Eye (1988) 2, 670-676

Corneo-Scleral Rim Cultures: Donor Contamination A Case of Fungal Endophthalmitis Transmitted by K-Sol Stored Cornea

LYE P. FONGt, D. GLADSTONEtt, T.A. CASEYt and Tunbridge Wells

Summary This retrospective study of 549 corneo-scleral rim cultures shows that gentamicin, used in MK and K-Sol medium storage at 4°C, has decreased donor contamination from 43% in whole­ globe storage to 13%, but failed to eliminate coagulase negative staphylococci (37%), streptococci (28%) and fungi (28%). Donor-to-host transmitted staphylococcal and streptococcal endophthalmitis have been reported previously. We present the first documented case of donor-to-recipient transmitted fungal endophthalmitis following using corneas stored in MK or K-Sol solution at 4°C; Candida albicans was isolated. Recommendations are made to assess critically the true incidence of donor fungal contamination and the necessity of adding anti-mycotic agents to preservation medium for 4°C storage. In the absence of ideal antimicrobial cover for corneal preservation solutions, stringent prophylactic measures to reduce contamination and continued monitoring of corneo-scleral rim cultures are warranted, if the poor visual consequences of donor-to-host transmitted endophthalmitis are to be avoided.

The conversion of corneal storage from the retrospective review of 549 corneo-scleral whole eye moist chamber technique to donor rim cultures obtained from corneas McCarey-Kaufman (MK) and K-Sol tissue stored in MK and K-Sol medium or as whole culture media has coincided with increased globes. The objectives of this study were to: interest in the risk of postoperative (i) identify the spectrum of microbial endophthalmitis from contaminated donor contamination of donor corneas tissue and storage medium.1,2 Gentamicin, (ii) suggest recommendations for reducing which is used alone in MK and K-Sol donor corneal contamination in storage medium, is known to be highly effective media at 4°C. against staphylococcus and Pseudomonas (Ps.) aeruginosa but recent reports have Materials and Methods questioned its efficacy against certain strains Over an eight year period, between 1980 and of streptococcus3-5 in this storage medium. February 1988, 549 penetrating keratoplasty These recent reports, together with our cases performed at the Corneo-Plastic Unit, experience of two donor-related post­ Queen Victoria , East Grinstead, keratoplasty intraocular infections, one of were reviewed. Donor details were recorded, which was fungal in origin, initiated this including the donor age, cause of death, time

From tThe Corneo-Plastic Unit, Queen Victoria Hospital, East Grinstead, Sussex. ttMicrobiology Unit, Kent and Sussex Hospital, Tunbridge Wells, Sussex.

Correspondence to: Lye P. Fong, Corneo-Plastic Unit, Queen Victoria Hospital, East Grinstead, Sussex RH19 3DZ. CORNEO-SCLERAL RIM CULTURES: DONOR CONTAMINATION. FUNGAL ENDOPHTHALMITIS 671 from death to enucleation and method and transplantation, all residual corneo-sceral duration of storage_ The charts of the rims were swabbed and plated on to the recipients were searched for the pre­ culture media: blood, chocolate and modified operative corneal diagnosis, the presence of Sabouraud's agar at 37°C; Sabouraud's agar pre-existing ocular infections, possible ocular was also incubated at 25°C.6 Culture plates or systemic factors predisposing to infections for bacterial isolates were held for at least and the postoperative course for evidence of three days and fungal cultures for a minimum ocular infection. The minimum review period of one week, before any negative results was two months but most patients' charts were confirmed. Organisms isolated were yielded information for extended periods of identified by standard diagnostic criteria and time, often over a course of years, either sensitivities were perfomed for the assumed until the present date or referral to the local pathogenic strains of bacteria using antibiotic ophthalmologist. disc criteria.6 The standardised procedures for corneal preparation were as follows: whole eyes were Results enucleated, within 12 hours of death, in an Of the 549 cases reviewed (Table I), 302 aseptic manner by someone medically corneas were preserved in storage media, 251 qualified, using sterile instruments and in MK and 51 in K-Sol solutions; 14% were drapes. The globes were supported in sterile culture positive and all, except one, were containers, irrigated with 5 ml of Neosporin single isolates. Of 247 whole-globe stored ophthalmic solution (polymixin B-neomycin­ corneas, 107 (43%) were culture positive and gramicidin, Burroughs Wellcome Co.) and 11 were polymicrobial. Analysis of the packed in ice for transport to the , microbial spectrum showed that there were within one to two hours. Further preparation similar ratios of major pathogens in both the was as follows: according to the method of· whole-globe and medium stored corneas. corneal storage: Gram-positive organisms accounted for (i) Whole globes were stored at 4°C for up approximately 70% of cases and of these, to 24 hours and then immersed in two-thirds were staphylococci, most being soframycin solution for 30 minutes prior coagulase-negative staphylococci (CNS) and to transplantation one-third comprised streptococci. One­ (ii) MK and K-Sol corneas were immediately quarter of the cultures involved fungi. prepared by dissecting the corn eo-scleral During the study period, two patients discs with sterile instruments for developed post-keratoplasty endopthalmitis immersion in storage medium at 4°C. secondary to donor corneal contamination Most MK and K-Sol corneas were used and their details are presented in Table II. within twelve to forty-eight hours of death Patient 1, who had a complex past ocular and the longest period of storage was four history and disorganised anterior segment days. The MK and K-Sol stored corneas were from multiple surgical procedures, developed then warmed to room temperature for half to a Candida albicans keratitis and one hour before transplantation. endophthalmitis; the same organism had No routine topical antibiotics were used been recovered from the K-Sol stored preoperatively. At the end of surgery, sub­ corneo-sceral rim three weeks previously. On conjunctival injections of either gentamycin the same operative list, the paired donor or tobramycin (20 mg) and dexamethasone cornea, which was also stored in K-Sol (4 mg) were given and Oc. chloramphenicol solution and cultured Candida albicans, was 1 % was instilled. G. chloramphenicol 1 % used for a patient with pseudophakic and G. dexamethasone 0.1% were used keratopathy; his postoperative course was routinely, on a q.i.d. dosage, for one month uneventful. postoperatively, after which the antibiotic Ps. aeruginosa was isolated from the was discontinued and steroid drops tapered corneo-scleral rim of a whole globe in patient off. 2, who had not received routine subcon­ After excision of the donor disc for junctival antibiotics at the time of surgery 672 L. P. FONG ET AL.

Table I. Culture results of corneo -scleral rims.

Corneal storage MKand K-Sol Whole globe

Total number of cases 302 247 Number with culture positive rims 43 106 Polymicrobial cultures 1 11 Percentage of cases with positive cultures 14 43 Gram positive bacteria Coagulase negative staphylococcus 16 (37%)* 36(31%) Staphylococcus aureus 1 (2%) 8 (7%) Viridans streptococcus 5 (12%) 6 (6%) Streptococcus pneumoniae 5 (12%) 1 (1 %) Beta-haemolytic streptococcus o 6 (5%) Streptococcus faecalis 2 (5%) 8 (7%) Unidentifiedstreptococcus o 10 (9%)

Total number streptococcus 12 (28%) 32 (27%)

Corynebacteria spp 1 (2%) 7 (6%) Clostridia spp o 1 (1 %) Gram negative bacteria Pseudomonas aeruginosa 2 (5%) 6 (6%) Proteus spp o 2 (2%) Escherichia spp o 6 (6%) Fungi (total number) 12 (28%) 33 (31%) Candida albicans 7 (16%) 26 (22%) Aspergillus spp 2 (5%) 3 (3%) Cladosporium spp 2 (5%) o Penicillium spp 1 (2%) 4 (3%)

*Denominator = cases with positive cultures. because of gentamicin hypersensitivity. Candida albicans, Aspergillus spp and CNS, Another two patients (unpublished data), were unrelated to the donor infective pro­ whose MK-stored donor rims cultured CNS, cess; and none of the recipients developed developed hypopyons post-operatively but post-operative infections. microbiological investigations were not insti­ gated as the infection/inflammation resolved Discussion rapidly on intensive topical antibiotics and Although the use of high dose gentamicin steroids. (100 microgmlml) in storage media has sub­ The median age of the donors was 48 years stantially reduced the donor contamination (range four days to 82 years). The dis­ rate of whole-globe stored corneas (43%) to tribution of causes of death was uniform 14 %, the residual donor contamination still throughout both study groups and so com­ poses a significant risk for donor-to-host bined figuresare presented: acute myocardial transmitted infections. Leveille et aF had infarction 34% carcinoma 30%, road traffic shown a 22-fold increased incidence of endo­ accident 20% and cerebrovascular accident phthalmitis in patients with positive donor 18%. Of eight donors with systemic infec­ rim cultures when compared to those with tion, no organisms were isolated from two negative results. Although the incidence for whole-globe specimens; three MK-stored such donor-related infections is low, 0.4% corneas were culture negative; another three for whole-globe stored corneas and 0.3% of MK-preserved corneas, which cultured MK and K-Sol storage in our series, the CORNEO-SCLERAL RIM CULTURES: DONOR CONTAMINATION. FUNGAL ENDOPHTHALMITIS 673 extremely poor visual outcome4 justifies recorded by Mathers,S Mascerella8 and their every precautionary measure to reduce and, co-authors (Table III). They also found that ideally, eliminate all donor contaminants. eNS and streptococci were the predominant If aerobic cultures alone are considered, potential pathogens in corneo-sceral cultures the 14% incidence of positive cultures in our following MK storage. Although no series is comparable to 17% and 20% infections were reported from their smaller

Table II. Summary of cases with post-keratoplasty endophthalmitis and culture-positive donor rim corneas

Case History' Indication @ Post-operative Corneal Culture details Treatmentt Outcome AgelSexlDate forPK Course Storage IV Sc Top Sys

6O,M,2188 4/871CCE& ACIOL CoS staphyloma 3wkspost-PK K-Sollday Donor rim,corneal A A Mi - Hand 9/87 AV& IOLexchange Deep keratitis, suture, cornea& xl xl movements 12187prolapsed iris ACmass& conjunctiva: abscissed, IOLremoved, A V& PK anterior vitreous Candida albicam ftuffballs Vitreous aspirate: no growth

59,M,7/83 Herpetic keratitis FailedPK 3dayspost-PK Whole-globe Donorrim, cornea T T C No light ECCE&PK Cornealabscess 1 day & vitreous: M M perception & endophthalmitis Pseudomonas ae. Mez Mez Mez Perforation Phthisis

'ICCE = Intracapsular cataract extraction, AC = Anterior chamber, IOL = Intraocular lens, AV =

Anterior vitrectomy, ECCE = Extracapsular cataract extraction, PK = Penetrating keratoplasty @ CoS = Corneo-scleral t A = Amphotericin B, IV 10 microgm, SC 0.3 mg; Mi = Miconazole; T = Tobramycin; C Cephalexin; M = Methicillin; Mez = Mezlocillin; IV = Intravenous; SC = Subconjunctival; Top = Topical; Sys = Systemic.

Table III. Cornea-scleral rim cultures of corneas stored in MK and K-sol medium

Mascarella Mathers Fang 1979 1987 1988 N=200 N=291 N=302

Percentage of positive cultures 12 39 14 Propionibacterium spp 36%' 32% ND@ Coagulase-negativestaphylococcus 14% 30% 37% Staphylococcus aureus 7% 3% 2% Streptococci 11% 9% 28% Corynebacteria spp 14% 8% 2% Pseudomonas aeruginosa 0 0 5% Proteus 4% 1% 0 Candida albicans 4% 1% 16% Aspergillus spp 0 0 5% Cladosporium spp 0 0 5% Fusarium solani 4% 0 0 Torulopsis glabrata 4% 0 0 Penicillium spp 0 0 2%

, Denominator = cases with positive cultures @ ND = Not done 674 L. P. FONG ET AL. series of cases, Leveille et aF found a 0.2% addition of a second antibiotic, such as a incidence of donor-related post-operative cephalosporin or penicillin, to eliminate a endophthalmitis in 1,876 penetrating greater range of organisms, but others have keratoplasties with MK stored corneas. cautioned against the ineffectiveness of these Review of our data has highlighted the antibiotics at 4°C. 4,16 Antibiotic activity is need for a re-appraisal of our current eye dependent on the rate of metabolism of bank techniques. Donors need to be pre­ bacteria, and therefore, on temperature. selected to exclude potentially contaminated With "warming-up" from 4°C, those material from septic donors.9,l0 Further antimicrobials affecting more "essential" modifications of corneal preparation tech­ metabolism, e.g. aminoglycosides on ribo­ niques to reduce the donor contamination somes and DNA synthesis, would be expec­ rate include: mechanical irrigation with ted to be effective sooner than cell-wall saline solution, chemical preparation of the agents, such as penicillins and cephalo­ external ocular surface with povidine-iodine, sporins. The newer fluoroquinolene anti­ immersion rather than irrigation with anti­ biotics, which affect DNA gyrase activity in a biotics, the use of a laminar flowhood and a broader spectrum of bacteria, are potentially standardised period of warming of corneas usefup7 but more work is required to and storage media from 4°C.1l-13 ascertain their efficacyand safety. Gentamicin is effective against a wide The microbiological data in this study may range of organisms but, as our study shows, be improved by the following technical eNS, streptococci and fungi are resistant to modifications: this aminoglycoside, which is currently used (i) anaerobic cultures for both MK and K-Sol media. Streptococci (ii) elimination of pre-soaked antibiotics are particularly resistant, even with extended (iii) cultures in brain heart infusion broth periods of warming-up time.4 Clinical reports medium by Matoba and Baer and their co-workers3,4 (iv) maintaining fungal cultures for one have emphasised the emergence of month. streptococci as major aetiological agents in The high incidence of fungal recovery in the transmission of intra-ocular infections our study (28%) differs from the 0.5% through contaminated MK stored tissue. Of incidence reported by Badenoch and co­ the 15 cases where bacteria were recovered authors(18) and 11 % in Mascarella and from either corneo-scleral rims or MK Cavanagh's series.7 One explanation for the medium, one-third were due to discrepancies may be short culture periods. streptococci.1 ,3,4 Although most fungi grew within four days, Ps. aeruginosa is an infrequent external O'Day and his co-workers(19) showed that ocular contaminant. 14 Two of our MK and K­ one-quarter were positive two to three weeks Sol stored corneas cultured Ps. aeruginosa, after inoculation and they recommended that albeit with no transmission of infection. One .both liquid and solid media be kept for four case of post-keratoplasty endophthalmitis weeks before discarding. They also secondary to MK-stored donor cornea con­ recommended the use of multiple media taminated with Ps. aeruginosa was reported suitable for fungal growth, for which brain­ by Leveille et aU It cannot be determined if heart infusion medium was the most suitable. these infrequent contaminations were due to The true incidence of fungal contamina­ particularly virulent strains, large inoculae, tion, under effective eye bank decontamina­ in vivo resistance to gentamicin or in­ tion procedures and laboratory conditions adequate antimicrobial activity because of favourable to fungal growth, is unknown. insufficient warming of the MK medium. 4 Identification of this contamination risk, Although the organism was sensitive to following 4°C MK and K-Sol storage, will gentamicin in all three instances, the emer­ help determine the necessity of supplement­ gence of gentamicin-resistant strains remain ing anti-mycotic agents in storage media.20,21 an area of concern. IS Saggau and his co-authors22 observed a A number of authors3-s have suggested the natural attrition of fungal organisms at 4°C. CORNEO-SCLERAL RIM CULTURES: DONOR CONTAMINATION. FUNGAL ENDOPHTHALMITIS 675

Although the donor-to-host transmission References of fungal infections from whole-globe and 1 Girard LJ: Bacterial endophthalmitis following mgan cultured corneas is well docu­ the use of contaminated preserved corneal mented,8,21 our patient with Candida albicans tissue. Cornea 1982; 1: 255-7. is the first reported case of fungal infection 2 LeFrancois M, Baum JL: Flavobacterium endophthalmitis following keratoplasty. Arch transmitted through corneas stored in culture Ophthalmol1976; 94: 1907-9. medium at 4°C. The host immune defence 3 Matoba A, Moore MB, Merten JL, McCulley system in our patient was likely to have been JP: Donor-to-host transmission of strepto­ compromised, both from multiple surgical coccal infection by corneas stored in insults to the eye and longterm administra­ McCarey-Kaufman medium. Cornea 1984; 3: tion of topical .antibiotics and steroids. It is 105-8. interesting to note that no complications 4 Baer JC, Nirankari VS, Glaros DS: Strepto­ developed with the paired donor cornea coccal endophthalmitis from contaminated which was grafted into a relatively healthy donor corneas after keratoplasty. Clinical and eye. laboratory investigations. Arch Ophthalmol 1988; 106:517-20. Identification of a culture-positive donor 5 Mathers WD, Lemp MA. Corneal rim cultures. may aid in therapeutic guidelines for post­ Cornea1987; 6(3): 231-3. keratoplasty endophthalmitis, but micro­ 6 Jones DB, Liesegang TJ, Robinson NM: biological confirmation of infection is still Laboratory diagnosis of ocular infections. required. In the majority of our patients with American Academy of positive donor rim cultures, the routine post­ Manuals Program1981. operative management was not altered, 7 Leveille AS, McMullan D, Cavanagh HD: except for a few select patients, whqse Endophthalmitis following penetrating 90: antibiotics were changed according to sensi­ keratoplasty. Ophthalmol1983; 38-9. tivity data, depending on the inclination of 8 Mascarella K, Cavanagh HD: Penetrating keratoplasty using McCarey-Kaufman pre­ the individual surgeon. Nevertheless, routine served corneal tissue. South Med J 1979; 72: donor corneo-scleral cultures are required 1268-71. for monitoring quality control and the selec­ 9 Gandhi SS, Lamberts DW, Perry HD: Donor to tion of antibiotics in culture media and peri­ host transmission of disease via corneal trans­ operative care. plantation. Surv Ophthalmol 1981; 25: 306-- This study has confirmed that coagulase 11. negative staphylococci (37%), streptococci 10 Olson RJ, McMain, Slappey TE: Donor eye (28%) and fungi (28%) are resistant to contamination. Ann Ophthalmol 1979; 11: gentimicin prophylaxis in MK and K-Sol 1875-8. corneal storage at 4°C. Our report of the first 11 Pardos GJ, Gallagher MA: Microbial contami­ nation of donor eyes. A retrospective study. case of fungal infection transmitted by the Arch Ophthalmol1982; 100:1611-13. donor cornea after MK or K-Sol sorage 12 Keates RH, Mishler KE, Riedinger D: Bacterial suggests that the true incidence of fungal contamination of donor eyes. Am J Ophthal­ contamination needs to be established to mol 1977;84: 617-9. determine the necessity of adding anti­ 13 Apt L, Isenberg S, Yoshimori R, Paez JH: mycotic agents to the storage media for 4°C. Chemical preparation of the eye in ophthal­ Despite the low incidence of donor-related mic surgery. III. Effect of povidone-iodine on post-keratoplasty infections, it is imperative the conjunctiva. Arch Ophthalmoll984: 102: to mInImISe the possibility of such 728-9. devastating infections by relying on stringent 14 Polack FN, Locatcher-Khorazo D, Gutierrez E: decontamination procedures in the eye bank, Bacteriologic study of "donor" eyes. Arch Ophthalmol1967; 78: 219-25. continual monitoring of corneo-scleral rim 15 Poole TG, Insler MS: Contamination of donor cultures and the development of new anti­ cornea by gentamicin-resistant organisms. Am microbials either as a replacement or J OphthalmoI1984; 97: 560-4. supplement to gentimicin. 16 Baum J, Barza M, Kane A: Efficacyof penicillin G, cefazolin, and gentamicin in M-K medium at 4°C. Arch Ophthalmol1978; 96: 1262-4. 676 L. P. FONG ET AL.

17 Viti A, Seetharama S, Snyder I, Schwab IR: 20 Kowalski RP, Sundar Raj CV, Stuart JC, Dunn Bacterial growth and the use of ciprofioxacin DS: Antifungal synergism: A proposed dos­ in MK media. Invest Ophthalmol Vis Sci Supp age for corneal storage medium. Arch 1988; 29: 444. Ophthalmol1985; 103: 250-6. 18 Badenoch PR, Alfrich SJ, Wedding TR, Coster 21 Nelson JD, Mindrup EA, Chung CK, DJ: Effectiveness of a decontamination Lindstrom RL, Doughman DJ: Fungal con­ method for donor corneas. Br J Ophthalmol tamination in organ culture. Arch Ophthalmol 1988; 72: 225-7. 1983; 101: 280-3. 19 Q'Day DM, Akrabawi PL, Head WS, Ratner 22 Saggau DD, Bourne WM, Sinkeldam IR, HB: Laboratory techniques in human and Roberts GD: Replication of fungi in K-Sol experimental fungal infections. Am J corneal preservation medium at 4°C. Arch Ophthalmol1979; 87: 688-93. Ophthalmol1986; 104: 1362-3.