Objectives: Prevention of sepsis-related organ dysfunction in septic donors is crucial. In this study, septic donors were followed-up based on donor-Sequential Organ Failure Assessment criteria.
Materials and Methods:Between January 2014 and 2020 at our center, 29 primary kidney transplant recipients received organs from 20 septic donors. All donors received either pathogen-specific or broad-spectrum antibiotics at least 48 hours before procurement, and all recipients received similar treatment posttransplant for an average of 7 to 14 days. Donor eligibility was determined according to the sum of donor-Sequential Organ Failure Assessment scores obtained from 6 parameters: Pao2/Fio2ratio; platelet count; serum bilirubin, creatinine, and lactate levels; and presence of hypotension. The cut-off value for bacteremic donor acceptance was below 12 points.
Results: Fever (≥38 °C) persisted in 5 donors in the last 24 hours before organ removal. However, in these 5 donors, the mean donor-Sequential Organ Failure Assessment score was 6.5 ± 1.1, mean arterial pressure was >70 mm Hg, and serum lactate levels were <2 mmol/L. Fifteen donors had systemic inflammatory response syndrome scores of ≤2 with corresponding donor-Sequential Organ Failure Assessment scores of 7.9 ± 1.2; none had systemic inflammatory response syndrome scores >3, which would have indicated severe organ failure. In 28 recipients (97%), no donor-related infections were observed in the perioperative first month and afterwards.
Conclusions: Treatment of donors and recipients with a common protocol greatly reduced the risk of donor-induced infection transmission. In addition, we found the donor-Sequential Organ Failure Assessment criteria to be a helpful tool in predicting organ failure in infected donors.
Key words : Donor infection, Organ failure, SOFA score
Publications from 2010 and earlier have commonly reported the risk of contamination in recipients of organs transplanted from donors with documented bacteremia and have emphasized the importance of blocking organ donation from donors with a definitive diagnosis of sepsis.1-3 However, despite preventive strategies, such as screening donors for infection and the use of prophylactic antimicrobials guided by susceptibility testing, some centers have still reported significant morbidity (for example, perirenal abscess, peritonitis, rupture of arterial anastomosis) and mortality in cases where patients have clearly received organs from infected donors.4-6
Each year, the gap between organ supply and demand increases disproportionately worldwide. New strategies, including the utilization of infected donors, are being developed to expand the donor organ pool. Although the cases are rare and dependent on length of hospitalization, potential donors admitted to the intensive care units (ICU) can become infected with multidrug-resistant (MDR) organisms, which can occur in as little as 2 days.7,8 “MDR organisms” is the international terminology used to describe bacterial species with in vitro resistance to more than 1 antimicrobial agent and often refers to gram-negative and gram-positive bacteria, including carbapenemase-producing Klebsiella pneumoniae and Acinetobacter spp., extended-spectrum beta-lactamase-producing (ESBL+) Enterobacteriaceae, and methicillin-resistant Staphylococcus aureus.9
When specific treatment for the causative microorganism is given at least 48 hours before organ procurement, successful transplants using organs from donors with bacterial septic shock, pneumonia, or meningitis have been reported.10,11 The information available on the use of donor organs colonized or infected with MDR gram-negative bacteria, which are more difficult to manage both before organ retrieval and after transplant, is very limited.5,12,13 Few original articles in the literature are available that focus on MDR Pseudomonas aeruginosa and Acinetobacter baumannii infections in recipients.14 Nonetheless, these 2 microorganisms are the major causes of serious nosocomial infections in the ICU and have proved resistant to most broad-spectrum antibiotics for many years.15
Before considering a bacteremic donor for organ procurement, we need to understand how infection transmission can be minimized or, best of all, how it can be prevented. One important step is the early diagnosis of sepsis and septic shock in such donors. The definition of sepsis in this context is essentially based on infection and usually emphasizes the response of the host.16 To ensure the safety of the transplantation, it is also crucial to demonstrate the safe clinical course of the donor in terms of sepsis and sepsis-related organ dysfunction, with the use of reliable and practical criteria. For this purpose, we developed a modified Sequential (Sepsis-related) Organ Failure Assessment (SOFA) scoring system to evaluate organ failure in the donor (Table 1).17,18
The SOFA scoring system is a valuable tool in predicting mortality rates based on the degree of failure in various organ systems in ICU patients. The different SOFA scores and corresponding mortality rates are as follows: between 0 and 6, a mortality rate of <10% should be expected and scores between 6 and 9 or 10 and 12 predict ≤33% expected mortality or 34% to 50% mortality, respectively. All donors constituting our study group had a definitive diagnosis of brain and brainstem death and unexceptionally had a Glasgow coma score of ≤4. Because the Glasgow coma score was not expected to change in any donors and became a mathematical constant for all SOFA score calculations, we eliminated this parameter and replaced it with serum lactate levels. There is a direct relationship between blood lactate levels and sepsis and sepsis-related mortality. Recent studies have shown that incorporating blood lactate level measurements into the SOFA system generates an increased sensitivity in predicting mortality.19 After making these modifications, we named this new evaluation system the donor-SOFA (d-SOFA). Because a d-SOFA score of 12 and above indicated severe failure in at least 2 organ systems and, consequently, poor survival outcomes, we decided our threshold (the cut-off point) for bacteremic donor acceptance into our study should be <12 points.
In addition to this, we regularly monitored serum lactate levels as a tissue perfusion parameter and high-sensitivity C-reactive protein (CRP) levels with systemic inflammatory response syndrome (SIRS) criteria to follow the clinical course of dysregulated immune response of donors to sepsis. Systemic inflammatory response syndrome criteria included the following: a temperature higher than 38 °C or lower than 36 °C, a heart rate > 90 beats/minute, Paco2 <32 mm Hg, white blood cell count >12 000 /mm3 or <4000/mm3, or greater than 10% immature (band) forms.
In this study, we report the clinical outcomes of patients who received kidneys from septic donors with a history of infection by MDR organisms and other bacteria. Furthermore, we prepared a follow-up scheme, essentially based on d-SOFA criteria and serum lactate levels, to detect changes in the intensive care process and to reveal which septic donors could be used safely.
Materials and Methods
Patients who underwent deceased donor kidney transplant at our center between January 2014 and January 2020 were evaluated. Data on demographic characteristics, ICU records, and microbiological information of donors, along with any related medical treatment, were obtained from the donor hospital registration system. Nine of the deceased donors were from our own hospital, and 11 were from other medical facilities. Approval for the use of septic donors was decided jointly by the senior nephrologist and organ transplant surgeon at our center. An infectious disease specialist was also included in the decision-making phase regarding the choice of antibiotics to be administered in both the donor and recipient. Given that meningoencephalitis may be mistaken for a stroke or may not be recognized during standard screening, the organs of one particular donor, in whom the etiology of brain and brainstem death meningoencephalitis could not be excluded, were rejected. All septic donors in the ICU of our institution were treated with pathogen-specific antibiotics, and those from other hospitals received either pathogen-specific or broad-spectrum antibiotics at least 48 hours before procurement.
Infection in donors in the ICU was defined as fever >38 °C that was not present at admission but that developed 48 hours or more after admission and associated with a positive culture (blood, urine, sputum, etc).20 Diagnosis of pneumonia was based on chest computed tomography findings and clinical signs. In line with current consensus, we defined sepsis as life-threatening organ dysfunction caused by the host’s irregular response to infection. Septic shock was defined as poor tissue perfusion reflected by serum lactate level greater than 2 mmol/L (>18 mg/dL) and vasopressor requirement to maintain a mean arterial pressure (MAP) of at least 70 mm Hg.21
From all potential donors whose febrile periods started in the ICU, we drew at least 2 blood samples (1 from the central venous catheter), 1 endotracheal aspiration, and 1 urinary catheter sample for culture and susceptibility testing. All of these donors received ampicillin-sulbactam or cefoperazone-sulbactam prophylaxis until culture susceptibility results were available. Antibiotic doses were adjusted according to the current kidney function.
Most donors in the ICU and all subsequent recipients of kidneys from donors infected by carbapenem-resistant (CR) or carbapenem-intermediate (CI) category gram-negative bacteria were managed as much as possible with a combination therapy that included tigecycline (100 mg twice per day for the first 3 days, followed by 50 mg twice per day) and prolonged infusion carbapenems (either meropenem 3 g/day or imipenem cilastatin sodium 2 g/day) regardless of kidney function. The same regimen was applied to the recipients beginning on the day of transplant for an average of 7 to 14 days. Unfortunately, according to their preference, ICUs of some different hospitals only applied carbapenem to donors infected by CR or CI category gram-negative bacteria. Other kidney transplant patients received prophylactic antibiotherapy based on positive culture isolates and the sensitivity results of the donor. Serial blood and urine cultures were taken from all patients on the first day, first week, and whenever clinically indicated after surgery.
In deceased donors, acute kidney injury before donation was described according to the Kidney Disease Improving Global Outcomes criteria and was defined as an absolute increase in serum creatinine of at least 0.3 mg/dL within 48 hours or by a 50% increase in serum creatinine from baseline within 7 days or urine volume of less than 0.5 mL/kg/hour.22Antithymocyte globulin (ATG) dose of 2 to 4 mg/kg/day (ATG-Fresenius S 20 mg/mL, Fresenius Biotech) was used for induction immunosuppression. The average total ATG dose was 744 ± 236 mg. Maintenance immunosuppression consisted of tacrolimus, an antimetabolite (mycophenolate mofetil or mycophenolic acid), and methylprednisolone. Optimal trough blood concentrations of tacrolimus were kept between 5.0 and 10.0 ng/mL for the first year after transplant.
As an established protocol of our center, an implantation biopsy was carried out on all patients in the operating room after the revascularization of the graft, with a protocol biopsy, by preference in the first 6 months posttransplant, performed unless the patient refused permission.
Those with delayed graft function (DGF) underwent protocol biopsy between day 5 and day 7 after transplant. Patients with sudden deterioration of graft function during follow-up received an indication biopsy.
All statistical analyses were performed using SPSS version 16.0. P < .05 was considered to be statistically significant. For normally distributed data, we compared means by t tests; when data were not normally distributed, we used the Mann-Whitney U test. Fisher exact tests were used for comparison of proportions.
Over a 6-year period, we performed a total of 346 kidney transplants at our clinic. Of these, 209 were from deceased donors. Twenty donors met the inclusion criteria for this study. The causes of brain death were hemorrhagic stroke in 12 donors, ischemic stroke in 3 donors, and hypoxic encephalopathy in 5 donors. Duration of stay in the ICU was 9.9 ± 5.0 days (range, 4-22 days). Mean age of donors was 49 ± 13 years (range, 23-69 years). Both kidneys from the 9 donors at our hospital and 1 kidney from each of the 11 donors from other hospitals were implanted in 29 separate recipients. Seven kidneys (24%) were retrieved from expanded criteria donors. Kidneys of 3 donors, who developed acute kidney injury during ICU stay, were implanted in 5 recipients. Table 2 presents the demographic characteristics and laboratory results of kidney donors.
Despite antibiotic treatment, fever (≥38 °C) persisted in 5 donors in the last 24 hours before organ removal, with 4 having SIRS scores ≥3. Table 3 shows the length of ICU stay, foci of infection, isolated microorganisms, and the treatment applied for these donors. However, in these 5 donors, the final d-SOFA score was <8 (mean of 6.5 ± 1.1) and the MAP was >70 mm Hg. Serum lactate levels were <2 mmol/L in 4 donors and 3.0 mmol/L in 1 donor. These results clearly indicated that all 5 of these donors had received adequate fluid resuscitation and had no microperfusion defect or serious organ damage.
Average CRP values of 20 donors during ICU admission and in the last 24 hours before donor operation were 39.6 ± 64.6 mg/L and 297.0 ± 252.0 mg/L, respectively (t test = -4.6, P < .001). In fact, there were no donors with normal CRP values in the last 24 hours before organ procurement. The mean SIRS score of the 20 donors in the last 24 hours before organ retrieval was 1.9 ± 1.0. In the final assessment, 15 donors had a SIRS score of ≤2 (mean of 1.5 ± 0.7). The corresponding d-SOFA score of these patients in the last 24 hours before organ retrieval was 7.9 ± 1.2 (range, 5-10); none had SIRS score of >3, which would indicate severe organ failure. In this group, the mean serum lactate level was 1.87 ± 0.66 mmol/L (range, 0.7-2.7 mmol/L) and the mean MAP was 79.0 ± 13.8 mm Hg (range, 50-107 mm Hg) (Table 2). There was only one donor with a MAP value <65 mm Hg in this group.
Bacterial species were identified in the tracheal aspirates of 11 donors, the urine cultures of 9 donors, and the blood cultures of 7 donors. In 6 donors, there was more than 1 site of infection. Eleven donors were diagnosed with ventilator-associated pneumonia (2 in simultaneous blood and urine cultures, 2 in blood cultures, and 1 in urine culture growth), whereas 9 had urinary tract infections (3 in either blood or tracheal aspirate or both blood and tracheal aspirate culture growths). Serial chest radiographies or computed tomography scans of the 11 donors with positive tracheal aspirate cultures were suggestive of pneumonia.
Twenty donors had gram-negative, 9 had gram-positive, and 3 had both gram-negative and gram-positive isolates. More than 1 bacterial strain was isolated in 7 cases.
Among gram-negative bacteria, Acinetobacter bauman-nii was the most common, being isolated in 10 cases of which 9 were CR. In addition to this bacteria, ESBL+ Escherichia coli was found in 5 cases and carbapenem-sensitive (CS) and ESBL+ Klebsiella pneumonia were found in 3 cases. In addition, inducible beta-lactamase-producing (IBL+) Morganella morganii and CS Pseudomonas aeruginosa were isolated in 1 case each.
Among gram-positive bacteria, the isolated strains were ampicillin-sensitive Enterococcus faecalis in 3 donors and Staphylococcus aureus in 6 donors (3 methicillin-resistant and 3 methicillin-sensitive). No donors had vancomycin-resistant gram-positive growth. Overall, we isolated a total of 29 bacterial growths comprising 8 different bacteria in 3 different culture sites. These data are shown in Table 4.
Clinical characteristics of recipients
The mean age of the 29 primary kidney transplant recipients was 46 ± 11 years (range, 25-64 years) and 18 (62%) were men. Panel-reactive antibody tests for all patients were negative, and the mean HLA mismatch was 3.6 ± 0.7 (range, 2-5). Cold ischemia time was 14.8 ± 3.5 hours, and DGF was observed in 10 recipients (34%). The average duration of DGF was 9.7 ± 7.1 days. The average length of hospitalization was 14.1 ± 6.3 days (range, 7-32 days). As prophylactic antibiotherapy, carbapenems were administered in 18 (6 received simultaneous tigecycline), piperacillin-tazobactam in 7, cefoperazone-sulbactam in 2, and vancomycin and cefazolin sodium combination in 2 recipients. Three recipients received additional glycopeptide antibiotics. Mean duration of antibiotherapy was 8.8 ± 3.0 days (range, 7-14 days).
None of the recipients had histological features of graft infection in their implantation biopsy. The mean baseline and final serum creatinine values of recipients were 1.6 ± 0.4 mg/dL and 1.5 ± 0.6 mg/dL, respectively. The mean follow-up was 21.9 ± 16.1 months. In 28 recipients (97%), no donor-related infections were observed in the first perioperative month. In 1 recipient with an initially functioning graft, we observed an increase in serum creatinine in the second week posttransplant. The recipient was asymptomatic; he had no fever and no pyuria or bacteriuria in the urinalysis. However, the indication biopsy on day 14 posttransplant revealed graft infection, despite administration of meropenem 3 g/day to the donor for 2 days before organ donation and to the recipient for 7 days after implantation. The primary disease of this recipient was chronic pyelonephritis due to vesicoureteral reflux. The donor of this patient had ESBL+ Klebsiella pneumonia growth in the endotracheal aspiration culture and IBL+ Morganella morganii growth in the urine culture. We did not isolate any bacterial species in consecutive urine cultures of the recipient before and after the biopsy. Thus, we considered this case as a possible donor-derived infection. The SIRS and d-SOFA scores of the donor were 2 and 7, respectively. The recipient recovered uneventfully and achieved a baseline serum creatinine level (1.8 mg/dL) with a 10-day course of meropenem (3 g/day) and tigecycline (100 mg/day) treatment. This recipient should have received meropenem and tigecycline for 7 to 14 days as per protocol. However, he had meropenem prophylaxis for only 7 days, and this may have caused graft infection. After 3 weeks, he underwent a protocol biopsy where the histopathological features were consistent with subclinical T-cell-mediated rejection (i1, t1, v0, ptc0, C4d negative). He then received pulse steroid treatment (methylprednisolone 500 mg/day for 3 days). At the 10-month follow-up, the patient had uneventful findings and renal functions had remained stable.
Apart from this single case, no donor-related infections were encountered in any recipient in the early postoperative period (first 30 days posttransplant) or in the first year of follow-up. Recipients from donors with CR Acinetobacter baumannii did not develop any infections. This can be explained by the fact that all isolates from these donors were sensitive to tigecycline in in vitro susceptibility tests. Moreover, including the 8 recipients who had transplants from 6 donors with urinary sepsis, no histopathological features of graft infection were observed in any of the indications or protocol biopsies applied to the 13 recipients during the first 3 months and 10 recipients in the first year posttransplant.
In our study, the threshold value for d-SOFA score was determined as 12, and a (possible) donor-related infection was detected in only 1 of the 29 recipients after transplant of organs from septic donors with values lower than this score.
During follow-up, 1 recipient with excellent kidney function lost his graft because of acute coronary syndrome-induced renal cortical necrosis. In addition, 1 other patient died with a functional graft due to a cardiovascular event that developed in the second month after kidney transplant. One-year graft and patient survival rates were 93% and 97%, respectively.
In this study, we described the clinical courses of 29 patients who received kidneys from 20 septic donors. In all donors, both the site of infection and the causal organism were precisely determined subsequent to organ procurement. All of these donors received either broad-spectrum antibiotics or antibiotics that could be adjusted according to the susceptibility testing. The same treatment protocols were applied prophylactically to the corresponding recipients on the day of implantation and afterwards. However, antibiotic treatment was necessarily modified in recipients of donors with fever before donation. With this method, we encountered only 1 transmission of donor-induced infection (3%) in the early period posttransplant. This patient was anuric before transplant and had no previous urinary culture report indicating a specific isolate during the hemodialysis period. However, the indication biopsy revealed graft infection. Although there was no bacterial growth in the successive urine cultures taken after the biopsy, the patient benefited greatly from a regimen of antibiotics effective against the bacteria(s) detected in the infected donor. Consequently, the most likely explanation for this graft infection was the transmission of donor infection. Our results indicated that organ transplant can be performed safely with the application of effective prophylactic antibiotic therapy both to the bacteremic donor and to the corresponding recipient.
Similar results have been obtained in many studies, providing there was no septic shock in the donor. Zibari and colleagues23 administered broad-spectrum antibiotics to both donor and recipient and reported favorable survival outcomes with no sepsis in 16 kidney transplant recipients. In a report of 19 patients who received kidneys from infected donors, Outerelo and colleagues24 found no transmissions of donor-derived infection and reported 1-year graft and patient survival rates of 90% and 100%, respectively. Most interestingly, in their report of outcomes in 14 recipients who received transplants from donors with MDR-bacteremia, Mularoni and colleagues12 emphasized the importance of appropriate antibiotic therapy, citing that transmission did not occur in 10 of 14 recipients who received effective prophylactic antibiotics on the day of transplant and for at least 7 days posttransplant. Unfortunately, all of these studies encompass very few cases, and it is not clear how sepsis, septic shock, and organ failure were monitored in their bacteremic donors. In our study, we applied SIRS criteria to monitor the immune response and the d-SOFA scoring system to monitor sepsis-related organ failure in donors. In 20 donors, the average d-SOFA score in the last 24 hours before transplant was 7.9 ± 1.2. Although such a value predicts that approximately one-third of these donors may develop sepsis-related death, we did not observe any deaths in the whole group, possibly due to the low serum lactate levels (mean of 1.87 ± 0.7 mmol/L; range, 0.7-3.0 mmol/L) and the low number of donors with MAP <70 mm Hg (1 patient, 5%) in the final analysis.
Donor-induced septic complications that may develop in the recipient are frequently encountered early after organ transplant, especially when immunosuppressive therapy is at its peak.1,3,7 In our study, we showed that organs from bacteremic donors whose culture and antibiotic susceptibility test results are clearly known can be used safely with appropriate antibiotic treatment. It seems that the procedure of using the same treatment for both donor and recipient reduced the risk of transmission from the donor to below 5%. The same principle also applied to donors infected with MDR-gram negative bacteria in our series. The observation of only 1 donor-derived infection transmission from a total of 20 donors with MDR gram-negative bacteria (9 CR- and 9 beta-lactamase-producing) supports our interpretation. Similarly, in another study, implementation of an antibiotic regimen compatible with in vitro susceptibility tests yielded excellent results in 4 patients who received organs from a donor with carbapenemase-producing Klebsiella pneumonia sepsis.5Furthermore, the effectiveness of our carbapenem and tigecycline combination regimen against CI and CS Acinetobacter species isolates was apparent, even without antibiotic susceptibility testing results. In a previous study, after in vitro antibiotic combination testing, the same combination therapy was administered to 5 solid-organ recipients with MDR Acinetobacter baumannii infection, with successful treatment in 4 of 5 recipients (80%).25 In our study, sepsis-associated organ failure was predicted using our d-SOFA method. Interestingly, all of the 29 kidneys procured from septic donors functioned properly. This can be explained by the achievement of adequate microperfusion and hemodynamic stability in almost all donors, reflected by low serum lactate and MAP scores (<2) in the final check-up.
Herein, we intended to establish that the value of d-SOFA (that is, the severity of the infection) is a factor that predicts the transmission of the infection from the donor to the recipient in kidney transplantation. However, we did not have a comparator group of patients with values greater than 12 points. This and the relatively low number of cases are the most important limitations of this comprehensive study, where the d-SOFA score system was used for the first time in follow-up of infected donors. A recent study stated that the use of organs from donors with MDR gram-negative bacteria growth was not preferred and reported that the number of organs utilized from these donors had been gradually decreasing.26 Therefore, our results deserve to be prospectively validated to remedy this problem, which is still valid today. Finally, the availability of donor information and the notification of recipient hospitals are prerequisites for the use of infected donors. In conclusion, treatment of donors and recipients with a common protocol greatly reduced the risk of donor-induced infection transmission.10-12,21,22,27 In addition, the d-SOFA criteria may be helpful in predicting organ failure in infected donors.
Volume : 20
Issue : 3
Pages : 61 - 68
DOI : 10.6002/ect.MESOT2021.O25
From the 1Department of General Surgery and Transplantation,University of Health Sciences, the 2Department of Pathology, University of Health Sciences, and the 3Department of Nephrology and Transplantation, University of Health Sciences, Izmir Bozyaka Education and Research Hospital, Izmir, Turkey
Acknowledgements: Native-speaker English editing was performed by Claire Olmez, MEd, MSc, ELT. The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article and have received no financial support for the research, authorship, and/or publication of this article.
Corresponding author: Cenk Simsek, University of Health Sciences, Izmir Bozyaka Education and Research Hospital, Department of General Surgery and Transplantation, Karabaglar, Izmir 35100, Turkey
Phone: +90 232 2505050
Table 1. Donor-Sequential Organ Failure Assessment Score
Table 3. Donors With Persistant Fever in the Last 24 Hours Before Procurement
Table 4. Isolated Organisms and Culture Sites