Donor Infection and Transmission to the Recipient of a Solid Allograft
This paper was presented in part in Programs and Abstracts of the 45th Interscience Conference on Antimicrobial Agents and Chemotherapy, Washington, D.C., American Society for Microbiology, 2005: abstract K-1805, p. 359.
* Corresponding author: Oscar Len Abad, firstname.lastname@example.org
Transmission of infection from donor to recipient is a potential complication of transplantation. More data on this issue are needed to expand the insufficient donor pool. This study evaluates the incidence of donor nonviral infection, transmission from infected donors and the effect of donor infection on 30-day recipient survival. Data from 211 infected donors contributing to 292 (8.8%) of 3322 consecutive transplant procedures within RESITRA (Spanish Research Network for the Study of Infection in Transplantation) were prospectively compiled and analyzed. Lung was the most likely transplanted organ carried out with an infected donor and Staphylococcus aureus was the most commonly isolated microorganism. In more than a half of donors, the lung was the site of infection. Donor-to-host transmission was documented in 5 patients out of 292 (1.71%), 2 of whom died of the acquired infection (40%). Nonetheless, there was no difference in 30-day patient survival when comparing transplant procedures performed with organs from infected or uninfected donors. In conclusion, donor infection is not an infrequent event, but transmission to the recipient is quite low. Hence, with careful microbiological surveillance and treatment, the number of organs available for transplantation may be increased.
The shortage of suitable organ donors is currently the most important obstacle in organ transplantation (1). The disparity between demand and supply has led the transplant community to look beyond relatively young, relatively stable deceased organ donors toward more marginal candidates. These donors include those at extremes of age (2), those with severe hemodynamic instability or prolonged ICU stays and donors who may potentially transmit diseases to their recipients. In fact, nonviral infection is not unusual in potential organ donors. The use of medical devices, treatment of patients in hospital areas that have significant rates of bacterial contamination and, most importantly, the presence of certain medical conditions in the potential organ donor are important factors that facilitate bacterial infection. These infections have long been regarded as a contraindication to deceased organ donation due to the risk of transmitting pathogens. Numerous reports have documented isolated instances of transmission of infectious agents from donors to recipients, sometimes with disastrous results (3–5). In addition, severe compromise of initial transplant function has been reported in small series when organs from infected donors are used for desperate recipients (6). In contrast, some authors have reported low rates of infectious complications in recipients of potentially infected grafts, which suggests that a less restrictive organ donation policy may be feasible (7–13); nonetheless, the information is still scarce and controversial.
We undertook a prospective comparative study to determine the incidence of nonviral donor infection, the frequency with which nonviral infections are transmitted from donor organs to recipients and the impact on 30-day patient survival following transplantation in a consecutive series of transplant recipients. In addition, the survival results of these patients were compared to those of recipients of organs procured from donors without bacterial or fungal infection during the same time period.
Material and Methods
From September 2003 to August 2005, 3322 consecutive solid organ transplant recipients were prospectively included in an online data base within the Spanish Research Network for the Study of Infection in Transplantation (RESITRA, Red de Estudio de Infeccion en el Trasplante), which includes 16 transplant centers all over Spain. The data base contains information on several pretransplantation, perioperative and posttransplantation variables, as well as factors related to rejection and infectious episodes.
Data referring to donor infection included the microorganism and type of infection, divided into lung colonization, lung infection (pneumonia or empyema), bacteremia, central nervous system infection, endocarditis, pyelonephritis, intraabdominal infection and preservation fluid contamination according to the criteria of the Centers for Disease Control and Prevention (CDC) (14). For the purpose of the study the diagnosis of lung colonization was based on isolation of at least 103 cfu/mL of bacteria or of fungi from respiratory samples after selective and protected bronchial washing of the graft before its removal (15). Information related to age, gender and cause of death was also collected. Donor data from the time of admission up to the donation procedure were considered for the analysis. The diagnosis of infection was based on a combination of clinical, microbiological and imaging findings. Donor samples obtained in the ICU on the days before organ recovery included blood, tracheal aspirate, urine and any other site of clinically suspected infection, as well as the preservation fluids at organ recovery following a standard procedure recommended by the ONT (Spanish National Organization of Transplantation) (16). However, not all possible sample sites were cultured in all donors. All donors with positive cultures, regardless of the timing of the culture report relative to the donation, were included in the analysis. Donors considered to have an infectious disease at the time of organ procurement in whom the etiologic agent could not be identified were also included in the analysis. Organs with a documented infection were excluded from donation. Data on antimicrobial treatment prior to donation were not recorded. Possible pathogens/contaminants defined as Staphylococcus epidermidis, other coagulase-negative staphylococci, Propionibacterium acnes, nonpathogenic Neisseria and Corynebacterium spp were also considered for the study.
Recipient data included the prophylaxis received and the type and etiology of infection. Samples from biological fluids were cultured only in cases of clinically suspected infection. Infections were defined according to the criteria of the CDC (14). In summary, we considered a true bacterial infection as an isolation of a microrganism from a sterile site or from another site in the presence of clinical symptoms or signs of infection. We did not consider true infection those febrile episodes without microorganism isolation and resolved without empirical antibiotic treatment. We also did not analyze episodes considered as colonization or contamination (e.g. common skin organisms isolated only once and not accompanied by clinical signs of infection). When pathogens were isolated from donor cultures, the standard prophylactic regimen (not uniform among the 16 transplant centers) was changed to specific antibiotic therapy against the donor's microorganism and infection; the decision regarding its final duration was based on the judgment of the attending physician. In nonsurvivors, the date and cause of death were recorded. Recipient data were collected from the time of transplantation up to 30 days following the procedure. We also obtained data from the RESITRA data base for all transplants performed with organs procured from donors who did not have positive cultures or infection during the study period.
The only gut transplantation performed during the study period was not considered for the analysis.
For the purposes of the study, donor-to-host transmission was established when a positive isolate from a donor matched any positive culture in the respective recipient during the first 7 days after the procedure in the presence of clinical signs of infection and the same antibiotic susceptibility pattern.
The endpoints of the study were estimation of the incidence of nonviral donor infection, donor-to-host infection transmission rates and comparative patient survival between recipients of organs from infected or noninfected donors. Thirty-day patient survival was used as a measure of overall recipient outcome.
All data base entries and statistical analyses were performed with SPSS v12.0 (SPSS, Inc., Chicago, IL). Continuous variables are expressed as the median (range), and the Mann–Whitney-U test was used to investigate differences between groups. Categorical variables are expressed as fractions, and the chi-square test was used to compare groups. The p-values of less than 0.05 were considered significant.
Among 3322 consecutive solid organ transplantations, 211 infected donors contributed to 292 (8.8%) transplant procedures. Donor characteristics are summarized in Table 1. Among the transplanted organs, lung (29 out of 199, 14.6%) was the most likely to be carried out with an infected donor and pancreas the least likely (4 out of 91, 4.4%).
Table 1. Transplantation characteristics according to donors
|Transplant procedures||292 (8.8) ||3030 (91.2)|
|Median (range) donor age, years|| 47 (1–83)|| 49 (6–84)|
|Cause of brain death|
| Trauma||80 (27.4)||1020 (33.7)|
| Stroke||167 (57.2) ||1717 (56.7)|
| Others||45 (15.4)||293 (9.6)|
|Type of organ procedure|
| Heart||31 (10.6)||278 (9.2)|
| Kidney||145 (49.7) ||1432 (47.3)|
| Liver||82 (28.1)||1063 (35.1)|
| Lung||29 (9.9) ||170 (5.6)|
| Pancreas||4 (1.4)|| 87 (2.9)|
| Gut||1 (0.3)||0|
Microorganisms isolated from the 211 infected donors are summarized in Table 2; S. aureus was the most frequently isolated microorganism, followed by Enterobacter cloacae and Escherichia coli. There were no differences in the type of microorganism isolated in the infected donor among the various types of transplantation. As is shown in Table 2, gram-positive cocci were most commonly isolated from sputum, blood and CSF, whereas gram-negative bacilli were most frequent in urine and preservation fluid. Possible pathogens/contaminants were found in 6.1% (13/211) of donor isolates, all corresponding to plasma coagulase-negative staphylococci (12 were from plasma and one was from preservation fluid). Although all donors were considered to have an infectious disease at the time of organ procurement, cultures were negative in 21.8% (46/211), mainly because of antibiotic administration at the time the samples were taken (pneumonia, 33 cases; meningitis, 8 cases; pyelonephritis, 3 cases; and peritonitis, 2 cases). In more than half the donors, the lung was the site of infection, either as colonization or pneumonia. Bacteremia was the second most frequent donor infection. Distribution of donor infection was uniform among the various types of transplant, with the exception of respiratory colonization, which predominated in lung transplantation.
Table 2. Microbiology of isolates obtained from infected donors by type of infection
|Staphylococcus aureus||22 (33.3)||10 (18.2)|| 9 (23.7)|| ||1 (5) ||1 (10)|| ||5 (23.8)||48 (20.1)|
|Streptococcus pneumoniae|| 8 (12.1)||1 (1.8)|| || ||9 (45)|| || || ||18 (7.5) |
|S. viridans||1 (1.5)||1 (1.8)||2 (5.2)|| || || || ||3 (14.3)||7 (2.9)|
|S. pyogenes|| || ||1 (2.6)|| || || || || ||1 (0.4)|
|Enterococcus faecalis|| || ||1 (2.6)|| 5 (20.8)|| || || ||4 (19) ||10 (4.2) |
|E. faecium||1 (1.5)|| || ||1 (4.2)|| || || || ||2 (0.8)|
|Coagulase negative S.|| || ||12 (31.6)|| || ||1 (10)|| || ||13 (5.4) |
|Enterobacter cloacae||6 (9.1)||4 (7.3)||3 (7.9)||2 (8.3)||1 (5) ||1 (10)||3 (60)||2 (9.5) ||22 (9.2) |
|Escherichia coli||3 (4.5)||3 (5.4)||2 (5.2)||10 (41.7)|| ||1 (10)|| ||3 (14.3)||22 (9.2) |
|Klebsiella pneumoniae||1 (1.5)||1 (1.8)||2 (5.2)||2 (8.3)|| || || ||3 (14.3)||9 (3.8)|
|K. oxytoca||2|| || || || ||3 (30)|| || ||5 (2.1)|
|Serratia marcescens||2 (3) || || || || || || || ||2 (0.8)|
|Morganella morgagni|| || || || || ||1 (10)|| || ||1 (0.4)|
|Pseudomonas aeruginosa||6 (9.1)||1 (1.8)||3 (7.9)||1 (4.2)|| ||1 (10)|| ||1 (4.7) ||13 (5.4) |
|Acinetobacter baumannii||5 (7.6)|| ||2 (5.2)|| || ||1 (10)|| || ||8 (3.3)|
|Haemophilus influenzae||6 (9.1)|| || || || || || || ||6 (2.5)|
|Neisseria meningitidis|| || || || ||1 (5) || || || ||1 (0.4)|
|Candida albicans|| || ||1 (2.6)|| || || || || ||1 (0.4)|
|Aspergillus fumigatus||3 (4.5)|| || || || || || || ||3 (1.2)|
|Mycobacterium tuberculosis|| ||1 (1.8)|| || || || || || ||1 (0.4)|
|Unknown|| ||33 (60) || || 3 (12.5)||8 (40)|| ||2 (40)|| ||46 (19.2)|
|Total||66 (27.6)||55 (23) ||38 (15.9)||24 (10) ||20 (8.4) ||10 (4.2)|| 5 (2.1)||21 (8.8) ||239|
Specific antibiotic therapy was administered for 14 days in 41.8% of recipients (122/292). There were 70 episodes of infection at any site in 56 of 292 (19.2%) recipients during the first 30 days after transplantation. Donor-transmitted infection occurred in five recipients (1.7%), and two (40%) died from the transmitted disease. Data on these patients is shown in Table 3.
Table 3. Donor-to-host transmission
|Kidney2||Preservation fluid||Wound infection||Escherichia coli, Enterobacter cloacae, Morganella morgagni||2||First-generation cephalosporin3||Death|
|Kidney4||Pneumonia||Pyelonephritis||Klebsiella pneumoniae||6||First-generation cephalosporin||Cure|
|Liver5||Bacteremia||Pneumonia||Pseudomonas aeruginosa||2||Fourth-generation cephalosporin||Cure|
|Liver5||Pneumonia||Bacteremia||E. coli||7||Amoxicillin + clavulanic acid||Cure|
|Lung6||Lung colonization||Pneumonia||Staphylococcus aureus||6||Fourth-generation cephalosporin||Death|
During the study period, 3030 organs were transplanted from uninfected donors for whom complete data were available. A total of 722 episodes of infection at any site were documented in 620 (20.5%) of these recipients during the first 30 days after transplantation.
Lung recipients performed with an infected donor showed the highest rate of infection after the procedure. Nevertheless, comparison between recipients receiving organs from nonviral-infected donors and recipients from uninfected donors showed no differences in 30-day survival or incidence of infection (Table 4). In addition, when individual organ transplants were compared, there were no significant differences in patient survival according to organ type. Patients receiving infected organs showed no obvious reduction in early survival after transplantation and all organ types had similar recipient outcomes, regardless of the infectious status of the donor.
Table 4. Distribution and 30-day outcome of transplants from infected donors versus noninfected donors
|Total||292||56 (19.2)||284 (97.3)||3030||620 (20.5)||2939 (90) |
|Liver|| 82 (28.1)||17 (20.7)|| 80 (97.6)||1063 (35.1)||225 (21.2)||1017 (95.7)|
|Lung||29 (9.9)||11 (37.9)|| 26 (89.7)||170 (5.6)|| 42 (24.7)|| 151 (88.8)|
|Heart|| 31 (10.6)|| 5 (16.1)|| 29 (93.5)||278 (9.2)|| 41 (14.7)|| 259 (93.2)|
|Kidney||145 (49.7)||22 (15.2)||144 (99.3) ||1432 (47.3)||290 (20.3)||1425 (99.5)|
|Pancreas|| 4 (1.4)||1 (25)|| 4 (100)|| 87 (2.9)|| 22 (25.3)|| 87 (100)|
There was one case of active tuberculosis in a donor whose kidneys were used as allografts. The donor was a 44-year-old man in whom autopsy confirmed disseminated tuberculosis. The kidney recipients began treatment on day 7 after transplantation with isoniazid, ethambutol and pyrazinamide for 2 months, and continued with isoniazid and ethambutol for another 16 months. After 1 year of follow-up, one of the grafts has developed chronic rejection, but no infection. In the other case, both the graft and patient are doing well.
The shortage of suitable donors of organs for transplantation is a widely recognized, worldwide problem. The number of patients awaiting solid organ transplantation is continuously rising, while the number of donors does not show a parallel increase. This situation has created a growing discrepancy, in which larger numbers of patients are waiting for a transplantation, waiting times are longer and deaths on national transplant waiting lists are increasing (17).
One approach to remedy the deficit in organ availability is to extend the donor pool by accepting marginal donors. Although the criteria for marginal donors is not well defined, it is recognized that one or more of the following conditions results in increased donor risk: age, adverse background (alcohol, drugs, cardiovascular disease), prolonged hypotension or high inotropic support, prolonged intensive care stay and infection (17).
Even assuming that identifying the same organism with the same susceptibility pattern only suggests transmission of nonviral infection from the donor to the recipient, in our study this fact was low (1.7%). The idea that donor-to-host transmission of bacterial infections is unusual has been suggested in some studies (11), even those restricted to bacteremic donors (7,8) or those with meningitis (9,18,19). These studies have shown that transmission can be prevented with the use of appropriate antimicrobial agents in recipients and donors prior to organ procurement. Based on this information, some authorities have begun to consider organ donations from hemodynamically stable patients without multiorgan failure when a sufficient course of prior antibiotic therapy has been administered (20), thus allowing expansion of the donor pool.
All recipients included in our study received perioperative antimicrobial prophylaxis, and when donor infection was documented, pathogen- and infection-specific antimicrobial therapy. Freeman et al. (7) reported that transmission of bacterial and fungal microorganisms from donor to recipient is averted when recipients receive antimicrobial therapy. In the authors' experience, however, most donors also underwent treatment because a majority of organs were procured after receiving positive blood culture results. Nonetheless, similar results were reported by Lumbreras et al. in recipients of organs from donors whose bacteremia was revealed following transplantation (8). Recently, Caballero et al. published the results of successful liver and kidney transplantation from deceased donors with left-sided endocarditis (21). The organs were considered suitable for transplantation after ultrasonographic study, and macroscopic observation ensured their correct function and the absence of abscesses. Antimicrobial therapy proven to be adequate for the sensitivity pattern was maintained in all recipients following transplantation. Infection by the pathogen causing endocarditis was not transmitted to any of the seven recipients. In a recent cohort of lung recipients (15), our group found a 6.59% incidence of donor-to-host transmission of bacterial or fungal infection. Nonetheless, this figure is low compared to the infection rate observed in the donors (52%). Two of the eight (25%) patients who received a graft from a bacteriemic patient developed infection. This finding contrasts with previous data (7), reporting no evidence of infection transmission in a large solid organ transplantation series including lung transplantation. However, comparisons cannot be made because of the small number of bacteriemic patients in that cohort.
The design of this study did not allow us to determine whether antibiotics effectively prevented transmission of nonviral organisms from organ donors to recipients. Unfortunately, information regarding antibiotics prescribed for the donor prior to donation and its duration was not collected. Therefore, the data cannot be used to justify the need for ‘treatment’ of donor organisms in recipients; however, in the light of our results (two deaths due to infection likely transmitted from the donor), prudence dictates that treatment with effective agents is warranted as soon as the infection becomes known. In fact, recipients of organs from bacteremic donors in most of the larger series were aggressively treated with antibacterial prophylaxis, usually longer than 48 h after transplantation. There are no controlled trials indicating the optimal duration of antibacterial treatment for recipients of organs from bacteremic donors, but 5 to 7 days of appropriate therapy seems to be the most frequently cited regime (22). Hence, control mechanisms should be implemented to ensure rapid recognition of positive cultures from organ procurement procedures and transplantation. Even though it may be rare, donor-to-host transmission of bacterial infection can be lethal at short term or even after an year of follow-up (11).
The issue of whether to consider active tuberculosis in a donor as a contraindication to solid organ transplantation is a matter of debate. The information in the literature is scarce and universal recommendations are not available at this time. Therefore, the approach to use in each case should be individualized. In a large study of 55 patients with posttransplant tuberculosis, 31% died despite administration of adequate treatment and nearly one-third developed hepatic toxicity, which was severe in half of them (23). The prognosis seems to be better in more recent studies (24). In one report (25), a patient received a liver transplant despite isolation of Mycobacterium tuberculosis in sputum. Treatment was started and no reactivation was reported after 4 months of follow-up. In another report, tuberculous spondylodiscitis showed a favorable response following liver transplantation performed to treat fulminant hepatic failure due to isoniazid therapy (26). The idea of preventing recipient tuberculosis infection acquired from the donor is well supported in our study because treatment proved to be effective, although the GESITRA guidelines (Spanish Group for the Study of Infection in Transplant Recipients) (16) consider this infection a total contraindication for transplantation.
This study includes all donors with infection regardless of the timing of the culture result relative to the organ procurement procedure. Although the risk of transmission may vary according to the length of time before the infection is diagnosed and treated in the donor, the risk of transmission is still present. We have no data on the reasons why transplant centers chose to proceed with transplantation in the presence of known infection. Although published guidelines for selecting infection-related organs are available, many donor and recipient factors have a bearing on the final decision to accept an organ from a given donor for a given recipient.
Some degree of selection bias on the part of the participating transplant centers may have been introduced in the study; it is possible that potential donors with infection causing profound signs of systemic sepsis may not have been considered for organ recovery. Thus, our data cannot be used to justify the use of organs from donors with severe systemic sepsis. Nonetheless, the fact that the organs from donors in our group functioned reasonably well raises the possibility that even severely infected donors might yield acceptable organs.
In this study, 30-day survival of patients receiving organs from donors with infection was equal to that of recipients of organs from uninfected donors. This suggests that there was no significant compromise of the preservability or subsequent function of organs from infected donors. Previous studies have suggested that a possible link between donor infection and primary nonfunction might be explained by bacterial product breakdown rather than acquisition of actual bacterial infection from the donor (27), and that high endotoxin levels transferred with the graft may correlate with a higher incidence of primary nonfunction (28). The consistent patient outcome across all organ types suggests that no specific organ type is more or less likely to transmit an infection or function poorly when recovered from an infected donor. In previous studies, the use of organs from donors with bacteremia (7,8) or meningitis (9) was not associated with a higher incidence of recipient mortality, poor graft function, or immunologic or surgical problems in the recipients (8,11).
Our data do not provide a rationale for excluding organs from infected donors. Many other factors influence decisions regarding the suitability of specific organs for specific recipients. However, it is possible that at least some organs from these donors are not used for fear of transmitting infections to, or compromising the recipient. The data presented here indicate that it may be possible to successfully transplant more of these organs. Thus, donors with any ‘treatable’ infection should not be ruled out as possible donors. This policy may increase the total number of available organs for transplantation and help improve the organ shortage. The design of our study does not warrant the assertion that organs from donors with infection are safe in all circumstances, but it shows a low rate of transmission of infection from the organ donor to the recipient, and no deleterious impact on recipient survival, thus implying that these organs are perfectly adequate for transplantation. In conclusion, we do not know what influence antimicrobial therapy had on this favorable outcome, but because of the potential lethal consequences of infection in these patients, all efforts should be made to establish the diagnosis as soon as possible and aggressively administer pathogen-specific antimicrobial therapy to organ recipients.
This work was funded by grants from the Spanish Ministry of Health and Consumer Affairs (Ministerio de Sanidad y Consumo), Instituto de Salud Carlos III, Spanish Research Network for the Study of Infection in Transplantation (RESITRA G03/075) and Spanish Network for Research in Infectious Diseases (REIPI RD06/0008). The authors thank Celine Cavallo for English language editing and Jesus Vicente for statistical assistance.
Conflicts of interest statement
The authors declare no conflicts of interest.