Background

The genus Acinetobacter comprises a heterogeneous group of non-motile, aerobic, oxidase negative, non-fermentative, gram-negative coccobacilli [1, 2]. They are widespread in natural moist and hospital environments, and are associated with skin colonization of hospitalized patients [3]. Although they were thought to have low pathogenicity, the Acinetobacter species have been recognized as opportunistic nosocomial pathogens that mainly affect immune-compromised patients and patients hospitalized in intensive care units (ICUs) [4]. It has emerged as one of the most troublesome pathogens for health care institutions globally over the past 2 decades, owing to its increasing prevalence and rapid development of drug resistance.

The genus Acinetobacter comprises 39 genomic species (http://www.bacterio.net/acinetobacter.html) [5]. While Acinetobacter species such as A baumannii, A. nosocomialis, and A. pittii are frequently isolated as human pathogens [6–11]; other species, such as A. ursingii, are rarely reported as pathogens [12, 13]. The low incidence of A. ursingii infection may be further complicated by the inaccurate identification tools used in clinical laboratories. In this study, we aimed to describe the incidence and clinical characteristics of A. ursingii bacteremia, the performance of two phenotypic identification systems and one matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometer, and the antimicrobial susceptibilities of the isolates. Owing to the predominance of A. baumannii in clinical settings, we also compared the clinical features of A. ursingii and A. baumannii bacteremia.

Methods

Subjects

Patients who were admitted to the Taipei Veterans General Hospital (T-VGH) from January 2000 to December 2008, were included. T-VGH is a 2980-bed medical center that serves about 120 thousand person-times pear year. It serves not only veterans but also their families and other individuals. The charts were reviewed from all patients with symptoms and signs of infection who had at least one positive blood culture for A. ursingii and A. baumannii. If patients had two or more positive blood cultures, only the first blood culture was included. The source of infection was determined as recommended by the Centers of Disease Control guidelines [14, 15]. Patients under 18 years of age and those with incomplete medical records were excluded. The protocol was approved by the T-VGH Institutional Review Board (approval number: 2011-10-012IC), with a waiver for informed consent.

Data collection

Medical records were reviewed to obtain clinical information, including demographic characteristics; underlying diseases; severity of illness; the presence of a ventilator, central venous catheters, a nasogastric tube, or a Foley catheter at the time of onset of bacteremia; intensive care unit (ICU) hospitalization; and survival. Chronic kidney disease was defined as an estimated glomerular filtration rate <60 mL/min/1.73 m2. Neutropenia was defined as an absolute neutrophil count of <0.5 × 109 neutrophils/L. Recent surgery was defined as any operation performed within 4 weeks prior to the onset of bacteremia. Shock was defined as hypotension (systolic blood pressure [SBP] <90 mmHg, mean arterial pressure <70 mmHg, or a SBP decrease > 40 mmHg) with evidence of end organ dysfunction. Bacteremia cases without a definite identified source were defined as primary bacteremia. The severity of illness was evaluated using the Acute Physiology and Chronic Health Evaluation II (APACHE II) score [16] within 24 h prior to bacteremia onset.

Appropriate antimicrobial therapy was defined as administration of at least one antimicrobial agent to which the causative pathogen was susceptible within 48 h of the onset of bacteremia by an approved route and at a dosage consistent with end organ(s) function. Antimicrobial therapy that did not meet this definition was considered inappropriate. Monotherapy with an aminoglycoside was not considered the appropriate therapy. All-cause 14-day and 28-day mortality rates were recorded.

Bacterial isolates, genotypic and phenotypic identification, pulsed-field gel electrophoresis analysis, and determination of antimicrobial minimal inhibitory concentration

From January 2000 to December 2008, 616 clinical isolates of Acinetobacter were isolated from blood samples at T-VGH. All isolates were presumed to be Acinetobacter species, as determined using phenotypic methods with the 32GN system or the Vitek 2 system (bioMérieux, Marcy l’Etoile, France). These isolates were included in our study for further identification. A multiplex-polymerase chain reaction method was then used to identify A. baumannii at the genomic species level [17]. Isolates belonging to non-A. baumannii species were identified as A. ursingii using 16S rRNA gene sequence [18] and confirmed by 16S-23S rRNA internal transcribed spacer (ITS) sequence analysis [19]. Pulsed-field gel electrophoresis (PFGE) was performed to determine the clonality of the isolates [20]. These A. ursingii isolates were then used to determine the performance of the Vitek 2 (bioMérieux), Phoenix (Becton Dickinson, NJ, USA), and matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometer systems (Bruker Daltonics, Billerica, MA) in the identification of this species. The antimicrobial minimal inhibitory concentrations (MICs) for the isolates were determined by using the Vitek 2 system (bioMérieux). The tested antimicrobials were ampicillin-sulbactam, ceftazidime, ceftriaxone, cefepime, imipenem, amikacin, gentamicin, ciprofloxacin, levofloxacin, and colistin. The breakpoint interpretation was determined according to the recommendations of the Clinical Laboratory Standards Institute (CLSI) [21].

Statistical analysis

To assess differences, the Student’s t-test or the Mann–Whitney rank sum test was used to analyze continuous variables, while the chi-square test with Yate’s correction or Fisher’s exact test was used to compare discrete variables. Time to mortality was analyzed using the Kaplan–Meier survival analysis and the long-rank test. A p-value <0.05 was considered statistically significant. All analyses were processed with the Statistical Package for the Social Sciences (SPSS) software version 18.0 (SPSS, Chicago, IL, USA).

Results

Incidence and clinical features of A. ursingii bacteremia

During the study period, 616 patients were found to have Acinetobacter species bacteremia and were included in our study. Among the isolates, 19 (3.1 %) were identified as A. ursingii by16S rRNA gene sequence analysis and confirmed by ITS sequence analysis (similarity: 98-99 % to reference strains) and 252 (40.9 %) as A. baumannii. For the PFGE analysis, two isolates had smeared DNA, two had 93 % similarity, and 15 had similarity less than 80 % (Fig. 1). The annual incidence of A. ursingii among Acinetobacter species bacteremia in this study was 1.5–5.2 %.

Fig. 1
figure 1

Pulse-field gel electrophoresis patterns of the Acinetobacter ursingii isolates

Full size image

The clinical data of the first three A. ursingii bacteremia patients were incomplete and they were therefore excluded from further analysis. The comparison of demographic features, underlying diseases, APACHE II scores, and outcomes of A. ursingii and A. baumannii bacteremia patients included in this study are summarized in Table 1.

Table 1 Demographic data, clinical features, and outcomes of patients with Acinetobacter ursingii and Acinetobacter baumannii bacteremia
Full size table

The gender of the patients with A. ursingii bacteremia was similar, while most patients with A. baumannii bacteremia were male. Primary bacteremia was mostly noted among those with A. ursingii infection (62.5 %), while respiratory tract infection (51.6 %) was the major source of A. baumannii bacteremia. The comorbidity of these two groups was similar, except that A. ursingii bacteremia tended to occur in patients with hematologic malignancies (p value < 0.01) or neutropenia who had undergone chemotherapy in the past month (p value <0.01). Patients with A. ursingii bacteremia had lower APACHE II scores (p value < 0.01), and less often acquired infection in the intensive care unit than patients with A. baumannii bacteremia (p value < 0.01). Consequently, patients with A. ursingii bacteremia underwent fewer invasive produces, including endotracheal tubing or tracheostomy (p value < 0.01), nasogastric tubing (p value < 0.01), and ventilator support (p value = 0.02).

About half of the patients with A. ursingii (50.8 %) and A. baumannii bacteremia (62.5 %) had received inappropriate antimicrobial therapy within 48 h of the onset of bacteremia. However, the 14-day (p value = 0.04) and 28-day (p value = 0.02) mortality rates of the A. ursingii group were significantly lower than those of the A. baumannii group. The Kaplan-Meier survival curves also showed that patients with A. ursingii had a higher cumulative survival rate than those with A. baumannii (Fig. 2).

Fig. 2
figure 2

The Kaplan-Meier survival curves of patients with bacteremia caused by Acinetobacter ursingii and Acinetobacter baumannii. The 30-day mortality rate of A. ursingii bacteremia was significantly lower than that of A. baumannii bacteremia (p-value = 0.0352)

Full size image

Identification

The identifications of these 19 clinical isolates of A. ursingii under the Vitek 2, Phoenix, and MALDI-TOF mass spectrometer systems are listed in Table 2. According to the ID-GNB card of the Vitek 2 system, 9 isolates (47.4 %) were correctly identified as A. ursingii. The other 10 isolates (52.6 %) were incorrectly identified as A. lwoffii. The Phoenix system incorrectly identified all 19 isolates. Among them, 15 isolates (78.9 %) were misidentified as Alcaligenes faecalis, 3 isolates (15.8 %) as A. lwoffii/haemolyticus, and 1 isolate (5.3 %) as Moraxella species. All the 19 isolates (100 %) were correctly identified as A. ursingii by using MALDI-TOF mass spectrometer analysis.

Table 2 Identifications obtained with the Phoenix, Vitek 2 systems, and matrix-assisted laser desorption ionization time-of-flight mass spectrometer for the Acinetobacter ursingii isolates
Full size table

Antimicrobial susceptibility

The antimicrobial susceptibility results of the 19 isolates are summarized in Table 3. All the A. ursingii isolates were resistant or had intermediate susceptibility to ceftriaxone and ceftazidime, and all were susceptible to levofloxacin and imipenem. About half of the A. ursingii isolates were resistant or had intermediate susceptibility to ciprofloxacin (47.4 %) and cefepime (42.1 %). A small number of the isolates were resistant or had intermediate susceptibility to amikacin (10.5 %), gentamicin (15.8 %), ampicillin-sulbactam (21.1 %), and colistin (15.8 %).

Table 3 Antimicrobial susceptibility profiles of the 19 Acinetobacter ursingii isolates
Full size table

Discussion

Acinetobacter ursingii is a rare pathogen that mostly causes bacteremia in patients with hematologic malignancies. In this study, most cases were of primary bacteremia, and patients had milder disease severity and underwent fewer invasive procedures than patients with A. baumannii bacteremia. Although more than half of the patients with A. ursingii and A. baumannii bacteremia had undergone inappropriate antimicrobial therapy within 48 h of the onset of bacteremia, the 14-day and 28-day mortality rates of patients with A. ursingii bacteremia were significantly lower than those of patients with A. baumannii bacteremia.

As in previous studies [12, 22], a low incidence of A. ursingii bacteremia was noted in our study (1.5–5.2 % during the study period). Compared to the risk factors and clinical characteristics of patients with A. baumannii bacteremia, patients with A. ursingii bacteremia are believed to be more immunosuppressed than patients with A. baumannii bacteremia, due to the higher concurrence rate in patients with hematologic malignancy, neutropenia, and chemotherapy treatment. Compared to the condition of patients with A. baumannii bacteremia, that of patients with A. ursingii seemed less severe, as indicated by lower APACHE II scores, fewer ICU admissions and invasive procedures, and lower mortality rates. The results indicated a lower virulence of A. ursingii, and this may account for the lower incidence of A. ursingii bacteremia.

It is unclear why most of the A. ursingii cases were primary bacteremia without an obvious source of infection. Among the patients with A. ursingii bacteremia, central venous catheters were placed in 6 patients (37.5 %), arterial catheters in 6 patients (37.5 %), and total parenteral nutrition in 1 patient (6.25 %). One patient (6.25 %) required dialysis and 4 patients (25 %) needed ventilator support at the onset of bacteremia. The intravascular device may serve as a port of entry for A. ursingii bacteremia.

Phenotypic schemes are generally insufficient to accurately identify the Acinetobacter isolates at the species level [23–25]. Phenotypic identification by commercial colorimetric systems is also unsatisfactory [6, 12, 26, 27]. Using systems such as the Vitek 2, API20NE systems (bioMérieux, Marcy l’Etoile, France) and the Phoenix system, the clinically relevant species of the A. calcoaceticus–A. baumannii complex are frequently uniformly identified as A. baumannii, and many other species are not identified [6, 12, 26, 27]. On comparison, the Vitek 2 systems in our study could correctly identify about half of the isolates, and Phoenix systems failed to correctly identify any. Protein fingerprinting using a MALDI-TOF mass spectrometer is a promising molecular method for rapid identification of Acinetobacter species with high-throughput capability. A previous study revealed a 72.4 % correct identification rate for Acinetobacter species by using MALDI-TOF mass spectrometer [28]. The correct rate of A. ursingii identification was 93.3 % in that study. As the database of the MALDI-TOF mass spectrometer expanded, 100 % of the A. ursingii isolates were correctly identified in our study.

In our study, all the A. ursingii isolates exhibited high resistance to third- and fourth-generation cephalosporins. These results suggest the presence of some extended spectrum β-lactamase, which deserves further study. More than 80 % of the A. ursingii isolates were susceptible to imipenem, levofloxacin, amikacin, gentamicin, and colistin. While comparing the antimicrobial susceptibility of the A. ursingii isolates in our study with the isolates reported by Cattoir in 2006 [12], the resistance rate to ciprofloxacin, gentamicin, amikacin, and colistin seemed higher. Further studies are needed to elucidate the mechanisms underlying antimicrobial resistance.

Conclusion

In conclusion, A. ursingii is a rare pathogen with a low mortality rate. This pathogen mostly causes primary bacteremia in patients with malignancies.