In the United States, infection is the second leading cause of death in kidney transplant recipients after cardiovascular disease. Also, recipients have more than 40 times the annual rate of sepsis compared with the general population (1, 2).
Bloodstream infection (BSI) remains as an important cause of morbidity and mortality after solid-organ transplantation (3). Septic shock is the most severe complication in the context of BSI, increasing overall mortality (4). Case fatality rate from BSI causing organ dysfunction is still high, ranging from 25% to 50% (4–6). BSI affects 19% to 35% of liver transplant patients, 5% of kidney recipients, 25% of lung transplant patients, and 10% of heart transplant patients (6).
Although infection is a leading cause of death in kidney transplant (5), little is known about the incidence, epidemiology, and clinical significance of BSI in this population. Previous studies have suggested that BSI often occurs in other solid-organ transplant recipients, especially in liver allograft recipients, and is associated with deceased donor, acute rejection, diabetes mellitus, and hemodialysis before transplant (7, 8). However, few studies in the literature analyzed BSI infections in kidney transplant recipients (6). In addition, most studies report the whole spectrum of infection in transplant patients and not specifically BSI (6). The aim of this analysis was to study the epidemiology, causative factor, demographic characteristics, risk factors, and outcome of BSI after kidney transplantation.
MATERIALS AND METHODS
We retrospectively reviewed a cohort of kidney transplant patients diagnosed with BSI from 2000 to 2006 in two teaching hospitals. Patients were identified based on the daily listings supplied by the Microbiology Department that reports blood culture results to Hospital do Rim e Hipertensao and Hospital Sao Paulo. Data from cases and controls were collected from medical records. We stratified the study by classifying BSI as early and late before and after 6 months from transplantation, respectively. A case-control study was used to assess available risk factors for BSI occurrence and mortality. The controls were matched by age (±10 years) and time of the transplantation (±1 month). For each patient (case) with BSI, one matched control without BSI during the whole posttransplant period was selected. The variables analyzed were as follows: gender, age, type of last dialysis, transfusion, immunosuppression (cyclosporine, azathioprine, mycophenolate mofetil, tacrolimus, rapamicine, prednisone, antithymocyte antibodies, or anti interleukin-2 receptors), source of the allograft (living or deceased donor), crossmatching for human leukocyte antigen (HLA) compatibility (HLA I was defined as a renal graft from living haploidentical donors, HLA II from identical donor, and HLA III from distinct donor), donor (D) and recipient (R) cytomegalovirus serostatus before transplantation and cytomegalovirus treatment previous to BSI, hepatitis B, C viral serology (recipient only), acute and chronic rejection and treatment employed, Charlson Score (9), Acute Physiology and Chronic Health Evaluation (APACHE) II Score (10), total length of hospitalization, admission diagnoses using International Code of Diseases 10, placement of a ureteric stent after transplantation, type of ureteric implantation, mechanical ventilation, central venous catheter use, shock, and outcome. To evaluate risk factors for mortality, a nested case control comparing survivors and nonsurvivors at 30 days after a BSI episode was used.
This study was conducted at Universidade Federal de São Paulo, where the world's largest active kidney transplant center is located, comprising the Hospital do Rim e Hipertensao and Hospital Sao Paulo, both units dedicated to kidney transplantation and related procedures.
Transplant procedures followed classical techniques. The standard techniques more common for the management of the ureter during renal transplantation with ureteroneocystostomy were as follows: (1) Lich-Gregoir (extravesical technique), where the ureter is identified in the retrovesical space, bluntly dissected, and secured. Once the ureter is mobilized, a detrusor incision is made in a cephalad direction in a 458 angle for a distance of 2 to 3 cm. The ureter is then laid in, and the muscle is closed loosely over the ureter with interrupted, absorbable sutures. (2) Politano-Leadbetter (intravesical technique), where incision is made in the bladder and retractor flattens the trigone. Ureteral dissection is performed. The ureter is inserted into the neohiatus, and a submucosal tunnel is created. Then the distal ureter is sutured with absorbable sutures. Finally, the bladder is closed in two layers of absorbable running sutures. In both techniques, the use of a ureteric stent was decided by the surgical team (usually staying in place for 4 weeks) and the bladder catheter being routinely removed on day 4. Antibiotic prophylaxis included 1 g cephalotin at anesthetic induction and another 500 mg postoperatively each 8 hr for 2 days (with dose adjustment according to renal function), followed by prophylactic trimethoprim-sulfamethoxazole at 80 to 400 mg, two tablets per day for 6 months. Acute allograft rejection was diagnosed through biopsy or based on clinical criteria. Episodes of rejection were initially managed with 500 mg methylprednisolone per day during 3 days and, in case of steroid resistance, 1 to 1.5 mg/kg thymoglobulin for 7 to 10 days.
The primary diagnosis was defined according to the International Classification of Diseases, 10th Revision. Comorbidities were rated according to the 19-category score, where each category is weighed based on the adjusted 1-year mortality risk, as proposed by Charlson et al. (9). The severity of the patients' condition was measured with the APACHE II Score (10).
Patients younger than 18 years, contaminants, and polymicrobial infections were excluded.
For each blood culture, 20 mL of blood were drawn under sterile conditions; 10 mL were injected into a BACTEC Aerobic/F bottle, and the remaining 10 mL were injected into a BACTEC Myco/F Lytic bottle (Bactec 9240, Becton Dickinson Diagnostic Instruments Systems, Sparks, MD). Isolates were identified through standard microbiological tests (11).
BSI was considered significant according to the Centers for Disease Control criteria (12). A symptomatic BSI was considered on the isolation of a pathogenic microorganism (other than skin contaminants such as diphteroids, Bacillus spp, or coagulase-negative Staphylococcus) in one culture in the presence of signs of infection (chills, fever, and hypotension) or isolation of the contaminant in two consecutive cultures, with signs of infection. Culture specimens were obtained when clinically indicated, and no surveillance was performed routinely (12). BSI was considered community acquired if the first positive blood culture specimen was taken in outpatients or within the first 48 hr after admission (12). After this period, the infection was considered a healthcare-associated infection, according to the Centers for Disease Control (12). Multiple antibiotic resistance was defined as enterococci resistant to vancomycin; Staphylococcus sp resistant to methicillin; enterobacteriaceae resistant to at least two of the following: extended spectrum penicillins, quinolones, third-generation cephalosporins, or aminoglycosides (13). The source of BSI was considered documented if there were focal signs or symptoms of infection and the same microorganism was isolated from blood and the infected site (12). For example, urinary tract infection (UTI) was defined as the presence of bacteriuria (>105 colony-forming units/mL) or pyuria (10 leukocytes/mm3) with symptoms of infection and was also considered when the attending physician instituted treatment for a positive urine culture irrespective of specific clinical findings.
Adequate empirical antimicrobial treatment was defined as therapy administered within 24 hr of sampling for blood culture including any antimicrobial agent to which the organism was susceptible (14).
Clinical condition associated to BSI was classified as systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis, or septic shock (12). SIRS was defined as two or more of the following: (1) temperature of more than 38°C or less than 36°C, (2) heart rate of more than 90 beats per minute, (3) respiratory rate of more than 20 breaths per minute or partial arterial CO2 pressure of less than32 mm Hg, or (4) white blood cell count of more than 12×109/L or less than 4×109/L or the presence of more than 10% immature neutrophils. Sepsis was defined as SIRS associated with an agent isolated from at least one blood culture (14). Severe sepsis was defined as organ dysfunction, hypotension, or systemic manifestations of hypoperfusion (14). Septic shock was defined as severe sepsis associated with hypotension unresponsive to intravenous fluid challenge or the need for treatment with a vasopressor agent (14). The maximal inflammatory response was defined as severe sepsis, septic shock, or death. We considered 30-day mortality for the analyses of risk factors for mortality from BSI.
Sequential univariate and multivariate analyses were performed to compare the characteristics of controls and cases after transplantation and to identify the risk factors for BSI and mortality. Continuous variables were compared by using the Student's t test for normally distributed variables and the Mann-Whitney U test for nonnormally distributed variables. Differences in proportions were compared by a chi-square test or Fisher's exact test, when appropriate. α was set equal to 0.05, and all tests of significance were two tailed. When colinearity was identified between two variables in a correlation matrix, the one with the greatest clinical relevance associated with mortality was included in the multivariate analysis. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for all variables. Variables found to be significant by univariate analysis were then entered into a multivariate model. All statistical analyses were performed using the Statistical Package for the Social Sciences software (version 16.0, SPSS Inc., Chicago, IL).
During the assessed period, 3308 transplant procedures were performed at Hospital do Rim e Hipertensao and at Hospital Sao Paulo, 2322 from living donors and 986 from deceased donors, 2129 in male and 1179 in female recipients. Overall, 200 patients were initially enrolled, and 15 were excluded because they did not meet criteria for true infection (pseudobacteremia). Of the remaining 185 patients, all were included for risk factors analyses, but for the estimation of BSI incidence, we included only those patients transplanted during the period of data collection (2000–2006). As a result, 49 patients were excluded from this calculation, because they were transplanted before the study period. The overall incidence of BSI was 4.11% in this period, and the incidence of BSI in patient-years was one BSI per 145.9 patient-years. The median time for the occurrence of BSI after transplantation was 235 days, with 62% BSI episodes occurring after 6 months posttransplant. Considering all patients, 54%, 44%, 21%, and 5.9% were free form infection at 1, 2, 5, and 10 years, respectively, after transplant (Fig. 1).
The primary source of infection was the urinary tract in 37.8% of patients, and the 30-day mortality rate was 24.3% (Table 1). No deaths occurred among control patients. Gram-negative BSI originated mainly from the urinary tract, whereas gram-positive BSI was related mainly to the central venous catheter (Table 1). When pathogens associated with BSI were stratified by source, Escherichia coli was associated with UTI and Staphylococcus sp with central venous catheter infection (Table 2).
E. coli was the most common agent causing posttransplant BSI, causing 56 (30.3%) infections, of which 32 (57.1%) showed concordance between urine cultures and blood isolates; Klebsiella pneumoniae was recovered from 21 (11.4%), Enterobacter aerogenes from 19 (10.3%), and Pseudomonas aeruginosa from 17 (9.2%) BSI episodes. Coagulase-negative Staphylococcus was the most common gram-positive agent in posttransplant BSI, occurring in 12 (6.5%) patients, whereas Staphylococcus aureus was present in 11 (5.9%) patients. Candida albicans was the most prevalent agent of fungal BSI (2.7%). BSI due to Mycobacterium tuberculosis was recorded in four (2.2%) patients (Table 3).
Infections because of multiresistant agents were caused by E. coli in five (2.7%) patients, K. pneumoniae in eight (4.3%) patients, and E. aerogenes in nine (4.8%) patients. Coagulase-negative methicillin-resistant Staphylococcus was found in five (2.7%) patients and S. aureus in six (3.2%) patients.
Risk Factors Associated With BSI
Analyzing the two time points of BSI (early and late) through multivariate analysis, the independent risk factors for BSI in the early period were acute rejection (OR=3.69 and 95% CI 1.36–9.97, P=0.03), followed by insertion of a ureteric stent after transplantation (OR=3.60 and 95% CI 1.50–7.52, P=0.002) and being a recipient of a deceased donor (OR=3.16 and 95% CI 1.39–7.17, P=0.001). For late BSI, related factors were acute rejection (OR=4.04 and 95% CI 1.78–9.17, P=0.001), Charlson Score (9) more than or equal to 3 (OR=6.98 and 95% CI 2.43–20.01, P=0.01), and being a recipient of a deceased donor (OR=3.37 and 95% CI 1.73–6.55, P=0.001; Table 4).
The following covariates were considered as risk factors for mortality in the univariate analysis: the presence of two or more comorbidities, ureteric stent, APACHE II Score (10) more than or equal to 20, intensive care unit (ICU) hospitalization, acute rejection, deceased donor, shock, respiratory failure, and inadequate antibiotic therapy. In the multivariate analysis, the risk factors independently related to mortality were APACHE II Score (10) more than or equal to 20 (OR=6.39; 95% CI 2.07–19.73, P=0.001), presence of shock at diagnosis (OR=9.87; 95% CI 2.95–32.94, P=0.001), and respiratory failure (OR=7.66; 95% CI 2.56–22.87, P=0.001; Table 5).
Our study includes the largest series of BSI episodes in kidney transplant patients, with an incidence rate of 4.11%. This rate is similar to what is usually reported (6%–11%) (4, 8, 15), but we cannot rule out underestimation because some patients may have been treated in other medical facilities at the time BSI occurred. Conversely, the rate is lower when compared with other groups of solid organ transplants (10%–35%) (6, 16), where ICU stay and use of central venous devices are more frequent than the kidney transplantation setting.
We have shown that BSI was associated with UTI in 37.8% of the cases, and gram-negative pathogens were responsible for most cases of BSI in kidney transplant recipients (9, 11, 15, 16). This has been shown in earlier studies and reflects the vulnerability of the urinary tract to infection in these patients, because of surgical manipulation, chronic local inflammation, and the need for prosthetic devices. Central venous catheter-related infection was also recognized as an important occurrence in our series, consistent with reports of other solid-organ transplant recipients (17, 18). However, different from what is reported in other groups of transplant patients, such infections do not occur early after transplant, where a central venous line is not usually employed. They occur late after transplant during readmissions for clinical complications that require a central line or ICU admission.
M. tuberculosis was isolated in four patients, three with pulmonary and one with urinary infection. Tuberculosis is endemic in Brazil, with a 2.4% incidence in transplant recipients, that is, a prevalence 5-fold higher than the 0.5% estimated for the Brazilian general population (19).
Less than 40% of the patients developed BSI within the first 6 months after transplant. This reflects, as has been stated above, a less complicated postoperative course when compared with other transplants, dispensing ICU stay and a central venous line. During this period, immunologic factors may play a role, because more immunosuppressed patients (i.e., those with an acute rejection episode and those receiving an organ from a deceased donor) were shown to be at increased risk according to the multivariate analysis (20, 21). The anatomy of the urinary tract may also play a role, because the ureteric stent was disclosed as an independent risk factor (22–24). The use of a ureteric stent may reflect a group of patients with a less favorable urinary tract anatomy or constitute a risk factor per se, although this may be a matter of debate (22–24).
Late BSI infections comprise more than 60% of all BSI in this group of patients and general clinical conditions (as reflected by a higher Charlson Score ), and immunologic factors may be directly related (21).
We have shown a higher mortality rate when compared with other studies of kidney transplanted patients (25). The possible explanations are that we have considered mortality because of all causes and not mortality directly related to BSI, sometimes used in other studies (26). Also, a high proportion of our BSI episodes occurred in the course of severe clinical complications after transplantation in the ICU setting, where mortality is higher. In fact, the risk factors associated to death were those related to severity of disease, such as use of mechanical ventilation, vasopressors, and high APACHE Score.
Previous reports suggest that inappropriate antibiotic therapy is an independent risk factor for death in patients with BSI (6, 15). In our analysis, we did not find an evident correlation between the adequate use of antimicrobial therapy and the risk of mortality. Probably, an earlier adjustment of antimicrobial therapy may have contributed to lower the overall mortality rate.
Our study has the major limitation of the retrospective nature of its design, relying on information on clinical definitions and outcomes from medical records and not from direct contact with the patients. We tried to minimize such shortcomings using standard definitions and securing consistency on the analysis of the patients' records.
Some practical implications may arise from the analysis of our data. For instance, the development of guidelines for the removal of the ureteric stent as early as possible may be discussed. Also, a more close follow-up of patients at increased risk (for instance after rejection episodes or recurrent severe UTI), eventually with surveillance blood cultures requisition, may be of benefit. Finally, understanding the variables associated to BSI will help to prioritize resources and plan for strategies to decrease BSI-related mortality in kidney transplant patients (27).
The authors thank Marcello Sampaio Di Pietro for helpful statistical analysis.
1.Abbott KC, Oliver JD, Hypolite I, et al. Hospitalizations for bacterial septicemia after renal transplantation in the United States. Am J Nephrol
2001; 21: 120.
2.Alangaden GJ, Thyagarajan R, Gruber SA, et al. Infectious complications after kidney transplantation: Current epidemiology and associated risk factors. Clin Transplant
2006; 20: 401.
3.Candel FJ, Grima E, Matesanz M, et al. Bacteremia and septic shock after solid-organ transplantation. Transplant Proc
2005; 37: 4097.
4.Harbarth S, Ferrière K, Hugonnet S, et al. Epidemiology and prognostic determinants of bloodstream infections in surgical intensive care. Arch Surg
2002; 137: 1353.
5.Fishman JA. Infection in solid-organ transplant recipients. N Engl J Med
2007; 357: 2601.
6.Palmer SM, Alexander BD, Sanders LL, et al. Significance of blood stream infection after lung transplantation: Analysis in 176 consecutive patients. Transplant
2000; 69: 2360.
7.Foley JM, Paunio M, Lyytikäinen O, et al. Bacteremia among kidney transplant recipients: A case-control study of risk factors and short-term outcomes. Scand J Infect Dis
2000; 32: 69.
8.McClean K, Ktneteman N, Taylor G. Comparative risk of bloodstream infection in organ transplant recipients. Infect Control Hosp Epidemiol
1994; 15: 582.
9.Charlson ME, Sax FL, Mackenzie CR, et al. Assesing illness severity: Does clinical judgment work? J Chronic Dis
1986; 39: 439.
10.Knaus WA, Draper EA, Wagner DP, et al. APACHE II: A severity of disease classification system. Crit Care Med
1985; 13: 818.
11.Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing
. 2006; M2-A9. Wayne, PA.
12.Garner JS, Jarvis WR, Emori TC, et al. CDC definitions for nosocomial infections. Am J Infect Control
1988; 16: 128.
13.Husain S, Chan KM, Palmer SM, et al. Bacteremia in lung transplant recipients in the current era. Am J Transplant
2006; 6: 3000.
14.Kollef MH, Micek ST. Strategies to prevent antimicrobial resistance in the intensive care unit. Crit Care Med
2005; 33: 1845.
15.Silveira FP, Marcos A, Kwak EJ, et al. Bloodstream infections in organ transplant recipients receiving alemtuzumab: No evidence of occurrence of organisms typically associated with profound T cell depletion. J Infect
2006; 53: 241.
16.Dantas SRPE, Kuboyama RH, Mazzali M, et al. Nosocomial infections in renal transplant patients: Risk factors and treatment implications associated with urinary tract and surgical site infections. J Hosp Infect
2006; 63: 117.
17.Clesca P, Dirlando M, Park SI, et al. Thymoglobulin and rate of infectious complications after transplantation. Transplant Proc
2007; 39: 463.
18.Singh N, Wagner MM, Obman A, et al. Bacteremias in liver transplant recipients: Shift toward gram-negative bacteria as predominant pathogens. Liver Transplant
2004; 10: 844.
19.Biz E, Pereira CAP, Moura LAR, et al. The use of cyclosporine modifies the clinical and histoplathological presentation of tuberculosis after renal transplantation. Rev Inst Med Trop S Paulo
2000; 42: 225.
20.Brayman KL, Stephanian E, Matas AJ, et al. Analysis of infectious complications occurring after solid-organ transplantation. Arch Surg
1992; 127: 38.
21.Gill JS, Abichandani R, Kausz AT, et al. Mortality
after kidney transplant failure: The impact of non-immunologic factors. Kidney Int
2002; 62: 1875.
22.Sansalone CV, Maione G, Aseni P, et al. Advantages of short time ureteric stening for prevention of urological complications in kidney transplantation: An 18-year experience. Transplant Proc
2005; 37: 2511.
23.Rabkin DG, Stifelman MD, Birkhoff J, et al. Early catheter removal decreases incidence of urinary tract infections in renal transplant recipients. Transplant Proc
1998; 30: 4314.
24.Osman Y, Ali-El-Dein B, Shokeir AA, et al. Routine insertion of ureteral stent in live-donor renal transplantation: Is it worthwhile. Urol
2005; 65: 867.
25.Lumbreras C, Sanz F, González A, et al. Clinical significance of donor-unrecognized bacteremia in the outcome of solid-organ transplant recipients. Clin Infect Dis
2001; 33: 722.
26.Moreno A, Cervera C, Gavaldá J, et al. Bloodstream infections among transplant recipients: Results of a nationwide surveillance in Spain. Am J Transplant
2007; 7: 2579.
27.Diekema DJ, Beekmann SE, Chapin KC, et al. Epidemiology and outcome of nosocomial and community-onset bloodstream infection. J Clin Microbiol
2003; 41: 3655.