Statistics from Altmetric.com
Editor,—Transmission of infection by donor tissue is a well known, albeit rare, complication of corneal transplantation.1 We have found two reports2 3 of Candida glabrataendophthalmitis, one with an organ cultured cornea when systemic antifungal treatment contributed to the patient’s death.2
A 43 year old man with keratoconus had two previous right corneal grafts. The first, in 1974, became vascularised and oedematous after several months; the second, in 1990, failed after two rejection episodes. In February 1995 he underwent further right penetrating keratoplasty using tissue typed material, stored in organ culture for 20 days and transport medium (5% dextran) for 2 days. On the first postoperative day UK Transplant Support Service Authority reported culture of a yeast from the transport medium, subsequently identified as Candida glabrata. Clinically the graft was satisfactory but topical econozole 1% (in arachis oil) was given prophylactically six times daily.
Increasing anterior uveitis was treated by increasing topical prednisolone acetate 1% from 2 hourly to hourly at postoperative day 4 and adding oral prednisolone 80 mg daily at day 11. At this stage a tiny white endothelial deposit developed at the graft interface. Over 4 days this became a white plaque of 2 mm diameter (Fig 1) but remained asymptomatic. The lesion was aspirated by paracentesis; microscopy showed a yeast and subsequently Candida glabrata was cultured. Amphotericin B 5 μg in 0.1 ml was injected into the anterior chamber immediately and 3 days later, and topical amphotericin B 0.15% commenced 1 hourly. The graft remained clear and the uveitis settled. Oral steroid was withdrawn completely over 7 weeks.
Further threatened rejection with anterior uveitis associated with loose corneal sutures occurred 5 months postoperatively. This was treated with hourly drops of prednisolone acetate 1%. Three weeks later an enlarging white mass, initially 0.75 mm in diameter, appeared elsewhere on the host endothelium (Fig 2). Aspiration and subsequent culture again confirmed Candida glabrata.
Amphotericin B 10 μg in 0.1 ml was injected into the anterior chamber, with subconjunctival injection of amphotericin B 300 μg in 0.3 ml. Topical amphotericin B 0.15% 1 hourly and prednisolone 1% 2 hourly were given. Both injections were repeated at 3 and 20 days. The uveitis settled, topical amphotericin was discontinued, and the graft remains clear 2.5 years later on prednisolone drops once daily.
No toxic effects have been noted from intracameral or topical amphotericin B.
Endophthalmitis flared up twice associated with intensive topical and/or oral steroid. Candida glabrata was cultured from the transport medium and the recipient anterior chamber, implicating the donor cornea as the source of infection. Candida glabrata is a relatively resistant organism and treatment was guided by antifungal susceptibility testing of the isolate. This suggested that the isolate was resistant to fluconazole (minimum inhibitory concentration, MIC >128 mg/l) and itraconazole (MIC 16 mg/l), but was of intermediate sensitivity to econazole (MIC 1 mg/l and 4 mg/l variously on different tests) (Table 1). The latter may explain econazole’s failure to prevent infection becoming established. Although the isolate was sensitive to amphotericin (MIC <0.25 mg/l) this was avoided as prophylaxis because of potential toxicity; however, no toxic effects occurred when it was subsequently used.
Failure of amphotericin B in the organ culture medium to eradicateCandida glabrata, despite its apparent sensitivity, has been described.4 A recent report5 showed 1.1% bacterial or fungal contamination of organ cultured corneas accepted for transplantation despite use of antimicrobial agents in the medium. Over 2.5 years Bristol Eye Bank records showed four occasions, including the present incident, when corneas were grafted and the transport medium was subsequently found to be infected. Two cases developed endophthalmitis (Candida spp andglabrata), while two remained clear of infection (Pseudomonas sp and Penicillium sp). Over 4500 corneas from the Bristol Eye Bank were grafted during this time, thus the incidence of known graft transmitted endophthalmitis is 0.04%. While accepting that some cases may go unreported, this can be compared with an overall incidence of postoperative endophthalmitis in the literature of 0.1–0.8%6-9 after corneal grafting with corneas stored at 4°C, and of 0.09%10 following extracapsular extraction.
This patient remained asymptomatic throughout; close monitoring allowed early detection and successful treatment of the infection.