Journal of Microbiology, Immunology and Infection 55 (2022) 107e113

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.e-jmii.com

Original Article

Emergence of concurrent levofloxacin- and

trimethoprim/sulfamethoxazole-resistant
Stenotrophomonas maltophilia: Risk factors
and antimicrobial sensitivity pattern analysis
from a single medical center in Taiwan
Rui-Xin Wu a, Ching-Mei Yu b, Sung-Teng Hsu c,
Ching Hsun Wang d,*

a
Division of Infectious Diseases and Tropical Medicine, National Defense Medical Center, Tri-Service
General Hospital Penghu Branch, Penghu, Taiwan
b
Department of Clinical Pathology, National Defense Medical Center, Tri-Service General Hospital, Taiwan
c
Infection Control Office, Tri-Service General Hospital, National Defense Medical Center, Taipei,
Taiwan
d
Division of Infectious Diseases and Tropical Medicine, Department of Internal Medicine, Tri-Service
General Hospital, National Defense Medical Center, Taipei, Taiwan

Received 27 September 2020; received in revised form 17 November 2020; accepted 15 December 2020
Available online 13 January 2021

KEYWORDS Abstract Background: The emergence of concurrent levofloxacin- and trimethoprim/sulfa-

Risk; methoxazole (TMP/SMX)-resistant Stenotrophomonas maltophilia (LTSRSM) in Taiwan is
Resistance; becoming a serious problem, but clinical data analysis on this has not been reported.
Stenotrophomonas Methods: A matched case-control-control study was conducted to investigate risk factors for
maltophilia, LTSRSM occurrence in hospitalized patients. For patients with LTSRSM infection/colonization
levofloxacin; (the case group), two matched control groups were used: control group A with levofloxacin-
Trimethoprim/ and TMP/SMX-susceptible S. maltophilia (LTSSSM) and control group B without S. maltophilia.
sulfamethoxazole Besides, tigecycline, ceftazidime, cefepime, ciprofloxacin, gentamicin, amikacin, and colistin
susceptibilities in collected LTSRSM and levofloxacin- and TMP/SMX-susceptible S. maltophilia
(LTSSSM) isolates were compared.
Results: From January 2014 to June 2016, 129 LTSRSM from cultured 1213 S. maltophilia iso-
lates (10.6%) were identified. A total of 107 LTSRSM infected patients paired with 107
LTSSSM-, and 107 non-S. maltophilia-infected ones were included. When compared with con-
trol group A, previous fluoroquinolone and TMP/SMX use was found to be independently

* Corresponding author. Division of Infectious Diseases and Tropical Medicine, Department of Internal Medicine, Tri-Service General
Hospital, National Defense Medical Center, No. 325, Section 2, Cheng-Kung Road, Neihu, 114, Taipei, Taiwan. Fax: þ886 2 87927258.
E-mail address: sasak0308@gmail.com (C.H. Wang).

https://doi.org/10.1016/j.jmii.2020.12.012
1684-1182/Copyright ª 2021, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 R.-X. Wu, C.-M. Yu, S.-T. Hsu et al.

associated with LTSRSM occurrence. When compared with control group B, mechanical venti-
lation, cerebrovascular disease, and previous fluoroquinolone use were risk factors for LTSRSM
occurrence. Eighty-five LTSRSM and 85 LTSSSM isolates were compared for antibiotic suscepti-
bilities; the resistance rates and minimum inhibitory concentrations of tigecycline and cefta-
zidime were significantly higher for LTSRSM than for LTSSSM isolates.
Conclusion: The emergence of LTSRSM showing cross resistance to tigecycline and ceftazidime
would further limit current therapeutic options. Cautious fluoroquinolone and TMP/SMX use
may be helpful to limit such high-level resistant strains of S. maltophilia occurrence.
Copyright ª 2021, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).

Introduction antibiotic susceptibilities of such resistant strains revealed

that TMP/SMX-resistant S. maltophilia isolates often
Stenotrophomonas maltophilia, a non-fermenting gram- exhibited concurrent levofloxacin resistance and vice
negative bacterium, is an important nosocomial pathogen versa.16e19 This feature of S. maltophilia, where mutation-
in hospital settings.1 In recent years, several worldwide driven resistance to one antibiotic results in cross-
surveillance studies have detected increasing infection resistance to others, has also been observed in experi-
rates of this bacterium in an expanding population of pa- mental evolution studies under laboratory settings.20 The
tients who were immunocompromised due to advances in emergence of such strains showing both levofloxacin and
medical technologies and treatments.2 Although of low TMP/SMX resistance would further limit therapeutic options
virulence, S. maltophilia can cause a wide range of in- in treating S. maltophilia infections, and this may become a
fections including the respiratory system, bloodstream, public health concern in the future. This alarming phe-
skin and soft tissue, bone and joints, biliary tract, and nomenon underscores an urgent need for effective control
urinary tract in immunocompromised patients.3 Treatment and prevention measures that could combat further
of these infections is difficult because S. maltophilia ex- dissemination. We therefore initiated a matched case-
hibits extensive resistance to a variety of antibiotics.2 control-control study to identify risk factors for concur-
Different resistance mechanisms, including inducible b- rent levofloxacin- and TMP/SMX-resistant S. maltophilia
lactamases L1 and L2, aminoglycoside-modifying enzymes, (LTSRSM) infections among hospitalized patients. Moreover,
and overexpression of multidrug efflux pumps, render S. the antibiotic susceptibilities of resistant and susceptible
maltophilia resistant to multiple structurally unrelated strains were analysed and compared.
antibiotics.2 As a consequence, the World Health Organi-
zation has listed S. maltophilia as one of the multidrug-
resistant bacteria in hospital settings.4 Trimethoprim/sul- Methods
famethoxazole (TMP/SMX) is considered the drug of choice
for treating susceptible S. maltophilia infections and has Study setting, design, and patient identification
been widely used for many years based upon reported
in vitro activity and favourable clinical outcomes.5,6 As This study was conducted at the Tri-Service General Hos-
alternatives to TMP/SMX, fluoroquinolones may be consid- pital, which is a medical center with 1800 beds located in
ered for patients infected with TMP/SMX-resistant S. mal- northern Taiwan. The study period was from January 2014
tophilia or those intolerant to TMP/SMX due to adverse drug to June 2016. The approval of Institutional Review Board of
effects. Recent studies comparing treatments with fluo- the hospital (TSGHIRB approval number: 2-101-05-074). To
roquinolones and TMP/SMX have suggested that fluo- assess risk factors for LTSRSM occurrence, a 1:1:1 matched
roquinolones have similar efficacy but fewer adverse drug case-control-control design was used. Adult inpatients (>18
effects than TMP/SMX.7,8 Nevertheless, the clinical efficacy years old) infected/colonized with LTSRSM and those
of fluoroquinolones other than levofloxacin still needs to be infected/colonized with levofloxacin- and TMP/SMX-
validated due to the limited number of clinical studies re- susceptible S. maltophilia (LTSSSM) were identified via a
ported and the current lack of clinical breakpoints. computerized medical records system and classified as the
Although the resistance rates to levofloxacin or TMP/SMX case group and control group A, respectively. Patients
vary geographically, longitudinal global surveillance reports infected/colonized with S. maltophilia that showed inter-
have revealed that this resistance has generally remained mediate resistance to levofloxacin were categorized in the
less than 10%.9 In recent years, increased resistance rates case group for analysis. For patients with multiple episodes
of S. maltophilia to TMP/SMX or levofloxacin have been of S. maltophilia infection/colonization, only the first was
observed in several countries from local surveillance included. The control group B patients were selected from
reports.10e15 In Taiwan, a more worrisome phenomenon the inpatient population without S. maltophilia during their
was noted in which resistance rates of S. maltophilia to hospital stay. Patients with S. maltophilia isolated less than
both TMP/SMX and levofloxacin were elevated in nation- 48 h after admission or those younger than 18 years of age
wide surveillance reports.16 Furthermore, analysing were excluded. The case group was matched to control

108
 Journal of Microbiology, Immunology and Infection 55 (2022) 107e113

group A by age (within 5 years), sex, and the site of isola- Statistical analysis
tion. The criteria for matching the case group to control
group B included age (within 5 years), sex, and time at risk. Continuous variables are presented as means  standard
Once several eligible controls were identified, they were deviations (SDs), and we used t-tests or ManneWhitney U
randomly chosen using Microsoft Excel 2013 software tests for comparisons as appropriate. Categorical variables
(Microsoft Corp., Redmond, WA, USA). are presented as numbers and percentages and compared
using the chi-square test or Fisher’s exact test. Variables
Relevant clinical data and definitions with p values < 0.05 on bivariate analysis were included in
a forward stepwise conditional logistic regression model for
The clinical information from identified patients was multivariate analysis. Two simultaneous multivariate
retrieved from a computerized medical records system. models were produced from the data obtained. The first
Possible risk factors were recorded including the following: model used the pairs of the case group and matched control
age, sex, time at risk, recent admission and intensive care group A, while the second model used the pairs of the case
unit (ICU) records, recent chemotherapy and surgery, group and matched control group B. All tests were 2-tailed,
comorbidities, previous antimicrobial exposure, and and a p value of <0.05 was considered statistically signifi-
indwelling medical devices. In the case group and control cant. All results were analysed using a commercially
group A, the time at risk was defined as the number of days available software package (SPSS, version 16.0; SPSS Inc.,
elapsed from patient admission to the date of the first S. Chicago, IL, USA).
maltophilia isolation. In control group B, the time at risk
was defined as the number of days from patient admission
to discharge. Previous antibiotic exposure in the case group
Results
and control group A was defined as at least 24 h of therapy
within two weeks before S. maltophilia isolation. In control During the study period from January 2014 to June 2016,
group B, previous antibiotic exposure was defined as 24 h of 1213 patients were identified as having cultures positive for
therapy within two weeks prior to discharge. Recent ther- S. maltophilia. Of these patients, 593 were hospitalized in
apeutic measures were defined as having occurred 30 days the ICU and the others were admitted to the general ward.
before S. maltophilia isolation in patients in the case group The most common isolation source was from the respiratory
and control group A; in control group B, recent therapeutic tract (1038/1213, 85.5%), followed by the blood (62/1213,
measures were defined as having occurred 30 days before 5.1%). One hundred and twenty-nine of the 1213 S. mal-
index discharge. Indwelling invasive medical devices used tophilia isolates exhibited concurrent levofloxacin- and
before S. maltophilia isolation in the case group and con- TMP/SMX resistance (10.6%). Among the 129 patients with
trol group A were recorded. For control group B, indwelling LTSRSM, we excluded 15 patients with LTSRSM isolated
invasive medical devices used before discharge were within 48 h after admission, and seven unsuccessfully
recorded. matched cases. Consequently, the case group consisted of
107 patients with LTSRSM. Among 107 cases in the case
group with LTSRSM, the isolation source were from the
Microbiologic methods
respiratory tract (98/107, 91.5%), followed by the wound
tissue (6/107, 5.6%), blood (2/107, 1.8%) and the urinary
We analysed the antibiotic susceptibilities of 85 consecu- tract (1/107, 0.9%). After matching, control group A con-
tive, non-duplicated preserved isolates of LTSRSM from the sisted of 107 patients with LTSSSM and control group B
respiratory tract and preserved 85 LTSSSM isolates consisted of 107 patients randomly identified from the
randomly chosen from the same source as LTSRSM in hos- inpatient population without S. maltophilia infections were
pitalized patients during the study period for comparison. included for comparisons to determine independent risk
The identification of S. maltophilia isolates was performed factors for LTSRSM occurrence.
by matrix-assisted laser desorption/ionization time-of-
flight mass spectrometry (bioMérieux Inc., Marcy-l’Etoile,
Rhône, France). The minimum inhibitory concentrations Risk factors for LTSRSM occurrence among
(MICs) of identified S. maltophilia isolates were determined hospitalized patients
using the VITEK 2 automated system (bioMérieux Inc.). The
breakpoints for TMP/SMX, levofloxacin, and ceftazidime for As shown in Table 1, when the case group was compared to
S. maltophilia were established according to 2020 Clinical the control group A, there were no statistically significant
and Laboratory Standards Institute (CLSI) criteria. The between-group differences in time at risk, recent admission
clinical breakpoints for tigecycline were established ac- records, recent ICU admission, recent chemotherapy and
cording to the 2020 European Committee on Antimicrobial surgery, comorbidities, and indwelling medical devices
Susceptibility Testing for Enterobacterales. The break- (p > 0.05). Previous antibiotic exposure revealed that the
points for other tested antibiotics including ciprofloxacin, case group was more likely than the controls to have been
cefepime, colistin, gentamicin, and amikacin were estab- exposed to fluoroquinolone and TMP/SMX (43.0% vs. 7.5%
lished according to the 2020 CLSI criteria for Pseudomonas and 12.1% vs. 1.9%, p < 0.001 and 0.003, respectively).
aeruginosa. When the case group was compared to control group B, the

109
 R.-X. Wu, C.-M. Yu, S.-T. Hsu et al.

Table 1 Risk factors associated with levofloxacin- and trimethoprim/sulfamethoxazole (TMP/SMX)-resistant Steno-
trophomonas maltophilia isolates determined by bivariate analysis.
Risk factor, n (%) Case group Control group A Pa Control group B Pb
(n Z 107) (n Z 107) value (n Z 107) value
Age in yearsc 77.0 (22.0) 77.0 (22.0) 0.964 75.0 (25.0) 0.999
Male. 73 (68.2) 73 (68.2) e 73 (68.2) 1.000
Time at risk (days)c,d 23 (23) 18 (19) 0.085 23 (23) 1.000
Recent admission records in the last 3 28 (26.2) 35 (32.7) 0.294 38 (35.5) 0.139
months
Prior ICU admission 89 (83.2) 87 (81.3) 0.721 32 (29.9) <0.001
Chemotherapy in the last month 5 (4.7) 2 (1.9) 0.445 14 (13.1) 0.031
Surgery in the last month 59 (55.1) 53 (49.5) 0.412 40 (37.4) 0.009
Comorbidities
Cerebrovascular disease 43 (40.2) 34 (31.8) 0.200 24 (22.4) 0.005
Dementia 15 (14.0) 20 (18.7) 0.355 10 (9.3) 0.287
Heart failure 19 (17.8) 17 (15.9) 0.715 11 (10.3) 0.115
Chronic renal insufficiency 40 (37.4) 29 (27.1) 0.108 28 (26.2)) 0.078
Chronic lung disease 13 (12.1) 12 (11.2) 0.831 9 (8.4) 0.368
Liver cirrhosis 4 (3.7) 4 (3.7) 1.000 9 (8.4) 0.152
Diabetes mellitus 38 (35.5) 41 (38.3) 0.671 39 (36.4) 0.887
Autoimmune disease 8 (7.5) 6 (5.6) 0.580 3 (2.8) 0.122
Cancer 22 (20.6) 29 (27.1) 0.261 34 (31.8) 0.062
Peripheral vascular disease 6 (5.6) 3 (2.8) 0.498 3 (2.8) 0.498
Peptic ulcer disease 24 (22.4) 22 (20.6) 0.739 21 (19.6) 0.615
Previous antibiotics exposure
Macrolide 11 (10.3) 12 (11.2) 0.825 4 (3.7) 0.061
Aminoglycoside 9 (8.4) 7 (6.5) 0.603 2 (1.9) 0.030
penicillin/b-lactamase inhibitor 44 (41.1) 46 (43.0) 0.782 37 (34.6) 0.324
3rd generation cephalosporin 23 (21.5) 31 (29.0) 0.208 17 (15.9) 0.293
4th generation cephalosporin 37 (34.6) 37 (34.6) 1.000 11 (10.3) <0.001
Carbapenem 34 (31.8) 41 (38.3) 0.316 15 (14.0) 0.002
Glycopeptide 28 (26.2) 31 (29.0) 0.646 6 (5.6) <0.001
Fluoroquinolone 46 (43.0) 8 (7.5) <0.001 23 (21.5) 0.001
TMP/SMX 13 (12.1) 2 (1.9) 0.003 3 (2.8) 0.009
Invasive medical devices
Central venous catheter insertion 61 (57.0) 62 (57.9) 0.890 35 (32.7) <0.001
Nasogastric tube insertion 99 (92.5) 98 (91.6) 0.800 56 (52.3) <0.001
Foley catheter insertion 86 (80.4) 80 (74.8) 0.325 54 (50.5) <0.001
Percutaneous surgical wound drainage 20 (18.7) 26 (24.3) 0.318 24 (22.4) 0.499
use
Mechanical ventilation 87 (81.3) 77 (72.0) 0.106 20 (18.7) <0.001
a
Comparison of patients in the case group and control group A.
b
Comparison of patients in the case group and control group B.
c
Data are presented as means (standard deviations).
d
Days of stay prior to isolation of S. maltophilia.
ICU, intensive care unit.

patients in the case group had higher rates for recent ICU catheter, and ventilator, than those in control group B
admission (83.2% vs. 29.9%, p < 0.001), recent surgery (57.0% vs. 32.7%, 92.5% vs. 52.3%, 80.4% vs. 50.5% and 81.3%
(55.1% vs. 37.4%, p Z 0.009), cerebrovascular disease vs. 18.7%, all p < 0.001).
(40.2% vs. 22.4%, p Z 0.005), and exposure to multiple The results of multivariate logistic regression are sum-
antibiotics including aminoglycoside (8.4% vs. 1.9%, marized in Table 2. In this analysis comparing the case
P Z 0.030), fourth-generation cephalosporin (34.6% vs. group to control group A, previous usage of fluoroquinolone
10.3%, p < 0.001), carbapenem (31.8% vs. 14.0%, (odds ratio [OR] 22.824; 95% confidence interval [CI]
p Z 0.002), glycopeptide (26.2% vs. 5.6%, p < 0.001), flu- 4.984e104.524) and previous usage of TMP/SMX (OR 17.724;
oroquinolone (43.0% vs. 21.5%, p < 0.001), and TMP/SMX 95% CI 1.495e210.097) were independent risk factors for
(12.1% vs. 2.8%, p Z 0.009). Moreover, patients in the case LTSRSM isolation. When analysed using control group B,
group were more likely to have invasive medical devices, mechanical ventilation (OR 59.471; 95% CI
including a central venous catheter, nasogastric tube, Foley 2.698e1311.079), cerebrovascular disease (OR 12.371; 95%

110
 Journal of Microbiology, Immunology and Infection 55 (2022) 107e113

Table 2 Multivariate analysis (logistic regression) of risk factors for levofloxacin- and trimethoprim/sulfamethoxazole (TMP/
SMX)-resistant Stenotrophomonas maltophilia isolates.
Risk factor, n (%) OR (95% CI) P value
Case group vs. control group A
a b

Fluoroquinolone 22.824 (4.984e104.524) <0.001

Trimethoprim/sulfamethoxazole 17.724 (1.495e210.097) 0.02
Case group vs. control group Bc
Mechanical ventilation 59.471 (2.698e1311.079) 0.010
Cerebrovascular disease 12.371 (1.943e78.748) 0.008
Fluoroquinolone 13.075 (1.502e113.819) 0.020
CI, confidence interval; OR, odds ratio.
a
Patients with levofloxacin- and TMP/SMX-resistant S. maltophilia (case group).
b
Patients with levofloxacin- and TMP/SMX-susceptible S. maltophilia (control group A).
c
Patients without S. maltophilia infection (control group B).

CI 1.943e78.748), and previous usage of fluoroquinolone mL and 1 mg/mL, respectively). The other antibiotics tested
(OR 13.075; 95% CI 1.502e113.819) were independently (cefepime, gentamicin, amikacin, and colistin) revealed
associated with LTSRSM isolation. Only one factor, previous limited in vitro activity and resistant rates were not
antibiotic exposure to fluoroquinolone, was found to be significantly different between the LTSRSM and LTSSSM
independently associated with occurrence of LTSRSM when isolates according to the CLSI MIC breakpoints for P. aeru-
the case group was compared to both control groups A and ginosa (p > 0.05).
B, respectively.

Antimicrobial susceptibility comparisons between Discussion

LTSRSM and LTSSSM
To our knowledge, the current study was the first to
The MIC ranges, MIC50, MIC90 values, and the resistant investigate risk factors and antibiotic susceptibilities for
percentages to the tested antibiotics of the 85 LTSRSM and LTSRSM from hospitalized patients. A matched case-
85 LTSSSM isolates are shown in Table 3. All LTSRSM isolates control-control design was used and compared the case
consistently showed resistance to ciprofloxacin, another group of patients with LTSRSM to two control groups of
type of fluoroquinolone, with MIC50/90 values > 4 mg/mL. patients, one with LTSSSM and one comprised of randomly
Among the 170 S. maltophilia isolates tested, including the selected inpatients without S. maltophilia infection to
85 LTSRSM and 85 LTSSSM isolates, 31% were resistant to eliminate the potential biased risk estimates from tradi-
ceftazidime, and the LTSRSM isolates were significantly tional caseecontrol studies.21 The main findings of our
more likely to exhibit resistance to ceftazidime than the study were that previous antibiotic use of fluoroquinolone
LTSSSM isolates (43.5% vs. 25.9%, p Z 0.016). With regard to and TMP/SMX, patients with recent mechanical ventilation
tigecycline susceptibility, the resistance rates were signif- and cerebrovascular diseases were independent risk factors
icantly higher among the LTSRSM isolates than among the associated with subsequent LTSRSM occurrence. In addi-
LTSSSM isolates (90.6% vs. 11.8%, p < 0.001). Moreover, the tion, the MICs and resistance rates of tigecycline and cef-
values of MIC50 and MIC90 were 4 mg/mL and >8 mg/mL for tazidime were higher in LTSRSM than LTSSSM isolates.
the LTSRSM isolates, respectively, which were at least 8- Inpatients with multiple comorbidities frequently sub-
fold higher than those for the LTSSSM isolates (<0.5 mg/ jected to indwelling medical devices such as ventilators are

Table 3 Minimal inhibitory concentration and susceptibility comparisons of antibiotics between levofloxacin- and trimetho-
prim/sulfamethoxazole (TMP/SMX)-resistant Stenotrophomonas maltophilia (LTSRSM) and levofloxacin- and TMP/SMX-
susceptible S. maltophilia (LTSSSM).
Antibiotic MIC range (mg/mL) MIC50 (mg/mL) MIC90 (mg/mL) No. (%) resistant isolates
LTSRSM LTSSSM LTSRSM LTSSSM LTSRSM LTSSSM LTSRSM LTSSSM P value
(n Z 85) (n Z 85) (n Z 85) (n Z 85) (n Z 85) (n Z 85) (n Z 85) (n Z 85)
Ceftazidime <1e>64 <1e>64 16 4 >64 >64 37 (43.5) 22 (25.9) 0.016
Cefepime <1e>64 <1e>64 16 32 >64 >64 63 (74.1) 64 (75.2) 1.000
Tigecycline <0.5e>8 <0.5e4 4 <0.5 >8 1 77 (90.6) 10 (11.8) <0.001
Colistin <0.5e>16 <0.5->16 <0.5 <0.5 >16 >16 34 (40.0) 24 (28.2) 0.106
Ciprofloxacin >4 <0.12e4 >4 0.5 >4 2 85 (100) 12 (14.1) <0.001
Gentamicin 2e>16 <1e>16 >16 8 >16 >16 43 (50.6) 41 (48.2) 0.878
Amikacin 4e>64 <2e>64 >64 >64 >64 >64 43 (50.6) 44 (51.8) 1.000

111
 R.-X. Wu, C.-M. Yu, S.-T. Hsu et al.

at risk of acquiring multidrug-resistant strains.22e24 It may Results of current study from clinical data analysis
be that such patients with medical advices who need more revealing consistent results strengthened findings from past
nursing care during hospitalization are more likely to be basic studies but molecular mechanism characterisation
infected from cross transmission and that infected resistant using clinical S. maltophilia isolates to correlate clinical
organisms are more difficult to be eradicated with anti- data results was warranted for further validation. Cautious
biotic treatment than susceptible ones. Consistently, our use of TMP/SMX and fluoroquinolone to reduced selective
study revealed that ventilator use was associated with pressure may be important to limit LTSRSM emergence in
subsequent LTSRSM occurrence. For the same reason, pa- the hospital. Other antibiotics tested in our study, including
tients with poor functional status showing increased activ- colistin, all revealed limited in vitro activity against LTSRSM
ities of daily living (ADL) score may need increased bedside isolates. The emergence of multidrug-resistant S. malto-
care and are more likely to be infected with resistant philia, as presented here, would be a great concern for
bacteria.25e27 This may explain why being a patient with current and future medical practice. New drug develop-
cerebrovascular disease who may have a high ADL score was ment would be important to provide clinicians more ther-
identified as another risk factor for LTSRSM occurrence in apeutic options when faced with infections caused by such
our study. The risk factors identified for LTSRSM occurrence resistant strains.
mentioned above indicate that implementation of strict There were some limitations to this study. First, this
hand hygiene protocols in caring for inpatients may be study was retrospective in nature; hence, our results were
helpful for preventing LTSRSM from spreading in the hos- susceptible to potential selection biases in the study.
pital. In our study, we also found that previous fluo- Second, molecular epidemiology of the collected isolates
roquinolone and TMP/SMX use were associated with was not carried out. We therefore could not assess if
subsequent LTSRSM occurrence, which was consistent with patient-to-patient transmission occurred during the study
a previous study found that both antibiotics increased the period, which may have influenced our results in the
likelihood of subsequent multidrug-resistant S. maltophilia clinical data analysis. Third, this was a single medical
isolation in cancer patients.28 Besides past and our present center in northern Taiwan and caution was necessary
clinical studies, we were able to observe similar results when extrapolating these results to other regions in
from basic studies of S. maltophilia in the laboratory. In Taiwan or other countries.
experimental evolution studies reported, S. maltophilia
exhibited simultaneous resistance to quinolone and TMP/ Conclusion
SMX after either TMP/SMX or quinolone exposure. The
resistance mechanism was resulted from overexpression of
The current study identified that mechanical ventilation
resistance-nodulation-cell division (RND) efflux pumps,
use, patients with cerebrovascular disease, and previous
mainly SmeDEF.20,29 Since overexpression of the RND efflux
fluoroquinolone and TMP/SMX exposure were independent
pump was also reported the major resistance mechanism
risk factors for subsequent LTSRSM occurrence. LTSRSM
for tigecycline in S. maltophilia, cross resistance to tige-
isolates exhibited higher rates of resistance against tige-
cycline may be anticipated.30 Agree with our inference, we
cycline and ceftazidime than LTSSSM. Future multicenter
also noted elevated MICs and higher resistance rates of
studies with prospective evaluations and molecular char-
tigecycline in LTSRSM than LTSSSM in our study. Tigecycline
acterizations of clinical LTSRSM isolates to corroborate our
has broad activity against gram-positive and -negative or-
study results were necessary.
ganisms, including multidrug-resistant organisms. From
recent global surveillance studies, tigecycline also demon-
strated good in vitro activity against S. maltophilia.31,32 Declaration of competing interest
Although the clinical breakpoints of tigecycline for S.
maltophilia have not been determined, clinical studies The authors declare no conflict of interest.
have suggested that tigecycline has efficacy equivalent to
that of TMP/SMX in treating S. maltophilia infections.33
Consequently, tigecycline has been considered an alterna- Acknowledgements
tive therapeutic option. The cross-resistance of LTSRSM to
tigecycline we presented here may further restrict already This work was supported by a grant from the Tri-Service
limited therapeutic choices in treating S. maltophilia in- General Hospital Penghu Branch (TSGH-PH-107-02 and
fections. The MIC50 values and resistance rates of ceftazi- TSGH-PH-109-5) and the Tri-Service General Hospital, Na-
dime were also higher in LTSRSM than in LTSSSM. The tional Defense Medical Center (TSGH-C103-187, TSGH-
resistance mechanism of S. maltophilia for ceftazidime C104-195).
may generally be considered to correlate with intrinsic
beta-lactamases L1 and L2 but overexpression of efflux
pump transporters was also reported.34 Whether there was
References
difference between LTSSSM and LTSRSM on resistance
1. Senol E. Stenotrophomonas maltophilia: the significance
mechanisms to ceftazidime warranted future investigations
and role as a nosocomial pathogen. J Hosp Infect 2004;57:
for further exploration. Study results on mechanisms of 1e7.
concurrent TMP/SMX and fluoroquinolone resistance re- 2. Chang YT, Lin CY, Chen YH, Hsueh PR. Update on infections
ported before were from basic studies based on laboratory caused by Stenotrophomonas maltophilia with particular
S. maltophilia strains under the laboratory environment, attention to resistance mechanisms and therapeutic options.
which may not be applicable to real clinical settings. Front Microbiol 2015;6:893.

112
 Journal of Microbiology, Immunology and Infection 55 (2022) 107e113

3. Looney WJ, Narita M, Muhlemann K. Stenotrophomonas mal- 19. Pien CJ, Kuo HY, Chang SW, Chen PR, Yeh HW, Liu CC, et al.
tophilia: an emerging opportunist human pathogen. Lancet Risk factors for levofloxacin resistance in Stenotrophomonas
Infect Dis 2009;9:312e23. maltophilia from respiratory tract in a regional hospital. J
4. Brooke JS. New strategies against Stenotrophomonas malto- Microbiol Immunol Infect 2015;48:291e5.
philia: a serious worldwide intrinsically drug-resistant oppor- 20. Sanchez MB, Martinez JL. Overexpression of the efflux pumps
tunistic pathogen. Expert Rev Anti Infect Ther 2014;12:1e4. SmeVWX and SmeDEF is a major cause of resistance to Co-
5. Andelkovic MV, Jankovic SM, Kostic MJ, Zivkovic Zaric RS, trimoxazole in Stenotrophomonas maltophilia. Antimicrob
Opancina VD, Zivic MZ, et al. Antimicrobial treatment of Agents Chemother 2018;62:e00301e18.
Stenotrophomonas maltophilia invasive infections: systematic 21. Rafailidis PI, Bliziotis IA, Falagas ME. Case-control studies
review. J Chemother 2019;31:297e306. reporting on risk factors for emergence of antimicrobial
6. Muder RR. Optimizing therapy for Stenotrophomonas malto- resistance: bias associated with the selection of the control
philia. Semin Respir Crit Care Med 2007;28:672e7. group. Microb Drug Resist 2010;16:303e8.
7. Cho SY, Kang CI, Kim J, Ha YE, Chung DR, Lee NY, et al. Can 22. van Loon K, Voor In ’t. A systematic review and meta-analyses
levofloxacin be a useful alternative to trimethoprim- of the clinical epidemiology of carbapenem-resistant
sulfamethoxazole for treating Stenotrophomonas maltophilia enterobacteriaceaeHolt AF, Vos MC, editors. Antimicrob
bacteremia? Antimicrob Agents Chemother 2014;58:581e3. Agents Chemother 2018;62:e01730. 17.
8. Wang YL, Scipione MR, Dubrovskaya Y, Papadopoulos J. Mono- 23. Mittal G, Gaind R, Kumar D, Kaushik G, Gupta KB, Verma PK,
therapy with fluoroquinolone or trimethoprim-sulfamethoxazole et al. Risk factors for fecal carriage of carbapenemase pro-
for treatment of Stenotrophomonas maltophilia infections. ducing Enterobacteriaceae among intensive care unit patients
Antimicrob Agents Chemother 2014;58:176e82. from a tertiary care center in India. BMC Microbiol 2016;16:
9. Gales AC, Seifert H, Gur D, Castanheira M, Jones RN, Sader HS. 138.
Antimicrobial susceptibility of Acinetobacter calcoaceticus- 24. Drinka P, Niederman MS, El-Solh AA, Crnich CJ. Assessment of
Acinetobacter baumannii complex and Stenotrophomonas risk factors for multi-drug resistant organisms to guide empiric
maltophilia clinical isolates: results from the SENTRY antimi- antibiotic selection in long term care: a dilemma. J Am Med Dir
crobial surveillance program (1997-2016). Open Forum Infect Assoc 2011;12:321e5.
Dis 2019;6:S34e46. 25. Brito V, Niederman MS. Healthcare-associated pneumonia is a
10. Herrera-Heredia SA, Pezina-Cantu C, Garza-Gonzalez E, Bocane- heterogeneous disease, and all patients do not need the same
gra-Ibarias P, Mendoza-Olazaran S, Morfin-Otero R, et al. Risk broad-spectrum antibiotic therapy as complex nosocomial
factors and molecular mechanisms associated with trimethoprim- pneumonia. Curr Opin Infect Dis 2009;22:316e25.
sulfamethoxazole resistance in Stenotrophomonas maltophilia in 26. El Solh AA, Pietrantoni C, Bhat A, Bhora M, Berbary E. In-
Mexico. J Med Microbiol 2017;66:1102e9. dicators of potentially drug-resistant bacteria in severe nursing
11. Hu LF, Chen GS, Kong QX, Gao LP, Chen X, Ye Y, et al. Increase home-acquired pneumonia. Clin Infect Dis 2004;39:474e80.
in the prevalence of resistance determinants to trimetho- 27. Trick WE, Weinstein RA, DeMarais PL, Kuehnert MJ,
prim/sulfamethoxazole in clinical Stenotrophomonas malto- Tomaska W, Nathan C, et al. Colonization of skilled-care fa-
philia isolates in China. PloS One 2016;11:e0157693. cility residents with antimicrobial-resistant pathogens. J Am
12. Neela V, Rankouhi SZ, van Belkum A, Goering RV, Awang R. Geriatr Soc 2001;49:270e6.
Stenotrophomonas maltophilia in Malaysia: molecular epide- 28. Ansari SR, Hanna H, Hachem R, Jiang Y, Rolston K, Raad I. Risk
miology and trimethoprim-sulfamethoxazole resistance. Int J factors for infections with multidrug-resistant Steno-
Infect Dis 2012;16:e603e7. trophomonas maltophilia in patients with cancer. Cancer 2007;
13. Lai CC, Chen YS, Lee NY, Tang HJ, Lee SS, Lin CF, et al. Sus- 109:2615e22.
ceptibility rates of clinically important bacteria collected from 29. Pak TR, Altman DR, Attie O, Sebra R, Hamula CL, Lewis M,
intensive care units against colistin, carbapenems, and other et al. Whole-genome sequencing identifies emergence of a
comparative agents: results from Surveillance of Multicenter quinolone resistance mutation in a case of Stenotrophomonas
Antimicrobial Resistance in Taiwan (SMART). Infect Drug Resist maltophilia bacteremia. Antimicrob Agents Chemother 2015;
2019;12:627e40. 59:7117e20.
14. Zhanel GG, Adam HJ, Baxter MR, Fuller J, Nichol KA, 30. Blanco P, Corona F, Martinez JL. Mechanisms and phenotypic
Denisuik AJ, et al. Antimicrobial susceptibility of 22746 path- consequences of acquisition of tigecycline resistance by Sten-
ogens from Canadian hospitals: results of the CANWARD 2007- otrophomonas maltophilia. J Antimicrob Chemother 2019;74:
11 study. J Antimicrob Chemother 2013;68:7e22. 3221e30.
15. Chung HS, Hong SG, Kim YR, Shin KS, Whang DH, Ahn JY, et al. 31. Farrell DJ, Sader HS, Jones RN. Antimicrobial susceptibilities of
Antimicrobial susceptibility of Stenotrophomonas maltophilia a worldwide collection of Stenotrophomonas maltophilia iso-
isolates from Korea, and the activity of antimicrobial combi- lates tested against tigecycline and agents commonly used for
nations against the isolates. J Kor Med Sci 2013;28:62e6. S. maltophilia infections. Antimicrob Agents Chemother 2010;
16. Wu H, Wang JT, Shiau YR, Wang HY, Lauderdale TL, Chang SC, 54:2735e7.
et al. A multicenter surveillance of antimicrobial resistance on 32. Pfaller MA, Flamm RK, Duncan LR, Mendes RE, Jones RN,
Stenotrophomonas maltophilia in Taiwan. J Microbiol Immunol Sader HS. Antimicrobial activity of tigecycline and cefoper-
Infect 2012;45:120e6. azone/sulbactam tested against 18,386 Gram-negative organ-
17. Wang CH, Yu CM, Hsu ST, Wu RX. Levofloxacin-resistant Sten- isms from Europe and the Asia-Pacific region (2013-2014).
otrophomonas maltophilia: risk factors and antibiotic suscep- Diagn Microbiol Infect Dis 2017;88:177e83.
tibility patterns in hospitalized patients. J Hosp Infect 2020; 33. Tekce YT, Erbay A, Cabadak H, Sen S. Tigecycline as a thera-
104:46e52. peutic option in Stenotrophomonas maltophilia infections. J
18. Wang CH, Lin JC, Chang FY, Yu CM, Lin WS, Yeh KM. Risk factors Chemother 2012;24:150e4.
for hospital acquisition of trimethoprim-sulfamethoxazole 34. Blanco P, Corona F, Martinez JL. Involvement of the RND efflux
resistant Stenotrophomonas maltophilia in adults: a matched pump transporter SmeH in the acquisition of resistance to
case-control study. J Microbiol Immunol Infect 2017;50: ceftazidime in Stenotrophomonas maltophilia. Sci Rep 2019;9:
646e52. 4917.

113