciting inflammation or endotheliitis).4 A viral etiology in corneal endotheliitis is further supported by reports that endothelial lesions and inflammation resolve rapidly with the application of topically applied and systemic antiviral treatment.2 Cytomegalovirus is increasingly recognized as the most common virus implicated in corneal endotheliitis, especially in Asian patients.5 Although the exact pathogenesis of CMV corneal endotheliitis is not yet understood, it is postulated that anterior chamber–associated immune deviation (ACAID) normally prevents the proliferation of CMV within the anterior chamber (AC), leading to viral infection of the endothelium.6 However, it remains unclear whether CMV endotheliitis is attributable to a primary CMV infection or a local reactivation of CMV in the anterior segment secondary to dendritic cell activation provoked by a separate trigger, seen in other parts of the body.5 Although medical therapy remains the mainstay of treatment, persistent corneal edema or irreversible corneal decompensation often ensues because of the cytopathic effect on corneal endothelial cells, leading to the need for a corneal graft.7 In this study, we describe the outcome of various forms of corneal transplantation, but mainly Descemet stripping automated endothelial keratoplasty (DSAEK) performed in eyes that presented with endotheliitis and corneal decompensation. We also compared the outcomes between eyes that were ‘‘CMV positive’’ and ‘‘CMV negative,’’ as assessed using real-time PCR preoperatively.
METHODS STUDY COHORT:
consecutive
ELSEVIER INC. ALL
We conducted a retrospective review of patients who presented with corneal
RIGHTS RESERVED.
0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.03.020
endotheliitis (ie, localized corneal edema with mild AC inflammation and KPs1) who subsequently underwent corneal transplantation. All patients underwent corneal transplantation at the Singapore National Eye Centre (SNEC) between January 1, 2008 and December 1, 2009. We obtained clinical data from our ongoing Singapore Corneal Transplant Study cohort, an audited longitudinal prospective study that contains preoperative, intraoperative, and yearly postoperative follow-up clinical data on all corneal transplants performed by corneal surgeons at the SNEC.8 We conducted this study with approval from our institution’s Singapore Health Services Institutional Review Board and followed the principles of the Declaration of Helsinki. INVESTIGATIONS:
At presentation, all eyes with suspected corneal endotheliitis with localized corneal edema, KPs, and mild AC inflammation had an AC tap performed preoperatively at the slit lamp using an aseptic technique. A minimum of 100 mL of aqueous was obtained and the samples were either delivered to the laboratory at 18 C-25 C within 1 hour, or stored and transported at 2 C-4 C within 24 hours. The samples were aliquoted immediately upon arrival at the laboratory and were either used for DNA extraction immediately or kept at 20 C and extracted for DNA within 1 week. The DNA was extracted using Qiagen QIAmp DNA Extraction Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions and tested for CMV DNA by PCR. The aqueous samples were sent for a validated, nested multiplex PCR for HSV, varicella zoster virus (VZV), CMV, and Toxoplasma gondii DNA as previously described.3 The lower limits of detection for HSV-1, HSV-2, CMV, VZV, and Toxoplasma gondii are 360, 120, 50, 40, and 3 copies per PCR reaction, respectively, with a sensitivity exceeding 95% for all organisms.9 Patients who were CMV positive were also analyzed for CMV DNA using real-time PCR with a sensitivity of 7 copies/mL with a 95% confidence interval (CI).10 All patients with a positive aqueous CMV DNA also underwent tests for CMV antigen (cytospin), CMV antibodies, and human immunodeficiency virus (HIV) antibodies in the serum. MANAGEMENT:
All patients with endotheliitis were co-managed by the SNEC Cornea Service and the SNEC Ocular Inflammation and Immunology Service. Patients included in this study underwent either a penetrating keratoplasty (PK) (4 eyes), DSAEK (27 eyes), or Descemet membrane endothelial keratoplasty (DMEK) (1 eye) using techniques as previously described.11,12 Preoperatively, patients were treated with topical corticosteroids and topical antiglaucoma medications if intraocular pressure (IOP) was raised. Patients who were CMV positive were co-managed with the infectious disease physician at the Singapore General Hospital and treated with oral valganciclovir (900 mg) twice daily for 6 weeks followed by 450 mg twice daily for a further 6 weeks. Topical ganciclovir
VOL. 156, NO. 2
ophthalmic gel 0.15% (Virgan; Laboratoires The´a, Clermont-Ferrand, France) was also prescribed 5 times a day. A week before completion of therapy, a repeat aqueous analysis for CMV using real-time PCR was done. Quiescent ocular inflammation for at least 6 months and undetectable CMV DNA titer with repeat aqueous PCR analysis for CMV-positive eyes were conditions required before corneal transplantation was performed. Our main outcome measure was a recurrence of endotheliitis, presenting clinically as localized corneal edema with AC inflammation and KPs. Recurrence was also suspected when patients presented with an unexplained sudden decrease in endothelial cell count (performed routinely at 6 months, then yearly postoperatively) in an uninflamed eye and normal graft appearance or intraocular inflammation that failed to clear or worsened with conventional high-dose corticosteroid treatment regimens.13 Postoperatively, visual acuity (VA), IOP, graft clarity, and anterior segment inflammation were monitored, with a high index of suspicion for any recurrence of endotheliitis. We used similar postoperative topical corticosteroid medication regimens in all patients: a single drop of prednisolone acetate 1% every 3 hours for 1 week, 3 times a day for 6 months, twice daily for 3 months, then once a day for up to 1 year. All patients received topical levofloxacin 0.5% every 3 hours for 1 week, followed by 3 times a day for 6 months. Patients who suffered from recurrent endotheliitis postoperatively underwent a repeat aqueous tap, and the aqueous sample was reanalyzed with the tetraplex PCR as described above, with real-time PCR analysis for CMV-positive eyes. Patients who were CMV positive were restarted on topical ganciclovir ophthalmic gel 0.15% every 3 hours and were watched closely for clinical resolution of endotheliitis. Patients who were negative for all viruses tested on PCR analysis were presumed to have immunologic rejection and were treated with only topical prednisolone acetate 1% every 3 hours. STATISTICAL ANALYSIS: Statistical analysis included descriptive statistics, where the mean and standard deviation (mean 6 SD) were calculated for the continuous variables, while frequency distribution and percentages were used for categorical variables. Comparisons between categorical variables were conducted using x2/Fisher exact tests as appropriate. The independent t test and 1-way analysis of variance test were used for continuous measures group comparisons. P value <.05 was considered statistically significant. All analyses were performed using STATA version 11 (StataCorp LP, College Station, Texas, USA).
RESULTS A TOTAL OF 32 EYES OF 31 PATIENTS UNDERWENT CORNEAL
grafts for irreversible corneal decompensation secondary to
CORNEAL GRAFTS FOR ENDOTHELIITIS
261
endotheliitis. Five eyes were found to have evidence of CMV from aqueous PCR analysis preoperatively. The remaining 27 eyes did not have any evidence of CMV or any other viruses tested by the nested multiplex PCR— and all of these eyes were not found to be CMV positive during the study period. All patients were found to be immunocompetent and negative for CMV antigen and antibody tests as described above. Table 1 compares the clinical characteristics between these CMV-positive and CMV-negative eyes. CYTOMEGALOVIRUS-POSITIVE EYES: Patients who were CMV positive were all successfully treated with anti-CMV treatment as described above, until ocular inflammation was quiescent for at least 6 months with an undetectable CMV DNA titer, with repeat aqueous PCR analysis before corneal transplantation. Postoperatively, a significantly higher proportion of patients who were CMV positive had recurrence of endotheliitis after corneal transplantation within 1 year postoperatively (3/5, 60%), as compared with those who were CMV negative (2/27, 7.4%) (P ¼ .01). However, there was no significant difference in terms of graft failure between CMV-negative and CMV-positive eyes (3.7% vs 0%, respectively, P ¼ .65). Visual outcomes at 1 year (best-corrected visual acuity 20/40 or better) were also comparable between CMV-positive (3/5) and CMVnegative eyes (13/27) (P ¼ .84). The 3 CMV-positive eyes presenting with recurrence, at a median of 10 months (range 3-11 months) after corneal transplantation, were seen to develop low-grade AC inflammation, development of new keratic precipitates on the graft endothelium, and reduction in graft clarity. A repeat aqueous sample was obtained in these eyes, which demonstrated the presence of CMV DNA on real-time PCR analysis (Table 2). Figure 1 illustrates the clinical presentation of Patient 2. Preoperatively, this patient presented with clinical signs of corneal endotheliitis, with fine to medium KPs, corneal edema, and low-grade AC inflammation, and was also noted to be CMV positive (Figure 1, Top left). He underwent an uncomplicated DSAEK (Figure 1, Top right) but developed recurrent corneal endotheliitis 3 months after the surgery, leading to graft edema, and the presence of KPs and mild AC inflammation (Figure 1, Bottom left). After institution of topical ganciclovir and oral valganciclovir, the endotheliitis and inflammation resolved and the graft remained clear (Figure 1, Bottom right). Similarly, Patient 4 (Figure 2) was CMV positive preoperatively (Figure 2, Top left) and developed significant corneal decompensation (Figure 2, Top right) despite anti-CMV therapy. DSAEK was successfully performed, but the patient suffered recurrence of corneal endotheliitis and inflammation postoperatively (Figure 2, Bottom left) and required reinstitution of anti-CMV therapy (Figure 2, Bottom right). The graft remained clear with resolution of inflammation after treatment.
262
TABLE 1. Baseline Demographics of Subjects With Corneal Endotheliitis Comparing Cytomegalovirus-Positive and Cytomegalovirus-Negative Eyes
Mean age (years 6 SD) Sex (male) Mean follow-up (months 6 SD) Race Chinese Malay Indian Other Previous surgery Trabeculectomy Tube Pseudophakic Type of graft DSAEK PK DMEK
Cytomegalovirus Negative (n ¼ 27)
Cytomegalovirus Positive (n ¼ 5)
P Valuea
65 6 12 15 18 6 11
62 6 5 5 25 6 6
.53 .13 .16
18 5 1 3
5 0 0 0
.78
4 0 15
2 1 4
.23 .16 .63
23 3 1
4 1 0
.86
DMEK ¼ Descemet membrane endothelial keratoplasty; DSAEK ¼ Descemet stripping automated endothelial keratoplasty; PK ¼ penetrating keratoplasty. a 2 x /Fisher exact tests as appropriate.
All patients with recurrent corneal endotheliitis and found to be CMV positive were restarted on oral ganciclovir and topical valganciclovir and monitored closely for further signs of inflammation (as stated in Methods section above). After successful treatment, all 5 CMVpositive eyes had continued to have clear grafts for a median duration of 21 months (range 13-44 months). However, 1 eye (Patient 5) developed late graft failure after 18 months and required a repeat DSAEK. This repeat graft remained clear for 6 months, then suffered a recurrence of CMV endotheliitis (CMV real-time PCR DNA count: 3.5-7.5 3 106 viral copies) and eventually failed 14 months later, despite oral valganciclovir and topical ganciclovir therapy. In the remaining patients, repeat AC taps for real-time PCR analysis revealed reduction to undetectable CMV levels after 6 months of therapy. CYTOMEGALOVIRUS-NEGATIVE EYES: Only 2 of 27 CMV-negative eyes (7.4%) developed an episode of inflammation after corneal transplantation within 1 year (at 3-6 months); these eyes were CMV negative on repeat aqueous sampling and were treated successfully with increased frequency of topical prednisolone acetate 1% for 3 months, leading to a resolution of the inflammation. Of note, we did not observe any increased incidence of de novo glaucoma between CMV-positive and
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
923 6.0 3 106 11 1.6 3 106 20/160 71/Chinese/male 5
BCVA ¼ best-corrected visual acuity; CMV ¼ cytomegalovirus; DSAEK ¼ Descemet stripping automated endothelial keratoplasty; ECC ¼ endothelial cell count; HM ¼ hand motions; PCR ¼ polymerase chain reaction analysis; PK ¼ penetrating keratoplasty.
HM Graft failure/DSAEK 32
20/25 20/30 20/25 20/40 Clear graft/PK Clear graft/DSAEK Clear graft/DSAEK Clear graft/DSAEK 58/Chinese/male 60/Chinese/male (right eye) 60/Chinese/male (left eye) 61/Chinese/male 1 2 3 4
20/2000 20/60 20/400 20/400
NA 3 NA 10
Nondetectable 1.8 3 106 Nondetectable 1.0 3 103
2817 3067 2538 1318
Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Oral valganciclovir 900 mg once daily Topical ganciclovir gel 0.15% 7.7 3 103 770 1.8 3 106 1.1 3 103
21 16 13 44
of Graft (Months) (1 Year) PCR (Copies) Recurrence (Months) Postoperative Treatment Demographics (Age/Ethnicity/Sex) Patient
BCVA
CMV PCR
Final
BCVA
Final Outcome/Type Follow-up Duration ECC Postoperative Real-Time CMV Time to Corneal Endotheliitis Preoperative Real-Time Preoperative
TABLE 2. Outcomes of Patients With Cytomegalovirus-Positive Corneal Endotheliitis who Underwent Corneal Transplantation
VOL. 156, NO. 2
CMV-negative eyes 1 year after corneal transplantation (0/5 vs 2/27, P ¼ .65).
DISCUSSION IT IS OFTEN DIFFICULT TO DISTINGUISH CLINICALLY
between endotheliitis attributable to immunologic rejection or viral infection in patients who have undergone corneal transplantation.14 Both conditions present with the same nonspecific, heterogeneous clinical signs, which include any combination of mild AC inflammation, KPs that may be linear in orientation, localized endothelial decompensation, unexplained reduction in endothelial cell count, or raised IOP.15 Hence, in our center, all patients who presented with these nonspecific signs underwent an aqueous tap to exclude viral endotheliitis before they were presumptively treated for immunologic rejection. It is particularly important to differentiate viral endotheliitis from graft rejection because the treatment for allograft rejection involves intensive topical and often systemic immunosuppression—which can exacerbate latent viral infections such as CMV and cause increased inflammation with subsequent graft failure.15 In this study, we found that patients with previously diagnosed CMVpositive endotheliitis are more likely to have a recurrence of endotheliitis within a year following corneal transplantation—despite preoperative anti-CMV treatment and evidence of successful eradication of CMV and ocular inflammation before corneal transplantation was performed. This is in contrast to patients with corneal endotheliitis in the absence of detectable CMV preoperatively, who are less likely to develop signs of endotheliitis postoperatively even while receiving similar postoperative topical corticosteroid therapy. It has been suggested that endotheliitis secondary to graft rejection in DSAEK usually occurs between 12 and 18 months postoperatively,16 whereas in our study recurrent endotheliitis presented earlier, between 3 and 11 months postoperatively. There have been few isolated case reports of CMV endotheliitis after keratoplasty.14,15,17–19 In our study, most patients with endotheliitis underwent DSAEK, as it is currently the most common corneal transplantation surgery that is performed for endothelial disease in our center. Of note, the few eyes that underwent PK (n ¼ 4) or DMEK (n ¼ 1) did not develop a recurrence of endotheliitis secondary to infection or graft rejection. We previously described a case series of CMV endotheliitis following DSAEK, and suggested that preoperative corneal decompensation might be attributable to undiagnosed bullous keratopathy secondary to CMVinduced endothelial damage, as preoperative aqueous PCR analysis was not performed in these cases.15 In this study, we excluded all subjects who did not have any
CORNEAL GRAFTS FOR ENDOTHELIITIS
263
FIGURE 1. Patient 2 with cytomegalovirus (CMV)-positive corneal endotheliitis. (Top left) Anterior segment photograph demonstrating medium-sized keratic precipitates (KPs), corneal edema, and low-grade anterior chamber (AC) inflammation at presentation; he was found to be CMV positive preoperatively. (Top right) Postoperative slit-lamp photograph 1 month after uncomplicated Descemet stripping automated endothelial keratoplasty (DSAEK). (Bottom left) After 3 months, he developed recurrent corneal endotheliitis postoperatively, leading to graft edema, with KPs and mild AC inflammation. (Bottom right) Slit-lamp photograph demonstrating resolution of endotheliitis and of inflammation, with a clear DSAEK graft 16 months after surgery.
preoperative aqueous PCR analysis; hence, only subjects who had aqueous samples tested for viruses were included. This allowed us to examine the difference between CMVpositive and CMV-negative eyes with corneal endotheliitis confirmed preoperatively. Of note, we had excluded 7 eyes that did not have preoperative AC taps (because of lack of consent or other reasons) but underwent corneal transplantation. Of these, 4 eyes subsequently developed recurrent endotheliitis with CMV DNA detected postoperatively within 1 year, and had successful treatment with antiCMV therapy as described above. This supports our opinion that a high index of suspicion of CMV infection is required, and that patients with a possible presentation of corneal endotheliitis should have preoperative AC taps to exclude a viral etiology, so that treatment and closer monitoring can be instituted postoperatively. Cytomegalovirus may be harbored latently in the anterior segment and, upon reactivation by the use of topical corticosteroids postoperatively following any surgical procedure, can clinically manifest as recurrence of the viral endotheliitis.20 Patients with CMV endotheliitis may present with focal or diffuse corneal edema, with a variety of clinical presentations, such as KPs in a linear pattern, immune-ring formation, or coin-shaped lesions.6,21 Our study results seem to support this, as patients who were CMV positive preoperatively were treated with antiCMV therapy and were rendered CMV negative before 264
corneal transplantation. Moreover, these patients clinically responded well to anti-CMV treatment, which coincided with reduction of CMV DNA on real-time PCR.22 As patients with high viral loads of CMV were found in a previous study of anterior uveitis to have greater endothelial cell damage,4 we aggressively treated with anti-CMV therapy these patients with recurrent endotheliitis following transplantation who were CMV positive. In our study, most patients’ corneal grafts remained clear with no additional recurrences of endotheliitis for a mean duration of 2 years postoperatively, with the exception of 1 eye that eventually suffered recurrent endotheliitis with graft failure despite anti-CMV therapy. The results of our study may have important implications on clinical practice. Patients with corneal decompensation secondary to presumed endotheliitis should be fully investigated before corneal transplantation to exclude CMV— which may actually preclude the necessity for transplantation if they subsequently recover after antiviral treatment. Patients whose aqueous samples are CMV positive preoperatively may benefit from anti-CMV treatment until ocular inflammation resolves and aqueous PCR analysis confirms that the eye is CMV negative before corneal transplantation. After corneal transplantation, these patients still require close monitoring, with a high index of suspicion for recurrence of CMV corneal endotheliitis. This usually occurs earlier postoperatively, before 12 months—which
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
FIGURE 2. Patient 4 with cytomegalovirus (CMV)-positive corneal endotheliitis. (Top left) Anterior segment photograph demonstrating the typical signs of corneal endotheliitis: mild anterior chamber inflammation, medium keratic precipitates, raised intraocular pressure, and corneal edema. (Top right) CMV DNA was detected on polymerase chain reaction preoperatively, and the eye developed significant corneal decompensation despite anti-CMV therapy. (Bottom left) After resolution of corneal endotheliitis, Descemet stripping automated endothelial keratoplasty was successfully performed. However, there was recurrence of corneal endotheliitis and inflammation at 10 months postoperatively. (Bottom right) The graft remained clear with resolution of inflammation after treatment.
may help to differentiate from corneal graft rejection, which usually occurs from 12-18 months. Nonetheless, we recommend that patients who suffer recurrence of endotheliitis post keratoplasty should have a repeat AC tap and treatment according to the results (ie, CMV-positive results with antiCMV treatment; CMV-negative results with topical corticosteroids). We recognize that despite anti-CMV treatment, 1 corneal graft still failed among the CMV-positive eyes. However, the risk factors for graft failure and the assessment of this current anti-CMV treatment regime in such cases are beyond the scope of this current study. We recognize the limitations of our study, which include its small sample size and retrospective nature. However, we present the first and largest series, to our knowledge, of outcomes in patients who have undergone keratoplasty specifically for corneal endotheliitis. Moreover, for reasons yet unclear, there is a higher incidence of CMV-positive corneal endotheliitis among Asian patients. We recognize the possibility of our donor corneas harboring CMV, as we
do not routinely screen donors for CMV and CMV serum– positive donors are allowed to donate corneas. However, these are in conformance with current eye-banking guidelines. We also did not detect any other viruses from the aqueous sampling in our study population, such as HSV or VZV—although we did not exclude patients with a history of herpes or zoster infections. In our case series, we found that Asian patients with corneal endotheliitis who underwent corneal transplantation were more likely to have a recurrence of endothelial inflammation if they were CMV positive preoperatively, despite successful anti-CMV treatment before surgery. Our study suggests that these patients with corneal endotheliitis may benefit from preoperative aqueous PCR analysis before corneal transplantation—and that there should be a high index of suspicion for recurrent CMV endotheliitis in patients who were CMV positive preoperatively, and who present with corneal decompensation within 1 year after corneal transplantation.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. The authors indicate no funding support. Contributions of authors: design and conduct of the study (M.A., C.S., S.P.C., D.T., J.S.M.); collection (M.A., C.S., S.P.C., D.T., J.S.M.), management (M.A., C.S., S.P.C., D.T., J.S.M.), analysis (M.A., C.S., S.P.C., D.T., J.S.M.), and interpretation of the data (M.A., C.S., S.P.C., D.T., J.S.M.); and preparation of the manuscript (M.A., C.S., S.P.C., D.T., J.S.M.). The authors acknowledge Lynette Lin Ean Oon, Senior Consultant, Department of Pathology, Singapore General Hospital; and Evelyn Siew-Chuan Koay, Director, Molecular Diagnosis Centre, National University Hospital, Singapore, for providing relevant laboratory data, and for contributions with regard to details on viral polymerase chain reaction techniques and interpretation of results.
VOL. 156, NO. 2
CORNEAL GRAFTS FOR ENDOTHELIITIS
265
REFERENCES 1. Khodadoust AA, Attarzadeh A. Presumed autoimmune corneal endotheliopathy. Am J Ophthalmol 1982;93(6): 718–722. 2. Chee SP, Jap A. Treatment outcome and risk factors for visual loss in Cytomegalovirus endotheliitis. Graefes Arch Clin Exp Ophthalmol 2012;250(3):383–389. 3. Dabil H, Boley ML, Schmitz TM, Van Gelder RN. Validation of a diagnostic multiplex polymerase chain reaction assay for infectious posterior uveitis. Arch Ophthalmol 2001;119(9): 1315–1322. 4. Miyanaga M, Sugita S, Shimizu N, et al. A significant association of viral loads with corneal endothelial cell damage in cytomegalovirus anterior uveitis. Br J Ophthalmol 2010; 94(3):336–340. 5. Carmichael A. Cytomegalovirus and the eye. Eye (Lond) 2012;26(2):237–240. 6. Koizumi N, Suzuki T, Uno T, et al. Cytomegalovirus as an etiologic factor in corneal endotheliitis. Ophthalmology 2008;115(2):292–297.e3. 7. Suzuki T, Ohashi Y. Corneal endotheliitis. Semin Ophthalmol 2008;23(4):235–240. 8. Tan DT, Janardhanan P, Zhou H, et al. Penetrating keratoplasty in Asian eyes: the Singapore Corneal Transplant Study. Ophthalmology 2008;115(6):975–982.e1. 9. Chee SP, Bacsal K, Jap A, Se-Thoe SY, Cheng CL, Tan BH. Corneal endotheliitis associated with evidence of cytomegalovirus infection. Ophthalmology 2007;114(4):798–803. 10. Boeckh M, Huang M, Ferrenberg J, et al. Optimization of quantitative detection of cytomegalovirus DNA in plasma by real-time PCR. J Clin Microbiol 2004;42(3):1142–1148. 11. Ang M, Htoon HM, Cajucom-Uy HY, Tan D, Mehta JS. Donor and surgical risk factors for primary graft failure following Descemet’s stripping automated endothelial keratoplasty in Asian eyes. Clin Ophthalmol 2011;5:1503–1508. 12. Mehta JS, Por YM, Poh R, Beuerman RW, Tan D. Comparison of donor insertion techniques for Descemet stripping
266
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
automated endothelial keratoplasty. Arch Ophthalmol 2008; 126(10):1383–1388. Design and methods of The Collaborative Corneal Transplantation Studies. The Collaborative Corneal Transplantation Studies Research Group. Cornea 1993;12(2):93–103. Chu HY, Sun CC, Chuang WY, et al. Cytomegalovirus associated corneal endotheliitis after penetrating keratoplasty in a patient with Fuchs corneal endothelial dystrophy. Br J Ophthalmol 2012;96(2):300–301. Anshu A, Chee SP, Mehta JS, Tan DT. Cytomegalovirus endotheliitis in Descemet stripping endothelial keratoplasty. Ophthalmology 2009;116(4):624–630. Li JY, Terry MA, Goshe J, Shamie N, Davis-Boozer D. Graft rejection after Descemet’s stripping automated endothelial keratoplasty: graft survival and endothelial cell loss. Ophthalmology 2012;119(1):90–94. Lusthaus JA, Kim P, Steller AK, et al. Successful corneal autograft after clearance of anterior chamber cytomegalovirus with oral valganciclovir in a patient with multiple failed corneal allografts. Cornea 2011;30(9):1054–1057. Sonoyama H, Araki-Sasaki K, Osakabe Y, et al. Detection of cytomegalovirus DNA from cytomegalovirus corneal endotheliitis after penetrating keratoplasty. Cornea 2010;29(6): 683–685. Suzuki T, Hara Y, Uno T, Ohashi Y. DNA of cytomegalovirus detected by PCR in aqueous of patient with corneal endotheliitis after penetrating keratoplasty. Cornea 2007;26(3):370–372. Zamir E, Stawell R, Jhanji V, Vajpayee RB. Corneal endotheliitis triggered by cataract surgery in a Chinese patient with cytomegalovirus anterior uveitis. Clin Experiment Ophthalmol 2011;39(9):913–915. Chee SP, Jap A. Immune ring formation associated with cytomegalovirus endotheliitis. Am J Ophthalmol 2011;152(3): 449–453.e1. Hwang YS, Shen CR, Chang SH, et al. The validity of clinical feature profiles for cytomegaloviral anterior segment infection. Graefes Arch Clin Exp Ophthalmol 2011;249(1): 103–110.
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
Biosketch Dr Marcus Ang is a Registrar at the Singapore National Eye Center. He attained a Masters in Clinical Investigation at the National University of Singapore. His research interests include Cornea and Refractive surgery, Ocular inflammation and Uveitis.
VOL. 156, NO. 2
CORNEAL GRAFTS FOR ENDOTHELIITIS
266.e1
Biosketch Assoc Prof. Jodhbir S. Mehta is Head of the Tissue Engineering and Stem Cell Group at the Singapore Eye Research Institute and Head of the Corneal Service and Senior Consultant in the Refractive Service of the Singapore National Eye Center. He has academic affiliations with Duke-National University of Singapore Graduate Medical School. He has won 24 awards including those at the American Academy of Ophthalmology, and the Association for Research in Vision and Ophthalmology.
266.e2
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
To describe outcomes of corneal transplantation for irreversible corneal decompensation from corneal endotheliitis in Asian eyes. DESIGN: Retrospective, observational case series. METHODS: We reviewed consecutive patients with corneal endotheliitis (32 eyes of 31 subjects) who underwent keratoplasty (January 1, 2008-December 1, 2009). All eyes had preoperative aqueous polymerase chain reaction (PCR) analysis for viruses, including cytomegalovirus (CMV). CMV-positive patients were treated preoperatively with topical corticosteroids and antiCMV treatment (oral valganciclovir 900 mg twice daily, topical ganciclovir 0.15% 5 applications per day, for 6 weeks) with complete resolution of ocular inflammation, and quiescence for at least 6 months before corneal transplantation. Our main outcome measure was recurrence of endotheliitis within 1 year after corneal transplantation. RESULTS: Five eyes were CMV positive; the remaining 27 eyes were negative for all viruses on PCR analysis. CMV-positive patients had a higher rate of recurrence of endotheliitis within 1 year after corneal transplantation, compared with CMV-negative eyes (60% vs 7.4%, P [ .01). The CMV-positive eyes had recurrent endotheliitis at a median of 10 months (range 3-11 months) after corneal transplantation. After successful antiCMV treatment, all 5 CMV-positive eyes then continued to have clear grafts for a median duration of 21 months (range 13-44 months). CONCLUSION: Our study suggests that Asian patients with corneal endotheliitis may benefit from preoperative aqueous PCR analysis before corneal transplantation. Such patients were more likely to have a recurrence of endothelial inflammation if they were CMV positive preoperatively, despite successful anti-CMV treatment before surgery. (Am J Ophthalmol 2013;156: 260–266. Ó 2013 by Elsevier Inc. All rights reserved.) Accepted for publication Mar 14, 2013. From Singapore National Eye Centre (M.A., C.C.A.S., S.P.C., D.T.H.T., J.S.M.); Singapore Eye Research Institute (M.A., C.C.A.S., S.P.C., D.T.H.T., J.S.M.); Department of Ophthalmology, National University Health System (C.C.A.S., S.P.C., D.T.H.T., J.S.M.); and Duke-National University of Singapore Graduate Medical School (D.T.H.T., J.S.M.), Singapore. Inquiries to A/Professor Jodhbir S. Mehta, Singapore National Eye Centre, 11 Third Hospital Avenue, Singapore 168751; e-mail: [email protected]
260
Ó
2013 BY
C
ORNEAL ENDOTHELIITIS IS AN ENTITY CHARACTER-
ized by localized corneal edema with the presence of mild iritis and keratic precipitates (KPs).1 Although corneal endotheliitis was once considered ‘‘autoimmune’’ or ‘‘idiopathic,’’ there have been increasing reports that viruses such as cytomegalovirus (CMV) and herpes simplex virus (HSV) may play a role as etiologic agents.2 Polymerase chain reaction (PCR) analysis of aqueous humor has been shown to be capable of detecting the virus3; however, real-time PCR and quantification of viral DNA may be useful in distinguishing an etiologic virus, as opposed to a ‘‘bystander’’ virus (the latter does not play a causative role in inciting inflammation or endotheliitis).4 A viral etiology in corneal endotheliitis is further supported by reports that endothelial lesions and inflammation resolve rapidly with the application of topically applied and systemic antiviral treatment.2 Cytomegalovirus is increasingly recognized as the most common virus implicated in corneal endotheliitis, especially in Asian patients.5 Although the exact pathogenesis of CMV corneal endotheliitis is not yet understood, it is postulated that anterior chamber–associated immune deviation (ACAID) normally prevents the proliferation of CMV within the anterior chamber (AC), leading to viral infection of the endothelium.6 However, it remains unclear whether CMV endotheliitis is attributable to a primary CMV infection or a local reactivation of CMV in the anterior segment secondary to dendritic cell activation provoked by a separate trigger, seen in other parts of the body.5 Although medical therapy remains the mainstay of treatment, persistent corneal edema or irreversible corneal decompensation often ensues because of the cytopathic effect on corneal endothelial cells, leading to the need for a corneal graft.7 In this study, we describe the outcome of various forms of corneal transplantation, but mainly Descemet stripping automated endothelial keratoplasty (DSAEK) performed in eyes that presented with endotheliitis and corneal decompensation. We also compared the outcomes between eyes that were ‘‘CMV positive’’ and ‘‘CMV negative,’’ as assessed using real-time PCR preoperatively.
METHODS STUDY COHORT:
consecutive
ELSEVIER INC. ALL
We conducted a retrospective review of patients who presented with corneal
RIGHTS RESERVED.
0002-9394/$36.00 http://dx.doi.org/10.1016/j.ajo.2013.03.020
endotheliitis (ie, localized corneal edema with mild AC inflammation and KPs1) who subsequently underwent corneal transplantation. All patients underwent corneal transplantation at the Singapore National Eye Centre (SNEC) between January 1, 2008 and December 1, 2009. We obtained clinical data from our ongoing Singapore Corneal Transplant Study cohort, an audited longitudinal prospective study that contains preoperative, intraoperative, and yearly postoperative follow-up clinical data on all corneal transplants performed by corneal surgeons at the SNEC.8 We conducted this study with approval from our institution’s Singapore Health Services Institutional Review Board and followed the principles of the Declaration of Helsinki. INVESTIGATIONS:
At presentation, all eyes with suspected corneal endotheliitis with localized corneal edema, KPs, and mild AC inflammation had an AC tap performed preoperatively at the slit lamp using an aseptic technique. A minimum of 100 mL of aqueous was obtained and the samples were either delivered to the laboratory at 18 C-25 C within 1 hour, or stored and transported at 2 C-4 C within 24 hours. The samples were aliquoted immediately upon arrival at the laboratory and were either used for DNA extraction immediately or kept at 20 C and extracted for DNA within 1 week. The DNA was extracted using Qiagen QIAmp DNA Extraction Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions and tested for CMV DNA by PCR. The aqueous samples were sent for a validated, nested multiplex PCR for HSV, varicella zoster virus (VZV), CMV, and Toxoplasma gondii DNA as previously described.3 The lower limits of detection for HSV-1, HSV-2, CMV, VZV, and Toxoplasma gondii are 360, 120, 50, 40, and 3 copies per PCR reaction, respectively, with a sensitivity exceeding 95% for all organisms.9 Patients who were CMV positive were also analyzed for CMV DNA using real-time PCR with a sensitivity of 7 copies/mL with a 95% confidence interval (CI).10 All patients with a positive aqueous CMV DNA also underwent tests for CMV antigen (cytospin), CMV antibodies, and human immunodeficiency virus (HIV) antibodies in the serum. MANAGEMENT:
All patients with endotheliitis were co-managed by the SNEC Cornea Service and the SNEC Ocular Inflammation and Immunology Service. Patients included in this study underwent either a penetrating keratoplasty (PK) (4 eyes), DSAEK (27 eyes), or Descemet membrane endothelial keratoplasty (DMEK) (1 eye) using techniques as previously described.11,12 Preoperatively, patients were treated with topical corticosteroids and topical antiglaucoma medications if intraocular pressure (IOP) was raised. Patients who were CMV positive were co-managed with the infectious disease physician at the Singapore General Hospital and treated with oral valganciclovir (900 mg) twice daily for 6 weeks followed by 450 mg twice daily for a further 6 weeks. Topical ganciclovir
VOL. 156, NO. 2
ophthalmic gel 0.15% (Virgan; Laboratoires The´a, Clermont-Ferrand, France) was also prescribed 5 times a day. A week before completion of therapy, a repeat aqueous analysis for CMV using real-time PCR was done. Quiescent ocular inflammation for at least 6 months and undetectable CMV DNA titer with repeat aqueous PCR analysis for CMV-positive eyes were conditions required before corneal transplantation was performed. Our main outcome measure was a recurrence of endotheliitis, presenting clinically as localized corneal edema with AC inflammation and KPs. Recurrence was also suspected when patients presented with an unexplained sudden decrease in endothelial cell count (performed routinely at 6 months, then yearly postoperatively) in an uninflamed eye and normal graft appearance or intraocular inflammation that failed to clear or worsened with conventional high-dose corticosteroid treatment regimens.13 Postoperatively, visual acuity (VA), IOP, graft clarity, and anterior segment inflammation were monitored, with a high index of suspicion for any recurrence of endotheliitis. We used similar postoperative topical corticosteroid medication regimens in all patients: a single drop of prednisolone acetate 1% every 3 hours for 1 week, 3 times a day for 6 months, twice daily for 3 months, then once a day for up to 1 year. All patients received topical levofloxacin 0.5% every 3 hours for 1 week, followed by 3 times a day for 6 months. Patients who suffered from recurrent endotheliitis postoperatively underwent a repeat aqueous tap, and the aqueous sample was reanalyzed with the tetraplex PCR as described above, with real-time PCR analysis for CMV-positive eyes. Patients who were CMV positive were restarted on topical ganciclovir ophthalmic gel 0.15% every 3 hours and were watched closely for clinical resolution of endotheliitis. Patients who were negative for all viruses tested on PCR analysis were presumed to have immunologic rejection and were treated with only topical prednisolone acetate 1% every 3 hours. STATISTICAL ANALYSIS: Statistical analysis included descriptive statistics, where the mean and standard deviation (mean 6 SD) were calculated for the continuous variables, while frequency distribution and percentages were used for categorical variables. Comparisons between categorical variables were conducted using x2/Fisher exact tests as appropriate. The independent t test and 1-way analysis of variance test were used for continuous measures group comparisons. P value <.05 was considered statistically significant. All analyses were performed using STATA version 11 (StataCorp LP, College Station, Texas, USA).
RESULTS A TOTAL OF 32 EYES OF 31 PATIENTS UNDERWENT CORNEAL
grafts for irreversible corneal decompensation secondary to
CORNEAL GRAFTS FOR ENDOTHELIITIS
261
endotheliitis. Five eyes were found to have evidence of CMV from aqueous PCR analysis preoperatively. The remaining 27 eyes did not have any evidence of CMV or any other viruses tested by the nested multiplex PCR— and all of these eyes were not found to be CMV positive during the study period. All patients were found to be immunocompetent and negative for CMV antigen and antibody tests as described above. Table 1 compares the clinical characteristics between these CMV-positive and CMV-negative eyes. CYTOMEGALOVIRUS-POSITIVE EYES: Patients who were CMV positive were all successfully treated with anti-CMV treatment as described above, until ocular inflammation was quiescent for at least 6 months with an undetectable CMV DNA titer, with repeat aqueous PCR analysis before corneal transplantation. Postoperatively, a significantly higher proportion of patients who were CMV positive had recurrence of endotheliitis after corneal transplantation within 1 year postoperatively (3/5, 60%), as compared with those who were CMV negative (2/27, 7.4%) (P ¼ .01). However, there was no significant difference in terms of graft failure between CMV-negative and CMV-positive eyes (3.7% vs 0%, respectively, P ¼ .65). Visual outcomes at 1 year (best-corrected visual acuity 20/40 or better) were also comparable between CMV-positive (3/5) and CMVnegative eyes (13/27) (P ¼ .84). The 3 CMV-positive eyes presenting with recurrence, at a median of 10 months (range 3-11 months) after corneal transplantation, were seen to develop low-grade AC inflammation, development of new keratic precipitates on the graft endothelium, and reduction in graft clarity. A repeat aqueous sample was obtained in these eyes, which demonstrated the presence of CMV DNA on real-time PCR analysis (Table 2). Figure 1 illustrates the clinical presentation of Patient 2. Preoperatively, this patient presented with clinical signs of corneal endotheliitis, with fine to medium KPs, corneal edema, and low-grade AC inflammation, and was also noted to be CMV positive (Figure 1, Top left). He underwent an uncomplicated DSAEK (Figure 1, Top right) but developed recurrent corneal endotheliitis 3 months after the surgery, leading to graft edema, and the presence of KPs and mild AC inflammation (Figure 1, Bottom left). After institution of topical ganciclovir and oral valganciclovir, the endotheliitis and inflammation resolved and the graft remained clear (Figure 1, Bottom right). Similarly, Patient 4 (Figure 2) was CMV positive preoperatively (Figure 2, Top left) and developed significant corneal decompensation (Figure 2, Top right) despite anti-CMV therapy. DSAEK was successfully performed, but the patient suffered recurrence of corneal endotheliitis and inflammation postoperatively (Figure 2, Bottom left) and required reinstitution of anti-CMV therapy (Figure 2, Bottom right). The graft remained clear with resolution of inflammation after treatment.
262
TABLE 1. Baseline Demographics of Subjects With Corneal Endotheliitis Comparing Cytomegalovirus-Positive and Cytomegalovirus-Negative Eyes
Mean age (years 6 SD) Sex (male) Mean follow-up (months 6 SD) Race Chinese Malay Indian Other Previous surgery Trabeculectomy Tube Pseudophakic Type of graft DSAEK PK DMEK
Cytomegalovirus Negative (n ¼ 27)
Cytomegalovirus Positive (n ¼ 5)
P Valuea
65 6 12 15 18 6 11
62 6 5 5 25 6 6
.53 .13 .16
18 5 1 3
5 0 0 0
.78
4 0 15
2 1 4
.23 .16 .63
23 3 1
4 1 0
.86
DMEK ¼ Descemet membrane endothelial keratoplasty; DSAEK ¼ Descemet stripping automated endothelial keratoplasty; PK ¼ penetrating keratoplasty. a 2 x /Fisher exact tests as appropriate.
All patients with recurrent corneal endotheliitis and found to be CMV positive were restarted on oral ganciclovir and topical valganciclovir and monitored closely for further signs of inflammation (as stated in Methods section above). After successful treatment, all 5 CMVpositive eyes had continued to have clear grafts for a median duration of 21 months (range 13-44 months). However, 1 eye (Patient 5) developed late graft failure after 18 months and required a repeat DSAEK. This repeat graft remained clear for 6 months, then suffered a recurrence of CMV endotheliitis (CMV real-time PCR DNA count: 3.5-7.5 3 106 viral copies) and eventually failed 14 months later, despite oral valganciclovir and topical ganciclovir therapy. In the remaining patients, repeat AC taps for real-time PCR analysis revealed reduction to undetectable CMV levels after 6 months of therapy. CYTOMEGALOVIRUS-NEGATIVE EYES: Only 2 of 27 CMV-negative eyes (7.4%) developed an episode of inflammation after corneal transplantation within 1 year (at 3-6 months); these eyes were CMV negative on repeat aqueous sampling and were treated successfully with increased frequency of topical prednisolone acetate 1% for 3 months, leading to a resolution of the inflammation. Of note, we did not observe any increased incidence of de novo glaucoma between CMV-positive and
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
923 6.0 3 106 11 1.6 3 106 20/160 71/Chinese/male 5
BCVA ¼ best-corrected visual acuity; CMV ¼ cytomegalovirus; DSAEK ¼ Descemet stripping automated endothelial keratoplasty; ECC ¼ endothelial cell count; HM ¼ hand motions; PCR ¼ polymerase chain reaction analysis; PK ¼ penetrating keratoplasty.
HM Graft failure/DSAEK 32
20/25 20/30 20/25 20/40 Clear graft/PK Clear graft/DSAEK Clear graft/DSAEK Clear graft/DSAEK 58/Chinese/male 60/Chinese/male (right eye) 60/Chinese/male (left eye) 61/Chinese/male 1 2 3 4
20/2000 20/60 20/400 20/400
NA 3 NA 10
Nondetectable 1.8 3 106 Nondetectable 1.0 3 103
2817 3067 2538 1318
Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Topical ganciclovir gel 0.15% Oral valganciclovir 900 mg once daily Topical ganciclovir gel 0.15% 7.7 3 103 770 1.8 3 106 1.1 3 103
21 16 13 44
of Graft (Months) (1 Year) PCR (Copies) Recurrence (Months) Postoperative Treatment Demographics (Age/Ethnicity/Sex) Patient
BCVA
CMV PCR
Final
BCVA
Final Outcome/Type Follow-up Duration ECC Postoperative Real-Time CMV Time to Corneal Endotheliitis Preoperative Real-Time Preoperative
TABLE 2. Outcomes of Patients With Cytomegalovirus-Positive Corneal Endotheliitis who Underwent Corneal Transplantation
VOL. 156, NO. 2
CMV-negative eyes 1 year after corneal transplantation (0/5 vs 2/27, P ¼ .65).
DISCUSSION IT IS OFTEN DIFFICULT TO DISTINGUISH CLINICALLY
between endotheliitis attributable to immunologic rejection or viral infection in patients who have undergone corneal transplantation.14 Both conditions present with the same nonspecific, heterogeneous clinical signs, which include any combination of mild AC inflammation, KPs that may be linear in orientation, localized endothelial decompensation, unexplained reduction in endothelial cell count, or raised IOP.15 Hence, in our center, all patients who presented with these nonspecific signs underwent an aqueous tap to exclude viral endotheliitis before they were presumptively treated for immunologic rejection. It is particularly important to differentiate viral endotheliitis from graft rejection because the treatment for allograft rejection involves intensive topical and often systemic immunosuppression—which can exacerbate latent viral infections such as CMV and cause increased inflammation with subsequent graft failure.15 In this study, we found that patients with previously diagnosed CMVpositive endotheliitis are more likely to have a recurrence of endotheliitis within a year following corneal transplantation—despite preoperative anti-CMV treatment and evidence of successful eradication of CMV and ocular inflammation before corneal transplantation was performed. This is in contrast to patients with corneal endotheliitis in the absence of detectable CMV preoperatively, who are less likely to develop signs of endotheliitis postoperatively even while receiving similar postoperative topical corticosteroid therapy. It has been suggested that endotheliitis secondary to graft rejection in DSAEK usually occurs between 12 and 18 months postoperatively,16 whereas in our study recurrent endotheliitis presented earlier, between 3 and 11 months postoperatively. There have been few isolated case reports of CMV endotheliitis after keratoplasty.14,15,17–19 In our study, most patients with endotheliitis underwent DSAEK, as it is currently the most common corneal transplantation surgery that is performed for endothelial disease in our center. Of note, the few eyes that underwent PK (n ¼ 4) or DMEK (n ¼ 1) did not develop a recurrence of endotheliitis secondary to infection or graft rejection. We previously described a case series of CMV endotheliitis following DSAEK, and suggested that preoperative corneal decompensation might be attributable to undiagnosed bullous keratopathy secondary to CMVinduced endothelial damage, as preoperative aqueous PCR analysis was not performed in these cases.15 In this study, we excluded all subjects who did not have any
CORNEAL GRAFTS FOR ENDOTHELIITIS
263
FIGURE 1. Patient 2 with cytomegalovirus (CMV)-positive corneal endotheliitis. (Top left) Anterior segment photograph demonstrating medium-sized keratic precipitates (KPs), corneal edema, and low-grade anterior chamber (AC) inflammation at presentation; he was found to be CMV positive preoperatively. (Top right) Postoperative slit-lamp photograph 1 month after uncomplicated Descemet stripping automated endothelial keratoplasty (DSAEK). (Bottom left) After 3 months, he developed recurrent corneal endotheliitis postoperatively, leading to graft edema, with KPs and mild AC inflammation. (Bottom right) Slit-lamp photograph demonstrating resolution of endotheliitis and of inflammation, with a clear DSAEK graft 16 months after surgery.
preoperative aqueous PCR analysis; hence, only subjects who had aqueous samples tested for viruses were included. This allowed us to examine the difference between CMVpositive and CMV-negative eyes with corneal endotheliitis confirmed preoperatively. Of note, we had excluded 7 eyes that did not have preoperative AC taps (because of lack of consent or other reasons) but underwent corneal transplantation. Of these, 4 eyes subsequently developed recurrent endotheliitis with CMV DNA detected postoperatively within 1 year, and had successful treatment with antiCMV therapy as described above. This supports our opinion that a high index of suspicion of CMV infection is required, and that patients with a possible presentation of corneal endotheliitis should have preoperative AC taps to exclude a viral etiology, so that treatment and closer monitoring can be instituted postoperatively. Cytomegalovirus may be harbored latently in the anterior segment and, upon reactivation by the use of topical corticosteroids postoperatively following any surgical procedure, can clinically manifest as recurrence of the viral endotheliitis.20 Patients with CMV endotheliitis may present with focal or diffuse corneal edema, with a variety of clinical presentations, such as KPs in a linear pattern, immune-ring formation, or coin-shaped lesions.6,21 Our study results seem to support this, as patients who were CMV positive preoperatively were treated with antiCMV therapy and were rendered CMV negative before 264
corneal transplantation. Moreover, these patients clinically responded well to anti-CMV treatment, which coincided with reduction of CMV DNA on real-time PCR.22 As patients with high viral loads of CMV were found in a previous study of anterior uveitis to have greater endothelial cell damage,4 we aggressively treated with anti-CMV therapy these patients with recurrent endotheliitis following transplantation who were CMV positive. In our study, most patients’ corneal grafts remained clear with no additional recurrences of endotheliitis for a mean duration of 2 years postoperatively, with the exception of 1 eye that eventually suffered recurrent endotheliitis with graft failure despite anti-CMV therapy. The results of our study may have important implications on clinical practice. Patients with corneal decompensation secondary to presumed endotheliitis should be fully investigated before corneal transplantation to exclude CMV— which may actually preclude the necessity for transplantation if they subsequently recover after antiviral treatment. Patients whose aqueous samples are CMV positive preoperatively may benefit from anti-CMV treatment until ocular inflammation resolves and aqueous PCR analysis confirms that the eye is CMV negative before corneal transplantation. After corneal transplantation, these patients still require close monitoring, with a high index of suspicion for recurrence of CMV corneal endotheliitis. This usually occurs earlier postoperatively, before 12 months—which
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
FIGURE 2. Patient 4 with cytomegalovirus (CMV)-positive corneal endotheliitis. (Top left) Anterior segment photograph demonstrating the typical signs of corneal endotheliitis: mild anterior chamber inflammation, medium keratic precipitates, raised intraocular pressure, and corneal edema. (Top right) CMV DNA was detected on polymerase chain reaction preoperatively, and the eye developed significant corneal decompensation despite anti-CMV therapy. (Bottom left) After resolution of corneal endotheliitis, Descemet stripping automated endothelial keratoplasty was successfully performed. However, there was recurrence of corneal endotheliitis and inflammation at 10 months postoperatively. (Bottom right) The graft remained clear with resolution of inflammation after treatment.
may help to differentiate from corneal graft rejection, which usually occurs from 12-18 months. Nonetheless, we recommend that patients who suffer recurrence of endotheliitis post keratoplasty should have a repeat AC tap and treatment according to the results (ie, CMV-positive results with antiCMV treatment; CMV-negative results with topical corticosteroids). We recognize that despite anti-CMV treatment, 1 corneal graft still failed among the CMV-positive eyes. However, the risk factors for graft failure and the assessment of this current anti-CMV treatment regime in such cases are beyond the scope of this current study. We recognize the limitations of our study, which include its small sample size and retrospective nature. However, we present the first and largest series, to our knowledge, of outcomes in patients who have undergone keratoplasty specifically for corneal endotheliitis. Moreover, for reasons yet unclear, there is a higher incidence of CMV-positive corneal endotheliitis among Asian patients. We recognize the possibility of our donor corneas harboring CMV, as we
do not routinely screen donors for CMV and CMV serum– positive donors are allowed to donate corneas. However, these are in conformance with current eye-banking guidelines. We also did not detect any other viruses from the aqueous sampling in our study population, such as HSV or VZV—although we did not exclude patients with a history of herpes or zoster infections. In our case series, we found that Asian patients with corneal endotheliitis who underwent corneal transplantation were more likely to have a recurrence of endothelial inflammation if they were CMV positive preoperatively, despite successful anti-CMV treatment before surgery. Our study suggests that these patients with corneal endotheliitis may benefit from preoperative aqueous PCR analysis before corneal transplantation—and that there should be a high index of suspicion for recurrent CMV endotheliitis in patients who were CMV positive preoperatively, and who present with corneal decompensation within 1 year after corneal transplantation.
ALL AUTHORS HAVE COMPLETED AND SUBMITTED THE ICMJE FORM FOR DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST and none were reported. The authors indicate no funding support. Contributions of authors: design and conduct of the study (M.A., C.S., S.P.C., D.T., J.S.M.); collection (M.A., C.S., S.P.C., D.T., J.S.M.), management (M.A., C.S., S.P.C., D.T., J.S.M.), analysis (M.A., C.S., S.P.C., D.T., J.S.M.), and interpretation of the data (M.A., C.S., S.P.C., D.T., J.S.M.); and preparation of the manuscript (M.A., C.S., S.P.C., D.T., J.S.M.). The authors acknowledge Lynette Lin Ean Oon, Senior Consultant, Department of Pathology, Singapore General Hospital; and Evelyn Siew-Chuan Koay, Director, Molecular Diagnosis Centre, National University Hospital, Singapore, for providing relevant laboratory data, and for contributions with regard to details on viral polymerase chain reaction techniques and interpretation of results.
VOL. 156, NO. 2
CORNEAL GRAFTS FOR ENDOTHELIITIS
265
REFERENCES 1. Khodadoust AA, Attarzadeh A. Presumed autoimmune corneal endotheliopathy. Am J Ophthalmol 1982;93(6): 718–722. 2. Chee SP, Jap A. Treatment outcome and risk factors for visual loss in Cytomegalovirus endotheliitis. Graefes Arch Clin Exp Ophthalmol 2012;250(3):383–389. 3. Dabil H, Boley ML, Schmitz TM, Van Gelder RN. Validation of a diagnostic multiplex polymerase chain reaction assay for infectious posterior uveitis. Arch Ophthalmol 2001;119(9): 1315–1322. 4. Miyanaga M, Sugita S, Shimizu N, et al. A significant association of viral loads with corneal endothelial cell damage in cytomegalovirus anterior uveitis. Br J Ophthalmol 2010; 94(3):336–340. 5. Carmichael A. Cytomegalovirus and the eye. Eye (Lond) 2012;26(2):237–240. 6. Koizumi N, Suzuki T, Uno T, et al. Cytomegalovirus as an etiologic factor in corneal endotheliitis. Ophthalmology 2008;115(2):292–297.e3. 7. Suzuki T, Ohashi Y. Corneal endotheliitis. Semin Ophthalmol 2008;23(4):235–240. 8. Tan DT, Janardhanan P, Zhou H, et al. Penetrating keratoplasty in Asian eyes: the Singapore Corneal Transplant Study. Ophthalmology 2008;115(6):975–982.e1. 9. Chee SP, Bacsal K, Jap A, Se-Thoe SY, Cheng CL, Tan BH. Corneal endotheliitis associated with evidence of cytomegalovirus infection. Ophthalmology 2007;114(4):798–803. 10. Boeckh M, Huang M, Ferrenberg J, et al. Optimization of quantitative detection of cytomegalovirus DNA in plasma by real-time PCR. J Clin Microbiol 2004;42(3):1142–1148. 11. Ang M, Htoon HM, Cajucom-Uy HY, Tan D, Mehta JS. Donor and surgical risk factors for primary graft failure following Descemet’s stripping automated endothelial keratoplasty in Asian eyes. Clin Ophthalmol 2011;5:1503–1508. 12. Mehta JS, Por YM, Poh R, Beuerman RW, Tan D. Comparison of donor insertion techniques for Descemet stripping
266
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
automated endothelial keratoplasty. Arch Ophthalmol 2008; 126(10):1383–1388. Design and methods of The Collaborative Corneal Transplantation Studies. The Collaborative Corneal Transplantation Studies Research Group. Cornea 1993;12(2):93–103. Chu HY, Sun CC, Chuang WY, et al. Cytomegalovirus associated corneal endotheliitis after penetrating keratoplasty in a patient with Fuchs corneal endothelial dystrophy. Br J Ophthalmol 2012;96(2):300–301. Anshu A, Chee SP, Mehta JS, Tan DT. Cytomegalovirus endotheliitis in Descemet stripping endothelial keratoplasty. Ophthalmology 2009;116(4):624–630. Li JY, Terry MA, Goshe J, Shamie N, Davis-Boozer D. Graft rejection after Descemet’s stripping automated endothelial keratoplasty: graft survival and endothelial cell loss. Ophthalmology 2012;119(1):90–94. Lusthaus JA, Kim P, Steller AK, et al. Successful corneal autograft after clearance of anterior chamber cytomegalovirus with oral valganciclovir in a patient with multiple failed corneal allografts. Cornea 2011;30(9):1054–1057. Sonoyama H, Araki-Sasaki K, Osakabe Y, et al. Detection of cytomegalovirus DNA from cytomegalovirus corneal endotheliitis after penetrating keratoplasty. Cornea 2010;29(6): 683–685. Suzuki T, Hara Y, Uno T, Ohashi Y. DNA of cytomegalovirus detected by PCR in aqueous of patient with corneal endotheliitis after penetrating keratoplasty. Cornea 2007;26(3):370–372. Zamir E, Stawell R, Jhanji V, Vajpayee RB. Corneal endotheliitis triggered by cataract surgery in a Chinese patient with cytomegalovirus anterior uveitis. Clin Experiment Ophthalmol 2011;39(9):913–915. Chee SP, Jap A. Immune ring formation associated with cytomegalovirus endotheliitis. Am J Ophthalmol 2011;152(3): 449–453.e1. Hwang YS, Shen CR, Chang SH, et al. The validity of clinical feature profiles for cytomegaloviral anterior segment infection. Graefes Arch Clin Exp Ophthalmol 2011;249(1): 103–110.
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013
Biosketch Dr Marcus Ang is a Registrar at the Singapore National Eye Center. He attained a Masters in Clinical Investigation at the National University of Singapore. His research interests include Cornea and Refractive surgery, Ocular inflammation and Uveitis.
VOL. 156, NO. 2
CORNEAL GRAFTS FOR ENDOTHELIITIS
266.e1
Biosketch Assoc Prof. Jodhbir S. Mehta is Head of the Tissue Engineering and Stem Cell Group at the Singapore Eye Research Institute and Head of the Corneal Service and Senior Consultant in the Refractive Service of the Singapore National Eye Center. He has academic affiliations with Duke-National University of Singapore Graduate Medical School. He has won 24 awards including those at the American Academy of Ophthalmology, and the Association for Research in Vision and Ophthalmology.
266.e2
AMERICAN JOURNAL OF OPHTHALMOLOGY
AUGUST 2013