Infectious diseases, especially bloodstream infection (BSI), are a major cause of morbidity and mortality after solid organ transplantation (SOT),1,2 including pediatric liver transplantation (LT).3
There are several reports that have evaluated posttransplant BSI, most of which have focused on early-onset BSI because infections and other complications occur predominantly during the early posttransplantation period.2,4–10 In adult LT, Sganga et al4 reported that 28% of recipients developed BSI in the first 60 days after LT. Furthermore, in Japanese adult living donor LT, 34.3% of LT recipients developed BSI in the first 90 days after LT and had a higher mortality rate than recipients without BSI.5 Pediatric LT recipients have also been reported to develop BSI at a rate of 21–25% during the early periods after LT.6,7 BSIs that occur during the early period after LT are predominately related to intraabdominal infection and catheter-related bloodstream infection (CRBSI).4,6,7,11
On the other hand, few studies have evaluated BSI in the late period after LT. Although the prevalence of BSI appears to decrease with time after LT, recipients remain at a high risk of infection (0.3 per 1000 transplant-days: > 6 months post-LT),1,12 and late-onset infection after SOT has been reported to be an independent risk factor for mortality (hazard ratio, 9.12; P < 0.001).13 Furthermore, 90-day all-cause mortality rates after late-onset BSI in adult LT have been reported to be nearly as high as those of early onset BSI (17.6% versus 12.7%).14 However, the etiology and microbiology of late-onset BSI appears to differ from that of the early period.1,2,12,14,15 Thus, it is important to understand the characteristics and risk factors for late-onset BSI after LT to correctly direct empiric management. In this study, we evaluated the characteristics and risk factors of BSI during the late period after pediatric LT.
MATERIALS AND METHODS
This is a retrospective cohort study conducted at the National Center for Child Health and Development (NCCHD) in Tokyo, the largest pediatric LT center in Japan, from November 2005 to March 2016. We evaluated the characteristics of late-onset BSI and of recipients with late-onset BSI.
Demographic data at the time of LT (age, sex, body weight, underlying disease, donor type, ABO blood group match, and pediatric end-stage liver disease [PELD] score, vaccine status before LT), information related to surgery (graft to recipient weight ratio, blood loss [per body weight] and operative time), and clinical course during the first 6 months after LT (Epstein-Barr virus [EBV] infection, cytomegalovirus [CMV] infection, early-onset BSI, duration of hospital stay, biliary stenosis, immunosuppressive agents and rejection) were obtained from the LT database and medical records. Data regarding late-onset BSI (isolated organisms, the site of infection, days after LT, immunosuppressive therapy and in- or outpatient) and prognosis (retransplantation and death) were also collected.
We included recipients who underwent LT at NCCHD during the study period. Recipients older than 18 years of age, those with transplantation for graft failure after LT and cases of death or retransplantation within the first 6 months were excluded.
This study was approved by the Institutional Review Board at the NCCHD.
Our primary analysis aimed to define the clinical characteristics of late-onset BSI, such as the causative organisms and site of BSI.
The secondary analysis focused on defining the risk factors for developing late-onset BSI after 6 months post-LT. We also evaluated whether BSI after 6 months post-LT was associated with long-term outcomes such as retransplantation and death.
Early-onset BSI was defined as BSI that occurred within the first 6 months after LT, and late-onset BSI was defined as onset beyond 6 months after LT.
BSI was defined as an infection that met at least one of the following criteria: (1) isolation of a recognized pathogen cultured from one or more blood cultures and (2) isolation of organisms that are known contaminants (ie, diphtheroids [Corynebacterium spp], Bacillus [not B. anthracis] spp, Propionibacterium spp, coagulase-negative staphylococci [including S epidermidis], viridans group streptococci, Aerococcus spp, Micrococcus spp) from two or more blood cultures drawn on separate occasions16 in the presence of at least one of the following signs or symptoms: fever, chills or hypotension (if patients were < 1 year of age: fever, hypothermia, apnea or bradycardia). Isolations of organisms known as skin contaminants from one set of blood cultures were excluded. BSI due to the same pathogen at the same site of infection within 30 days was considered a single episode.
CRBSI was defined as BSI with isolation of the same organism from both the catheter tip and the blood culture or the same organisms from two blood cultures taken from central and peripheral blood with no other apparent source of the BSI.17
Intraabdominal infection was defined as BSI with at least one of the following criteria: (1) isolation of the same organisms from purulent material obtained from the intraabdominal space; (2) abscess or other evidence of intraabdominal infection observed during surgical operation or histopathologic examination; or (3) at least two of the following signs or symptoms with no other recognized causes: fever, nausea, vomiting, abdominal pain or jaundice and radiographic evidence of infection (eg, abdominal findings on ultrasound, computed tomography scan, magnetic resonance imaging or abdominal x-ray).16 Cholangitis, peritonitis and intraabdominal abscess were included in this category.16
Other sites of infection of BSI were defined using criteria proposed by the U.S. Centers for Disease Control and Prevention.16 Cases that did not meet criteria were classified as BSI from an unknown focus.
EBV infection was defined as the detection of EBV-DNA in the peripheral blood cells by polymerase chain reaction as previously described.18 CMV infection was defined as the detection of CMV antigenemia, with or without CMV disease. In our institution, preemptive therapy was used to prevent CMV disease in all recipients including high-risk serostatus groups.19,20
We examined the vaccine status before LT for pneumococcal conjugate vaccine (PCV) 7 or 13 and Haemophilus influenzae type b (Hib). One or more inoculations were considered a history of positive vaccine history. In Japan, Hib vaccine was licensed in 2008, PCV7 in 2010, and PCV13 in 2013. Meningococcus vaccine is not included in the recommended Japanese vaccine schedule.
The standard immunosuppressive regimen after LT consisted of tacrolimus and steroids.7 Tacrolimus was started one day after LT, and the dose was adjusted to maintain a trough level of 10–15 mg/L for the first 2 weeks, followed by 8–10 mg/L (for days 15–28 post-LT) and was reduced thereafter. Methylprednisolone was started intraoperatively (10 mg/kg/dose) and continued with tapering. Steroids were weaned off when the recipients’ liver function became stable. Recipients older than or equal to one year of age with ABO-incompatible donors received Rituximab 2 weeks before LT and mycophenolate mofetil (MMF) after LT. When acute rejection was suspected, recipients were generally treated with high-dose methylprednisolone. Recipients showing adverse effects of immunosuppressive agents, steroid-resistant rejection, or refractory rejection were managed on a case-by-case basis. Rejection was defined based on the results of liver biopsy and histologic scores (Rejection Activity index) of 3 or higher, graded according to the Banff criteria.21
Ampicillin (120 mg/kg/day, q6 hrs) and cefotaxime (120 mg/kg/day, q6 hrs) were administered intravenously within 1 hour before LT and continued for 48 hours after surgery. Perioperative prophylactic regimen was also modified according to the patient’s history of infection and colonization. Trimethoprim-sulfamethoxazol was prescribed from 1 week to at least 3 months after LT to prevent Pneumocystis jirovecii pneumonia.
In our hospital, when LT recipients developed fever or abnormal vital signs during the late-periods, blood examination and blood cultures are routinely performed. Other investigations such as urine and other cultures, ultrasonographic examination were performed according to symptoms. An infectious diseases specialist chose antibiotics based on disease severity, history of past infections, and past culture results.
Data were analyzed using SPSS Statistics 23.0 software (SPSS, Chicago, IL). We evaluated the risk factors for developing late-onset BSI 6 months after LT. Categorical variables were compared using χ2 test or Fisher exact test as needed. Continuous variables were described as the median and interquartile range (IQR) and were compared using Mann-Whitney U test. Multivariable logistic regression analysis was performed, adjusting for covariates if variables showed an association with P < 0.1 in the univariate analysis. The covariate “history of early-onset BSI” was also introduced in the multivariate models regardless of statistical significance because of the known associations.22
Kaplan-Meier analysis and log-rank test were used to analyze the retransplantation and mortality rates.
P values less than 0.05 were considered statistically significant.
In total, 392 LTs were performed in 379 recipients at the NCCHD during the study period. Fifty-two recipients were excluded: 10 recipients were older than 18 years of age, 15 underwent LT for graft failure after LT, and 31 died or required retransplantation within the first 6 months (four LT overlapped). Ultimately, we evaluated 340 LT recipients. Median follow-up duration was 1571 days (IQR 711–2608 days). Patients’ demographics are shown in Table 1.
Etiology of Late-onset BSI
After LT, 188 blood cultures were positive, including 33 contaminants. Of the 155 BSI episodes, 117 developed during the first 6 months after LT, most of which occurred during hospitalization (Fig., Supplemental Digital Content 1, https://links.lww.com/INF/C833). The remaining 38 episodes were considered late-onset BSI. Median postoperative days of late-onset BSI was 811 days (IQR, 366–1405 days). Twenty-nine (9%) LT recipients developed late-onset BSI, including six recipients with more than one late-onset BSI. Twenty (53%) of the late-onset BSIs episodes occurred in hospitalized patients, and 18 (47%) were in outpatients. The prevalence of late-onset BSI after LT was high during 2005–2007 but decreased after that (Fig., Supplemental Digital Content 2, https://links.lww.com/INF/C834). There were 42 organisms (nine Gram-positive cocci, 33 Gram-negative rods [GNR]) isolated from the blood cultures of recipients with late-onset BSI (Table 2). There were no vancomycin-resistant enterococci in this study. Among 33 GNR, there were eight antimicrobial resistant organisms.
The most frequent sites of late-onset BSI was intraabdominal infection (18, 47%), including 14 cholangitis and four peritonitis. Three BSI classified as intraabdominal infection during hospitalization developed after percutaneous transhepatic cholangio drainage (PTCD) for biliary stenosis. One developed cholangitis due to E. faecium (vancomycin susceptible) 2 days after clamping the PTCD tube inserted 4 days before clamping. She had received one dose of cefoperazon/sulbactam for prophylaxis before PTCD. The others developed bacteremia within 12 hours of PTCD; one due to extended-spectrum beta-lactamase producing E. coli (cefmetazole susceptible) after receiving cefmetazol for prophylaxis, and another due to metallo-beta-lactamase producing K. pneumoniae after receiving piperacillin/tazobactam. All intraabdominal infection with BSI in outpatients were cholangitis secondary to biliary stenosis.
There were also 14 (37%) episodes of BSI with an unknown focus. Ten of these were caused by GNR without cholangitis or any abdominal symptoms, whereas three were caused by Streptococcus pneumoniae. All of them underwent transplantation before 1-year of age and developed pneumococcal bacteremia without a focus at the age of 1, 2, and 4 after 273, 792, and 1234 postoperative days, respectively. Two had not received PCV because these episodes occurred before PCV7 was introduced to Japan, and one had received one dose of PCV7 after LT but was infected by a non-PCV7 serotype (serotype 15A).
Four recipients who had bacteremia with an unknown focus were maintained on only tacrolimus at a trough level of less than 3 mg/L. One patient with a urinary tract infection was receiving treatment for refractory rejection with prednisolone, MMF and rapamycin.
Six episodes of late-onset BSI occurred while the patient was still under trimethoprim-sulfamethoxazole prophylaxis.
Risk Factors for Developing Late-onset BSI
We compared the characteristics of recipients with and without late-onset BSI. Twenty-nine LT recipients (9%) developed late-onset BSI. In the univariate analysis, blood loss > 60 mL/kg during the operation (odds ratio [OR] =2.89; 95% confidence interval [CI], 1.24–6.73; P = 0.01), prolonged operative time > 12 hours (OR = 6.28; 95% CI, 2.32–17.00; P = 0.001), history of early-onset BSI (OR = 2.56; 95% CI, 1.15–5.70; P = 0.02), prolonged hospital stay more than 6 months (OR = 6.06; 95% CI, 1.71–21.53; P = 0.01), biliary stenosis (OR = 9.55; 95% CI, 3.67–24.73; P < 0.001) and histologic rejection during 6 months after LT (OR = 2.78; 95% CI, 1.25–6.22; P = 0.01) were significantly associated with the development of late-onset BSI (Table 3). In the multivariate analysis, a prolonged operative time > 12 hours (OR = 3.55; 95% CI, 1.06–11.91; P = 0.04) and biliary stenosis (OR = 4.60; 95% CI, 1.55–13.68; P = 0.006) were independent risk factors for developing late-onset BSI (Table 4). The association of the operative time and prevalence of BSI over the years is illustrated in Fig., Supplemental Digital Content 2, https://links.lww.com/INF/C834.
Long-Term Outcomes After Late-onset BSI
Kaplan-Meier analysis was performed to evaluate the long-term outcomes of recipients with and without late-onset BSI. Although there were no BSI episodes that directly resulted in retransplantation or death, late-onset BSI was significantly associated with retransplantation rate (P = 0.04) and increased mortality (P < 0.001; Fig. 1).
Eight children including four with late-onset BSI died beyond 6 months after pediatric LT. The causes of death of the four children with a history of late-onset BSI were chronic rejections in two, pulmonary hypertension in one, and respiratory failure in one. The causes of death of the four without a history of late-onset BSI were recurrences of tumor in two, pulmonary hypertension in one and liver failure in one. There were no children who died due to late-onset BSI.
This is the largest report to describe BSIs occurring beyond the first 6 months after pediatric LT. Late-onset BSI developed in 9% of the LT recipients. We found that the most frequent site of infection was intraabdominal infection, followed by bacteremia without an apparent focus.
GNRs were the major cause of late-onset BSI. Most were due to intraabdominal infection, but some individuals developed GNR bacteremia without any focus, regardless of whether they were inpatients or outpatients. In previous reports, 54–70% of infections that occurred 6 months after SOT were caused by GNR.13 Intraabdominal infection was the most common site of infection for BSI occurring more than 60 days after adult LT.14 In pediatric LT recipients, 11 bacteremia cases occurring more than 2 months after LT have been reported, including two due to biliary strictures and three with an undetermined focus.22 Thus, for patients presenting with signs of infection during the late periods after LT, empiric GNR coverage should be considered even in the absence of abdominal symptoms.
Three recipients (0.8%) developed BSI caused by S. pneumoniae. Although all these occurred in unvaccinated patients or were due to a nonvaccine serotype, the prevalence was high compared with the prevalence of pneumococcal bacteremia in Japanese children < 5 years old before the introduction of PCV7 in our country (11.2–60.2/100,000 person-years).23,24 As previously reported,25 LT recipients are at a higher risk of invasive pneumococcal disease.
Biliary stenosis and a prolonged operative time more than 12 hours were independent risk factors for developing late-onset BSI in the multivariate analysis. Biliary stenosis may predispose to cholangitis and BSI. Although prolonged operative time is known to be a risk factor for early bacterial infection after LT,2,26,27 we found an unexpected association with late-onset BSI. Interestingly, the prevalence of bacteremia decreased in parallel with the decrease in the number of surgeries exceeding 12 hours over time. Prolonged surgical time in pediatric LT have been reported to be associated with biliary complications.28 Biliary stenosis might be one of the complications after prolonged operation for LT. Since biliary stenosis and prolonged operative time were risk factor for late-onset BSI in multivariate analysis, we speculate that other anatomical complications can also predispose individuals to septic cholangitis well beyond the early postoperative period.
Late-onset BSI after pediatric LT was significantly associated with retransplantation rate and mortality. Early-onset BSI has been reported to be associated with an increased mortality in adult LT5,10 as well as in pediatric LT.7 Kim et al10 also reported that recipients with early-onset BSI were at a significantly higher risk of mortality compared with recipients who did not have an infection or those who had infection without bacteremia. Late-onset infection after SOT, including kidney, liver and heart transplantation, has been reported to be associated with higher mortality.13 The association between late-onset BSI and patient death might reflect baseline severity like persistent technical complication, rejection and increasing immunosuppressive therapy.
There are some limitations to this study. First, some LT recipients were followed up in other hospitals, and the number of BSI episodes may have been underreported. However, most received follow-up at our institution at least once a year, and serious adverse events including BSI were reported to us. Second, although antimicrobial use without obtaining blood cultures is rare in our hospital, children followed up in other hospital might have received oral antibiotics without blood cultures, which might have resulted in pseudo-negative blood cultures. Third, we used 6 months as the cutoff to distinguish between early- and late-onset BSI. Previous reports have used several different cutoff times (30, 60, 90 and 100 days, 6 months) to distinguish between early and late infection.12–15,22 In our institution, all the BSI episodes that occurred within the first 6 months occurred in inpatients, and the characteristics were similar to the BSI that occurred during the first 30 days. Thus, we evaluated the BSI that occurred after 6 months post-LT. However, the use of different cutoff times may change the prevalence and characteristics of late-onset BSI.
In conclusion, late-onset BSI developed in 9% of recipients after pediatric LT. GNR accounted for the majority of the late-onset BSI and were often due to intraabdominal infection. Prolonged operative time for LT and biliary stenosis was an independent risk factor for developing late-onset BSI. Late-onset BSI was significantly associated with increased retransplantation rate and mortality in LT recipients.
1. Fishman JA. Infection in solid-organ transplant recipients. N Engl J Med. 2007;357:26012614.
2. Kim SI. Bacterial infection after liver transplantation. World J Gastroenterol. 2014;20:62116220.
3. Nafady-Hego H, Elgendy H, Moghazy WE, et al. Pattern of bacterial and fungal infections in the first 3 months after pediatric living donor liver transplantation: an 11-year single-center experience. Liver Transpl. 2011;17:976984.
4. Sganga G, Spanu T, Bianco G, et al. Bacterial bloodstream infections in liver transplantation: etiologic agents and antimicrobial susceptibility profiles. Transplant Proc. 2012;44:19731976.
5. Iida T, Kaido T, Yagi S, et al. Posttransplant bacteremia in adult living donor liver transplant recipients. Liver Transpl. 2010;16:13791385.
6. Rhee KW, Oh SH, Kim KM, et al. Early bloodstream infection after pediatric living donor living transplantation. Transplant Proc. 2012;44:794796.
7. Shoji K, Funaki T, Kasahara M, et al. Risk factors for bloodstream infection after living-donor liver transplantation in children. Pediatr Infect Dis J. 2015;34:10631068.
8. Hashimoto M, Sugawara Y, Tamura S, et al. Bloodstream infection after living donor liver transplantation. Scand J Infect Dis. 2008;40:509516.
9. Kierzkowska M, Majewska A, Dobrzaniecka K, et al. Blood infections in patients treated at transplantation wards of a clinical hospital in Warsaw. Transplant Proc. 2014;46:25892591.
10. Kim SI, Kim YJ, Jun YH, et al. Epidemiology and risk factors for bacteremia in 144 consecutive living-donor liver transplant recipients. Yonsei Med J. 2009;50:112121.
11. Iinuma Y, Senda K, Fujihara N, et al. Surgical site infection in living-donor liver transplant recipients: a prospective study. Transplantation. 2004;78:704709.
12. Juan RS, Aguado JM, Lumbreras C, et al.; RESITRA Network, Spain. Incidence, clinical characteristics and risk factors of late infection in solid organ transplant recipients: data from the RESITRA study group. Am J Transplant. 2007;7:964971.
13. Cervera C, Fernández-Ruiz M, Valledor A, et al. Epidemiology and risk factors for late infection in solid organ transplant recipients. Transpl Infect Dis. 2011;13:598607.
14. Lee SO, Kang SH, Abdel-Massih RC, et al. Spectrum of early-onset and late-onset bacteremias after liver transplantation: implications for management. Liver Transpl. 2011;17:733741.
15. Singh N, Gayowski T, Wagener MM, et al. Predictors and outcome of early- versus late-onset major bacterial infections in liver transplant recipients receiving tacrolimus (FK506) as primary immunosuppression. Eur J Clin Microbiol Infect Dis. 1997;16:821826.
16. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control. 2008;36:309332.
17. Mermel LA, Allon M, Bouza E, et al. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 Update by the Infectious Diseases Society of America. Clin Infect Dis. 2009;49:145.
18. Ken-Ichi I, Fukuda A, Kawano F, et al. Effective control of Epstein—Barr virus infection following pediatric liver transplantation by monitoring of viral DNA load and lymphocyte surface markers. Pediatr Transplant. 2012;16:748757.
19. Saitoh A, Sakamoto S, Fukuda A, et al. A universal preemptive therapy for cytomegalovirus infections in children after live-donor liver transplantation. Transplantation. 2011;92:930935.
20. Furuichi M, Fujiwara T, Fukuda A, et al. Fulminant hepatic failure as a risk factor for cytomegalovirus infection in children receiving preemptive therapy after living donor liver transplantation. Transplantation. 2016;100:24042409.
21. Banff schema for grading liver allograft rejection an international consensus document. Hepatology. 1997;25:658663.
22. Quirós-Tejeira RE, Ament ME, McDiarmid SV, et al. Late-onset bacteremia in uncomplicated pediatric liver-transplant recipients after a febrile episode. Transpl Int. 2002;15:502507.
23. Ishiwada N, Hishiki H, Nagasawa K, et al. The incidence of pediatric invasive Haemophilus influenzae and pneumococcal disease in Chiba prefecture, Japan before and after the introduction of conjugate vaccines. Vaccine. 2014;32:54255431.
24. Nakamura R, Togashi T. Population-based incidence of invasive haemophilus influenzae and pneumococcal diseases before the introduction of vaccines in Japan. Pediatr Infect Dis J. 2013;32:13941396.
25. Dipchand A, Fecteau A, Richardson S, et al. Invasive pneumococcal disease in pediatric organ transplant recipients: a high-risk population. Pediatr Transplant. 2005;9:183186.
26. Hollenbeak CS, Alfrey EJ, Souba WW. The effect of surgical site infections on outcomes and resource utilization after liver transplantation. Surgery. 2001;130:388395.
27. George DL, Arnow PM, Fox AS, et al. Bacterial infection as a complication of liver transplantation: epidemiology and risk factors. Rev Infect Dis. 1991;13:387396.
28. Lüthold SC, Kaseje N, Jannot AS, et al. Risk factors for early and late biliary complications in pediatric liver transplantation. Pediatr Transplant. 2014;18:822830.