Natural prognosis of carbapenem-resistant Acinetobacter baumannii bacteremia in patients who did not receive appropriate antibiotic treatment: A retrospective multicenter study in Korea.

Carbapenem-resistant Acinetobacter baumannii (CRAB) infection is a major issues in current era. The aim of study was to investigate the natural prognosis and prognostic factors associated with 28-day mortality in patients with CRAB bacteremia who were not receiving appropriate antibiotic treatment.Adult patients with CRAB bacteremia were retrospectively identified between April 2012 and March 2015 at 5 tertiary hospitals in Republic of Korea. Patients who were transferred to another hospital within 28 days of onset of bacteremia and who receive appropriate antibiotics more than 48 hours were excluded. We investigated prognostic factors associated with 28-day mortality in patients with CRAB bacteremia without appropriate antibiotic treatment.Of enrolled 205 patients, 143 (69.8%) patients died within 28 days after blood culture. Of patients with 28-day mortality, 88.9% (127/143) of patients died within 5 days. Of 78 patients who survived more than 5 days, the 28-day mortality was 20.5% (16/78). Diabetes mellitus (adjusted odds ratio [aOR] 3.81, 95% confidence interval [95% CI] 1.19-12.20), immunocompromised (aOR 8.72, 95% CI 2.62-29.70), sequential organ failure assessment (SOFA) ≥ 10 (aOR 13.87, 95% CI 3.70-51.96), vasopressor use (aOR 7.03, 95% CI 1.79-27.60), and pneumonia (aOR 4.44, 95% CI 1.67-11.78) were found to be the factors independently associated with the 28-day mortality.The 28-day mortality in patients with CRAB bacteremia without appropriate treatment was high, although some patients could survive. Severity and underlying conditions were important prognostic factors in patients with CRAB bacteremia.

Introduction

Acinetobacter baumannii is one of the major pathogens of hospital-acquired infection, which causes various diseases such as pneumonia, catheter-related infection, and urinary tract infection.[ This nonferment gram-negative rod bacteria has been notorious for its strong ability to acquire resistance to various antibiotics.[ Especially, high rate of carbapenem-resistant clinical isolates has been reported worldwide.[ In Korea, the rate of carbapenem resistance of A baumannii identified in hospitals has been reported to reach 90%.[

Understanding of natural prognosis of Carbapenem-resistant A baumannii (CRAB) bacteremia is important for establishing diagnostic and treatment strategies. If patients have died before the diagnosis of CRAB bacteremia, and they cannot receive appropriate antibiotics, we need to use a method that can be quickly diagnosed CRAB bacteremia. Choosing antibiotics against CRAB as empirical treatment also should be considered in this case. If patients survive without appropriate antibiotics, they may not need to be given immediately empirical antibiotics than can result in antibiotics toxicity and resistance. Information on risk factors associated with mortality can be useful for development of individualized strategies. Also, knowledge on natural prognosis can be used as comparative data for studies on diagnostic and therapeutic intervention.

Therefore, we investigated the natural history of patients with CRAB bacteremia without receiving appropriate antibiotic treatment. In addition, we investigated prognostic factors associated with 28-day mortality.

Methods

Study population and design

This retrospective multicenter study was conducted at 5 tertiary teaching hospitals located in Seoul, Bucheon, Ilsan, and Cheonan, Republic of Korea. By reviewing the daily computerized reports of blood cultures, the patients aged ≥18 years with a CRAB-positive blood culture were identified between April 2012 and March 2015. Bacteremia was defined as ≥ a positive blood culture for A baumannii and the presence of the clinical features compatible with infection. If a patient had undergone recurrent episodes of A baumannii bacteremia during the study period, only the first episode was considered.

The algorithm for selecting the research subjects and the research plan is shown in Fig. 1. It was defined as an appropriate antibiotic therapy if at least 1 antimicrobial agent to which the causative pathogen was susceptible was administered for more than 48 hours.[ Patients who were transferred to another hospital within 28 days of onset of bacteremia were excluded from the study. Among patients not receiving appropriate antibiotic treatment, the case group was defined as a patient who died within 28 days after blood culture test was done and survivors were categorized into the control group. To identify factors associated with 28-day mortality in patients with CRAB bacteremia without appropriate antibiotic treatment, clinical characteristics of case group and control group were compared.

Figure 1

The study algorithm for the analysis of prognostic factors associated with 28-day mortality in patients with carbapenem-resistant Acinetobacter baumannii (CRAB) bacteremia who did not receive appropriate antibiotic treatment.

Data collection

Patient's demographic data and information on comorbidity (malignancy, neurologic diseases, chronic lung diseases, diabetes mellitus [DM], liver cirrhosis on Child B or C classification, heart failure, and chronic kidney diseases [CKD]) were collected. Patients were defined as immunocompromised if they had human immunodeficiency virus (HIV) or acquired immunodeficiency syndrome (AIDS), were on solid organ transplantation, hematopoietic stem cell transplantation, had received chemotherapy within 6 weeks, had received systemic steroid equivalent to or higher than 20 mg of prednisone for 2 weeks or other immunosuppressive agents within 2 weeks before hospitalization. The severity of comorbidities was classified according to the McCabe and Jackson Classification system.[ Source of infection was determined according to the guidelines issued by the Centres for Disease Control and Prevention.[ Mechanical ventilation, vasopressor use, dialysis, APACHE II score,[ and the Sequential Organ Failure Assessment (SOFA) score [ were chosen as the index of severity. Previous CRAB colonization was defined as the isolation of CRAB from any clinical specimens within 30 days of the CRAB blood culture without symptoms or signs. Polymicrobial bacteremia was defined as isolation of any organism from blood within 2 days of the CRAB blood culture. The practice of 3rd generation antibiotics such as cephalosporins, cefepime, piperacillin/tazobactam, fluoroquinolones, carbapenems, tigecycline, and colistin just before CRAB blood culture was investigated.

Microbiologic methods

A baumannii identification was performed using standard methods. At least 4 species, A calcoaceticus, A baumannii, A nosocomialis, and A pitti, are invariably reported as A baumannii by clinical microbiology laboratories because biochemical methods cannot further differentiate them to the species level. Susceptibility testing was done using the microdilution method (MicroScan system; Baxter Health Care, West Sacramento, CA), and results were interpreted according to the National Committee for Clinical Laboratory Standards guidelines.[ Imipenem resistance was defined as a minimal inhibitory concentration of 16 μg/mL or above. A baumannii isolates with imipenem resistance were considered CRAB.

Statistical analysis

Statistical analyses were performed with SPSS for Windows (version 14.0); SPSS Inc, Chicago, IL). Continuous variables were compared using the Mann–Whitney U test or Student t test, as appropriate. Categorical variables were compared using Pearson χ2 test or Fisher exact test. A binary logistic regression was used to identify variables significantly associated with 28-day mortality in patients with CRAB bacteremia who did not receive appropriate antibiotic treatment. The variables found to be statistically significant at a 5% level in the univariate analysis were included in the multivariate analysis. Among indices of severity, SOFA score ≥ 10 was chosen as a binary variable according to the calculation of receiver operating curve. All significance testing was 2-tailed, and P < .05 was considered statistically significant.

Results

A total of 332 patients with CRAB bacteremia were identified. One hundred eleven patients who received appropriate antibiotics and 10 patients who were transferred to other hospitals within 28 days after the blood culture test was done were excluded. Finally, 205 patients who did not appropriate antibiotics were enrolled in the study. It usually took 4 days (range, 2–8 days) to receive results after blood culture was done. Among patients who did not receive appropriate antibiotics, carbapenems (114, 55.6%) were most commonly prescribed as antibiotics for gram-negative bacteremia, followed by piperacillin/tazobactam (40, 19.5%) and cefepime (1.4%). Thirteen patients were categorized into the group who did not receive appropriate antibiotics despite of colistin usage, because they died before administration of colistin for more than 48 hours. The 28-day mortality in patients with CRAB bacteremia without appropriate antibiotic treatment was 69.8% (143/205). Of patients with 28-day mortality, 88.9% (127/143) of patients died within 5 days. Patients who died within 28 days were included in the case group and survivors were classified as the control group.

The differences in the clinical characteristics of case group and control groups are shown in Table 1. Patients with 28-day mortality were less likely to be included in the nonfatal McCabe classification (case group 53, 37.1% vs control group 43, 69.4%; P < .01). Malignancy (case group 51, 35.7% vs control group 13, 21.0%; P = .048), DM (case group 50, 35.0% vs control group 13, 21.0%; P = .049), CKD (case group 23, 16.0% vs control group 3, 4.8%; P = .04) and immunocompromised status (case group 41, 28.7% vs control group 8, 12.9%; P = .02) were underlying medical conditions that were more frequently found in patients who died within 28 days. CRAB bacteremia was more frequently developed during intensive care unit hospitalization in patients with 28-day mortality (case group 117, 81.8% vs control group 31, 50.0%; P < .01). Of severity indices, the SOFA score (median 13, interquartile range [IQR] 9–16 vs 5, 4–7; P < .001) and APACHE II score (median 22, IQR 14–31 vs 15, 4–21; P < .001) of patients who died within 28 days were higher than those who survived. Also, patients with 28-day mortality were more likely to receive mechanical ventilation (case group 95, 66.4% vs control group 17, 27.4%; P < .001), vasopressor use (case group 92, 64.3% vs 4, control group 6.5%; P < .01), or dialysis (case 46, 32.2% vs control group 3, 4.8%; P < .01). As shown in Table 2, variables of McCabe classification, malignancy, DM, CKD, immunocompromised status, ICU admission, SOFA ≥ 10, mechanical ventilation, vasopressor use, dialysis, and pneumonia as a source of bacteremia were included in the multivariate analysis. Finally, DM (adjusted odds ratio [aOR] 3.81, 95% confidence interval [95% CI] 1.19–12.20), immunocompromised (aOR 8.72, 95% CI 2.62–29.70), SOFA ≥ 10 (aOR 13.87, 95% CI 3.70–51.96), vasopressor use (aOR 7.03, 95% CI 1.79–27.60), and pneumonia (aOR 4.44, 95% CI 1.67–11.78) were found to be the factors independently associated with the 28-day mortality in patients with CRAB bacteremia who did not receive appropriate antibiotic treatment.

Table 1

Comparison of clinical characteristics between case and control groups with carbapenem-resistant Acinetobacter baumannii (CRAB) bacteremia in patients who did not receive appropriate antibiotic treatment.

Table 2

Prognostic factors associated with 28-day mortality in patients with carbapenem-resistant Acinetobacter baumannii (CRAB) bacteremia who did not receive appropriate antibiotic treatment.

The survival distribution of patients with CRAB bacteremia without adequate antibiotic treatment is shown in Fig. 2. Of patients who survived more than 5 days, the 28-day mortality was 20.5% (16/78). All of 23 patients who survived more than 5 days and did not have any of bad prognostic factors survived more than 28 days despite not receiving appropriate antibiotic treatment.

Figure 2

Kaplan–Meier survival curve of patients with carbapenem-resistant Acinetobacter baumannii (CRAB) bacteremia without appropriate antibiotic treatment. Figure B shows the comparison of survival between patients with any of risk factor and with none of risk factor.

Discussion

This study reports that patients with CRAB bacteremia who did not receive appropriate antibiotic treatment died early. Despite high mortality in patients with CRAB bacteremia, it is interesting to note that some patients survived from CRAB bacteremia even after not receiving appropriate antibiotic treatment. This is the first clinical study that shows natural prognosis of patients with CRAB bacteremia without appropriate antibiotics. The present study was expected to be used for anticipating prognosis of patients with CRAB bacteremia and establishing treatment strategies for CRAB bacteremia.

Major cases of mortality occurred within 5 days and this was also the case in patients without any risk factor associated with 28 day mortality (Fig. 2). However, none of patients who did not have any poor prognostic factors died if they survived from CRAB bacteremia for more than 5 days. This result suggests that diagnostics and therapeutic strategies for CRAB bacteremia should focus on reducing mortality of those who died within 5 days. As delayed treatment increases the mortality,[ empirical treatment including antibiotics against CRAB which can reach the therapeutic level quickly could be a possible option in endemic areas. To make use of early diagnostic tools such as matrix-assisted laser desorption ionization time-of-flight mass spectrometry,[ an effective therapeutic regimen is mandatory. However, colistin, the current drug of choice for CRAB, may not be suitable weapon owing to its pharmacokinetic/ pharmacodynamic characteristics.[ Hopefully, ongoing research will lead to the development of a monoclonal antibody that binds to the surface of A baumannii.[

Three aspects to determine the prognosis of patients with CRAB bacteremia include host factor, therapeutic factor and microbiologic factor. First, comorbidity of host is a well-known prognostic factor associated with mortality in patients with CRAB bacteremia. In a previous observation-based study, Charlson index (hazard ratio 1.16, 95% CI 1.02–1.32; P .028) was associated with 30-day crude mortality of patients with CRAB bacteremia.[ A prospective cohort study involving adult patients with CRAB infection also reported McCabe classification as a prognostic factor associated with overall mortality.[ The association of DM and immunocompromised status with higher 28-day mortality in our study is consistent with the previous studies. Second, it is evident that severity was associated with prognosis in CRAB bacteremia as SOFA ≥ 10 was associated with higher 28-day mortality in our study. In addition to the SOFA score,[ Pitt Bacteremia Score,[ septic shock,[ and APACHE II score [ have been reported as severity index associated with mortality in CRAB infection. The severity of bacteremia is not only affected by the host factor but also by the virulence of microorganisms. Biofilm formation, phospholipase C production, hemolytic activity, acinetobactin production, and rpoB gene encoding the β-subunit of the RNA polymerase have been known as virulence factors of A baumannii.[ Interestingly, there are reports stating that mortality varies depending on the clone.[ This phenomenon can be explained by the fact that only certain clone has a virulence factor.[ Unfortunately, in our retrospective study, microbiologic factors associated with severity of CRAB bacteremia could not be evaluated, because isolates of CRAB were not collected during the study period. This result emphasizes the necessity of microbiologic analysis in the future.

This study had several limitations. First, some variables that affected the outcomes in patients with CRAB bacteremia may have been omitted from the analysis due to the retrospective nature of the study. Second, death in some patients might not have been caused by CRAB bacteremia, but rather by alternative causes such as terminal cancer or impediments to care such as refusal of intensive care treatment. Third, the definition of appropriate empirical antibiotics can be controversial. In our previous analysis, appropriate empirical antimicrobial therapy for more than 48 hours, within 5 days of the onset of bacteremia, was observed as an associated factor with survival in patients with CRAB bacteremic pneumonia.[ Based on this analysis, we defined appropriate antibiotics as described in the method. Lastly, we studied A baumannii complex rather than A baumannii. A variable of species that could affect mortality was not evaluated,[ although previous studies reported that about 90% of A baumannii complex with multidrug or carbapenem resistance belonged to the genomic species of A baumannii. We are in the process of carrying out prospective studies on carbapenem-resistant gram-negative bacteremia including microbiological analysis.

In conclusion, most of the patients with CRAB bacteremia died early, although some patients with CRAB bacteremia can survive without appropriate antibiotic treatment. Severity and underlying conditions were important prognostic factors in patients with CRAB bacteremia. This study suggests that efforts should be made to find ways to reduce the early mortality in patients CRAB bacteremia.

Author contributions

Conceptualization: Tark Kim, Eun Jung Lee, Seong Yeon Park, Ki-Ho Park.

Data curation: Tark Kim, Eun Jung Lee, Seong Yeon Park, Shi Nae Yu, Yu Mi Lee, Ki-Ho Park, Se Yoon Park, Min Hyok Jeon, Eun Ju Choo, Tae Hyong Kim, Mi Suk Lee.

Formal analysis: Tark Kim, Eun Jung Lee.

Investigation: Tark Kim.

Methodology: Tark Kim, Seong Yeon Park, Ki-Ho Park.

Project administration: Tark Kim.

Supervision: Eun Jung Lee, Tae Hyong Kim.

Validation: Tark Kim, Eun Jung Lee.

Writing – original draft: Tark Kim.

Writing – review & editing: Tark Kim, Eun Jung Lee.