One of the most dreaded complications of penetrating keratoplasty is postoperative endophthalmitis, the incidence of which has been reported to be approximately 0.2% to 0.77%.1 Of these, 75% have been reported to be bacterial in origin whereas 20% have a fungal etiology.2 Several improvements have been made in both preoperative and intraoperative processes to reduce the rate of postoperative endophthalmitis. These have been standardized by the Eye Bank Association of America (EBAA) and have included:
- The addition of streptomycin to gentamicin in the corneal storage medium (July 1993, Optisol GS versus historical Optisol, Bausch and Lomb, Irvine, CA)
- A shift to in-situ excision of donor corneoscleral tissue rather than enucleation (this shift began in mid-1995)
- The addition of whole globe immersion in 5% povidone-iodine for 2 minutes followed by 2 rinses in sterile saline (January 1996)
- Serial irrigation of donor eyes with povidone-iodine eye drops (June 1997)
- The cold storage of tissue in its preservative medium at 4° C.3
However, even with these changes, postoperative endophthalmitis has not been eliminated and the surgeon continues to endure the question of whether to perform donor corneoscleral rim cultures at the time of surgery.
Background Literature Review
In 2006, 44,747 corneas were supplied by U.S. eye banks from which 33,674 corneal transplants were performed in the United States.4 Assuming the above incidence rate of 0.2% to 0.77%, between 90 and 345 cases of penetrating keratoplasty-associated endophthalmitis occur annually. Given the potential morbidity of this complication, corneal surgeons are appropriately prudent to desire early isolation and microbiologic susceptibility profiles of potential pathogens following corneal transplantation.
Numerous retrospective studies and meta-analyses have been performed to evaluate the utility of donor corneoscleral rim cultures in predicting endophthalmitis. A study at the authors' institution evaluated 774 corneoscleral rims obtained between January 1992 and November 1997.3 Of these, 41 (5.3%) yielded a positive culture result. None of these patients went on to develop postoperative endophthalmitis, however, 2 patients with negative corneoscleral rim cultures did develop an endophthalmitis within the postoperative period.
A recently published meta-analysis further investigated the prognostic role of donor corneoscleral rim cultures in predicting endophthalmitis.5 This study included 17, 614 corneal grafts, from studies published between 1975 and 2006, of which 2,459 (14%) had positive donor rim cultures. In aggregate, these studies reported 31 (0.2%) cases of postoperative endophthalmitis; however, only 21 (0.1%) of these had concordant recipient and donor microbiologic isolates. Ten of 10 eyes (100%) with candidal endophthalmitis had concordance between donor and recipient isolates, whereas concordance was demonstrated in only 11 of 20 (55%) of bacterial endophthalmitis cases.
The relative risk of endophthalmitis among eyes receiving a culture-positive rim was 12.2 times greater than that among those with negative cultures. However, this risk was 247 times greater when fungus was retrieved from the donor rim. Evaluating these data further demonstrates that a culture-positive rim was 5 times more frequent among recipients who developed endophthalmitis. When looking specifically at cases of fungal endophthalmitis, the study found the positive likelihood ratio of a donor rim fungal culture for post-keratoplasty fungal endophthalmitis to be 52.4%. This translates to a positive predictive value of fungal recovery for fungal endophthalmitis of 2.6% if an estimated pretest probability for fungal endophthalmitis of 0.05% is used.5 A lower association between fungal donor rim contamination and postoperative infection has been described by Keyhani et al.6 In their study the incidence of fungal endophthalmitis following penetrating keratoplasty was 0.16%, with a 14% concordance between fungus-positive rim cultures and fungal endophthalmitis.
These data must be evaluated keeping in mind that there is high interstudy heterogeneity with the percent of culture positive rims ranging from 5.3%3 to 36.8%.7 This variability is likely due to several differences in the processes of harvesting and storing donor tissue over the past 3 decades including preservation antibiotics, varied culture techniques between laboratories, and broader changes, as discussed above.
While the true value of obtaining donor corneal rim cultures has been debated in the literature as described above, the cost of routinely obtaining these cultures has been estimated to be between $2,000,000 and $6,000,000 per year in the United States.5,8 It should be noted that the performance of a true cost-benefit analysis dependends on the definition of cost, as this could be vastly different if defined in terms of cost to the patient, cost to the hospital, cost of blindness, or cost to society. Because of this and the truly low incidence of post-penetrating keratoplasty endophthalmitis, the authors do not believe that the benefit of routine donor rim culture outweighs the cost. In fact, for those patients who develop endophthalmitis, a positive or negative culture result does not tend to change our clinical decision-making. It is the culture results that are obtained by an anterior-chamber and vitreous tap in the setting of clinically suspected endophthalmitis that are the most valuable in continued.
While data from these prior studies demonstrate a wide range of positive donor corneoscleral rim cultures, they overwhelmingly demonstrate a relatively low concordance of postoperative bacterial endophthalmitis isolates with positive donor rim cultures. This demonstrates that preoperative donor rim cultures are somewhat unreliable predictors of bacterial endophthalmitis.3 However, in the case of fungal endophthalmitis, the utility of donor rim cultures has been shown to be greater but variable.
Aside from the low positive predictive value (found to be 2.6% by Wilhelmus et al.) of donor rim cultures, the corneal surgeon should also consider the timing of such results. Culture isolate and susceptibility results are often negative or not available during the critical 24- to 72-hour postoperative period in which an endophthalmitis is often clinically diagnosed, leaving the onus on the surgeon of beginning empiric treatment prior to receiving culture results. Thus the surgeon must consider intervention prior to receiving culture results, and this raises the specter of the utility of sending a test in which the results are not known soon enough to affect clinical decision-making. In fact, studies have suggested the lack of predictive value of donor rim cultures and the limited effect they have on clinical management.3,8
When faced with a possible post-keratoplasty endophthalmitis patient, the authors perform a vitreous tap for culture and sensitivity, begin empiric treatment with broad spectrum intravitreal antibiotics, and consider the addition of fortified topical antibiotics. A surgeon must also consider the quandary of receiving positive donor rim culture results in the setting of a quiet clinical picture. In this scenario, a decision of whether or not to initiate prophylactic treatment must be made. Given the known low positive predictive value of the test and the potential side effects of treatment, specifically of initiating systemic antifungal therapy, the authors do not initiate treatment until signs or symptoms of endophthalmitis are present.
Other solutions can be considered in an attempt to decrease postoperative endophthalmitis resulting from infectious donor tissue. These include increased selectivity in tissue screening, as other studies have demonstrated a higher incidence of microbial contamination in donors who died from systemic infections9 or those who have been on a respirator prior to death.10 Prior studies have also demonstrated decreased activity of antibiotics at 4° C, emphasizing the importance of warming corneal tissue and preservative solution to room temperature at least 1 hour prior to transplantation.2,11 Likewise, the addition of an antifungal agent to the preservative solution could be considered.6
Given that the results of microbiologic cultures do not present in a timely fashion, and that the actual incidence of postoperative endophthalmitis is acceptably low, the authors recommend following a patient's clinical examination after penetrating keratoplasty and treating accordingly, regardless of the outcome of corneoscleral rim cultures. Therefore, the authors do not perform routine rim cultures on patients undergoing penetrating keratoplasty.
- Krachmer J, Mannis M, Holland E. Cornea. Vol. 2. Surgery of the Cornea and Conjunctiva. Philadelphia, PA: Elsevier Mosby; 2005:1522.
- Kloess P, Stulting R, Waring G, Wilson L. Bacterial and fungal endophthalmitis after penetrating keratoplasty. Am J Ophthalmol. 1993;115(3):309-316.
- Everts R, Fowler W, Chang D, Reller L. Corneoscleral rim cultures: lack of utility and implications for clinical decision-making and infection prevention in the care of patients undergoing corneal transplantation. Cornea. 2001;20(6):586-589.
- Eye Bank Association of America statistics. Available at http://www.restoresight.org/. Accessed May 2008.
- Wilhelmus K, Hassan S. The prognostic role of donor corneoscleral rim cultures in corneal transplantation. Ophthalmology. 2007;114(3):440-445.
- Keyhani K, Seedor JA, Shah MK, Terraciano AJ, Ritterband DC. The incidence of fungal keratitis and endophthalmitis following penetrating keratoplasty. Cornea. 2005;24:288-291.
- Reed J, Bealer L, Sloop Cea. Questionable benefit of cultures at the time of penetrating keratoplasty [abstract]. Cornea. 1994;13:101.
- Wiffen S, West B, Maguire L, Bourne W. The value of routine donor corneal rim cultures in penetrating keratoplasty. Arch Ophthalmol. 1997;115(6):719-724.
- Hassan S, Wilhelmus K, Dahl G, et al. Infectious disease risk factors of corneal graft donors. Arch Ophthalmol. 2008;126(2):235-239.
- Karjalainen K, Vannas A. Bacterial contamination of donor corneas. Ophthalmic Surg. 1984;15(9):770-772.
- Baum J, Barza M, Kane A. Efficacy of penicillin G, cefazolin, and gentamicin in M-K medium at 4 degrees C. Arch Ophthalmol. 1978;96(7):1262-1264.
Drs. Awdeh and Koreishi state that they have no financial relationship with the manufacturer or provider of any product or service discussed in this article or with the manufacturer or provider of any competing product or service.