Total parenteral nutrition (TPN) is essential to maintain the nutritional status of patients who are unable to receive nutrients by the enteral route.1 However, several complications can result from the treatment of TPN, including catheter-related bacteremia (CRB),2 which can occur in 1.3–26.2% of the catheters used for TPN.3 In addition, patients who require TPN for nutritional support are often critically ill. They may have multiple conditions, such as malnutrition and hyperglycemia3 which can further predispose the patients to diverse infections that might complicate with bacteremia. In hospitalized patients, the most common pathogens of bacteremia in the patients receiving TPN are coagulase-negative Staphylococcus, S aureus, Enterococcus spp, Candida spp, Klebsiella pneumoniae and Pseudomonas aeruginosa.3 While Acinetobacter species have emerged as important nosocomial pathogens,4 the clinical characteristics of Acinetobacter bacteremia have been rarely described in patients receiving TPN.
Acinetobacter is one of the glucose nonfermenting Gram-negative bacilli. The genus comprises more than 30 genomic species, among which the most clinically relevant Acinetobacter spp. are Acinetobacter baumannii, Acinetobacter genomic species 3 (AGS 3) and Acinetobacter genomic species 13TU (AGS 13TU).4,5 They are indiscriminate by phenotypic methods used in clinical laboratory and are often grouped together and reported as A baumannii complex (Abc).6 However, the antimicrobial-resistant pattern among the A baumannii and nonbaumannii Abc (including AGS 3 and AGS 13TU) is different.7–10 Furthermore, the clinical characteristics and outcomes of the patients presenting with A baumannii and nonbaumannii Abc bacteremia are suggested to be different.6,11 The literature shows that the identification of the Abc to genomic species, or at least differentiation of the A baumannii from nonbaumannii Abc may indicate some clinical significance.6,11 However, current studies included patients with heterogeneous background, which may over- or underestimate the clinical significance among the Abc bacteremia in certain subgroups of patients, especially those with more homogenous background who may be unique in their clinical presentations.
In this report, we present and compare the clinical characteristics, microbiological data and outcomes of both bacteremia of A baumannii and nonbaumannii Abc in patients receiving TPN.
2.1. Hospital setting and study population
The study was conducted in Taipei Veterans General Hospital (TVGH), where there were 2900 beds, including 220-beds in intensive care units (ICUs), and 97,000 to 98,000 admission per year. Hospitalized patients who had Abc bacteremia between August 1998 and December 2007 were identified from microbiological records, and those who received TPN during the bacteremia onset were included in this study. The TPN was prepared, stored and infused by using standardized methods,3 e.g., it was prepared in clean rooms using a laminar-flow hood and infused for no more than 24 hours. The medical records of the patients were reviewed retrospectively. The demographic characteristics, underlying diseases, invasive procedures used, laboratory data at the time of the onset of the bacteremia, treatments and clinical outcomes were retrieved and analyzed. This study was approved by the institutional review board of TVGH.
2.2. Microbiological studies
The bacteria were phenotypically identified as Abc by using the API ID 32GN system (bioMerieux, Marcy l'Etoile, France). For patients with multiple blood cultures positive for Abc, the first isolate was chosen in the following studies. Genomic species identification of A baumannii was performed by using a multiplex polymerase chain reaction (PCR) method,12 and was verified by the presence of blaOxA-51-like gene in this isolate by using the previously described PCR detection method.7 Isolates recognized as nonbaumannii Abc were identified to genomic species level by using 16S-23S ribosomal DNA intergenic spacer sequence analysis described previously.13 The genetic relationship of the Abc isolates was determined by pulsed-field gel electrophoresis (PFGE),14 which included the use of restriction enzyme ApaI and a CHEF-Mapper electrophoresis system (Bio-Rad Laboratories, Richmond, CA, USA). Molecular Analyst Fingerprinting, Fingerprinting Plus, and Fingerprinting DST software (Bio-Rad Laboratories) with unweighted-pair group method and arithmetic average were used for cluster analysis. Isolates which had similarity >80% in PFGE were classified as the same clone. Antimicrobial susceptibility testing was undertaken using the disk diffusion method, and was interpreted according to the recommendation of Clinical Laboratory Standards Institute.15Escherichia coli (ATCC 25922) and Pseudomonas aeruginosa (ATCC 27853) were included in the test as quality control strains.
An episode of Abc bacteremia was defined as isolation of Abc from blood culture on one or more occasions and based on clinical manifestations compatible with sepsis syndrome.16 The date of the collection of the first positive blood culture was considered the date of the onset of bacteremia. Previous use of antibiotics was defined as the use of antibiotics in the preceding 30 days prior to the date of bacteremia. Postchemotherapy status was defined as having received chemotherapy within 7 days prior to the first episode of Abc bacteremia. The use of steroid and immunosuppressive agent was defined as medication use within 2 weeks prior to a bacteremic episode. Neutropenia was identified as an absolute neutrophil count <500 cells/mm3. Recent stay in ICU was characterized as the stay in ICU within 2 weeks before the onset of bacteremia. A patient was considered to be in shock if there was evidence of organ hypoperfusion, systolic blood pressure was <90 mmHg, systolic blood pressure decreased >30 mmHg compared with the baseline blood pressure, or if the patient required vasopressor agents to maintain blood pressure despite adequate fluid resuscitation. The focus of infection was identified according to the definitions of the Centers for Disease Control and Prevention17 by infectious diseases specialists. If no infectious focus could be identified, the bacteremia was classified as primary. Multidrug resistance (MDR) in Abc was defined as nonsusceptibility to more than or at least three classes of the following antimicrobial agents, including β-lactam/β-lactamase inhibitor combinations, extended-spectrum cephalosporins, carbapenems, aminoglycosides, and fluoroquinolones.18 Appropriate empirical antibiotic treatment was defined as the use of one or more active antibiotics to which the organism was susceptible in vitro within 72 hours after onset of bacteremia. The major outcome analysis was 14-day mortality, which was defined as death within 14 days after the diagnosis of Abc bacteremia. The all-cause in-hospital mortality, defined as death during hospitalization for any cause, was also recorded and analyzed.
2.4. Statistical analyses
SPSS for Windows (Version 17; SPSS, Chicago, IL, USA) was used for all data analysis. The Chi-square test with Yates correction or Fisher's exact test were used to compare categorical differences, as appropriate. Continuous variables were analyzed using the Mann-Whitney U test or two-sample t test. A multivariate analysis with logistic regression was performed to identify independent factors associated with A baumannii bacteremia. Survival curves were prepared using the Kaplan-Meier method, and the log rank test was used to compare univariate survival distribution between bacteremia due to A baumannii and nonbaumannii Abc. Multivariate analysis with logistic regression was performed to identify independent prognostic factors associated with 14-day mortality of Abc bacteremia. Variables with p values of <0.05 in the univariate analysis were entered into the multivariate analysis. A p value of <0.05 was considered statistically significant.
3.1. Clinical characteristics
During the 9-year study period, a total of 46 patients with Abc bacteremia were identified (Fig. 1). Based on the results of bacterial identification, 23 patients had bacteremia due to A baumannii, while the other 23 patients had bacteremia due to nonbaumannii Abc (including 15 AGS 13 TU and 8 AGS 3). There was no difference in gender (male: 15/23 vs. 11/23, p = 0.372) and age (65.3 ± 19.6 vs. 61.9 ± 17.2 years, p = 0.531) among patients with bacteremia due to A baumannii or nonbaumannii Abc. The underlying diseases, the invasive procedures used, as well as the purposes for initiating TPN in two-group patients are listed in Tables 1 and 2, respectively. The two groups of patients had very similar underlying conditions, except there was a higher proportion of patients with A baumannii bacteremia who had indwelling Foley catheter, nasogastric (NG) tube in place and ventilator use. In contrast, all of the seven patients with port-a-catheter in place had bacteremia due to nonbaumannii Abc. The mean duration between the initiation of TPN and the onset of bacteremia was significantly longer in patients with A baumannii bacteremia (21 ± 11.7 vs. 10.4 ± 7.9 days, p = 0.001). The patients with bacteremia due to A baumannii had a slightly higher Acute Physiology and Chronic Health Evaluation (APACHE) II score at the onset of bacteremia, however without significant difference (22. 7 ± 7.4 vs. 19.1 ± 8.2, p = 0.129). Primary bacteremia was found in 43.5% vs. 65.2% of patients with A baumannii and nonbaumannii bacteremia, respectively (p = 0.236). The other infectious foci of the bacteremia were CRB (21.7% vs. 17.4%, p = 1.000), pneumonia (21.7% vs. 17.4%, p = 1.000), intra-abdominal abscess (4.3% vs. 0, p = 1.000) and urinary tract infection (8.7% vs. 0, p = 0.489). The albumin (2.7 ± 0.4 vs. 2.8 ± 0.4 g/L, p = 0.577), C-reactive protein (9.2 ± 6.7 vs. 9.3 ± 8.8 mg/dL, p = 0.993) and glucose level [median (range): 115 (89–238) vs. 116 (42–389) mg/dL, p = 0.745) were comparable between the two groups of patients. Other laboratory data, including white blood cell counts, hemoglobin level, platelet count, aspartate aminotransferase level, and creatinine level were also similar between the two groups (data not shown).
3.2. Clonality and antimicrobial susceptibilities of the bacterial isolates
All the isolates belonged to different pulsotypes (data not shown). The antimicrobial-resistant rates of the bacterial isolates are shown in Table 3. Compared to nonbaumannii Abc isolates, A baumannii isolates were more resistant to all tested antibiotics except imipenem. More A baumannii isolates had the MDR phenotype (91.3% vs. 43.5%, p = 0.002).
3.3. Factors associated with acquisition of A baumannii bacteremia
A baumannii was resistant to more antibiotics, and therefore the identification of patients likely infected with A baumannii might be useful for the selection of an appropriate therapy. The univariate analysis showed that in the A baumannii bacteremia group, there were significantly more cases receiving NG insertion, use of ventilator or on Foley catheter, but less cases with port-a-catheter implantation (Table 1). Other factors associated with A baumannii bacteremia were MDR phenotype of the bacterial isolates, use of TPN for the purpose of sepsis-related ileus (Table 2), and the onset of bacteremia at ≥15 days after initiation of TPN (60.9% in patients with A baumannii vs. 21.7% in nonbaumannii Abc bacteremia, p = 0.017). However, the multivariate analysis revealed that the onset of bacteremia at ≥15 days after TPN use was the only independent factor that associated with A baumannii bacteremia [odds ration (OR) 7.214, 95% confidence interval (CI) 1.108–46.989, p = 0.039].
3.4. Treatment, outcome and prognosis factor
There was smaller number of patients with A baumannii bacteremia who received an appropriate empirical therapy when compared to those with nonbaumannii Abc bacteremia (39.1% vs. 65.2%), however the difference was not significant (p = 0.140). All the catheters that associated with CRB including the port-a-catheter were removed, and a new catheter was inserted. Shock developed in a comparable portion of patients in both groups (21.7% vs. 17.4%, p = 1.000) during the onset of bacteremia. None of the patients had metastatic complication of the bacteremia, such as osteomyelitis and endocarditis. The 14-day mortality (30.4% vs. 21.7%, p = 0.737) and all-cause in-hospital mortality (60.9% vs. 39.1%, p = 0.238) of the patients with A baumannii and nonbaumannii Abc bacteremia did not differ significantly. In addition, the cumulative survival rate at 14 days was also similar between the two groups by Kaplan-Meier method (Fig. 2). In the univariate analysis, the acquisition of A baumannii was not associated with 14-day mortality of the patients with Abc bacteremia (p = 0.737). Shock (p = 0.039) and APACHE II score ≥21 at the onset of bacteremia (p = 0.004) were associated with 14-day mortality in these patients. The multivariate analyses showed that the only independent factor for 14-day mortality in these patients was APACHE score >21 (OR 14.888, CI 1.668–132.868, p = 0.016).
Acinetobacter bacteremia is rarely described in patients receiving TPN. In TVGH, there were about 300 to 400 patients receiving TPN every year,19,20 however there were only 46 patients with Abc bacteremia identified in 9-year period. To our knowledge, however, this is the largest series of Abc bacteremia in patients receiving TPN. In the latest two studies that compared the clinical features and outcome of patients with A baumannii and nonbaumannii Abc bacteremia,6,11 the use of TPN was not mentioned. Although the use of TPN was found in 51 among 123 patients with A baumannii infection in a recent article,21 the exact number of patients who had bacteremia was not revealed. Similarly, another recent study attempted to characterize the microorganisms responsible for CRB in patients receiving TPN over 12 years. It found only five (2.3%) among the 222 isolates that were glucose nonfermenting Gram-negative bacilli (the Acinetobacter species was not specified).22
In this study, the patients with A baumannii and non baumannii Abc bacteremia were separately presented. There were some differences in clinical features between the two groups of the patients, e.g., the patients with A baumannii bacteremia had a longer use of TPN before their onset of bacteremia compared to the patients with nonbaumannii Abc bacteremia. However, the 14-day and in-hospital all-cause mortality rate did not differ significantly between these two groups of the patients.
In the univariate analysis, the underlying conditions of the two groups were largely similar, except the patients with A baumannii bacteremia were more likely to have invasive procedures, such as Foley catheter, NG tube and ventilator use. These findings were consistent with the finding of Lee et al.6 Interestingly, all the patients with port-a-catheter acquired nonbaumannii Abc bacteremia, however the reason is currently unknown. In this study, the length of time of the use of TPN prior to the onset of bacteremia was longer in the patients with A baumannii bacteremia. The literature suggests a similar finding, that A baumannii bacteremia occurs in a later period during hospitalization.6Acinetobacter isolates that were resistant to more classes of antibiotics may be isolated in the later period during hospitalization, as carbapenem-resistant A baumannii was also isolated at a longer duration after admission as compared to carbapenem susceptible A baumannii.21
In the multivariate analysis, the longer use of TPN prior to the onset of bacteremia (≥ 15 days) was the only independent factor to predict the acquisition of A baumannii bacteremia. Differentiation of A baumannii from nonbaumannii Abc may have therapeutic implication, as A baumannii displayed a resistance pattern to more classes of antibiotics, which had been repeatedly shown.7,8 As a result, the patients with A baumannii tended to receive inappropriate empirical therapy.6,11
Although our findings regarding the clinical features of patients with A baumannii and nonbaumannii Abc were quite similar with those found in the latest two studies with larger case numbers,6,11 the impact of A baumannii on patients' outcome was different between ours and that found in the previous two studies, in which the patients with A baumannii bacteremia were shown to indicate a poorer outcome. The discrepancy of the results might be due to the following reasons. First, although this is the largest case series of Abc bacteremia in patients with TPN, the case number was still relatively small, which might have obscured the difference. Second, in the previous studies, the patient populations were more heterogeneous and the patients with A baumannii bacteremia had presented with a higher level of illness severity as reflected by higher SOFA and Pitt. bacteremia scores11 or higher SAPS II and Pitt. bacteremia scores.6 Our patients in both groups had similar and high APHACHE scores. The mortalities of the bacteremic patients with severe illness are usually high,11 regardless of the pathogens. As a result, the difference between the pathogenecity of different Acinetobacter species might be obscured in the patients with a higher level of illness severity. Consequently, the only independent factor associated with mortality in our patients is high APACHE II score. Therefore, the impact of A baumannii on patient mortality should be verified by additional studies, in which other confounding factors that might associate with patients' mortality, such as host factors, disease severity and appropriateness of therapy, should be adjusted.
Of note, this study is adherent to several limitations of a retrospective study. The prevalence of Abc bacteremia in patients receiving TPN can only be identified with a prospective study. In addition, the small number of cases makes it difficult to draw a conclusion regarding the impact of A baumannii bacteremia or inappropriate empirical therapy on patients' outcome.
In conclusion, Acinetobacter bacteremia is still relatively rare in patients receiving TPN. In our study, while the patients with A baumannii bacteremia could be differentiated from those with nonbaumannii Abc bacteremia by longer use of TPN prior to the onset of bacteremia, the outcome of the two groups of patients did not differ significantly.
This study was supported by grants from Taipei Veterans General Hospital (V99C1-014 and V99S5-006) and the National Science Council (NSC98-2314-B-010-010-MY3).
1. Jeejeebhoy KN. Total parenteral nutrition
: potion or poison? Am J Clin Nutr
2. Chen HS, Wang FD, Lin M, Lin YC, Huang LJ, Liu CY. Risk factors for central venous catheter-related infections in general surgery. J Microbiol Immunol Infect
3. Opilla M. Epidemiology of bloodstream infection associated with parenteral nutrition. Am J Infect Control
4. Dijkshoorn L, Nemec A, Seifert H. An increasing threat in hospitals: multidrug-resistant Acinetobacter baumannii
. Nat Rev Microbiol
5. Peleg AY, Seifert H, Paterson DL. Acinetobacter baumannii
: emergence of a successful pathogen. Clin Microbiol Rev
6. Lee NY, Chang TC, Wu CJ, Chang CM, Lee HC, Chen PL, et al. Clinical manifestations, antimicrobial therapy, and prognostic factors of monomicrobial Acinetobacter baumannii
complex bacteremia. J Infect
7. Lee YT, Huang LY, Chiang DH, Chen CP, Chen TL, Wang FD, et al. Differences in phenotypic and genotypic characteristics among imipenem-non-susceptible Acinetobacter
isolates belonging to different genomic species in Taiwan. Int J Antimicrob Agents
8. Houang ET, Chu YW, Chu KY, Ng KC, Leung CM, Cheng AF. Significance of genomic DNA group delineation in comparative studies of antimicrobial susceptibility of Acinetobacter
spp. Antimicrob Agents Chemother
9. Lin YC, Sheng WH, Chen YC, Chang SC, Hsia KC, Li SY. Differences in carbapenem resistance genes among Acinetobacterbaumannii, Acinetobacter
genospecies 3 and Acinetobacter
genospecies 13TU in Taiwan. Int J Antimicrob Agents
10. Lim YM, Shin KS, Kim J. Distinct antimicrobial resistance patterns and antimicrobial resistance-harboring genes according to genomic species of Acinetobacter
isolates. J Clin Microbiol
11. Chuang YC, Sheng WH, Li SY, Lin YC, Wang JT, Chen YC, et al. Influence of genospecies of Acinetobacter baumannii
complex on clinical outcomes of patients with Acinetobacter
bacteremia. Clin Infect Dis
12. Chen TL, Siu LK, Wu RC, Shaio MF, Huang LY, Fung CP, et al. Comparison of one-tube multiplex PCR, automated ribotyping and intergenic spacer (ITS) sequencing for rapid identification of. Acinetobacter baumannii. Clin Microbiol Infect
13. Chang HC, Wei YF, Dijkshoorn L, Vaneechoutte M, Tang CT, Chang TC. Species-level identification of isolates of the Acinetobacter calcoaceticus
complex by sequence analysis of the 16S-23S rRNA gene spacer region. J Clin Microbiol
14. Huang LY, Chen TL, Lu PL, Tsai CA, Cho WL, Chang FY, et al. Dissemination of multidrug-resistant, class 1 integron-carrying Acinetobacter baumannii
isolates in Taiwan. Clin Microbiol Infect
15. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing: 20th informational supplement. CLSI document M100–S20. Wayne, PA: Clinical and Laboratory Standards Institute; 2010.
16. Chen HP, Chen TL, Lai CH, Fung CP, Wong WW, Yu KW, et al. Predictors of mortality in Acinetobacter baumannii
bacteremia. J Microbiol Immunol Infect
17. Garner JS, Jarvis WR, Emori TG, Horan TC, Hughes JM. CDC definitions for nosocomial infections. Am J Infect Control
18. Paterson DL. The epidemiological profile of infections with multidrug-resistant Pseudomonas aeruginosa
species. Clin Infect Dis
19. Kung SP, Lui WY. Correlation between serum transferrin level and prognosis in patients receiving total parenteral nutrition
. J Chin Med Assoc
20. Lin LY, Lin HC, Lee PC, Ma WY, Lin HD. Hyperglycemia correlates with outcomes in patients receiving total parenteral nutrition
. Am J Med Sci
21. Baran G, Erbay A, Bodur H, Onguru P, Akinci E, Balaban N, et al. Risk factors for nosocomial imipenem-resistant Acinetobacter baumannii
infections. Int J Infect Dis
22. Collins CJ, Fraher MH, Bourke J, Phelan D, Lynch M. Epidemiology of catheter-related bloodstream infections in patients receiving total parenteral nutrition
. Clin Infect Dis