Abstract
Hepatocyte transplantation is being used in patients with liver-based metabolic disorders and acute liver failure. Hepatocytes are isolated from unused donor liver tissue under GMP conditions. Cells must be free of microbiological contamination to be safe for human use. The experience of microbiological screening during 72 hepatocyte isolation procedures at one center is reported. Samples were taken at different stages of the process and tested using a blood culture bottle system and Gram stain. Bacterial contamination was detected in 37.5% of the UW organ preservative solutions used to transport the liver tissue to the Cell Isolation Unit. After tissue processing the contamination was reduced to 7% overall in the final hepatocyte product, irrespective of the presence of initial contamination of the transport solution. The most common organisms recovered were coagulase-negative staphylococci, a skin commensal. A total of 41 preparations of fresh or cryopreserved hepatocytes were used for cell transplantation in children with liver-based metabolic disorders without any evidence of sepsis due to infusion of hepatocytes. In conclusion, the incidence of bacterial contamination of the final product was low, confirming the suitability of the organs used, hepatocyte isolation procedure, and the environmental conditions of the clean room.
Introduction
Hepatocyte transplantation is being investigated as a treatment for patients with liver-based metabolic disorders and acute liver failure (8). The procedure requires isolation of hepatocytes from unused or rejected donor livers. As there is limited availability of donor tissue, it is important to isolate high-quality hepatocytes and ensure that preparations are not lost due to microbial contamination. Hepatocytes are isolated from donor tissues using a collagenase perfusion technique under strict sterile conditions.
The King's Cell Isolation Unit was established in 2001 to prepare human cells from donor organs for clinical use. It is a purpose-built sterile unit for the production of cells under Good Manufacturing Practice (GMP) conditions. The Cell Isolation Unit consists of interconnecting rooms: an aseptic room, a change room, a preparation room, and an air-lock. There is an air pressure gradient between the rooms of more than 10 kPa, with the highest pressure in the aseptic room, where tissue processing is performed in laminar flow cabinets. Air can only enter the unit through HEPA filters.
A Quality Control system is used to ensure that all processing is performed in clean conditions to the highest standards. The unit is microbiologically monitored each week to ensure that microbial counts are kept within proscribed limits and that appropriate actions are taken if these limits are exceeded (14).
The organs used for cell isolation are procured and screened as per UK Transplant (UKT) guidelines. Virological testing is routinely performed and includes hepatitis B and C viruses and human immunodeficiency virus screening; however, there is no standard procedure for bacteriological screening of whole organs. Contamination of organ transportation media has been reported previously for a range of different organs (16), including liver (2), kidney (3,9,18), and pancreas (19). For cell transplantation bacterial screening is usually carried out on the organ as it arrives and during the cell isolation procedure and product. The livers available for cell transplantation are likely to come from suboptimal donors. Although bacteraemia has been previously reported in 5% of whole organ liver donors (6), there is little evidence of a negative impact on recipient outcome (2,6). Clearly, the importance of avoiding microbial contamination—from both the initial source and subsequent manipulations—is high and must be minimized.
In this article we report the microbiological screening results for livers processed and hepatocytes isolated in our unit over a 5-year period.
Materials and Methods
Procurement
Organs for cell isolation were procured as per the UKT protocol for clinical transplantation and trans-ported to the Cell Isolation Unit in University of Wisconsin (UW) solution on ice. Seventy-two liver tissues were processed, 44 male, 28 female, median age 25 years (range 0.75–76 years), of whom 8 had mild or moderate steatosis. The liver tissue used was either unused segments from cut-down or split liver transplant procedures or livers rejected for organ transplantation due to steatosis, poor perfusion, or anatomical reasons.
Hepatocyte Isolation
Hepatocytes were isolated from livers using a collagenase perfusion technique (17). Briefly, livers or liver sections were first cannulated, perfused with buffer solutions, and digested with collagenase to release hepatocytes. The cell suspension obtained was then centrifuged to purify the hepatocytes. The isolated hepatocytes were either prepared for clinical administration on the same day or cryopreserved using a controlled rate freezer before storage in the vapor phase of a liquid nitrogen tank.
Criteria for Clinical Use
Our criteria for accepting fresh cells for transplantation is a viability greater than 60%, cell number greater than 5 × 108 cells (per preparation), and no organisms seen in a Gram stain preparation (10). With fresh cells there is not sufficient time for cultures to be performed. For cryopreserved cells, aliquots submitted for culture in blood culture bottles must also be culture negative (Fig. 1), as in this case there is sufficient time for testing.
Flow chart of criteria for accepting hepatocytes per isolation procedure for transplantation into patients.
Microbial Analysis During Hepatocyte Isolations
The BacT/ALERT (bioMérieux, Marcy l'Etoile, France) blood culture bottle system was used to test for aerobic and anaerobic microorganisms in samples taken during processing. The first sample was taken from the UW transport solution in which the liver was delivered, the second from the perfusate after collagenase digestion of the liver, and the third from the final supernatant of the washing centrifugation step. A Gram stain was performed on the final cell products as a rapid check.
More recently a further specimen of the final cell product has also been tested using BacT/ALERT blood cultures bottles, as an additional check to the Gram stain. Samples of 5 ml were inoculated into the BacT/ALERT blood culture bottles (aerobic and anaerobic) with a syringe and needle after wiping the bung with an alcohol wipe and allowing the alcohol to evaporate off. The bottles were labeled and delivered to the microbiology laboratory accompanied by an appropriate request form. With the BacT/ALERT system positive cultures are flagged automatically, Gram stains performed, and subcultures made for identification of the organisms present. Results were reported to the hepatocyte transplantation team as they became available, which was usually in 2–3 days.
Microbial Analysis of Cryopreserved Hepatocytes for Infusion
Further testing was carried out when cryopreserved cells were thawed and prepared for clinical administration. Culture specimens were taken from the supernatant when the cells were washed to remove DMSO (cryoprotectant used during cell freezing) before being transferred to syringes. A Gram stain was performed on the product immediately prior to administration.
Statistics
Fisher's exact test was used to analyze the incidence of contamination and Mann-Whitney U-test to compare groups with and without contamination.
Results
UW Transport Solution
A total of 72 samples of UW transport solution were analyzed with 27 (37.5%) found to be culture positive. There was no significant difference in the age of the donors with (median: 31 years, range: 1–76 years) or without (24 years, 0.75–68 years) contamination in the UW solution. No increased incidence of microbial contamination was found with non-heart-beating donor livers (n = 11) or livers that were steatotic (n = 8) compared to conventional donor livers. The cause of death of the donor was also analyzed and no significant difference in incidence of contamination was found between livers from donors that had died of natural causes (~60% in both groups) and those that had died as a result of accidents.
The cold ischemia time for liver tissue where UW solution contamination was detected was significantly higher than when no contamination was detected (median: 12 h, range: 8.1–30 h vs. 10.7 h, 1.2–22.3 h, p = 0.046). The presence of contamination in the UW solution had no significant effect on the outcome of the hepatocyte isolation procedure in terms of cell viability (65%, 24–89% with and 68%, 3.6–95% without contamination) and cell yield (2.68 × 106 cells/g liver, 0.41–13.7 × 106 cells/g with and 2.21 × 106 cells/g, 0.04–24.6 × 106 cells/g without contamination).
Processing and Products
Microbial growth was detected in 16.9% of the samples taken at the first stage of processing: the effluent from perfusion of the liver (Table 1). At the second stage of processing, during the final wash of the isolated cells, microbial growth was found in 7.1% of samples showing a reduction during processing. At the final stage, contamination detected by Gram stain was only found in 2.7% of products. Microbial growth was found in 3 of the 21 products where the final cell product was tested. All 5 of these Gram stain or culture positive preparations (cryopreserved) were rejected for clinical use.
Number of Positive Cultures and Gram Stain Found at Different Stages During the Hepatocyte Isolation Procedure
The incidence of contamination in the samples taken during processing was analyzed according to whether there was contamination in the UW solution before processing (Table 2). A significantly higher incidence of contamination (p = 0.046) was found in the effluent samples taken from tissues with contaminated UW solution. However, this carryover was not found at subsequent stages of processing including the final product. In the final products when culture was performed, 1 out of 9 was positive where the UW solution was contaminated and 2 out of 12 where contamination was not detected in the UW solution. There was thus no relationship between the culture results at the beginning and at the end of processing.
Incidence of Contamination in Subsequent Stages of the Hepatocyte Isolation Procedure in Relation to Contamination in the UW Solution in Which Liver Tissue Was Transported
p = 0.0462, significant difference between those with or without contamination in UW solution.
Microorganisms Recovered
UW Solution
In the 27 positive UW solution cultures a total of 39 isolates were recovered (Table 3). Nine cultures were found to have more than one microorganism present. Of the contaminated transport media 74.4% had Gram positive bacteria, 20.5% were Gram negative, and 5.1% were yeasts. The most frequent bacteria recovered were coagulase-negative staphylococci.
Microorganisms Recovered at Different Stages of Tissue Processing
Percent of organisms found shown in parentheses. Gram-positive organisms shown in bold.
Effluent
In the 12 culture-positive effluents there were a total of 18 different microorganisms; five yielded mixed cultures. All were found to be Gram positive: Coagulase-negative staphylococci (n = 7), Bacillus spp. (n = 4), Staphylococcus aureus (n = 1), other viridans streptococci (n = 3), Clostridium spp. (n = 1), and diphtheroids (n = 2).
Supernatant
In the 5 positive supernatant cultures a total of 6 different microorganisms were found; one yielded mixed growth. Sixty-six percent were Gram positive and 34% Gram negative. These were coagulase-negative staphylococci (n = 2), viridans streptococci (n = 1), diphtheroids (n = 1), Klebsiella aerogenes (n = 1), and Haemophilus influenzae (n = 1).
Final Product
In the 3 positive final product cultures a total of 4 different microorganisms were found; one yielded a mixed growth. These isolates were all Gram positive: coagulase-negative staphylococci (n = 2) and Bacillus spp. (n = 2). In the 2 preparations found to be Gram stain positive, no bacterial culture data for the final product was available as they were performed before the product was routinely cultured.
Clinical Administration
Of the 72 isolations performed in King's Cell Isolation Unit, 40 procedures gave cells that achieved criteria for clinical administration based on cell number, viability, and microbiological data. No contamination was detected when cryopreserved hepatocytes were thawed for clinical use either in the supernatants when cells were washed to remove DMSO or on Gram stain of the final products before administration.
To date 10 patients have been treated at King's with a total of 41 hepatocyte infusions of fresh (n = 15) or cryopreserved (n = 26) hepatocytes. All of the 41 infusions were free of microorganisms on Gram stain and subsequent culture analysis. None of the patients developed sepsis in the immediate period after hepatocyte transplantation.
Discussion
Over one third of the donor organs for isolation of hepatocytes were found to have bacterial contamination in the UW organ preservation solution used to transport the organs from the hospitals where they had been harvested. This is perhaps not surprising considering the nature of the donors who may have been supported for a number of days in intensive care units where invasive procedures and hospital-acquired infections are prevalent. Contamination was most frequent in the organ transport solution but lower during the hepatocyte isolation procedure, possibly the effect of washing and dilution. This has been reported with islet cell preparations (4,5,11,12). The contamination was reduced in the final product to the extent that there was no difference in incidence of contamination compared to organs received with no contamination in the UW solution.
A longer liver cold ischemia time was associated with more frequent contamination, suggesting that there may be bacterial growth in UW solution at 4°C. UW solution contains many constituents, including sugars, that support the growth of microorganisms so even small amounts of bacteria could grow. Also livers with longer cold ischemia times often come from donors with complications where there is a delay in decisions about harvest and suitability for transplantation.
No contamination was found in thawed cells after cryopreservation. The effect of cryopreservation on microbial growth is not known, although DMSO is considered to have an inhibitory effect. Studies have shown that many microorganisms including staphylococci can survive cryopreservation (13).
The most frequently recovered bacteria were coagulase-negative staphylococci, which is consistent with other groups (1). It is also the case for islet cell preparations (4). Staphylococci are common skin commensals and a likely source of contamination. Only a few of the organisms that were detected in the study are of concern. Potential overt pathogens included H. influenzae, S. aureus, and streptococci, particularly in immunosuppressed patients. Enterobacteriaceae and Gram-negative nonfermenting bacteria, which are widely distributed in the environment, may act as opportunistic pathogens. These bacteria frequently cause nosocomial infections and can be antimicrobial resistant. These organisms were predominantly recovered from UW solution.
Potential pathogens were found in the liver perfusion effluent: Klebsiella spp. and H. influenzae. They were found after digestion of the tissue but not found in previous steps. This could be due to the organisms being released when the tissue was disrupted during processing. There was no evidence of infection in the donor information available with these livers to explain their presence.
The microorganisms isolated could have originated from the environment, skin or mucosa of the operator or donor, or from the gut of the donor. No Escherichia coli and few enterococci were isolated, which are typical of gut/lumen organisms (Table 3). The absence of molds—which are classic environmental contaminants—confirms that gross environmental contamination throughout processing did not occur.
The livers are handled in a sterile facility under GMP conditions and the utmost care is taken with tissues. However, the operators, laboratory environment, equipment, and solutions still remain possible sources of contamination and steps are taken to avoid contamination from these. All equipment is sterilized before use and manufacturers certify all of the solutions used in processing. The environment is monitored; rooms are monitored weekly for microbial growth using agar plates to detect airborne and surface bacteria. If limits based on number of bacterial colonies detected are exceeded (e.g., no colonies allowed in a laminar flow hood), the source of the contamination is identified and the area decontaminated. The operator technique is monitored by “finger dabs” using agar plates during the isolation procedure. Despite these precautions and testing, de novo contamination is possible and was found in the cell product in 3 of the 72 procedures. It is not possible to determine the exact source of this small incidence of de novo contamination, but as the presumed origin was skin and the environment this could be related to the processing activity.
There is no control over the condition of incoming livers and this has been found to be the main source of microorganisms. With the limited numbers of organs available we should not reject products if at all possible. Unlike drug therapies, a method of sterilization is not available without damaging the cells. Antibiotics could be added during processing, but these might mask bacterial growth rather than eliminate it and might cause allergic reactions: sterile technique continues to be of paramount importance. Contaminated islet and bone marrow cells have been transplanted in other conditions without serious infectious complications with suitable antibiotic prophylaxis (6,15,20). This raises the possibility that contaminated hepatocytes kept cryopreserved in reserve could be used in the case of an emergency with appropriate antibiotics.
In the long term a more sensitive and instant detection method is needed as cells need to be transplanted soon after isolation. The Gram stain detects grossly contaminated samples and for this reason Gram stain-positive cells are rejected for clinical use. The Gram stain will not detect low numbers of bacteria or yeasts. New methods could solve this problem. One such technique is 16s rRNA amplification, which allows the broad spec-trum detection of even small amounts of bacterial rRNA even from nonviable organisms (7).
In summary, although bacterial contamination of donor liver tissue is not uncommon, most final cell preparations were microorganism free after the tissue processing procedure. All cells used clinically were microorganism free and were transplanted into children with metabolic liver disease without any adverse clinical events. Only a small number of products were found to be contaminated and were rejected for clinical use. The lack of contamination of the hepatocyte preparations confirms the sterility of the environment and procedures used.