CASE REPORT
Intrastromal Antifungal Injection With Secondary Lamellar Interface Infusion for Late-Onset Infectious Keratitis After DSAEK Elmer Y. Tu, MD, and Joshua Hou, MD
Purpose: The aim of this study was to report the successful medical management of 2 cases of late-onset endothelial keratoplasty–related stromal interface infections.
Methods: All cases of endothelial keratoplasty–related infections treated with intrastromal antifungal injections were compiled. The following information was collected: demographic data, surgical indications, donor rim cultures, donor mate outcomes, clinical course, diagnostic tests, and clinical outcome.
Results: Two cases of interface fungal keratitis diagnosed on clinical appearance and confocal microscopy were identified. Both patients refused to undergo further surgery and failed systemic and/ or topical therapy. Each received 3 to 4 intrastromal injections, with secondary infusion into the graft–host interface, which resulted in the complete involution of their interface opacities. Conclusions: Intrastromal antifungal injection may be an effective alternative to surgical intervention in late-onset fungal Descemet stripping endothelial keratoplasty interface keratitis. Early treatment may preserve graft viability and result in a good visual outcome without the need for either penetrating keratoplasty or potential pathogen exposure to the anterior chamber. Key Words: lamellar keratoplasty, infection, fungal, cornea, lamellar interface, amphotericin B, voriconazole, candida, yeast, endothelial keratoplasty
complications challenge the DSAEK surgeon, including graft dislocation, graft inversion, epithelial ingrowth, and optical interface irregularities. Another unique complication of all forms of lamellar keratoplasty is the development of late-onset interface infectious keratitis.1–4 The deep stromal location of the infiltrate limits access for microbiologic testing and reduces achievable drug levels at the site of infection. Although significant controversy exists regarding the value and form of prophylactic antifungal therapy,5 almost all established interface infections have failed systemic and/or topical antifungal therapy most commonly requiring penetrating keratoplasty or even leading to the loss of the eye.3,4,6–8 We present an alternative medical therapeutic technique that can avoid the longand short-term risks of urgent keratoplasty and preserve a potentially viable endothelial graft.
MATERIALS AND METHODS All cases of presumed infectious interface keratitis after endothelial keratoplasty seen at the University of Illinois Eye and Ear Infirmary (IEEI) were identified. Cases were reviewed with respect to demographic data, surgical indications, donor rim cultures, clinical course, diagnostic tests, and clinical outcome. Donor mate outcomes were obtained from the source eye bank.
(Cornea 2014;33:990–993)
D
escemet stripping endothelial keratoplasty (DSAEK) has largely supplanted penetrating keratoplasty as the procedure of choice in patients with primarily endothelial pathology such as Fuchs endothelial dystrophy or pseudophakic bullous keratopathy. DSAEK offers significant advantages in almost every aspect of transplant-related complications including intraoperative safety, late wound dehiscence, rejection, and speed of visual recovery. However, several unique
Received for publication April 22, 2014; revision received May 17, 2014; accepted May 20, 2014. Published online ahead of print July 22, 2014. From the Department of Ophthalmology and Visual Sciences, University of Illinois Eye and Ear Infirmary, Chicago, IL. The authors have no funding or conflicts of interest to disclose. Reprints: Elmer Y. Tu, MD, Cornea and External Disease Service, Department of Ophthalmology and Visual Sciences, University of Illinois College of Medicine, Chicago, IL 60612 (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
990
| www.corneajrnl.com
RESULTS Case Reports Patient 1 A 66-year-old man presented from an outside institution with a history of increasing intraocular inflammation 3 months after an uncomplicated DSAEK and formation of 2 discrete interface opacities (Figs. 1A, B). Topical difluprednate given twice daily had been unsuccessful in controlling his iritis and had been discontinued 2 days before presentation. The patient remained only on brimonidine/timolol combination drops. This was his third endothelial keratoplasty after a complicated cataract extraction 2 years earlier. On examination, his vision was found to have decreased from a peak postoperative value of 20/50 to 20/150 on presentation at the University of Illinois at Chicago. His vision was 20/60 while on topical corticosteroids just 2 days before. Intraocular pressure was 19, and his examination showed perilimbal injection, diffuse corneal edema, diffuse keratic precipitates, 3+ anterior chamber cells, and Cornea Volume 33, Number 9, September 2014
Cornea Volume 33, Number 9, September 2014
Intrastromal Antifungal Injection
FIGURE 1. A, External photograph of patient 1 showing 2 peripheral interlamellar infiltrates (arrows), diffuse corneal edema, and keratic precipitates. B, Slit-beam photograph of the 2 interlamellar infiltrates. C, Confocal microscopy showing hyperreflective, round, budding-like structures in the lamellar interface suggestive of the presence of yeast elements. D, Anterior segment ocular coherence tomography showing separation of the lamellar interface with an intact peripheral graft–host junction immediately after an intrastromal amphotericin B injection was given. E, Anterior segment ocular coherence tomography showing the resolution of the paracentral graft detachment 1 day after the injection. The interface lesion appears hyperreflective in the peripheral lamellar graft interface. F, External photograph taken 2 months after the initiation of treatment, which included 2 injections of intrastromal with secondary interlamellar infusion of amphotericin B and 1 injection of voriconazole. There is resolution of the more peripheral interface lesion and scarring/involution of the more central lesion. There was persistent corneal edema despite the resolution of the KP. G, Slit-beam photograph of the involuting interface lesion. The needle track is still visible below the lesion.
2 round interface plaques in the inferior temporal graft–host interface (Figs. 1A, B). The patient was initially given oral acyclovir for iritis of an unknown origin. At follow-up 4 days later, the patient’s vision had decreased further to 20/500. Confocal microscopy (Corneal Module; HRT Heidelberg, Germany) showed lesions suggestive of budding/paired yeast near the clinically localized interface lesions (Fig. 1C). A penetrating keratoplasty was recommended, but the patient refused further surgery. Because of cost, the patient did not receive prescribed oral voriconazole and continued to demonstrate crescendo inflammation and decreasing vision. After informed consent was obtained, an intrastromal injection of amphotericin B (5 mg/0.1 mL) was given in the corneal stroma anterior to the Ó 2014 Lippincott Williams & Wilkins
surgical intralamellar space on day 6, which separated the DSAEK interface without disturbing the peripheral graft– host junction (Fig. 1D). Oral fluconazole 200 mg given twice a day was initiated postoperatively. Remarkable improvement in the patient’s symptoms of pain and swelling were noted on his follow-up visit 4 days later, with resolution of the interface separation (Fig. 1E). There was resolution of the patient’s keratic precipitates and significant reduction in his anterior chamber inflammation, but the patient continued to have persistent corneal edema. The interface lesions shrank in size but did not completely resolve. Injections were then repeated at 2-week intervals with eventual involution and scarring of the interface opacities (Fig. 1F). After a final taper of antifungal medications, www.corneajrnl.com |
991
Tu and Hou
the patient’s cornea remained edematous but uninflamed limiting his vision to 20/500 (Fig. 1G).
Patient 2 A 70-year-old man presented with a complicated history of sequential cataract extraction, intraocular lens exchange, DSAEK for corneal decompensation, and an Ahmed valve shunt for a resultant uncontrolled intraocular pressure elevation, all performed at an outside institution. The patient presented to the Retina and Glaucoma Services at the IEEI with “kissing” choroidal detachments secondary to hypotony. After drainage of his hemorrhagic choroidal effusions, his vision returned to 20/ 60. However, over the following year, the patient’s vision deteriorated to 20/300 secondary to corneal decompensation and edema with tube-corneal touch. An uncomplicated repeat DSAEK combined with a shunt revision was performed at the IEEI. By postop week 7, the patient’s vision had improved to 20/ 70. The patient was otherwise asymptomatic; however, 2 new interface opacities in the inferonasal quadrant were noted at that time (Figs. 2A, B). No other anterior segment inflammation was seen. Confocal microscopy (Confoscan 3, Nidek, Japan) demonstrated small round opacities, approximately 4 mm in size, just peripheral to the intrastromal opacities (Fig. 2C). The patient was offered either penetrating keratoplasty or removal of the endothelial lenticule for a presumed fungal keratitis, but refused because of his improved vision. Oral voriconazole was started, and topical prednisolone acetate was reduced to twice daily with some improvement of the smaller peripheral lesion. However, the larger lesion progressively enlarged despite 3 weeks of treatment with oral voriconazole. Intrastromal injections of voriconazole (50 mg/0.1 mL), which infused and separated the stromal graft–host interface, were given weekly
Cornea Volume 33, Number 9, September 2014
for 3 weeks and the oral voriconazole was discontinued because of intolerable systemic side effects. Both lesions involuted and resolved, leaving coalescent, clumpy interface opacities. Six months later, the corrected distance visual acuity had improved to 20/60 with a clear graft and a corneal thickness of 578 (Fig. 2D).
DISCUSSION We have demonstrated that intrastromal/stromal interface infusion of antifungals can be used to successfully treat late-onset interface infectious keratitis without the need for performing therapeutic keratoplasty. Although DSAEK has revolutionized the treatment of corneal endothelial failure, increasing reports of late interface infections in DSAEK and in other lamellar procedures, such as deep anterior lamellar keratoplasty, have raised concerns about the safety of the tissue supply and the appropriate prophylaxis of fungal contaminated transplanted tissue.5 Once an infection has been established, however, successful treatment has been achieved primarily by full thickness corneal transplantation, although some have suggested the removal of the endothelial lenticule as an alternative.3,4,6–8 Results have generally been good, but these infections have resulted in the loss of an eye.4 Late-onset interface infections after DSAEK, specifically, have been almost exclusively caused by Candida species, usually occurring several weeks to months after surgery without any history of other trauma or intervention.3,4,6–8 Other organisms have been rarely reported.9 The yeast is presumably introduced at the time of surgery as contamination from the host flora or donor tissue. Studies have shown a significant correlation with positive fungal donor rim cultures in those patients who are infected, but interface infections have
FIGURE 2. A, External photograph of patient 2 showing 2 peripheral interlamellar infiltrates (arrows). B, Slit-beam photograph of the 2 interlamellar infiltrates. C, Confocal microscopy showing hyperreflective, round, budding-like structures (;4 mm in size) in the lamellar interface suggestive of fungal elements. D, Slit-beam photograph taken 6 months after the initiation of treatment, which included 3 injections of intrastromal voriconazole with secondary infusion of the interlamellar space. There is an almost complete resolution of both lesions.
992
| www.corneajrnl.com
Ó 2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 9, September 2014
also been noted in patients with negative donor rim cultures.3,4,6–8,10 Both of our patients had donor rim cultures that were negative. One mate cornea was not placed, and the other mate did not have cultures taken, but their postoperative course was otherwise uncomplicated. Confocal microscopy and clinical appearance were supportive of a diagnosis of a yeast pathogen in a location inaccessible to standard culturing methods. Penetration of topical and oral antifungals into the deep stroma varies widely. However, regardless of the agent used, antifungal therapy without surgical intervention has seldom resulted in a cure in late-onset interface keratitis in DSAEK.3,4,6–8 Most case reports or series of DSAEK-related keratitis have, therefore, recommended either penetrating keratoplasty or, uncommonly, removal of the DSAEK lenticule with intracameral antifungal irrigation. A large penetrating keratoplasty to encompass the entire endothelial graft has been the procedure of choice to prevent exposure of the pathogen to the intraocular space. If treated promptly, these interventions are successful, but subject the patient permanently to the short- and long-term disadvantages of penetrating keratoplasty, including heightened rates of graft rejection, irregular astigmatism, and wound dehiscence. Intrastromal injections of antifungals have been shown to be effective for other forms of deep fungal keratitis, and deliver high, persistent concentrations of antifungal directly into the corneal stroma.11–14 A significant concern for intrastromal injection in a recent endothelial keratoplasty is the inadvertent dissection of the peripheral graft–host interface, potentially resulting in graft dislocation and/or exposure of the pathogen to the anterior chamber. Despite our attempts to restrict the injections to the anterior stroma, all our intrastromal injections resulted in some separation of the graft–host stromal interface. This is likely analogous to injection of air in deep anterior lamellar keratoplasty surgery where fluid or air will preferentially accumulate in the plane of least resistance. However, in both cases, peripheral graft–host scar formation prevented graft dislocation and further extravasation of the injected fluid beyond the margin of the graft. Given the late-onset nature of these yeast infections, the firm adhesion of the peripheral graft–host junction was not unexpected. Intrastromal injections in these cases can thus result in both maximized antifungal exposure to the interface lesions and persistent stromal drug levels. Nevertheless, the volume of the injection should be limited to avoid peripheral graft–host dehiscence and its advisability for earlier onset DSAEK infections, before peripheral scar formation, is unknown. It should be noted that although both our patients received oral or topical antifungals as part of their management, significant improvement was only noted after intrastromal injection of either amphotericin B or voriconazole. Patient 1 received only fluconazole, which penetrates poorly into the eye. Dramatic resolution of his ocular inflammation was noted after the first intrastromal injection of amphotericin B was given. The cornea remained edematous after resolution of the infection, but the chronicity of the infection and restricted corticosteroid use both negatively impacted his graft viability. The second patient seemed to have a partial response to oral voriconazole, a drug with good tissue penetration. Ó 2014 Lippincott Williams & Wilkins
Intrastromal Antifungal Injection
Unfortunately, the larger lesion enlarged while the patient was on oral medications, and he experienced intolerable systemic side effects. Involution of the larger lesion began only after the first injection of voriconazole was given. The patient’s cornea remains clear 6 months after treatment. The presumptive diagnosis of a yeast keratitis was supported by confocal microscopy and response to antifungal drugs in both cases. Intrastromal injection with secondary interlamellar infusion of either amphotericin B and voriconazole seems to be an effective therapeutic option in late fungal keratitis largely confined to the interface and immediately adjacent stroma, especially with preservation of graft function elsewhere. Although curative, therapeutic penetrating keratoplasty or endothelial lenticule removal exposes the patient to significant irreversible risks in a situation where a viable graft may be preserved or repeat keratoplasty delayed until a lower risk surgery can be performed including a repeat DSAEK. Intrastromal injection with secondary interlamellar infusion antifungal infusion should, therefore, be considered as a therapeutic option for late-onset, presumed fungal keratitis after lamellar corneal surgery. REFERENCES 1. Panda A, Pushker N, Nainiwal S, et al. Rhodotorula sp. infection in corneal interface following lamellar keratoplasty—a case report. Acta Ophthalmol Scand. 1999;77:227–228. 2. Fontana L, Parente G, Di Pede B, et al. Candida albicans interface infection after deep anterior lamellar keratoplasty. Cornea. 2007;26: 883–885. 3. Kitzmann AS, Wagoner MD, Syed NA, et al. Donor-related Candida keratitis after Descemet stripping automated endothelial keratoplasty. Cornea. 2009;28:825–828. 4. Koenig SB, Wirostko WJ, Fish RI, et al. Candida keratitis after Descemet stripping and automated endothelial keratoplasty. Cornea. 2009;28: 471–473. 5. Aldave AJ, DeMatteo J, Glasser DB, et al. Report of the Eye Bank Association of America Medical Advisory Board subcommittee on fungal infection after corneal transplantation. Cornea. 2013;32:149–154. 6. Lee WB, Foster JB, Kozarsky AM, et al. Interface fungal keratitis after endothelial keratoplasty: a clinicopathological report. Ophthalmic Surg Lasers Imaging. 2011;42 Online:e44–48. 7. Ortiz-Gomariz A, Higueras-Esteban A, Gutiérrez-Ortega ÁR, et al. Lateonset Candida keratitis after Descemet stripping automated endothelial keratoplasty: clinical and confocal microscopic report. Eur J Ophthalmol. 2011;21:498–502. 8. Sharma N, Agarwal PC, Kumar CS, et al. Microbial keratitis after Descemet stripping automated endothelial keratoplasty. Eye Contact Lens. 2011;37:320–322. 9. Miyamoto T, Eguchi H, Tserennadmid E, et al. Methicillin-resistant Staphylococcus aureus keratitis after Descemet’s stripping automated endothelial keratoplasty. Case Rep Ophthalmol. 2013;4:269–273. 10. Sedaghat MR, Hosseinpoor SS. Candida albicans interface infection after deep anterior lamellar keratoplasty. Indian J Ophthalmol. 2012; 60:328–330. 11. Sharma N, Chacko J, Velpandian T, et al. Comparative evaluation of topical versus intrastromal voriconazole as an adjunct to natamycin in recalcitrant fungal keratitis. Ophthalmology. 2013;120:677–681. 12. Qu L, Li L, Xie H. Corneal and aqueous humor concentrations of amphotericin B using three different routes of administration in a rabbit model. Ophthalmic Res. 2010;43:153–158. 13. Prakash G, Sharma N, Goel M, et al. Evaluation of intrastromal injection of voriconazole as a therapeutic adjunctive for the management of deep recalcitrant fungal keratitis. Am J Ophthalmol. 2008;146:56–59. 14. Garcia-Valenzuela E, Song CD. Intracorneal injection of amphotericin B for recurrent fungal keratitis and endophthalmitis. Arch Ophthalmol. 2005;123:1721–1723.
www.corneajrnl.com |
993
Intrastromal Antifungal Injection With Secondary Lamellar Interface Infusion for Late-Onset Infectious Keratitis After DSAEK Elmer Y. Tu, MD, and Joshua Hou, MD
Purpose: The aim of this study was to report the successful medical management of 2 cases of late-onset endothelial keratoplasty–related stromal interface infections.
Methods: All cases of endothelial keratoplasty–related infections treated with intrastromal antifungal injections were compiled. The following information was collected: demographic data, surgical indications, donor rim cultures, donor mate outcomes, clinical course, diagnostic tests, and clinical outcome.
Results: Two cases of interface fungal keratitis diagnosed on clinical appearance and confocal microscopy were identified. Both patients refused to undergo further surgery and failed systemic and/ or topical therapy. Each received 3 to 4 intrastromal injections, with secondary infusion into the graft–host interface, which resulted in the complete involution of their interface opacities. Conclusions: Intrastromal antifungal injection may be an effective alternative to surgical intervention in late-onset fungal Descemet stripping endothelial keratoplasty interface keratitis. Early treatment may preserve graft viability and result in a good visual outcome without the need for either penetrating keratoplasty or potential pathogen exposure to the anterior chamber. Key Words: lamellar keratoplasty, infection, fungal, cornea, lamellar interface, amphotericin B, voriconazole, candida, yeast, endothelial keratoplasty
complications challenge the DSAEK surgeon, including graft dislocation, graft inversion, epithelial ingrowth, and optical interface irregularities. Another unique complication of all forms of lamellar keratoplasty is the development of late-onset interface infectious keratitis.1–4 The deep stromal location of the infiltrate limits access for microbiologic testing and reduces achievable drug levels at the site of infection. Although significant controversy exists regarding the value and form of prophylactic antifungal therapy,5 almost all established interface infections have failed systemic and/or topical antifungal therapy most commonly requiring penetrating keratoplasty or even leading to the loss of the eye.3,4,6–8 We present an alternative medical therapeutic technique that can avoid the longand short-term risks of urgent keratoplasty and preserve a potentially viable endothelial graft.
MATERIALS AND METHODS All cases of presumed infectious interface keratitis after endothelial keratoplasty seen at the University of Illinois Eye and Ear Infirmary (IEEI) were identified. Cases were reviewed with respect to demographic data, surgical indications, donor rim cultures, clinical course, diagnostic tests, and clinical outcome. Donor mate outcomes were obtained from the source eye bank.
(Cornea 2014;33:990–993)
D
escemet stripping endothelial keratoplasty (DSAEK) has largely supplanted penetrating keratoplasty as the procedure of choice in patients with primarily endothelial pathology such as Fuchs endothelial dystrophy or pseudophakic bullous keratopathy. DSAEK offers significant advantages in almost every aspect of transplant-related complications including intraoperative safety, late wound dehiscence, rejection, and speed of visual recovery. However, several unique
Received for publication April 22, 2014; revision received May 17, 2014; accepted May 20, 2014. Published online ahead of print July 22, 2014. From the Department of Ophthalmology and Visual Sciences, University of Illinois Eye and Ear Infirmary, Chicago, IL. The authors have no funding or conflicts of interest to disclose. Reprints: Elmer Y. Tu, MD, Cornea and External Disease Service, Department of Ophthalmology and Visual Sciences, University of Illinois College of Medicine, Chicago, IL 60612 (e-mail: [email protected]). Copyright © 2014 by Lippincott Williams & Wilkins
990
| www.corneajrnl.com
RESULTS Case Reports Patient 1 A 66-year-old man presented from an outside institution with a history of increasing intraocular inflammation 3 months after an uncomplicated DSAEK and formation of 2 discrete interface opacities (Figs. 1A, B). Topical difluprednate given twice daily had been unsuccessful in controlling his iritis and had been discontinued 2 days before presentation. The patient remained only on brimonidine/timolol combination drops. This was his third endothelial keratoplasty after a complicated cataract extraction 2 years earlier. On examination, his vision was found to have decreased from a peak postoperative value of 20/50 to 20/150 on presentation at the University of Illinois at Chicago. His vision was 20/60 while on topical corticosteroids just 2 days before. Intraocular pressure was 19, and his examination showed perilimbal injection, diffuse corneal edema, diffuse keratic precipitates, 3+ anterior chamber cells, and Cornea Volume 33, Number 9, September 2014
Cornea Volume 33, Number 9, September 2014
Intrastromal Antifungal Injection
FIGURE 1. A, External photograph of patient 1 showing 2 peripheral interlamellar infiltrates (arrows), diffuse corneal edema, and keratic precipitates. B, Slit-beam photograph of the 2 interlamellar infiltrates. C, Confocal microscopy showing hyperreflective, round, budding-like structures in the lamellar interface suggestive of the presence of yeast elements. D, Anterior segment ocular coherence tomography showing separation of the lamellar interface with an intact peripheral graft–host junction immediately after an intrastromal amphotericin B injection was given. E, Anterior segment ocular coherence tomography showing the resolution of the paracentral graft detachment 1 day after the injection. The interface lesion appears hyperreflective in the peripheral lamellar graft interface. F, External photograph taken 2 months after the initiation of treatment, which included 2 injections of intrastromal with secondary interlamellar infusion of amphotericin B and 1 injection of voriconazole. There is resolution of the more peripheral interface lesion and scarring/involution of the more central lesion. There was persistent corneal edema despite the resolution of the KP. G, Slit-beam photograph of the involuting interface lesion. The needle track is still visible below the lesion.
2 round interface plaques in the inferior temporal graft–host interface (Figs. 1A, B). The patient was initially given oral acyclovir for iritis of an unknown origin. At follow-up 4 days later, the patient’s vision had decreased further to 20/500. Confocal microscopy (Corneal Module; HRT Heidelberg, Germany) showed lesions suggestive of budding/paired yeast near the clinically localized interface lesions (Fig. 1C). A penetrating keratoplasty was recommended, but the patient refused further surgery. Because of cost, the patient did not receive prescribed oral voriconazole and continued to demonstrate crescendo inflammation and decreasing vision. After informed consent was obtained, an intrastromal injection of amphotericin B (5 mg/0.1 mL) was given in the corneal stroma anterior to the Ó 2014 Lippincott Williams & Wilkins
surgical intralamellar space on day 6, which separated the DSAEK interface without disturbing the peripheral graft– host junction (Fig. 1D). Oral fluconazole 200 mg given twice a day was initiated postoperatively. Remarkable improvement in the patient’s symptoms of pain and swelling were noted on his follow-up visit 4 days later, with resolution of the interface separation (Fig. 1E). There was resolution of the patient’s keratic precipitates and significant reduction in his anterior chamber inflammation, but the patient continued to have persistent corneal edema. The interface lesions shrank in size but did not completely resolve. Injections were then repeated at 2-week intervals with eventual involution and scarring of the interface opacities (Fig. 1F). After a final taper of antifungal medications, www.corneajrnl.com |
991
Tu and Hou
the patient’s cornea remained edematous but uninflamed limiting his vision to 20/500 (Fig. 1G).
Patient 2 A 70-year-old man presented with a complicated history of sequential cataract extraction, intraocular lens exchange, DSAEK for corneal decompensation, and an Ahmed valve shunt for a resultant uncontrolled intraocular pressure elevation, all performed at an outside institution. The patient presented to the Retina and Glaucoma Services at the IEEI with “kissing” choroidal detachments secondary to hypotony. After drainage of his hemorrhagic choroidal effusions, his vision returned to 20/ 60. However, over the following year, the patient’s vision deteriorated to 20/300 secondary to corneal decompensation and edema with tube-corneal touch. An uncomplicated repeat DSAEK combined with a shunt revision was performed at the IEEI. By postop week 7, the patient’s vision had improved to 20/ 70. The patient was otherwise asymptomatic; however, 2 new interface opacities in the inferonasal quadrant were noted at that time (Figs. 2A, B). No other anterior segment inflammation was seen. Confocal microscopy (Confoscan 3, Nidek, Japan) demonstrated small round opacities, approximately 4 mm in size, just peripheral to the intrastromal opacities (Fig. 2C). The patient was offered either penetrating keratoplasty or removal of the endothelial lenticule for a presumed fungal keratitis, but refused because of his improved vision. Oral voriconazole was started, and topical prednisolone acetate was reduced to twice daily with some improvement of the smaller peripheral lesion. However, the larger lesion progressively enlarged despite 3 weeks of treatment with oral voriconazole. Intrastromal injections of voriconazole (50 mg/0.1 mL), which infused and separated the stromal graft–host interface, were given weekly
Cornea Volume 33, Number 9, September 2014
for 3 weeks and the oral voriconazole was discontinued because of intolerable systemic side effects. Both lesions involuted and resolved, leaving coalescent, clumpy interface opacities. Six months later, the corrected distance visual acuity had improved to 20/60 with a clear graft and a corneal thickness of 578 (Fig. 2D).
DISCUSSION We have demonstrated that intrastromal/stromal interface infusion of antifungals can be used to successfully treat late-onset interface infectious keratitis without the need for performing therapeutic keratoplasty. Although DSAEK has revolutionized the treatment of corneal endothelial failure, increasing reports of late interface infections in DSAEK and in other lamellar procedures, such as deep anterior lamellar keratoplasty, have raised concerns about the safety of the tissue supply and the appropriate prophylaxis of fungal contaminated transplanted tissue.5 Once an infection has been established, however, successful treatment has been achieved primarily by full thickness corneal transplantation, although some have suggested the removal of the endothelial lenticule as an alternative.3,4,6–8 Results have generally been good, but these infections have resulted in the loss of an eye.4 Late-onset interface infections after DSAEK, specifically, have been almost exclusively caused by Candida species, usually occurring several weeks to months after surgery without any history of other trauma or intervention.3,4,6–8 Other organisms have been rarely reported.9 The yeast is presumably introduced at the time of surgery as contamination from the host flora or donor tissue. Studies have shown a significant correlation with positive fungal donor rim cultures in those patients who are infected, but interface infections have
FIGURE 2. A, External photograph of patient 2 showing 2 peripheral interlamellar infiltrates (arrows). B, Slit-beam photograph of the 2 interlamellar infiltrates. C, Confocal microscopy showing hyperreflective, round, budding-like structures (;4 mm in size) in the lamellar interface suggestive of fungal elements. D, Slit-beam photograph taken 6 months after the initiation of treatment, which included 3 injections of intrastromal voriconazole with secondary infusion of the interlamellar space. There is an almost complete resolution of both lesions.
992
| www.corneajrnl.com
Ó 2014 Lippincott Williams & Wilkins
Cornea Volume 33, Number 9, September 2014
also been noted in patients with negative donor rim cultures.3,4,6–8,10 Both of our patients had donor rim cultures that were negative. One mate cornea was not placed, and the other mate did not have cultures taken, but their postoperative course was otherwise uncomplicated. Confocal microscopy and clinical appearance were supportive of a diagnosis of a yeast pathogen in a location inaccessible to standard culturing methods. Penetration of topical and oral antifungals into the deep stroma varies widely. However, regardless of the agent used, antifungal therapy without surgical intervention has seldom resulted in a cure in late-onset interface keratitis in DSAEK.3,4,6–8 Most case reports or series of DSAEK-related keratitis have, therefore, recommended either penetrating keratoplasty or, uncommonly, removal of the DSAEK lenticule with intracameral antifungal irrigation. A large penetrating keratoplasty to encompass the entire endothelial graft has been the procedure of choice to prevent exposure of the pathogen to the intraocular space. If treated promptly, these interventions are successful, but subject the patient permanently to the short- and long-term disadvantages of penetrating keratoplasty, including heightened rates of graft rejection, irregular astigmatism, and wound dehiscence. Intrastromal injections of antifungals have been shown to be effective for other forms of deep fungal keratitis, and deliver high, persistent concentrations of antifungal directly into the corneal stroma.11–14 A significant concern for intrastromal injection in a recent endothelial keratoplasty is the inadvertent dissection of the peripheral graft–host interface, potentially resulting in graft dislocation and/or exposure of the pathogen to the anterior chamber. Despite our attempts to restrict the injections to the anterior stroma, all our intrastromal injections resulted in some separation of the graft–host stromal interface. This is likely analogous to injection of air in deep anterior lamellar keratoplasty surgery where fluid or air will preferentially accumulate in the plane of least resistance. However, in both cases, peripheral graft–host scar formation prevented graft dislocation and further extravasation of the injected fluid beyond the margin of the graft. Given the late-onset nature of these yeast infections, the firm adhesion of the peripheral graft–host junction was not unexpected. Intrastromal injections in these cases can thus result in both maximized antifungal exposure to the interface lesions and persistent stromal drug levels. Nevertheless, the volume of the injection should be limited to avoid peripheral graft–host dehiscence and its advisability for earlier onset DSAEK infections, before peripheral scar formation, is unknown. It should be noted that although both our patients received oral or topical antifungals as part of their management, significant improvement was only noted after intrastromal injection of either amphotericin B or voriconazole. Patient 1 received only fluconazole, which penetrates poorly into the eye. Dramatic resolution of his ocular inflammation was noted after the first intrastromal injection of amphotericin B was given. The cornea remained edematous after resolution of the infection, but the chronicity of the infection and restricted corticosteroid use both negatively impacted his graft viability. The second patient seemed to have a partial response to oral voriconazole, a drug with good tissue penetration. Ó 2014 Lippincott Williams & Wilkins
Intrastromal Antifungal Injection
Unfortunately, the larger lesion enlarged while the patient was on oral medications, and he experienced intolerable systemic side effects. Involution of the larger lesion began only after the first injection of voriconazole was given. The patient’s cornea remains clear 6 months after treatment. The presumptive diagnosis of a yeast keratitis was supported by confocal microscopy and response to antifungal drugs in both cases. Intrastromal injection with secondary interlamellar infusion of either amphotericin B and voriconazole seems to be an effective therapeutic option in late fungal keratitis largely confined to the interface and immediately adjacent stroma, especially with preservation of graft function elsewhere. Although curative, therapeutic penetrating keratoplasty or endothelial lenticule removal exposes the patient to significant irreversible risks in a situation where a viable graft may be preserved or repeat keratoplasty delayed until a lower risk surgery can be performed including a repeat DSAEK. Intrastromal injection with secondary interlamellar infusion antifungal infusion should, therefore, be considered as a therapeutic option for late-onset, presumed fungal keratitis after lamellar corneal surgery. REFERENCES 1. Panda A, Pushker N, Nainiwal S, et al. Rhodotorula sp. infection in corneal interface following lamellar keratoplasty—a case report. Acta Ophthalmol Scand. 1999;77:227–228. 2. Fontana L, Parente G, Di Pede B, et al. Candida albicans interface infection after deep anterior lamellar keratoplasty. Cornea. 2007;26: 883–885. 3. Kitzmann AS, Wagoner MD, Syed NA, et al. Donor-related Candida keratitis after Descemet stripping automated endothelial keratoplasty. Cornea. 2009;28:825–828. 4. Koenig SB, Wirostko WJ, Fish RI, et al. Candida keratitis after Descemet stripping and automated endothelial keratoplasty. Cornea. 2009;28: 471–473. 5. Aldave AJ, DeMatteo J, Glasser DB, et al. Report of the Eye Bank Association of America Medical Advisory Board subcommittee on fungal infection after corneal transplantation. Cornea. 2013;32:149–154. 6. Lee WB, Foster JB, Kozarsky AM, et al. Interface fungal keratitis after endothelial keratoplasty: a clinicopathological report. Ophthalmic Surg Lasers Imaging. 2011;42 Online:e44–48. 7. Ortiz-Gomariz A, Higueras-Esteban A, Gutiérrez-Ortega ÁR, et al. Lateonset Candida keratitis after Descemet stripping automated endothelial keratoplasty: clinical and confocal microscopic report. Eur J Ophthalmol. 2011;21:498–502. 8. Sharma N, Agarwal PC, Kumar CS, et al. Microbial keratitis after Descemet stripping automated endothelial keratoplasty. Eye Contact Lens. 2011;37:320–322. 9. Miyamoto T, Eguchi H, Tserennadmid E, et al. Methicillin-resistant Staphylococcus aureus keratitis after Descemet’s stripping automated endothelial keratoplasty. Case Rep Ophthalmol. 2013;4:269–273. 10. Sedaghat MR, Hosseinpoor SS. Candida albicans interface infection after deep anterior lamellar keratoplasty. Indian J Ophthalmol. 2012; 60:328–330. 11. Sharma N, Chacko J, Velpandian T, et al. Comparative evaluation of topical versus intrastromal voriconazole as an adjunct to natamycin in recalcitrant fungal keratitis. Ophthalmology. 2013;120:677–681. 12. Qu L, Li L, Xie H. Corneal and aqueous humor concentrations of amphotericin B using three different routes of administration in a rabbit model. Ophthalmic Res. 2010;43:153–158. 13. Prakash G, Sharma N, Goel M, et al. Evaluation of intrastromal injection of voriconazole as a therapeutic adjunctive for the management of deep recalcitrant fungal keratitis. Am J Ophthalmol. 2008;146:56–59. 14. Garcia-Valenzuela E, Song CD. Intracorneal injection of amphotericin B for recurrent fungal keratitis and endophthalmitis. Arch Ophthalmol. 2005;123:1721–1723.
www.corneajrnl.com |
993
