Tigecycline-based versus sulbactam-based treatment for pneumonia involving multidrug-resistant Acinetobacter calcoaceticus-Acinetobacter baumannii complex

Background The treatment options for pneumonia involving multidrug-resistant Acinetobacter calcoaceticus-Acinetobacter baumannii (MDR Acb) complex are limited, and the optimal treatment has not been established. Methods To compare the efficacy of tigecycline-based with sulbactam (or ampicillin/sulbactam)-based therapy for pneumonia involving MDR Acb complex, we conducted a retrospective study comparing 84 tigecycline-treated adult patients during the period August 2007 to March 2010 with 84 sulbactam or ampicillin/sulbactam-treated adult patients during the period September 2004 to July 2007. Both groups had the matched Acute Physiology and Chronic Health Evaluation (APACHE) II score and received treatment for at least 7 days. Results The mean APACHE II score was 20.1 for both groups. More patients in sulbactam group had ventilator use (89.3 % versus 69.0 %), bilateral pneumonia (79.8 % versus 60.7 %) and combination therapy (84.5 % versus 53.6 %), particularly with carbapenems (71.4 % versus 6.0 %), while more patients in tigecycline group had delayed treatment (41.7 % versus 26.2 %) (P <0.05). At the end of treatment, more patients in sulbactam group had airway MDR Acb complex eradication (63.5 % versus 33.3 %, P <0.05). The clinical resolution rate was 66.7 % for both groups. The mortality rate during treatment was 17.9 % in sulbactam group, and 25.0 % in tigecycline group (P = 0.259). The multivariate analysis showed that bilateral pneumonia was the only independent predictor for mortality during treatment (adjusted odds ratio, 2.717; 95 % confidence interval, 1.015 to 7.272). Conclusions Patients treated with either tigecycline-based or sulbactam-based therapy had a similar clinical outcome, but tigecycline group had a lower microbiological eradiation rate.

Tigecycline is a glycylcycline with in vitro activity against MDR Acb complex [6]. The comparison analysis from the U.S. Food and Drug Administration showed that tigecycline treatment had a higher mortality rate than other antimicrobials in ventilator associated pneumonia (VAP) [7]. A recent study also reported a significantly lower cure rate in clinically evaluable patients with VAP treated with tigecycline when compared to imipenem (47.9 % versus 70.1 %) [8]. However, for pneumonia caused by MDR Acb complex resistant to carbapenems and other classes of antibiotics, off label use of tigecycline was common in clinical practice, and the clinical response rates ranged from 60 to 88 % in prior studies [9][10][11]. Sulbactam is a β-lactamase inhibitor with antimicrobial activity against Acinetobacter species [12]. It is available alone or in combination with ampicillin, and ampicillin doesn't contribute activity or synergism against A. baumannii [12]. Sulbactam or ampicillin/sulbactam had clinical response rates ranging from 67 to 75 % for pneumonia involving MDR A. baumannii (MDRAB) or MDR Acb complex in prior studies [13][14][15].
In our hospital, tigecycline was not available until August 2007. Before that, sulbactam or ampicillin/sulbactam might be the only treatment option with in vitro activity against MDR Acb complex. Thus, we conducted a retrospective study to compare the efficacy of tigecycline-based with sulbactam (or ampicillin/ sulbactam)-based treatment for pneumonia involving MDR Acb complex. With a match in the Acute Physiology and Chronic Health Evaluation (APACHE) II score for both groups, a comparison was made between tigecycline-treated adult patients during the period August 2007 to March 2010 and sulbactam (or ampicillin/sulbactam)-treated adult patients during the period September 2004 to July 2007. The clinical efficacy, outcomes and microbiological eradication were included for analyses.

Setting
Chang Gung Memorial Hospital (CGMH)-Linkou is a university-affiliated medical center providing both primary and tertiary health care in northern Taiwan. This retrospective study has been approved by institutional review boards of CGMH-Linkou (Number: 99-1478B and 100-0294B). The ethics committee granted a waiver for informed consent to be obtained.

Study design, patients and treatments
All hospitalized patients who were ≧ 18 years old and had pneumonia involving MDR Acb complex treated with tigecycline between August 2007 and March 2010, and sulbactam or ampicillin/sulbactam between September 2004 and July 2007, were reviewed. Each tigecycline-treated patient was matched to one sulbactam or ampicillin/sulbactam-treated patient based on identical values of APACHE II score and chart number sequence. Patients were excluded if they did not have a matched control or had a combination therapy with tigecycline and sulbactam (or ampicillin/sulbactam). Patients with initial bacteremia were also excluded since tigecycline treatment for bacteremia was controversial.
Pneumonia was diagnosed if the patient had a radiographic infiltrate that was new or progressive, along with at least two of the following clinical characteristics: new onset of fever (≧ 38°C) or hypothermia (< 35.5°C), leucocytosis (leucocyte count > 12000 cells/mm 3 ) or leucopenia (leucocyte count < 4000 cells/mm 3 ), decline in oxygenation (O2 saturation < 90 %), and increasing amount of purulent sputum [16]. Pneumonia involving MDR Acb complex was defined as clinical evidence of pneumonia with sputum or tracheal aspirate cultures positive for MDR Acb complex from 1 week before to 3 days after the first dose of tigecycline or sulbactam or ampicillin/sulbactam. Tracheal aspirate and sputum specimens were sent for bacterial culture only if their Gram's stains showed at least 25 neutrophils and less than 10 epithelial cells per low-power field. Growth was assessed semi-quantitatively. The etiologic pathogen of pneumonia was determined if the tracheal aspirate or sputum culture had an at least moderate growth, i.e., the growth confined up to primary streaking line and > 5 colonies in secondary streaking zone [17]. Polymicrobial pneumonia was defined as one or more additional etiologic bacterial species concurrently isolated from the respiratory tract during treatment.
All patients in tigecycline group received tigecycline for at least 7 days, with a 100-mg loading dose followed by 50 mg administered intravenously every 12 h. All patients in sulbactam group received intravenous sulbactam 1 g or ampicillin/sulbactam 3 g (at a ratio 2:1) every 6 or 8 h for at least 7 days. Dose and dosing interval were adjusted according to serum creatinine levels. Combination therapy was defined as simultaneous use of another class of antibiotics for at least 3 days. These antibiotics included carbapenems (meropenem or imipenem), fluoroquinolones (ciprofloxacin or levofloxacin), amikacin, cephalosporins (ceftazidime or cefepime), piperacillin, piperacillin-tazobactam, colistimethate, and aztreonam. Delayed treatment was defined as more than 3 days between the detection of airway MDR Acb complex isolates and the first dose of tigecycline or sulbactam or ampicillin/sulbactam.

Microbiology
Identification of Acb complex depended upon Gram staining and conventional biochemical tests [18]. Briefly, the isolates were identified as species of the genus Acinetobacter based on the following properties: aerobic, Gram-negative, nonmotile coccobacillary rods with a nonfermentative, catalase-positive and oxidase-negative reaction. Acinetobacter species with glucose-oxidizing non-haemolytic characteristics were classified as Acb complex. Antimicrobial susceptibility was determined and interpreted according to the criteria of Clinical and Laboratory Standards Institute by disk diffusion method [19]. Susceptibility to tigecycline was determined using disk diffusion method with Mueller-Hinton agar (BD Microbiology Systems, Cockeysville, MD) with the resistant breakpoint at ≧ 16 mm and susceptible breakpoint at ≦ 12 mm [20]. An isolate with full or intermediate resistance to amikacin, gentamicin, cefepime, ceftazidime, aztreonam, piperacillin, piperacillin-tazobactam, ciprofloxacin, imipenem and meropenem was defined as MDR Acb complex [21].
Cultures were collected from 1 week before the first dose of tigecycline or sulbactam (or ampicillin/sulbactam) to the discharge of patients. Pathogens, sites of growth and susceptibility testing were recorded. Microbial eradication of MDR Acb complex was defined as no growth of Acb complex or susceptibility change from MDR strains to susceptible strains in Acb complex in follow-up respiratory tract cultures before and 7 days after treatment cessation. Relapse was defined as new isolation of MDR Acb complex from the respiratory tract cultures within 2 weeks after initial eradication. Initial bacteremia was defined as bacteremia at the beginning of treatment, which meant at least one positive blood culture 1 week before to 3 days after the first dose of tigecycline or sulbactam or ampicillin/sulbactam. Bacteremia during treatment was defined as at least one positive blood culture 3 days after to the end of treatment.

Demography and comorbidity
Data on age, sex, surgery, and co-morbid illness were gathered by reviewing in-patient medical records. Comorbid illness included hepatic dysfunction of a serum total bilirubin level over 2.5 mg/dL or liver cirrhosis, renal insufficiency of a serum creatinine level above 2.0 mg/dL or requirement of dialysis, chronic pulmonary disease, heart disease, diabetes mellitus, immune compromise, and hematological or solid organ malignancy. Immune compromise was defined by at least one of the following: use of prednisone or equivalent over 20 mg per day for at least 2 weeks, organ transplant recipient, human immunodeficiency virus infection or acquired immunodeficiency syndrome, neutropenia (absolute neutrophil count less than 500 cells/mm3), use of immunosuppressive agents, and concurrent hematological malignancy.

Clinical conditions and outcomes
Ventilator use, vital signs, and infections other than pneumonia during treatment were recorded. Defervescence was defined as normal body temperature for at least 3 days at the end of treatment. Severity of illness was assessed by a modified APACHE II score, which was recorded within 48 h before or after the first dose of tigecycline or sulbactam or ampicillin/sulbactam. The 30-day mortality was defined as death occurring within 30 days after treatment. The chest radiographs were evaluated by at least two investigators. A series of chest radiographs were evaluated during treatment. Clinical resolution of pneumonia at the end of treatment was defined as (1) decreased pulmonary infiltrate, and (2) survival with stationary findings on chest radiographs and defervescence. Thus, patients with persistent fever or death during treatment would be defined as clinical failure if infiltrates were stationary. Progressing infiltrates were defined as clinical failure.

Statistical methods
All statistical analyses were performed using the Statistical Package for the Social Sciences for Windows (Version 15.0; SPSS Inc., Chicago, IL, USA). Categorical variables were compared using χ 2 test or Fisher exact test, as appropriate. Continuous variables were tested for normality of distributions by Kolmogorov-Smirnov test, and then compared by Student's t-test or the Mann-Whitney U test, as appropriate. Odds ratios (ORs) and 95 % confidence interval (CI) were calculated. The survival curve was plotted by means of the Kaplan-Meier method, and the log rank test was used to compare univariate survival distribution between tigecycline and sulbactam groups. Variables with a P value < 0.1 in univariate analysis and tigecycline use were included in a logistic regression model for multivariate analysis. All tests were two-tailed, and a P value of < 0.05 was considered significant.

Patients, demography and concomitant diseases
One hundred and sixteen tigecycline-treated episodes of pneumonia involving MDR Acb complex were identified in 112 patients, while 177 sulbactam or ampicllin/ sulbactam-treated episodes were identified in 173 patients. Finally, 84 tigecycline-treated patients were enrolled and matched to 84 patients treated with sulbactam (26 patients) or ampicillin/sulbactam (58 patients). The mean APACHE II score was 20.1 for both groups. In tigecycline group, 59 (70.2 %) and 25 (29.8 %) patients had positive MDR Acb complex cultures from tracheal aspirates and sputum, respectively. In sulbactam group, 73 (86.9 %) and 11 (13.1 %) patients had that from tracheal aspirates and sputum, respectively. There was no significant difference in age, gender, and concomitant diseases between the two groups (Table 1).

Clinical conditions
Patients in sulbactam group had more ventilator use (89.3 % versus 69.0 %) and bilateral pneumonia (79.8 % versus 60.7 %) than those in tigecycline group. There were no significant differences between these two groups in the overall rates of polymicrobial pneumonia and multisite infections. Pseudomonas aeruginosa and Methicillin-resistant Staphylococcus aureus were the most  Table 2). Other models of multivariate analysis including polymicrobial pneumonia, combination therapy, and combination with carbapenem or colistin also showed that bilateral pneumonia was the only independent predictor (Table 3).

Monotherapy of tigecycline and sulbactam
Thirty-nine (46.4 %) patients had tigecycline monotherapy and 13 (15.5 %) had sulbactam or ampicillin/sulbactam monotherapy. Tigecycline group had significant lower rates of ventilator use, bilateral pneumonia, and airway eradication of MDR Acb complex. Both groups had similar clinical resolution rates. However, tigecycline group had lower rates of 30-day mortality and mortality during treatment (25.6 % versus 53.8 %, 17.9 % versus 30.8 %, P >0.05). In the univariate analysis for the patients with monotherapy, both the survivors and the deceased during treatment had no significant difference in demography, concomitant diseases, clinical conditions, and treatment (Table 4).

Discussion
Prior case series studies reported clinical response rates ranging from 60 to 88 % in tigecycline treatment [9][10][11], and 67 to 75 % in sulbactam or ampicillin/sulbactam treatment for pneumonia involving MDRAB or MDR Acb complex [13][14][15]. There were only a few comparative studies investigating the efficacy of tigecycline or sulbactam, and usually they were compared with colistin or polymyxin, the other major treatment option for MDR Acb complex. Betrosian AP et al. reported that high-dose ampicillin/sulbactam monotherapy and colistin were comparably safe and effective treatment for critically ill patients with MDRAB VAP. The clinical success and improvement rate was 76.9 % for ampcillin/sulbactam group and 73.3 % for colistin group [22]. Oliveira MS et al. reported another study comparing ampicillin/ sulbactam with polymyxins in treating infections caused by carbapenem-resistant Acinetobacter spp. [23]. In the study, about half of the enrolled patients had Acinetobacter bacteremia, and quarter of them had pneumonia. The mortality rate during treatment was 33 % in ampicillin/sulbactam group and 50 % in polymyxin group, and polymyxin use was an independent factor associated with mortality during treatment [23]. Chuang YC et al. reported a study comparing tigecycline-based to colistinbased therapy for MDRAB pneumonia in intensive care units. The tigecycline group has an excess mortality of 16.7 % (60.7 % versus 44 %, P = 0.04). The excess mortality of tigecycline is significant only among those with minimal inhibitory concentration (MIC) > 2 μg/mL, but not for those with MIC ≦ 2 μg/mL [24].
To our knowledge, the study was the first comparative study of tigecycline-based versus sulbactam or ampicillin/sulbactam-based treatment for pneumonia involving MDR Acb complex. Our two patient groups were from different but successive time periods, and the major treatment for MDR Acb complex was different in each time period in our hospital. Before August 2007, sulbactam or ampicillin/sulbactam was the only option probably with in vitro activity against MDR Acb complex in our hospital. After that, tigecycline became the major treatment option because of its high susceptibility rate to MDR Acb complex. However, the clinical and microbiological diagnostic criteria and definition, and standards of care and infection control were similar in both time periods. Covariate adjustment with multivariate analyses and matching with disease severity were performed to reduce bias of the historically controlled comparison.
The patients from both groups were aged with complicated underlying diseases and high disease severity. Higher rates of ventilator use and bilateral pneumonia reflected that sulbactam group might have a higher severity of pneumonia than tigecycline group. A higher rate of delay use in tigecycline group might reflect the early policy of tigecycline use in our hospital: usually tigecycline was not used as empiric or first-line regimen for nosocomial infection. Both groups had similar clinical outcomes. Bilateral pneumonia was the only independent predictor for mortality during treatment in different models of multivariate analysis. Combination therapy did not stand out as an independent predictor, which might be due to difference of combination strategies and regimens between the two patient groups. Most patients in sulbactam group had combination with carbapenem for synergistic effect against MDR Acb complex; however, tigecycline group mainly had anti-pseudomonal cephalosporins and fluoquinolones to cover Pseudomonas aeruginosa.  All variables included in the final multivariable model are shown c The initial airway MDR Acb complex isolates with resistance to TG in TG group, or with resistance to SB in SB group d 26 patients in the deceased group and 126 in the survivors group had available data for evaluation Because most patients in sulbactam group had concurrent carbapenem use, the study results in them might mainly reflect the efficacy of combination of sulbactam and carbapenem. In the comparative analyses of monotherapy, the patients with sulbactam or ampicillin/sulbactam monotherapy had relatively higher mortality rates than the patients with tigecycline monotherapy or the overall sulbactam group. The results implied that combination with carbapenem might improve clinical outcomes of sulbactam-based treatment. Besides, more than half of the sulbactam group had sulbactam-resistant MDR Acb complex isolates during treatment. Combination with carbapenem might play a role giving a high airway eradication rate. The  The initial airway MDR Acb complex isolates with resistance to TG in TG group, or with resistance to SB in SB group d 35 patients in TG group and 9 in SB group; 7 patients in the deceased group and 37 in the survivors group had available data for evaluation patients with sulbactam monotherapy also had a high airway eradication rate, but most of them did not have sulbactam-resistant MDR Acb complex. Synergistic effect against MDRAB with the combination of sulbactam and carbapenem had been reported [4,25]. However, the synergistic effect was associated with the MICs of carbapenem and sulbactam. If the MICs exceeded achievable serum levels, the potential of sulbactam/carbapenem combination as treatment regimen for MDRAB infections might be limited [4]. In our study, full or intermediate sulbactam resistance was detected in 84.5 % of tigecycline-treated patients, therefore, physicians tended to use tigecycline for these patients with sulbactam-resistant MDB Acb complex. For these cases, the clinical outcomes of sulbactam group might not be achieved if they received sulbactam/carbapenem combination therapy.
There are some other limitations in this study. First, our respiratory specimens were clinical specimens from clinical practice, and they might not be obtained from deep sites in the lungs. Growths of etiologic pathogens were assessed semi-quantitatively if the specimens were qualified for culture. We cannot absolutely distinguish airway MDR Acb complex infections from colonization. However, our definition for pneumonia was practical, and our conclusion based on clinically relevant data and management could provide important information for clinical practice. Second, polymicrobial pneumonia and concomitant infections were common, and the clinical impact of other etiologic pathogens or extrapulmonary infections was not evaluated comprehensively. Third, we studied MDR Acb complex rather than MDRAB. Although prior studies reported that about 90 % of Acb complex with multidrug or carbapenem resistance was the genomic specie of A.baumannii, comparison with studies on A. baumannii isolates are not straightforward [26].

Conclusions
Tigecycline-based treatment had a similar clinical outcome to sulbactam or ampicillin/sulbactam-based treatment for pneumonia involving MDR Acb commplex, but tigecycline group had a lower microbiological eradiation rate. More comparison studies are essential to establish the optimal regimens for pneumonia involving MDR Acb complex.