Journal of Global Antimicrobial Resistance 24 (2021) 429–439

Contents lists available at ScienceDirect

Journal of Global Antimicrobial Resistance

journal homepage: www.elsevier.com/locate/jgar

What is the optimal antibiotic treatment strategy for

carbapenem-resistant Acinetobacter baumannii (CRAB)? A multicentre
study in Korea
Hyeri Seoka,1, Won Suk Choia,1, Shinwon Leeb , Chisook Moonc , Dae Won Parka ,
Joon Young Songa , Hee Jin Cheonga , Jieun Kimd , Jin Yong Kime , Mi Na Parkf ,
Yang Ree Kimg, Hyo-Jin Leeg , Bongyoung Kimd, Hyunjoo Paid , Yu Mi Joa , Jong Hun Kima ,
Jang Wook Sohna,*
a
Division of Infectious Diseases, Department of Internal Medicine, Korea University College of Medicine, Seoul, Republic of Korea
b
Department of Internal Medicine, Pusan National University School of Medicine and Medical Research Institute, Pusan National University Hospital, Busan,
Republic of Korea
c
Division of Infectious Diseases, Busan Paik Hospital, Inje University, Busan, Republic of Korea
d
Division of Infectious Diseases, Hanyang University College of Medicine, Seoul, Republic of Korea
e
Department of Internal Medicine, Incheon Medical Center, Incheon, Republic of Korea
f
Infection Control Office, Incheon Medical Center, Incheon, Republic of Korea
g
Division of Infectious Diseases, Department of Internal Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Article history: Objectives: The optimal treatment option for carbapenem-resistant Acinetobacter baumannii (CRAB) is
Received 5 November 2020 still limited. This study investigated the efficacy of three or more antibiotic types and regimens for
Received in revised form 13 January 2021 treatment of CRAB infection in high CRAB endemic areas.
Accepted 29 January 2021
Methods: A multicentre retrospective study was conducted to evaluate the efficacy of treatment types and
Available online 8 February 2021
regimens of CRAB infections in 10 tertiary hospitals in the Republic of Korea. The outcomes comprised 7-
day and 28-day mortality, and clinical and microbiological responses at 7 days, 28 days, and the end of
Keywords:
treatment. Nephrotoxicity and hepatotoxicity were evaluated as drug adverse reactions.
Acinetobacter baumannii
Acinetobacter infections
Results: A total of 282 patients were included in the study. Among the CRAB strains, the two most
Antibacterial agents susceptible antibiotics were colistin (99.6%) and minocycline (80.4%). A combination of colistin and
Drug resistance carbapenem significantly reduced 7-day mortality, and a sulbactam-containing regimen significantly
Combination drug therapy reduced 28-day mortality. Colistin monotherapy was significantly associated with increased 7-day and
28-day mortality. A minocycline-containing regimen showed the best microbiological responses at 7
days, 28 days, and the end of treatment. Colistin and tigecycline were associated with increased
nephrotoxicity and hepatotoxicity, respectively. Subgroup analysis of patients with pneumonia showed
similar results to the overall CRAB infection.
Conclusions: A combination of colistin and carbapenem and sulbactam-containing regimen may
contribute improved mortality in CRAB infections. Colistin monotherapy should be considered cautiously
in severe CRAB infections or CRAB pneumonia. A minocycline-containing regimen showed the best
microbiological responses, and further studies may be needed to evaluate improved mortality.
© 2021 The Authors. Published by Elsevier Ltd on behalf of International Society for Antimicrobial
Chemotherapy. This is an open access article under the CC BY-NC-ND license (http://creativecommons.
org/licenses/by-nc-nd/4.0/).

1. Background

The threat of Acinetobacter baumannii is increasing among

global antimicrobial resistance problems [1]. Acinetobacter bau-
* Corresponding author at: Division of Infectious Diseases, Department of mannii—one of the difficult-to-treat ESKAPE pathogens (Entero-
Internal Medicine, Korea University Anam Hospital, Korea University College of
coccus faecium, Staphylococcus aureus, Klebsiella pneumoniae,
Medicine, 73 Goryeodae-ro, Seongbuk-gu, Seoul 02841, Republic of Korea.
E-mail address: [email protected] (J.W. Sohn). Acinetobacter baumannii, Pseudomonas aeruginosa, and Enter-
1
These authors contributed equally to this article. obacter species)—is identified particularly in intensive care units

http://dx.doi.org/10.1016/j.jgar.2021.01.018
2213-7165/© 2021 The Authors. Published by Elsevier Ltd on behalf of International Society for Antimicrobial Chemotherapy. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

(ICUs) and contributes to high morbidity and mortality in critically rates, the ICU’s CRAB rates increased from 52.9% in 2006 to 89.8% in
ill patients [2–7]. Carbapenem-resistant Acinetobacter baumannii 2013 according to the Korean Nosocomial Infections Surveillance
(CRAB) has increased worldwide as antibiotic use increases, with System (KONIS) [11]. As CRAB causes not only the problem of high
up to 15% in northern European countries and 50% in southern resistance rate itself but also high mortality, CRAB ranked at the top
European countries including Spain, Italy, and Greece [8,9]. In the for critical-priority bacteria, which require development of new
United States, more than half the A. baumannii in hospital-related antibiotics analysed by the World Health Organization (WHO) [12].
infections were non-susceptible to carbapenem, accounting for The treatment options for CRAB infection are limited with
52% of mortality [10]. In South Korea, a country with high CRAB regards to the severity of disease, and the results of clinical or

Table 1
Overall characteristics of patients with infections caused by CRAB.

Variables Overall CRAB infection (n = 282) Pneumonia (n = 257)

Age, y 67.0  14.9 (59–77) 67.5  14.8 (60–78)
Gender, male 192/282 (68.1) 178/257 (69.3)
Underlying diseases, one or more 247/282 (87.6) 225/257 (87.5)
Hypertension 128/282 (45.4) 115/257 (44.7)
Diabetes mellitus 85/282 (30.1) 77/257 (30.0)
Cerebrovascular vascular diseases 69/282 (24.5) 66/257 (25.7)
Neuromuscular diseases 8/282 (2.8) 8/257 (3.1)
Dementia 15/282 (5.3) 15/257 (5.8)
Congestive heart failure 25/282 (8.9) 23/257 (8.9)
Ischaemic heart diseases 23/282 (8.2) 21/257 (8.2)
Valvular heart diseases 2/282 (0.7) 1/257 (0.4)
Chronic obstructive lung diseases/asthma 38/282 (13.5) 35/257 (13.6)
Chronic liver diseases 30/282 (10.6) 20/257 (7.8)
Chronic renal diseases 35/282 (12.4) 31/257 (12.1)
Haemodialysis/peritoneal dialysis 12/282 (4.3) 12/257 (4.7)
Malignant solid tumour 33/282 (11.7) 31/257 (12.1)
Hematologic malignancy 7/282 (2.5) 6/257 (2.3)
Stem cell transplantation 1/282 (0.4) 1/257 (0.4)
Solid organ transplantation 1/282 (0.4) 0/257 (0)
Taking immunosuppressive agents 3/282 (1.1) 3/257 (1.2)
Taken emergency operation within 1 month 38/282 (13.5) 34/257 (13.2)
Taken elective operation within 1 month 14/282 (5.0) 13/257 (5.1)
Age adjusted Charlson Comorbidity Index score 3.7  2.3 (2–5) 4.0  2.2 (2–5)
APACHE II score 19.3  6.9 (14–24) 19.3  6.8 (14–24)
SAPS II 48.6  15.3 (38–59) 48.6  14.9 (38–60)
Use of medical devices
Mechanical ventilator 215/282 (76.2) 200/257 (77.8)
Central catheter 218/282 (77.3) 197/257 (76.7)
Foley catheter 270/282 (95.7) 246/257 (95.7)
Nasogastric tube 244/282 (86.5) 223/257 (86.8)
Classification of infection
Pneumonia 257/282 (91.1)
With bacteraemia 68/257 (26.5) 68/257 (26.5)
Ventilator-associated pneumonia 180/257 (70.0) 180/257 (70.0)
Urinary tract infection 4/282 (1.4)
Primary bacteraemia 20/282 (7.1)
Susceptibility of CRAB
Susceptible to colistin 281/282 (99.6) 256/257 (99.6)
Susceptible to sulbactam 22/262 (8.4) 21/243 (8.6)
Susceptible to tigecycline 150/168 (53.2) 140/156 (89.7)
Susceptible to minocycline 205/255 (80.4) 191/237 (80.6)
Susceptible to TMP/SMX 26/282 (9.2) 25/257 (9.7)
Susceptible to amikacin 60/128 (46.9) 52/147 (47.3)
Mortality
7-day mortality 57/282 (20.2) 49/257 (19.1)
28-day mortality 108/267 (40.4) 97/246 (39.4)
Clinical response, success or improve
14 days 133/243 (54.7) 121/220 (55.0)
28 days 108/220 (49.1) 97/198 (49.0)
End of treatment 138/275 (50.2) 126/250 (50.4)
Microbiological response, success
14 days 84/162 (51.9) 74/150 (49.3)
28 days 79/112 (70.5) 70/102 (68.6)
End of treatment 87/162 (53.1) 77/151 (51.0)
Duration of the hospital stay, d 33.0  36.3 (11–42) 34.0  37.2 (11–44)
Alive, d 43.3  42.0 (16–58) 44.7  42.9 (17–60)
Expired, d 19.6  21.4 (6–24) 19.6  22.0 (6–24)
Antibiotic adverse reactions
Nephrotoxicity 156/282 (55.3) 144/257 (56.0)
Hepatotoxicity 36/282 (12.8) 33/257 (12.8)

Data are expressed as number of patients/total patients (%) or mean  standard deviation (interquartile range) unless otherwise indicated. CRAB = carbapenem-resistant
Acinetobacter baumannii; HIV = human immunodeficiency virus; AIDS = acquired immune deficiency syndrome; APACHE = acute physiology and chronic health evaluation;
SAPS = simplified acute physiology score; TMP/SMX = trimethoprim/sulfamethoxazole.

430
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

microbiological responses of antibiotics for CRAB vary depending urinary urgency, frequency, dysuria, suprapubic tenderness, and
on the studies. In previous studies, antibiotics used for CRAB costovertebral angle tenderness. Significant samples in culture
infection included sulbactam, colistin, aminoglycoside, tigecycline, were considered to be 103 to 105 colonies/mL by dipstick,
minocycline, rifampin, etc. The studies involved monotherapy, microscopy, and Gram stain. Central-line associated bloodstream
combination therapy with carbapenem or other antibiotics, and infection (CLABSI) was defined as the case where the central line
additional colistin nebuliser therapy [13–25]. Colistin is the most was inserted 48 h before the initial positive blood culture without
commonly used antibiotic for CRAB treatment, as monotherapy or any other cause of infection. All sulbactams were administered in
combination therapy with other antibiotics. One study reported the form of ampicillin/sulbactam. All minocycline used in this
that colistin monotherapy showed higher mortality rate compared study were oral formulations. The clinical response was classified
with sulbactam monotherapy, and other studies showed that as success, improvement, and failure. Success was defined as the
combination therapy of colistin and rifampin reduced ventilator- discontinuation of antibiotics because of clinical response to
associated pneumonia (VAP)-related mortality or higher microbi- antibiotics including defervescence, no need for vasopressor,
ological responses than colistin monotherapy [13–15]. Colistin- reversal of symptoms and signs, and normalization of laboratory
containing combination therapy has shown no significant im- findings. Improvement was defined as the maintenance of
provement in mortality compared with colistin monotherapy in antibiotics with clinical response. Failure was defined as death
other studies [16–19]. Tigecycline has been reported to improve or change to other antibiotics because of clinical deterioration.
the microbiological response in the treatment of CRAB, but had no Microbiological success was defined as the maintenance of no
significant effect on clinical improvement [20,21]. Previous studies growth in the specimen which was initially positive. Antibiotic-
compared two or three regimens, but in practice, various regimens induced nephrotoxicity was evaluated as creatinine according to
have been used for CRAB treatments, and no study has compared the RIFLE criteria [26]. Antibiotic-induced liver toxicity was
regimens together. This study investigated the clinical and evaluated according to Hy’s Law [27].
microbiological responses of various antibiotic regimens for CRAB
treatment in several tertiary hospitals in Korea, where the CRAB 2.3. Statistical analysis
rates are relatively high.
For comparison, Pearson χ2 tests and Fisher’s exact tests were
2. Methods used for categorical variables, and Student’s t test and Mann–
Whitney U tests were used for continuous variables, as appropri-
2.1. Study design, study population, and data collection ate. Cox proportional hazards regression analysis was used to
evaluate the association between the antibiotic regimens and the
A multicentre, retrospective cohort study was conducted to outcomes including mortality, clinical responses, and microbio-
evaluate treatment regimens for infections caused by CRAB logical responses. The types and regimens of antibiotics were
complex. The crude study population was composed of the adult analysed separately to avoid duplicate variables in multivariate
patients in intensive care units (ICU) diagnosed with CRAB analysis. The Kaplan–Meier curve was used to evaluate the
infections between November 2015 and November 2016 at 10 mortality according to the antibiotic regimens. All statistical
large Korean clinical centres. Superimposed infections during the analyses were performed using SPSS Statistics version 20.0 for
treatment of CRAB infections and infections caused by pathogens Windows (IBM Corp., Armonk, NY, USA).
other than CRAB in clinical specimens were excluded from this
study. The demographic data, comorbidity status based on the 3. Results
Charlson comorbidity index, and severity index such as acute
physiology and chronic health evaluation (APACHE) II score, and 3.1. Overall characteristics of patients in this study
simplified acute physiology score (SAPS) were collected. CRAB
infections were classified according to the site of infection and the During the study period, 429 patients were collected from 10
type and regimen of antibiotics used. Antibiotic treatment clinical centres. After 147 patients were excluded who did not
responses were assessed by clinical and microbiological responses maintain antibiotics targeting CRAB for more than 48 h because of
and were evaluated at 14 days and 28 days after treatment, and at referral to other hospitals, death, or being without infection, a total
the end of treatment. Clinical response to antibiotic therapy of 282 patients were finally enrolled in this study. The overall
consisted of mortality, duration of hospital stay, and treatment characteristics of the study population are shown in Table 1. The
success. Antibiotic adverse reactions were checked as serum mean age was 67 y, and 192 patients (68.1%) were male. Two-
creatinine and liver function tests. hundred and forty-seven (87.6%) patients had at least one
underlying disease. The most common comorbidities were
2.2. Definition hypertension (45.4%) and diabetes mellitus (30.1%), and the mean
age-adjusted Charlson Comorbidity Index was 4 (interquartile
CRAB was defined as A. baumannii with minimum inhibitory range [IQR] 0–13). More than 70% of patients had mechanical
concentration (MIC) >8 mg/L for imipenem or meropenem, or as A. ventilators, central catheters, foley catheters, or nasogastric tubes.
baumannii reported as resistance to imipenem or meropenem in Sites of infection caused by CRAB are pneumonia (91.1%), primary
automated systems such as VITEK or Microscan. CRAB pneumonia bacteraemia (7.1%), and UTI (1.4%) in the order of prevalence. Of the
was defined as when CRAB has been identified in respiratory CRAB pneumonia, 75% were VAP. Overall, 282 CRAB strains were
samples (such as sputum, bronchoalveolar lavage [BAL], or susceptible to colistin except one strain, and 205 patients (80.4%)
protected specimen brush [PSB] culture) with newly developed were susceptible to minocycline. The susceptibility rates of
or progressing infiltration, consolidation, or cavity on chest x-ray tigecycline and amikacin against CRAB were 53.2% and 46.9%,
while satisfying two or more of the following clinical symptoms; respectively. The baseline characteristics were not different
cough, purulent sputum, crackles, dyspnoea, hypoxia, or need for according to the pneumonia subgroup and antibiotic regimens
ventilator, and systemic inflammatory response syndrome (SIRS) compared with the overall study population.
score 2 or higher. CRAB urinary tract infection (UTI) included only The all-cause 7-day and 28-day mortality were 20.2% (57/281)
symptomatic UTI, defined as when CRAB was identified in more and 40.4% (108/267), respectively. The overall study population
than 105 colonies/mL in urine culture, along with fever above 38 ℃, showed treatment clinical success or improvement in

431
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

approximately 50% at 14 days, 28 days, and at the end of treatment. in order of frequency (Table 2). Among the types of antibiotics, a
In the concurrent microbiological evaluation, the treatment colistin-containing regimen showed the highest 7-day mortality
success rate was 51.9%, 70.5%, and 53.1% at 14 days, 28 days, and rate (22.8%) whereas a sulbactam-containing regimen showed
at the end of treatment, respectively. The mean lengths of hospital the lowest 7-day mortality rate (13.0%). The 28-day mortality
stay for all patients and survivors of CRAB infection were 33 days was the highest in a tigecycline-containing regimen (50.0%) and
(2–354) and 43 days (3–354), respectively. In the subgroup analysis the lowest in a minocycline-containing regimen (28.3%). Among
of patients with pneumonia, overall characteristics including the regimens of antibiotics, combination of colistin and rifampin
demographic and outcome data showed no significant differences showed the highest 7-day mortality. All patients treated with
among overall patients. sulbactam monotherapy survived on day 7, and combination
therapy of sulbactam and minocycline showed 7-day mortality of
3.2. Mortality rates of CRAB infections by the types and regimens of 7.1%. The 28-day mortality was highest in colistin monotherapy
antibiotics (50.0%) and lowest in the combination of colistin and rifampin
(10.0%). In the univariate analysis, combination therapy of
The antibiotics used in the treatment of CRAB were colistin, colistin and carbapenem decreased 7-day mortality, a sulbac-
carbapenem, sulbactam, minocycline, amikacin, and tigecycline tam-containing regimen decreased both 7-day and 28-day

Table 2
Types and regimens of antibiotics for CRAB infections and the resulting mortality rate.

Types and regimens of Overall CRAB 7-day Univariate analysis Multivariate 28-day Univariate analysis Multivariate analysis
antibiotics infection mortality analysis mortality

HR (95% CI) P HR (95% CI) P HR (95% CI) P HR (95% CI) P value

value value value
Age 0.99 (0.98– 0.952 1.00 (0.99– 0.970
1.02) 1.01)
Sex 0.95 (0.55– 0.839 0.94 (0.63– 0.756
1.63) 1.39)
Any underlying diseases 0.68 (0.32– 0.310 0.91 (0.50– 0.764
1.44) 1.67)
Charlson Comorbidity 1.00 (0.88– 0.966 0.98 (0.88– 0.638
Index 1.14) 1.08)
APACHE II score 1.04 (0.99– 0.070 0.99 (0.95– 0.828 1.04 (1.00– 0.015 1.02 (0.98– 0.306
1.08) 1.05) 1.06) 1.06)
SAPS II 1.02 (1.00– 0.012 1.02 (1.01– 0.002 1.02 (1.01– 0.003 1.02 (1.01– 0.001
1.04) 1.04) 1.03) 1.03)
Colistin-containing 171/282 (60.6) 39/171 1.24 (0.70– 0.463 73/166 1.30 (0.86– 0.210
regimen (22.8) 2.19) (44.0) 1.96)
Colistin monotherapy 58/282 (20.6) 20/58 2.05 (1.18– 0.011 1.66 (0.95– 0.075 28/56 (50.0) 1.53 (0.99– 0.053 1.79 (1.15– 0.011
(34.5) 3.54) 2.90) 2.37) 2.79)
Colistin + carbapenem 41/282 (14.5) 2/41 (4.9) 0.19 (0.05– 0.020 0.18 (0.04– 0.017 16/39 (41.0) 0.83 (0.48– 0.486
0.76) 0.73) 1.41)
Colistin + minocycline 22/282 (7.8) 4/22 (18.2) 1.03 (0.37– 0.957 7/22 (31.8) 0.95 (0.44– 0.903
2.84) 2.06)
Colistin + rifampin 17/282 (6.0) 7/17 (41.2) 1.69 (0.76– 0.196 4/18 (22.2) 1.71 (0.88– 0.112
3.72) 3.29)
Colistin + sulbactam 15/282 (5.3) 3/15 (20.0) 0.92 (0.29– 0.897 6/15 (40.0) 0.76 (0.33– 0.512
2.96) 1.73)
With colistin nebuliser 13/282 (4.6) 2/13 (15.4) 1.01 (0.25– 0.992 4/12 (33.3) 1.88 (0.68– 0.222
4.13) 5.17)
Carbapenem-containing 97/282 (34.4) 16/97 (16.5) 0.68 (0.38– 0.206 36/92 (39.1) 0.90 (0.60– 0.593
regimen 1.23) 1.34)
Carbapenem + sulbactam 18/282 (6.4) 2/18 (11.1) 0.49 (0.12– 0.319 4/18 (22.2) 0.51 (0.19– 0.184
2.00) 1.38)
Carbapenem + rifampin 10/282 (3.5) 1/10 (10.0) 1.51 (0.21– 0.684 1/10 (10.0) 0.44 (0.06– 0.418
10.90) 3.18)
Carbapenem + amikacin 9/282 (3.2) 3/9 (33.3) 3.39 (1.05– 0.041 2.22 (0.60– 0.234 3/8 (37.5) 3.43 (1.06– 0.039 3.04 (0.88– 0.078
10.93) 8.28) 11.04) 10.50)
Sulbactam-containing 69/282 (24.5) 9/69 (13.0) 0.53 (0.26– 0.081 0.52 (0.26– 0.074 21/66 (31.8) 0.53 (0.33– 0.011 0.53 (0.32– 0.009
regimen 1.08) 1.07) 0.86) 0.85)
Sulbactam monotherapy 9/282 (3.2) 0/9 0.05 (0.01– 0.366 2/8 (25.0) 0.44 (0.11– 0.253
35.37) 1.79)
Sulbactam + minocycline 14/282 (5.0) 1/14 (7.1) 0.62 (0.09– 0.629 2/14 (14.3) 0.42 (0.10– 0.231
4.44) 1.73)
Minocycline-containing 46/282 (16.3) 8/46 (17.4) 1.14 (0.54– 0.737 13/46 (28.3) 0.83 (0.46– 0.534
regimen 2.40) 1.49)
Amikacin-containing 22/282 (7.8) 3/22 (13.6) 1.11 (0.35– 0.865 6/18 (33.3) 1.36 (0.59– 0.469
regimen 3.54) 3.11)
Tigecycline-containing 12/282 (4.3) 2/12 (16.7) 0.78 (0.19– 0.730 5/10 (50.0) 0.86 (0.35– 0.737
regimen 3.20) 2.11)
Tigecycline 7/282 (2.5) 1/7 (14.3) 0.05 (0.01– 0.462 2/5 (40.0) 0.56 (0.14– 0.419
monotherapy 154.80) 2.28)

Data are expressed as number of patients/total patients (%) unless otherwise indicated. All sulbactams were administered in the form of ampicillin/sulbactam.
CRAB = carbapenem-resistant Acinetobacter baumannii; HR = hazard ratio; CI = confidence interval; APACHE = acute physiology and chronic health evaluation;
SAPS = simplified acute physiology score.

432
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

mortality, whereas colistin monotherapy and combination In Kaplan–Meier curves, patients treated with a sulbactam-
therapy of carbapenem and amikacin increased both 7-day containing regimen tended to have higher 7-day survival rates
and 28-day mortality. In multivariate analysis, combination without significance, and significantly higher 28-day survival rates
therapy of colistin and carbapenem decreased 7-day mortality (Fig. 1). The combination therapy of colistin and carbapenem had
(Adjusted hazard ratio [aHR] 0.18, 95% confidence interval [CI] significantly higher 7-day and survival rate, whereas 7-day and 28-
0.04–0.73, P value = 0.017), and sulbactam-containing regimen day mortality rates were significantly higher in patients treated
decreased 28-day mortality (aHR 0.53, 95% CI 0.32–0.85, with colistin monotherapy.
P = 0.009), whereas colistin monotherapy increased 28-day In CRAB pneumonia subgroup analysis, combination therapy of
mortality (HR 1.79, 95% CI 1.15–2.79, P = 0.011). Higher SAPS colistin and carbapenem (aHR 0.16, 95% CI 0.04–0.65, P = 0.011) and
increased both 7-day and 28-day mortality. Combination therapy carbapenem-containing regimen (aHR 0.48, 95% CI 0.24–0.94,
with colistin and carbapenem was not associated with decreased P = 0.033) significantly decreased 7-day mortality in multivariate
28-day mortality, and most patients died from CRAB infections analysis (Supplementary Table 1). A sulbactam-containing regi-
with high average APACHE scores and SAPS. men decreased 28-day mortality in CRAB pneumonia (aHR 0.53,

Fig. 1. Kaplan–Meier curves for 7-day and 28-day survival according to types and regimens of antibiotics for CRAB infection.

433
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

Table 3
Types and regimens of antibiotics for CRAB infections and the resulting clinical responses.

Types and regimens of 14 Univariate Multivariate 28 Univariate Multivariate End of Univariate Multivariate
antibiotics days analysis analysis days analysis analysis treatment analysis analysis

OR (95% P value OR P value OR P value OR P value OR P value OR P value

CI) (95% (95% (95% (95% (95%
CI) CI) CI) CI) CI)
Age 0.98 0.012 0.98 0.085 0.97 0.001 0.98 0.015 0.97 <0.001 0.98 0.010
(0.96– (0.97– (0.95– (0.96– (0.95– (0.96–
0.99) 1.00) 0.99) 0.99) 0.99) 0.99)
Sex 1.24 0.420 1.20 0.524 1.23 0.416
(0.73– (0.69– (0.74–
2.11) 2.07) 2.04)
Any underlying diseases 0.86 0.690 0.74 0.474 0.71 0.367
(0.39– (0.33– (0.34–
1.84) 1.67) 1.49)
Charlson Comorbidity 0.90 0.070
0.92 0.260 0.85 0.010 0.90 0.151 0.86 0.009 0.94 0.330
Index (0.80– (0.80– (0.74– (0.77– (0.77– (0.83–
1.01) 1.06) 0.96) 1.04) 0.96) 1.07)
APACHE II score 0.92 <0.001 0.99 0.786 0.92 <0.001 0.98 0.579 0.93 <0.001 1.00 0.871
(0.89– (0.94– (0.88– (0.93– (0.90– (0.94–
0.96) 1.05) 0.96) 1.04) 0.97) 1.05)
SAPS II score 0.95 <0.001 0.95 <0.001 0.95 <0.001 0.95 <0.001 0.96 <0.001 0.95 <0.001
(0.93– (0.94– (0.93– (0.93– (0.94– (0.94–
0.97) 0.97) 0.97) 0.71) 0.97) 0.97)
Colistin-containing 77/ 0.81 0.373 67/ 1.10 0.705 81/167 0.84 0.489
regimen 146 (0.50– 135 (0.67– (48.5) (0.52–
(52.7) 1.30) (49.6) 1.80) 1.37)
Colistin monotherapy 25/49 0.81 0.488 23/46 1.08 0.811 28/56 0.99 0.976
(51.0) (0.46– (50.0) (0.60– (50.0) (0.55–
1.46) 1.94) 1.78)
Colistin + carbapenem 16/36 0.68 0.261 14/33 0.81 0.555 17/41 0.66 0.228
(44.4) (0.35– (42.4) (0.41– (41.5) (0.34–
1.33) 1.63) 1.30)
Colistin + minocycline 12/19 1.38 0.472 11/19 1,68 0.244 12/21 1.35 0.508
(63.2) (0.58– (57.9) (0.70– (57.1) (0.55–
3.30) 4.02) 3.33)
Colistin + rifampin 6/12 0.59 0.317 5/11 0.66 0.440 5/16 0.43 0.128
(50.0) (0.213– (45.5) (0.22– (31.3) (0.15–
1.649) 1.92) 1.27)
Colistin + sulbactam 8/12 1.30 0.624 6/10 1.08 0.889 9/15 0.76 0.437
(66.7) (0.46– (60.0) (0.37– (60.0) (0.33–
3.68) 3.12) 1.73)
With colistin nebuliser 7/11 1.32 0.622 4/9 0.71 0.569 8/12 2.05 0.252
(63.6) (0.43– (44.4) (0.21– (66.7) (0.60–
4.04) 2.35) 6.96)
Carbapenem-containing 39/83 0.65 0.091 1.00 0.994 29/75 0.57 0.037 0.69 0.251 39/96 0.55 0.021 0.66 0.145
regimen (45.0) (0.40– (0.54– (38.7) (0.34– (0.37– (40.6) (0.33– (0.38–
1.07) 1.87) 0.97) 1.30) 0.91) 1.15)
Carbapenem + sulbactam 11/18 1.83 0.226 7/17 1.03 0.957 10/18 1.26 0.638
(61.1) (0.69– (41.2) (0.39– (55.6) (0.48–
4.86) 2.74) 3.29)
Carbapenem + rifampin 4/5 0.74 0.739 3/4 0.68 0.585 3/10 0.41 0.207
(80.0) (0.20– (75.0) (0.17– (30.0) (0.10–
2.68) 2.69) 1.63)
Carbapenem + amikacin 4/7 0.89 0.686 1/4 0.19 0.125 6/9 (66.7) 2.03 0.324
(57.1) (0.24– (25.0) (0.02– (0.50–
3.40) 1.57) 8.29)
Sulbactam-containing 40/64 1.78 0.040 1.69 0.107 28/54 1.14 0.654 38/68 1.36 0.279
regimen (62.5) (1.03– (0.89– (51.9) (0.65– (55.9) (0.78–
3.08) 3.19) 1.98) 2.35)
Sulbactam monotherapy 5/9 1.42 0.610 5/8 2.06 0.289 6/9 (66.7) 2.03 0.324
(55.6) (0.37– (62.5) (0.54– (0.50–
5.39) 7.86) 8.29)
Sulbactam + minocycline 10/13 2.95 0.074 3.12 0.101 6/9 1.22 0.719 8/12 2.05 0.252
(76.9) (0.90– (0.80– (66.7) (0.41– (66.7) (0.60–
9.63) 12.19) 3.62) 6.96)
Minocycline-containing 28/41 1.94 0.044 1.97 0.080 21/36 1.44 0.264 25/43 1.46 0.258
regimen (68.3) (1.02– (0.92– (58.3) (0.76– (58.1) (0.76–
3.70) 4.19) 2.72) 2.83)
Amikacin-containing 11/18 1.13 0.781 8/15 0.91 0.846 16/22 2.86 0.034 4.65 0.005
regimen (61.1) (0.47– (53.3) (0.37– (72.7) (1.09– (1.58–
2.70) 2.26) 7.55) 13.67)
Tigecycline-containing 5/11 0.79 0.697 4/10 0.80 0.718 6/12 0.99 0.990
regimen (45.5) (0.25– (40.0) (0.23– (50.0) (0.31–
2.56) 2.72) 3.16)

434
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

Table 3 (Continued)
Types and regimens of 14 Univariate Multivariate 28 Univariate Multivariate End of Univariate Multivariate
antibiotics days analysis analysis days analysis analysis treatment analysis analysis

OR (95% P value OR P value OR P value OR P value OR P value OR P value

CI) (95% (95% (95% (95% (95%
CI) CI) CI) CI) CI)
Tigecycline 4/6 1.51 0.595 3/5 1.21 0.802 5/7 (71.4) 2.54 0.271
monotherapy (66.7) (0.33– (60.0) (0.27– (0.48–
6.87) 5.53) 13.31)

Data are expressed as number of patients/total patients (%) unless otherwise indicated. All sulbactams were administered in the form of ampicillin/sulbactam.
CRAB = carbapenem-resistant Acinetobacter baumannii; OR = odds ratio; CI = confidence interval; APACHE = acute physiology and chronic health evaluation; SAPS = simplified
acute physiology score.

95% CI 0.32–0.87, P = 0.013). Colistin monotherapy (aHR 2.07, 95% colistin-containing regimen (aOR 4.21, 95% CI 1.88–9.40, P < 0.001)
CI 1.27–3.35, P = 0.003) and combination therapy of colistin and and the combination of colistin and minocycline (aOR 13.88, 95% CI
rifampin (aHR 2.01, 95% CI 1.02–3.97, P = 0.043) were related to 1.76–109.43, P = 0.013) were associated with increased microbio-
increased 28-day mortality. logical responses at 14 days after treatment. The combination of
carbapenem and sulbactam decreased microbiological responses
3.3. Clinical and microbiological responses of CRAB infections by the at 14 days (aOR 0.18, 95% CI 0.04–0.83, P = 0.028) and 28 days after
types and regimens of antibiotics treatment (aOR 0.28, 95% CI 0.08–0.95, P = 0.041).
Another outcome indicator—duration of hospital stay—did not
The clinical responses of antibiotics for CRAB infection varied differ depending on the types and regimens of antibiotics in the
depending on the types and regimens of antibiotics: 50–80% at 14 treatment of CRAB infection.
days, 40–75% at 28 days after treatment, and 30–72% at the end of
treatment (Table 3). A minocycline-containing regimen showed 3.4. Nephrotoxicity and hepatotoxicity by the types and regimens of
the most clinical responses at 14 days (68.3%) and 28 days (58.3%) antibiotics
of treatment, and an amikacin-containing regimen had the highest
clinical response rates (72.7%) at the end of treatment. Among the Nephrotoxicity occurred in 35–100% and hepatotoxicity in 8–
regimens of antibiotics, the combination of carbapenem and 42% of patients depending on the types and regimens of
rifampin showed the most clinical responses at 14 days (80.0%) and antimicrobial therapy (Table 5). In multivariate analysis, a
28 days (75.0%) after treatment. Sulbactam monotherapy (66.7%) colistin-containing regimen increased nephrotoxicity (aOR 2.42,
and a combination of sulbactam and minocycline (66.7%) had the 95% CI 1.44–4.06, P = 0.001) whereas a carbapenem-containing
highest clinical responses at the end of treatment. All types and regimen (aOR 0.55, 95% CI 0.32–0.94, P = 0.028), the combination of
regimens of antibiotics for CRAB infection did not show significant carbapenem and sulbactam (aOR 0.04, 95% CI 0.01–0.30, P = 0.002),
clinical responses except for amikacin-containing regimens. These and the combination of carbapenem and rifampin (aOR 0.14, 95% CI
were associated with more clinical improvement at the end of 0.03–0.68, P = 0.015) were associated with decreased nephrotoxi-
treatment (adjusted odds ratio [aOR] 4.65, 95% CI 1.58–13.67, city. The underlying disease was at risk of nephrotoxicity. In the
P = 0.005). Younger patients showed more clinical improvements pneumonia subgroup, the colistin-containing regimen was associ-
at 14 days and 28 days after treatment, and SAPS were associated ated with increased nephrotoxicity, whereas the carbapenem
with clinical improvement at 14 days and 28 days after treatment, regimen was associated with decreased nephrotoxicity (Supple-
and at the end of treatment. In the pneumonia subgroup, the mentary Table 4). In hepatotoxicity, tigecycline was the only
significant variables related to clinical improvements were the antibiotic associated with the risk of liver toxicity in the overall
same as for the overall study population (Supplementary Table 2). study population as well as the pneumonia subgroup.
A minocycline-containing regimen showed the highest micro-
biological responses at 14 days (80.6%) and 28 days (91.7%) after 4. Discussion
treatment, and at the end of treatment (92.0%) (Table 4). All
patients treated with sulbactam monotherapy had microbiological To our knowledge, this is the first study to compare the efficacy
responses at 14 days and 28 days after treatment, and at the end of of three or more types and regimens of antibiotics simultaneously
treatment. In multivariate analysis, colistin-containing therapy for CRAB infections through a multicentre study in a high endemic
(aOR 2.88, 95% CI 1.40–5.90. P = 0.004) and minocycline-containing area. The combination therapy of colistin and carbapenem
regimen (aOR 6.88, 95% CI 2.49–18.97, P < 0.001) were associated decreased 7-day mortality and a sulbactam-containing regimen
with increased microbiological responses at 14 days after decreased 28-day mortality, whereas colistin monotherapy
treatment. The combination of carbapenem and sulbactam were increased 7-day and 28-day mortality. An amikacin-containing
associated with decreased microbiological responses (aOR 0.25, regimen was associated with increased clinical responses at the
95% CI 0.07–0.86, P = 0.028), whereas the minocycline-containing end of treatment. A minocycline-containing regimen was associ-
regimen was associated with increased microbiological responses ated with increased microbiological responses at 14 days, 28 days,
(aOR 6.46, 95% CI 1.40–29.83, P = 0.017) at 28 days after treatment. and at the end of treatment.
A minocycline-containing regimen was also associated with Patients in this study were of relatively higher age (mean 67 y),
increased microbiological response (aOR 11.08, 95% CI 2.37– and had higher rates of bacteraemia (31%) than in previous studies,
51.72, P = 0.002) at the end of treatment. In the pneumonia and had one or more underlying diseases. This study may also
subgroup, a minocycline-containing regimen was also associated provide useful indications for CRAB pneumonia given that most
with increased microbiological responses at 14 days (aOR 7.98, 95% CRAB infections involve pneumonia and that 70% of pneumonia
CI 2.77–22.84, P < 0.001) and 28 days after treatment (aOR 6.09, cases are VAP. All CRAB strains except one were susceptible to
95% CI 1.30–28.58, P = 0.022) and at the end of treatment (aOR colistin and the susceptibility of minocycline remained relatively
11.35, 95% CI 2.41–53.33, P = 0.002) (Supplementary Table 3). A high at 80%. The susceptibility rates of ampicillin/sulbactam and

435
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

Table 4
Types and regimens of antibiotics for CRAB infections and the resulting microbiological responses.

Types and regimens of 14- Univariate Multivariate 28- Univariate Multivariate End of Univariate Multivariate
antibiotics day analysis analysis day analysis analysis treatment analysis analysis

OR (95% P OR (95% P value OR (95% P OR (95% P OR (95% P OR (95% P value

CI) value CI) CI) value CI) value CI) value CI)
Age 0.98 0.042 0.97 0.017 0.98 0.103 0.98 0.053 0.99 0.431
(0.96– (0.95– (0.95– (0.96– (0.96–
0.99) 0.99) 1.00) 1.00) 1.02)
Sex 1.49 0.245 1.58 0.290 1.34 0.380
(0.76– (0.68– (0.69–
2.89) 3.70) 2.61)
Any underlying diseases 0.57 0.238 0.33 0.158 0.53 0.226
(0.23– (0.07– (0.19–
1.45) 1.54) 1.49)
Charlson Comorbidity 0.88 0.102 0.83 0.054 0.82 0.042 0.79 0.003 0.78 0.005
Index (0.76– (0.69– (0.68– (0.67– (0.66–
1.03) 1.01) 0.99) 0.92) 0.93)
APACHE II score 0.96 0.088 0.99 0.735 0.96 0.289 0.96 0.162
(0.91– (0.94– (0.90– (0.91–
1.01) 1.05) 1.03) 1.02)
SAPS II score 0.99 0.590 1.01 0.608 1.00 0.888
(0.97– (0.98– (0.98–
1.02) 1.04) 1.03)
Colistin-containing 61/ 2.39 0.009 2.88 0.004 51/67 1.94 0.116 0.73 0.500 56/103 1.15 0.666
regimen 102 (1.25– (1.40– (76.1) (0.85– (0.29– (54.4) (0.61–
(59.8) 4.60) 5.90) 4.41) 1.84) 2.19)
Colistin monotherapy 13/30 0.66 0.303 10/17 0.54 0.255 13/31 0.57 0.169
(43.3) (0.30– (58.8) (0.19– (41.9) (0.26–
1.46) 1.56) 1.27)
Colistin + carbapenem 15/27 1.20 0.673 12/15 1.79 0.393 15/28 1.02 0.955
(55.6) (0.52– (80.0) (0.47– (53.6) (0.45–
2.74) 6.82) 2.32)
Colistin + minocycline 14/15 15.40 0.009 13/13 0.999 13/13 0.998
(93.3) (1.97– (100) (100)
120.16)
Colistin + rifampin 5/8 1.58 0.540 5/7 1.05 0.957 4/9 (44.4) 0.69 0.595
(62.5) (0.36– (71.4) (0.19– (0.18–
6.85) 5.69) 2.68)
Colistin + sulbactam 5/10 0.92 0.904 4/7 0.53 0.428 5/10 0.88 0.840
(50.0) (0.26– (57.1) (0.11– (50.0) (0.24–
3.23) 2.53) 3.15)
With colistin nebuliser 5/7 2.41 0.303 4/4 0.999 4/6 (66.7) 1.81 0.503
(71.4) (0.45– (100) (0.32–
12.77) 10.14)
Carbapenem-containing 19/54 0.36 0.003 0.70 0.370 19/33 0.43 0.055 0.73 0.500 20/58 0.30 0.001 0.49 0.051
regimen (35.2) (0.18– (0.32– (57.6)(0.18– (0.29– (34.5) (0.16– (0.24–
0.71) 1.53) 1.02) 1.84) 0.59) 1.01)
Carbapenem + sulbactam 2/14 0.13 0.010 5/12 0.25 0.028 0.25 0.028 4/14 0.32 0.065 0.42 0.172
(14.3) (0.03– (41.7) (0.07– (0.07– (28.6) (0.10– (0.12–
0.62) 0.86) 0.86) 1.07) 1.46)
Carbapenem + rifampin 0/4 0.999 1/2 0.41 0.533 0/6 0.999
(50.0) (0.03–
6.76)
Carbapenem + amikacin 1/3 0.46 0.527 1/2 0.41 0.533 1/4 (25.0) 0.29 0.283
(33.3) (0.04– (50.0) (0.025– (0.03–
5.15) 6.761) 2.81)
Sulbactam-containing 22/48 0.71 0.321 22/36 0.52 0.135 23/43 1.02 0.951
regimen (45.8) (0.36– (61.1) (0.22– (53.5) (0.51–
1.40) 1.22) 2.06)
Sulbactam monotherapy 6/6 0.999 6/6 0.999 5/5 (100) 0.999
(100) (100)
Sulbactam + minocycline 5/10 0.92 0.904 4/6 0.83 0.831 5/7 (71.4) 2.28 0.332
(50.0) (0.26– (66.7) (0.14– (0.43–
3.32) 4.75) 12.13)
Minocycline-containing 25/31 5.09 0.001 6.88 <0.001 22/24 5.98 0.020 6.46 0.017 23/25 13.51 0.001 11.08 0.002
regimen (80.6) (1.96– (2.49– (91.7) (1.32– (1.40– (92.0) (3.06– (2.37–
13.22) 18.97) 27.14) 29.83) 59.55) 51.72)
Amikacin-containing 4/12 0.44 0.192 5/9 0.49 0.312 7/13 1.03 0.954
regimen (33.3) (0.13– (55.6) (0.12– (53.8) (0.33–
1.52) 1.95) 3.22)
Tigecycline-containing 2/6 0.45 0.366 3/5 0.61 0.600 5/7 (71.4) 2.28 0.332
regimen (33.3) (0.08– (60.0) (0.10– (0.43–
2.54) 3.84) 12.13)

436
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

Table 4 (Continued)
Types and regimens of 14- Univariate Multivariate 28- Univariate Multivariate End of Univariate Multivariate
antibiotics day analysis analysis day analysis analysis treatment analysis analysis

OR (95% P OR (95% P value OR (95% P OR (95% P OR (95% P OR (95% P value

CI) value CI) CI) value CI) value CI) value CI)
Tigecycline 1/3 0.46 0.527 2/3 0.83 0.882 4/4 0.999
monotherapy (33.3) (0.04– (66.7) (0.07– (100.0)
5.15) 9.49)

Data are expressed as number of patients/total patients (%) unless otherwise indicated. All sulbactams were administered in the form of ampicillin/sulbactam.
CRAB = carbapenem-resistant Acinetobacter baumannii; OR = odds ratio; CI = confidence interval; APACHE = acute physiology and chronic health evaluation; SAPS = simplified
acute physiology score.

Table 5
Nephrotoxicity and hepatotoxicity by the types and regimens of antibiotics for CRAB infections.

Types and regimens of Nephrotoxicity Univariate analysis Multivariate analysis Hepatotoxicity Univariate analysis Multivariate analysis
antibiotics
OR (95% CI) P value OR (95% CI) P OR (95% CI) P OR (95% CI) P value
value value
Age 0.99 (0.98– 0.825 1.01 (0.98– 0.552
1.01) 1.03)
Sex 1.67 (1.01– 0.046 1.64 (0.95– 0.074 0.81 (0.39– 0.564
2.77) 2.81) 1.68)
Any underlying diseases 0.32 (0.14– 0.008 0.38 (0.16– 0.029 0.86 (0.31– 0.774
0.74) 0.90) 2.39)
Charlson Comorbidity Index 0.98 (0.88– 0.708 0.97 (0.83– 0.723
1.09) 1.14)
APACHE II score 0.95 (0.92– 0.006 0.96 (0.93– 0.055 0.98 (0.93– 0.350
0.99) 1.01) 1.028)
SAPS II score 0.99 (0.97– 0.106 0.98 (0.95– 0.078 0.98 (0.95– 0.064
1.01) 1.01) 1.01)
Colistin-containing regimen 60/171 (35.1) 2.71 (1.66– <0.001 2.42 (1.44– 0.001 23/171 (13.5) 1.17 (0.57– 0.669
4.44) 4.06) 2.42)
Colistin monotherapy 26/58 (44.8) 0.99 (0.56– 0.980 7/58 (12.1) 0.92 (0.38– 0.858
1.77) 2.23)
Colistin + carbapenem 17/41 (41.5) 1.17 (0.60– 0.654 5/41 (12.2) 0.94 (0.34– 0.906
2.28) 2.58)
Colistin + minocycline 0/22 0.998 3/22 (13.6) 1.09 (0.31– 0.899
3.87)
Colistin + rifampin 4/17 (23.5) 2.77 (0.88– 0.081 1.99 (0.61– 0.255 3/17 (17.6) 0.75 (0.09– 0.790
8.73) 6.56) 6.12)
Colistin + sulbactam 5/15 (33.3) 1.66 (0.55– 0.368 1/15 (6.7) 0.47 (0.06– 0.477
4.98) 3.71)
With colistin nebuliser 3/13 (23.1) 2.81 (0.76– 0.123 2/13 (15.4) 1.26 (0.27– 0.773
10.43) 5.92)
Carbapenem-containing 57/97 (58.8) 0.42 (0.25– 0.001 0.55 (0.32– 0.028 13/97 (13.4) 1.09 (0.53– 0.817
regimen 0.69) 0.94) 2.26)
Carbapenem + sulbactam 17/18 (94.4) 0.04 (0.01– 0.002 0.04 (0.01– 0.002 1/18 (5.6) 0.39 (0.05– 0.361
0.32) 0.30) 2.98)
Carbapenem + rifampin 8/10 (80.0) 0.19 (0.04– 0.039 0.14 (0.03– 0.015 1/10 (10.0) 0.75 (0.09– 0.790
0.92) 0.68) 6.12)
Carbapenem + amikacin 7/9 (77.8) 0.22 (0.05– 0.063 0.24 (0.05– 0.082 0/9 0.999
1.08) 1.20)
Sulbactam-containing 35/69 (50.7) 0.73 (0.42– 0.246 6/69 (8.7) 0.58 (0.23– 0.248
regimen 1.25) 1.46)
Sulbactam monotherapy 5/9 (55.6) 0.64 (0.17– 0.508 1/9 (11.1) 0.85 (0.10– 0.880
2.42) 7.00)
Sulbactam + minocycline 0/14 0.998 1/14 (7.1) 0.51 (0.07– 0.525
4.04)
Minocycline-containing 0/46 0.997 7/46 (15.2) 1.28 (0.52– 0.587
regimen 3.13)
Amikacin-containing regimen 16/22 (72.7) 0.28 (0.10– 0.009 0.53 (0.18– 0.247 0/22 0.998
0.73) 1.56)
Tigecycline-containing 8/12 (66.7) 0.39 (0.11– 0.130 4/12 (33.3) 3.72 (1.06– 0.040 4.59 (1.25– 0.021
regimen 1.32) 13.05) 16.82)
Tigecycline monotherapy 7/7 (100) 0.999 3/7 (42.9) 5.50 (1.18– 0.030 7.00 (1.35– 0.021
25.67) 36.34)

Data are expressed as number of patients/total patients (%) unless otherwise indicated. All sulbactams were administered in the form of ampicillin/sulbactam.
CRAB = carbapenem-resistant Acinetobacter baumannii; OR = odds ratio; CI = confidence interval; APACHE = acute physiology and chronic health evaluation; SAPS = simplified
acute physiology score.

437
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

trimethoprim/sulfamethoxazole were <10%. The most frequently combination of colistin and carbapenem, sulbactam-containing
used regimen in the treatment of CRAB infection was a colistin- regimen, and minocycline-containing regimen. Second, the opti-
containing regimen. Among colistin-containing regimens, colistin mal dose of sulbactam was not administered and an oral
monotherapy was the most commonly used, and carbapenems formulation of minocycline was used in this study. Nevertheless,
were the most frequently used antibiotics in combination therapy sulbactam and minocycline contributed to survival improvement
with colistin. and microbiological responses, respectively, which suggests the
The combination of colistin and carbapenem, which showed 7- need for further microbiological or clinical studies.
day survival improvement, was consistent with results which may Based on these results, a combination of colistin and
contribute to the reduction in mortality in previous studies [28]. This carbapenem, and a sulbactam-containing regimen may improve
combination has been proven to have a high synergistic effect against survival rate in treatment of CRAB infections. A minocycline-
CRAB in an in vitro study [29]. Although 28-day survival improvement containing regimen showed consistent significant improvement in
was not shown with the combination of colistin and carbapenem, the microbiological responses, and further studies may be needed to
potential result from higher APACHE score and SAPS in the death evaluate improved mortality. Colistin monotherapy should be
group should be considered in the interpretation. Considering the considered cautiously for severe CRAB infections given its
increased mortality, colistin monotherapy should be used cautiously significant association with increased 7-day and 28-day mortality.
in treatment for severe CRAB infection or pneumonia.
A sulbactam-containing regimen contributed significantly to Author contributions
the increased survival rate, which is notable for high survival
despite the low susceptibility of sulbactam. In sulbactam mono- WSC, HJC, HYK, YRK, HP, and JWS conceived the idea. WSC, SL,
therapy, all patients survived at day 7, despite a 22.2% resistance CM, DWP, JYS, JK, MNP, HJL, BK, YMJ, and JHK collected the data. JS,
rate. Most sulbactam-resistant strains were treated with a WSC, SL, CM, HJC, JYK, YRK, JP, and JWS analysed the data. HS, CWC,
combination of sulbactam and minocycline, which showed and JWS prepared the manuscript and final edition of the
relatively low mortality and good clinical and microbiological document. All authors read and approved the final manuscript.
responses without statistical significance. Research on the
combination of sulbactam and minocycline is limited, and Funding
additional basic and clinical studies may be needed based on this
result. In addition, all sulbactams used in this study were This work was supported by a grant of the Korea Health
ampicillin/sulbactam combinations, suggesting that the develop- Technology R&D Project through the Korea Health Industry
ment of a single sulbactam formulation may be required for use as Development Institute (KHIDI), funded by the Ministry of Health
high-dose sulbactam. & Welfare, Republic of Korea (grant number : HI19C0201).
Minocycline had the best microbiological responses at any time
and tended to have low mortality rate without statistical Conflict of interest
significance. Minocycline had bactericidal activity against CRAB
and synergistic effects with other antibiotics [25]. Previous studies All authors certify that there are no potential conflicts of
showed good clinical and microbiological responses in treatment interest to declare.
of CRAB infections when used intravenously alone or in combina-
tion, but results on improving survival are limited [25,30,31]. A Ethics approval
decrease in mortality without statistical significance may be
related to the inferiority of the oral formulation despite high oral The study protocol was approved by the institutional review
bioavailability of minocycline. Regarding the 80% susceptibility board of Korea University Anam Hospital (no. 2015AN0325). This
rate of minocycline, additional study may be needed to evaluate retrospective study was waived the need for informed consent.
the efficacy of intravenous minocycline. Tigecycline, an analogue of
tetracycline such as minocycline, showed no significant results in
Appendix A. Supplementary data
this study. It is presumed that most of the infections in this study
were pneumonia, and tigecycline levels in epithelial lining fluid
Supplementary material related to this article can be found, in the
and alveolar cells are relatively low [32].
online version, at doi:https://doi.org/10.1016/j.jgar.2021.01.018.
An amikacin-containing regimen showed good clinical re-
sponse at the end of treatment, but this result alone is insufficient
to judge the efficacy of amikacin for CRAB infections because of References
limited prior studies and low lung tissue penetration [33]. Further
studies on amikacin may be needed in that the possible [1] Antimicrobial Resistance: Global Report on Surveillance. World Health
Organization; 2014.
confounding variables such as demographic data, Charlson [2] Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, et al. Bad
comorbidity index, APACHE score, SAPS, mortality, and antibiotic bugs, no drugs: no ESKAPE! an update from the infectious diseases society of
duration were not significantly different between amikacin and America. Clin Infect Dis 2009;48:1–12, doi:http://dx.doi.org/10.1086/595011.
[3] Munoz-Price LS, Weinstein RA. Acinetobacter infection. N Engl J Med
non-amikacin groups. 2008;358:1271–81.
Colistin was significantly associated with the development of [4] Falagas ME, Kopterides P, Siempos II. Attributable mortality of Acinetobacter
any stage of nephrotoxicity and tigecycline was significantly baumannii infection among critically ill patients. Clin Infect Dis 2006;43:389.
[5] Jones CL, Clancy M, Honnold C, Singh S, Snesrud E, Onmus-Leone F, et al. Fatal
related to development of hepatotoxicity, as shown in previous outbreak of an emerging clone of extensively drug-resistant Acinetobacter
studies [34,35]. When colistin or tigecycline is used for treatment baumannii with enhanced virulence. Clin Infect Dis 2015;61:145–54.
of CRAB infections, careful monitoring is needed for side effects. [6] Cassini A, Högberg LD, Plachouras D, Quattrocchi A, Hoxha A, Simonsen GS,
et al. Attributable deaths and disability-adjusted life-years caused by
This study has several limitations. First, this study is retrospec- infections with antibiotic-resistant bacteria in the EU and the European
tive and non-randomized, so some biases may occur. The types, Economic Area in 2015: a population-level modelling analysis. Lancet Infect
number, and duration of antibiotics vary from group to group, and Dis 2019;19:56–66, doi:http://dx.doi.org/10.1016/s1473–3099(18)30605–4.
[7] Kwon KT, Oh WS, Song JH, Chang HH, Jung SI, Kim SW, et al. Impact of
the choices of antibiotics were determined by each clinician.
imipenem resistance on mortality in patients with Acinetobacter bacteraemia. J
Additional prospective studies may be needed for the antibiotics Antimicrob Chemother 2007;59:525–30, doi:http://dx.doi.org/10.1093/jac/
which were statistically significant in this study, such as the dkl499.

438
H. Seok, W.S. Choi, S. Lee et al. Journal of Global Antimicrobial Resistance 24 (2021) 429–439

[8] Towner KJ. Acinetobacter: an old friend, but a new enemy. J Hosp Infect infection. Yonsei Med J 2012;53:974–84, doi:http://dx.doi.org/10.3349/
2009;73:355–63, doi:http://dx.doi.org/10.1016/j.jhin.2009.03.032. ymj.2012.53.5.974.
[9] Surveillance of Antimicrobial Resistance in Europe 2018. European Centre for [22] Rattanaumpawan P, Lorsutthitham J, Ungprasert P, Angkasekwinai N,
Disease Prevention and Control; 2019. Thamlikitkul V. Randomized controlled trial of nebulized colistimethate
[10] Bulens SN, Yi SH, Walters MS, Jacob JT, Bower C, Reno J, et al. Carbapenem- sodium as adjunctive therapy of ventilator-associated pneumonia caused by
nonsusceptible Acinetobacter baumannii, 8 US metropolitan areas, 2012–2015. Gram-negative bacteria. J Antimicrob Chemother 2010;65:2645–9, doi:http://
Emerg Infect Dis 2018;24:727–34, doi:http://dx.doi.org/10.3201/ dx.doi.org/10.1093/jac/dkq360.
eid2404.171461. [23] Pachón-Ibáñez ME, Docobo-Pérez F, Jiménez-Mejias ME, Ibáñez-Martínez J,
[11] Choi JY, Kwak YG, Yoo H, Lee SO, Kim HB, Han SH, et al. Trends in the García-Curiel A, Pichardo C, et al. Efficacy of rifampin, in monotherapy and in
distribution and antimicrobial susceptibility of causative pathogens of device- combinations, in an experimental murine pneumonia model caused by
associated infection in Korean intensive care units from 2006 to 2013: results panresistant Acinetobacter baumannii strains. Eur J Clin Microbiol Infect Dis
from the Korean Nosocomial Infections Surveillance System (KONIS). J Hosp 2011;30:895–901, doi:http://dx.doi.org/10.1007/s10096–011–1173–6.
Infect 2016;92:363–71, doi:http://dx.doi.org/10.1016/j.jhin.2015.12.012. [24] Lee HJ, Bergen PJ, Bulitta JB, Tsuji B, Forrest A, Nation RL, et al. Synergistic
[12] Tacconelli E, Carrara E, Savoldi A, Harbarth S, Mendelson M, Monnet DL, et al. activity of colistin and rifampin combination against multidrug-resistant
Discovery, research, and development of new antibiotics: the WHO priority list Acinetobacter baumannii in an in vitro pharmacokinetic/pharmacodynamic
of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis model. Antimicrob Agents Chemother 2013;57:3738–45, doi:http://dx.doi.
2018;18:318–27, doi:http://dx.doi.org/10.1016/s1473–3099(17)30753–3. org/10.1128/aac.00703–13.
[13] Oliveira MS, Prado GV, Costa SF, Grinbaum RS, Levin AS. Ampicillin/sulbactam [25] Ritchie DJ, Garavaglia-Wilson A. A review of intravenous minocycline for
compared with polymyxins for the treatment of infections caused by treatment of multidrug-resistant Acinetobacter infections. Clin Infect Dis
carbapenem-resistant Acinetobacter spp. J Antimicrob Chemother 2014;59 Suppl 6:S374–380, doi:http://dx.doi.org/10.1093/cid/ciu613.
2008;61:1369–75, doi:http://dx.doi.org/10.1093/jac/dkn128. [26] Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P. Acute renal failure-
[14] Aydemir H, Akduman D, Piskin N, Comert F, Horuz E, Terzi A, et al. Colistin vs. definition, outcome measures, animal models, fluid therapy and information
The combination of colistin and rifampicin for the treatment of carbapenem- technology needs: the Second International Consensus Conference of the
resistant Acinetobacter baumannii ventilator-associated pneumonia. Epide- Acute Dialysis quality Initiative (ADQI) Group. Crit Care 2004;8:R204–212, doi:
miol Infect 2013;141:1214–22, doi:http://dx.doi.org/10.1017/ http://dx.doi.org/10.1186/cc2872.
s095026881200194x. [27] Temple R. Hy’s law: predicting serious hepatotoxicity. Pharmacoepidemiol
[15] Durante-Mangoni E, Signoriello G, Andini R, Mattei A, De Cristoforo M, Murino Drug Saf 2006;15:241–3, doi:http://dx.doi.org/10.1002/pds.1211.
P, et al. Colistin and rifampicin compared with colistin alone for the treatment [28] Poulikakos P, Tansarli GS, Falagas ME. Combination antibiotic treatment versus
of serious infections due to extensively drug-resistant Acinetobacter bau- monotherapy for multidrug-resistant, extensively drug-resistant, and pandrug-
mannii: a multicenter, randomized clinical trial. Clin Infect Dis 2013;57:349– resistant Acinetobacter infections: a systematic review. Eur J Clin Microbiol Infect
58, doi:http://dx.doi.org/10.1093/cid/cit253. Dis 2014;33:1675–85, doi:http://dx.doi.org/10.1007/s10096–014–2124–9.
[16] Florescu DF, Qiu F, McCartan MA, Mindru C, Fey PD, Kalil AC. What is the [29] Li J, Fu Y, Zhang J, Zhao Y, Fan X, Yu L, et al. The efficacy of colistin monotherapy
efficacy and safety of colistin for the treatment of ventilator-associated versus combination therapy with other antimicrobials against carbapenem-
pneumonia? A systematic review and meta-regression. Clin Infect Dis resistant Acinetobacter baumannii ST2 isolates. J Chemother 2020;32:359–67,
2012;54:670–80, doi:http://dx.doi.org/10.1093/cid/cir934. doi:http://dx.doi.org/10.1080/1120009x.2020.1764282.
[17] Batirel A, Balkan II, Karabay O, Agalar C, Akalin S, Alici O, et al. Comparison of [30] Greig SL, Scott LJ. Intravenous minocycline: a review in acinetobacter
colistin-carbapenem, colistin-sulbactam, and colistin plus other antibacterial infections. Drugs 2016;76:1467–76, doi:http://dx.doi.org/10.1007/s40265–
agents for the treatment of extremely drug-resistant Acinetobacter baumannii 016–0636–6.
bloodstream infections. Eur J Clin Microbiol Infect Dis 2014;33:1311–22, doi: [31] Lashinsky JN, Henig O, Pogue JM, Kaye KS. Minocycline for the treatment of
http://dx.doi.org/10.1007/s10096–014–2070–6. multidrug and extensively drug-resistant A. baumannii: a review. Infect Dis
[18] Kalin G, Alp E, Akin A, Coskun R, Doganay M. Comparison of colistin and Ther 2017;6:199–211, doi:http://dx.doi.org/10.1007/s40121–017–0153–2.
colistin/sulbactam for the treatment of multidrug resistant Acinetobacter [32] Gotfried MH, Horn K, Garrity-Ryan L, Villano S, Tzanis E, Chitra S, et al.
baumannii ventilator-associated pneumonia. Infection 2014;42:37–42, doi: Comparison of omadacycline and tigecycline pharmacokinetics in the plasma,
http://dx.doi.org/10.1007/s15010–013–0495-y. epithelial lining fluid, and alveolar cells of healthy adult subjects. Antimicrob
[19] Khawcharoenporn T, Pruetpongpun N, Tiamsak P, Rutchanawech S, Mundy LM, Agents Chemother 2017;61:, doi:http://dx.doi.org/10.1128/aac.01135–17.
Apisarnthanarak A. Colistin-based treatment for extensively drug-resistant [33] Fishbain J, Peleg AY. Treatment of Acinetobacter infections. Clin Infect Dis
Acinetobacter baumannii pneumonia. Int J Antimicrob Agents 2014;43:378–82, 2010;51:79–84, doi:http://dx.doi.org/10.1086/653120.
doi:http://dx.doi.org/10.1016/j.ijantimicag.2014.01.016. [34] Ordooei Javan A, Shokouhi S, Sahraei Z. A review on colistin nephrotoxicity. Eur
[20] Gordon NC, Wareham DW. A review of clinical and microbiological outcomes J Clin Pharmacol 2015;71:801–10, doi:http://dx.doi.org/10.1007/s00228–015–
following treatment of infections involving multidrug-resistant Acinetobacter 1865–4.
baumannii with tigecycline. J Antimicrob Chemother 2009;63:775–80, doi: [35] Wargo KA, Edwards JD. Aminoglycoside-induced nephrotoxicity. J Pharm Pract
http://dx.doi.org/10.1093/jac/dkn555. 2014;27:573–7, doi:http://dx.doi.org/10.1177/0897190014546836.
[21] Shin JA, Chang YS, Kim HJ, Kim SK, Chang J, Ahn CM, et al. Clinical outcomes of
tigecycline in the treatment of multidrug-resistant Acinetobacter baumannii

439