Major endothelial loss from corneas in organ culture: importance of second endothelial count

The aim of this study was to show that major losses can still occur on corneas judged suitable for grafting at the first count. In addition, we studied the frequency of these losses on 1992 corneas over a period of 4 years to evaluate the risk incurred.

We evaluated the incidence of these major losses and the associated risk factors. An Ishigawa diagram was created with the Cornea Bank team and the ophthalmologists involved in organ retrieval. Endothelial losses caused by bacterial or fungicidal contamination were excluded from the study. For the 29 corneas that suffered major losses, we analyzed the donor files for donor age, clinical file, geographical origins of the corneas, the person who did the retrieval, the length of time the cornea was stored, the data resulting from examining the endothelium at the bank by optical microscope, and the method used for sterilizing the material used. Specific analyses in cases of major loss of endothelial content: anatomopathologic examination of the corneas and search for the herpes simplex virus (HSV; type 1 or 2) by polymerase chain reaction (PCR). We carried out a statistical analysis using a chi(2) test on the 1992 corneas studied to see if the presence of diabetes (type 1 or 2) in the donor led to reduction levels different from those of corneas originating from nondiabetic donors.

The incidence was evaluated at between 0.4% and 3% of corneas sampled, and the associated risk factor was between 0.8% and 6% of grafted corneas. The occurrence of major losses was independent of donor age and was independent of the person who did the retrieval. The occurrence of major losses was independent of geographical origin. We tested our media for endotoxin before use and found levels from 0.22 to 3.9 UI/mL. We verified the absence of a chronological relationship between the batches of media used in the bank and the number of major losses observed, showing that the pyrogenicity limit was independent of cytotoxicity limits. Data analysis showed no difference in reduction levels between diabetic and nondiabetic donors (P < 0.05). Results on the detection of HSV-1 by PCR on the storage media were all negative, and these results agree with the anatomopathologic examinations that showed no signs of viral infection.

Total endothelial losses amounted to 1.4%/yr. Without the double endothelial counts, we would have had 29 primary graft rejections over that period. During storage, this loss has not been linked to a specific cause, but risk factors such as traumatic death, herpes infections, and badly controlled endotoxin levels should be considered when taking preventative actions. For the moment, a second endothelial count before grafting should be carried out, because all these problem grafts conformed to grafting criteria after the first count. The possibility of carrying out this second count is one of the recognized advantages of storage in organ culture.