Gram-Positive Bloodstream Infections in Liver Transplant Recipients: Incidence, Risk Factors, and Impact on Survival

Gram-Positive Bloodstream Infections in Liver Transplant Recipients: Incidence, Risk Factors, and Impact on Survival A. Bedini, M. Codeluppi, S. Cocchi, G. Guaraldi, F. Di Benedetto, C. Venturelli, M. Masetti, F. Prati, C. Mussini, V. Borghi, M. Girardis, G.E. Gerunda, F. Rumpianesi, and R. Esposito ABSTRACT The objective of the study was to assess the incidence, risk factors, and survival of gram-positive bloodstream infections (GP-BSIs) among liver transplant recipients during the first year after transplantation. Between October 2000 and September 2006, 42 episodes of GP-BSIs occurred in 205 patients with an overall incidence of 0.20 episodes/patient. Coagulase-negative staphylococci were detected in 45.2% of cases, Enterococcus species in 42.9% (E faecalis, eight; E faecium, seven; E avium, two; E gallinarum, one) and Staphylococcus aureus in 11.9%. Retransplantation was the only independent risk factor for GP-BSI (odds ratio [OR], 0.253; 95% confidence interval (CI), 0.089 to 0.715; P ⫽ .009). Thirty-day mortality rate was 28.5% and S aureus infections were related to a poorer outcome. It is noteworthy that all the isolates of S aureus were methicillin-resistant. Ampicillin was inactive against all the strains of E faecium and 50% of E avium isolates, but active against all E faecalis and E gallinarum strains. All the isolates were glycopeptide-susceptible. No significant differences in mortality rate were observed in relation to sex, etiologies of end-stage liver disease, cytomegalovirus infection/ reinfection, type of donor, rejection, or retransplantation. GP-BSI, the only independent risk factor for death (OR, 0.262; 95% CI, 0.106 to 0.643; P ⫽ .003), reduced the survival rate by 26% in the first year posttransplant. In conclusion, GP-BSIs impact significantly on morbidity and mortality posttransplant, particularly among retransplantations. Control measures are required to reduce the incidence of GP-BSIs in liver transplant recipients. These findings must be considered when empirical antimicrobial therapy is indicated while awaiting blood-culture results.

B

LOODSTREAM INFECTIONS (BSIs) are frequent among solid organ transplanted patients, and sepsis is a major cause of death in this setting.1–3 We reviewed the gram-positive BSIs (GP-BSIs) in people undergoing orthotopic liver transplantation (OLT). MATERIALS AND METHODS The study population included 205 consecutive patients who underwent OLT between October 2000 and September 2005. GPBSIs were defined by at least two positive blood cultures during a single hospitalization period. Coagulase-negative staphylococci accounted for BSIs if four or more blood cultures showed positive results and if fever or hypothermia was present. All blood cultures yielding gram-positive microorganisms in the first year after OLT were included (study period: October 2000 to September 2006). We calculated the mortality rate of the entire population and among patients with GP-BSIs, within 30 days from the first positive blood culture. Sex, age (stratified in three groups: ⬍45 years, 45 to 60 years, ⬎60 years), etiology of end-stage liver disease (ESLD),

cytomegalovirus (CMV) infection, and graft rejection were analyzed as risk factors for GP-BSI. Sex, age, etiology of ESLD, CMV infection, GP-BSI, and graft rejection were evaluated as risk factors

From the Department of Internal Medicine and Medical Specialties (A.B., M.C., S.C., G.G., F.P., C.M., V.B., R.E.), Infectious Diseases Clinic, University of Modena and Reggio Emilia, Department of General Surgery and Surgical Specialties (F.D.B., M.M., G.E.G.), Liver and Multivisceral Transplant Center, University of Modena and Reggio Emilia, Laboratory of Microbiology (C.V., F.R.), Department of General Surgery and Surgical Specialties (M.G.), Anaesthesiology and Reanimation, University of Modena and Reggio Emilia, Policlinico of Modena, Modena, Italy. Address reprint requests to Mauro Codeluppi, MD, Department of Internal Medicine and Medical Specialties, Infectious Diseases Clinic, University of Modena and Reggio Emilia, Via del Pozzo 71, 41100 Modena, Italy. E-mail: codeluppi.mauro@ unimore.it

© 2007 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710

0041-1345/07/$–see front matter doi:10.1016/j.transproceed.2007.05.055

Transplantation Proceedings, 39, 1947–1949 (2007)

1947

1948

BEDINI, CODELUPPI, COCCHI ET AL

for death. Univariate and multivariate analyses were performed with SPSS software 11.5. A P value of ⬍.05 was considered significant.

RESULTS

Between October 2000 and September 2005, 229 liver transplants were performed in 205 subjects. One hundred forty-five (70.7%) patients were men, and the overall median age was 52 years (range: 12 to 70 years). The reasons for OLT were: cirrhosis related to hepatitis viruses in 45% of cases, hepatocellular carcinoma (HCC) in 29%, alcohol-related cirrhosis in 7%, autoimmune cirrhosis in 5%, fulminant hepatitis in 3%, and other etiologies in 11%. Twenty-four subjects underwent retransplantation, 16 (66.6%) for primary graft nonfunction, 7 (29.1%) for arterial or portal vein thrombosis, and 1 (4.1%) for chronic rejection. During the study period, 42 episodes of GP-BSIs occurred in 28 patients with an overall incidence of 0.20 (42/205) episodes/patient. The mean follow-up after OLT was 10.5 months (range: 0 to 12 months). Maintenance immunosuppression consisted of tacrolimus or cyclosporine plus steroids; six patients received sirolimus. Perioperative anti-infectious prophylaxis was performed with cefotaxime and ampicillin administered for 3 to 5 days. Coagulase-negative staphylococci were detected among 19 episodes (45.2%) of BSIs, Enterococcus species (spp) in 18 (42.9%; eight strains of E faecalis, seven of E faecium, two of E avium, and one of E gallinarum) and Staphylococcus aureus in 5 (11.9%). Fourteen patients displayed two episodes of GP-BSIs in the first year after OLT. In Fig 1, the time distribution of GP-BSIs after transplantation is shown. In 54.7% of the cases, GP-BSIs occurred in the first 3 months posttransplant. In the opinion of the investigators, 91.6% of Staphylococcus spp infections and 55.5% of Enterococcus spp infections were possibly related to an indwelling central venous catheter. Patients who underwent retransplantation showed a greater risk of GP-BSIs (Table 1). This finding has been confirmed by logistic regression analysis with an odds ratio (OR) equal to 0.253 (95% confidence interval [CI], 0.089 to 0.715; P ⫽ .009). (Data not shown)

Table 1. Univariate Analysis of Risk Factors Associated with GP-BSI No. (%) of patients Variables

Sex Male Female Age (y) ⬍45 45–60 ⬎60 Hepatocellular carcinoma No Yes Hepatitis virus infection No Yes CMV infection/reinfection No Yes Transplant donor Cadaveric Living Transplant rejection No Yes Retransplantation No Yes

Without GP-BSI (n ⫽ 177)

With GP-BSI (n ⫽ 28)

125 (71) 52 (29)

20 (71) 8 (29)

.930

39 (22) 88 (50) 50 (28)

8 (29) 15 (53) 5 (18)

.474

127 (72) 50 (28)

19 (68) 9 (32)

.672

47 (27) 130 (73)

8 (29) 20 (71)

.823

151 (85) 26 (15)

21 (75) 7 (25)

.168

147 (83) 30 (17)

23 (82) 5 (18)

.906

135 (76) 42 (24)

18 (64) 10 (36)

.176

161 (91) 16 (9)

20 (71) 8 (29)

.003

P value

Survival rate in the first year after OLT was 79.5% and among patients with GP-BSIs, the 30-day mortality rate was 28.5%. Staphylococcus aureus was related to the highest 30-day mortality rate (60.0%), followed by coagulase-negative staphylococci (21.0%) and Enterococcus spp (16.6%). It is noteworthy that all isolates of S aureus were methicillin-resistant. Ampicillin was inactive against all strains of E faecium and the 50% of E avium isolates, but active against all E faecalis and E gallinarum strains. All isolates were susceptible to glycopeptides. Univariate analysis of the main risk factors for death are summarized in Table 2. No significant differences in mortality rate were observed in relation to sex, underlying diseases (HCC, hepatitis viruses), CMV infection/reinfection, donor (cadaveric or living), rejection, or retransplantation. Patients with GPBSIs displayed poorer outcomes (P ⫽ .002). Logistic regression analysis confirmed a death-related OR for GP-BSIs equal to 0.262 (95% CI, 0.106 to 0.643; P ⫽ .003). (Data not shown.) The survival rate after 1 year from OLT among patients without versus with GP-BSIs was 83.0% and 57.1%, respectively. DISCUSSION

Fig 1. Time distribution of GP-BSIs after OLTx.

Gram-positive BSIs incidence in our population was similar to that observed in larger studies.4,5 Retransplantation was the only independent risk factor for GP-BSI, probably for the higher complexity of surgical interventions required in

GRAM-POSITIVE BLOODSTREAM INFECTIONS

1949

Table 2. Univariate Analysis of Risk Factors Associated With 30-Day Mortality No. (%) of patients Variables

Sex Male Female Age (y) ⬍45 45–60 ⬎60 Hepatocellular carcinoma No Yes Hepatitis virus infection No Yes CMV infection/reinfection No Yes GP-BSI No Yes Transplant donor Cadaveric Living Transplant rejection No Yes Retransplantation No Yes

Survived (n ⫽ 163)

Died (n ⫽ 42)

P value

112 (69) 51 (31)

33 (79) 9 (21)

.210

38 (23) 83 (51) 42 (26)

9 (21) 20 (48) 13 (31)

.795

120 (73) 43 (27)

26 (62) 16 (38)

.135

45 (28) 118 (72)

10 (24) 32 (76)

.620

140 (86) 23 (14)

32 (76) 10 (24)

.127

147 (90) 16 (10)

30 (71) 12 (29)

.002

135 (83) 28 (17)

35 (83) 7 (17)

.937

124 (76) 39 (24)

29 (69) 13 (31)

.351

144 (88) 19 (12)

37 (88) 5 (12)

.964

these patients. High prevalence of GP-BSIs within the first 3 months from OLT may indicate that these infections are mainly related to the surgical intervention and the intensive

care unit management of these patients. Staphylococcus aureus infections were uncommon in our population but were related to high 30-day mortality rate (60.0%). This mortality rate was higher than that reported by Singh et al for methicillin-resistant S aureus bacteremia6 (21.0%), probably due to the small number of the cases. With regard to Enterococcus spp. infections, ampicillin-resistance strains did not influence 30-day mortality rate (16.6%). Univariate and multivariate analyses showed that GPBSIs affected liver transplant recipient survival, particularly during the first year. No other significant risk factor for death was observed. These results indicated that control measures are required to reduce the incidence of GP-BSIs. These findings must be considered when empirical antimicrobial therapy is indicated while awaiting blood-culture results. REFERENCES 1. Singh N, Paterson D, Gayowski T, et al: Predicting bacteremia and bacteremic mortality in liver transplant recipients. Liver Transpl 6:54, 2000 2. Candel FJ, Grima E, Matesanz M, et al: Bacteremia and septic shock after solid-organ transplantation. Transplant Proc 37:4097, 2005 3. Wagener MM, Yu VL: Bacteremia in transplant recipients: a prospective study of demographics, etiologic agents, risk factors, and outcomes. Am J Infect Control 20:239, 1992 4. Singh N, Squier C, Wannstedt C, et al: Impact of an aggressive infection control strategy on endemic Staphylococcus aureus infection in liver transplant recipients. Infect Control Hosp Epidemiol 27:122, 2006 5. Torre-Cisneros J, Herrero C, Canas E, et al: High mortality related with Staphylococcus aureus bacteremia after liver transplantation. Eur J Clin Microbiol Infect Dis 21:385, 2002 6. Singh N, Paterson DL, Chang FY, et al: Methicillin-resistant Staphylococcus aureus: the other emerging resistant gram-positive coccus among liver transplant recipients. Clin Infect Dis 30:322, 2000