Introduction
As declared by the World Health Organization (WHO), antimicrobial resistance is one of the biggest challenges and threats in modern Medicine. 1 Extensively drug-resistant organisms (XDRO), specifically non-glucose fermenting XDR Gram-negatives (eg, Pseudomonas aeruginosa, Acinetobacter baumannii), pose an urgent worldwide concern to human health, mainly due to serious lack in therapeutic options. 1 This leads to delay in initiation of appropriate antimicrobials, Reference Paul, Shani, Muchtar, Kariv, Robenshtok and Leibovici2 which is the strongest modifiable predictor for mortality in septic shock. Reference Paul, Shani, Muchtar, Kariv, Robenshtok and Leibovici2
Two additional non-glucose fermenting XDR Gram-negatives, which frequently display XDR phenotypes, Reference Jean, Hsueh and Lee3,Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4 are Stenotrophomonas maltophilia and Achromobacter xylosoxidans. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4,Reference Brooke5 These microorganisms were perceived in the past as commensals, Reference Jean, Hsueh and Lee3 but are increasingly reported as causative pathogens, associated with serious infections, particularly among intensive care unit (ICU) patients. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4,Reference Brooke5 The commonest reservoir for these organisms is the respiratory tract, and the commonest infectious syndromes are pneumonia (including ventilated-associated pneumonia) and blood stream infections [BSI]). Reference Brooke5 These pathogens have intrinsic mechanisms of resistances (both inducible and constitutive production Reference Jean, Hsueh and Lee3 ) to most antimicrobial agents, Reference Jean, Hsueh and Lee3–Reference Brooke5 including reports of pan-resistant isolates. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4–Reference Lobo, Tulu, Aris and Noone6 Due to baseline characteristics, Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4–Reference Lobo, Tulu, Aris and Noone6 and presumable frequent delays in initiation of appropriate therapy, Reference Paul, Shani, Muchtar, Kariv, Robenshtok and Leibovici2 devastating outcomes were reported from non-controlled case-series analyses, with high morbidity and mortality rates. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4,Reference Brooke5 The aim of this study was to explore the descriptive epidemiology, the independent predictors, and the outcomes, of S. maltophilia and of A. xylosoxidans invasive infections.
Methods
Retrospective, observational, matched nested case-control study was conducted for calendar years 2011–2022, at Shamir (Assaf Harofeh) Medical Center (895-bed university-affiliated acute-care facility), Israel. Data were obtained from all available records, and mortality data were retrieved from the Israeli Ministry of Interior records. The local ethic (“Helsinki”) committee approved the study prior to its initiation.
All adult (>18 years) hospitalized patients with S. maltophilia or with A. xylosoxidans bloodstream infection (BSI), per established BSI definition, Reference Singer, Deutschman and Seymour7 were enrolled. The time trend of infections over the years were analyzed by the Cochran-Armitage Test. Each BSI case was matched to three control (uninfected) patients, based on pre-defined matching criteria (depicted in order of importance): 1) ‘time at risk’ (number of days from admission to infection, ±5 days), 2) type of admitting ward (eg, Medicine, Surgery, ICU), and 3) age (per decade). Matched analyses of predictors to S. maltophilia and of predictors to A. xylosoxidans BSI, and non-matched outcomes analyses (nine different morbidity and mortality outcome parameters), were queried by logistic or Cox regressions, respectively.
Results
A total of 66 patients with S. maltophilia BSI and 8 patients with A. xylosoxidans BSI were enrolled. Twenty-three patients (31%) had a polymicrobial infection. There were no significant elevations in the incidences of S. maltophilia BSI (p for trend 0.45) and of A. xylosoxidans BSI (p for trend 0.57) during the 12-year study’s period. Five patients with S. maltophilia BSI could not be matched to control patients due to extreme length of ‘time a risk’ and were excluded from the analyses that compared cases to controls (Table 1), but not from the case-series descriptive analysis. Patients with S. maltophilia BSI or A. xylosoxidans BSI had a median age of 74 years (interquartile range [IQR] 67–82 years), 67% were males, and 16% were chronic residents of long-term care facilities (LTCF).
Predictors and outcomes of patients with Stenotrophomonas maltophilia bloodstream infections (BSI), Shamir Medical Center, 2011–2020
Note. IQR = interquartile range; OR = odds ratio; CI = confidence interval; MIC = minimal inhibitory concentration; LTCF = long-term care facility; ICU = intensive care unit; SD = standard deviation; TNF = tumor necrosis factor.
1 Valid percent implies the percent after removing missing values from the denominator.
2 Chronic lung disease refers to bronchiectasis, chronic obstructive pulmonary disease (COPD), asthma, restrictive lung disease.
3 Malignancy refers to a patient that had in the past or currently having a malignancy, hematological or solid-organ.
4 Chronic skin ulcers refer to lower limb diabetic foot wounds, decubitus ulcers, dwelling wound surrounding PEG insertion, surgical-site wounds, wounds surrounding catheters. Wounds involving soft tissue/mucosa without skin involvement (eg, in mouth, GI tract, genital-urinary tract) were not included.
5 Systemic glucocorticoid/steroid use of at least three doses in the past month.
6 Immunosuppressants and/or immunomodulators includes for example anti-TNF (infliximab [Remicade], adalimumab [Humira], certolizumab pegol [Cimzia], golimumab [Simponi], etanercept [Enbrel]) and others.
7 Immunosuppression include any of the following: 1) Neutropenia at culture date (<500 neutrophils/mm Reference Jean, Hsueh and Lee3 ), 2) glucocorticoids exposure in the past month, 3) chemotherapy in the past 3 months, 4) radiotherapy, 5) post-transplantation of any kind, 6) anti-TNF therapy in past 3 months, or 7) HIV carrier.
8 Isolation in the past 6 months of any of the following: Staphylococcus aureus resistant to oxacillin (MRSA), ampicillin and/or vancomycin resistant enterococci (VRE), Enterobacterales which is resistant to any 3rd or 4th generation cephalosporin (ESBL), Enterobacterales with meropenem MIC > 1 (CRE), Acinetobacter baumannii, or Pseudomonas aeruginosa.
9 Hemodialysis units, weekly visits at sub-specialty clinics.
10 Endoscopies, biopsies, permanent central line insertions, percutaneous interventions, and any type of surgery, brain/cardiac catheterization, intubation and mechanical ventilation.
11 Permanent/chronic internal foreign devices that were placed at least 48 hours prior to bacterial isolation from blood culture, such as any drain or catheter are included (dialysis Perm-A-Cath). Internal foreign devices such as heart valves, prosthetic joints, pacemakers, subcutaneous insulin pumps were not included. Empiric regimen = Antibiotic treatment given 2 days prior to 3 days following culture date; Consolidative (main) therapeutic treatment = antibiotic therapy given 3 days to 14 days following the culture date.
12 Time from last antibiotic use refers to the days from patient’s last antibiotic day to the day of the index bacterial culture.
13 Invasive procedure includes any surgery, endoscopy, biopsy, percutaneous intervention, catheterization and intubation.
Of the S. maltophilia offending isolates, six (9.8%) were resistant to trimethoprim/sulfamethoxazole (TMP/SMX), three (4.9%) to levofloxacin, and four (6.6%) to tigecycline. Among the A. xylosoxidans isolates, one strain (12.5%) was resistant to TMP/SMX, seven (87.5%) to levofloxacin, and all isolates were susceptible to meropenem.
The Table displays the univariable comparisons between 61 S. maltophilia BSI and 183 matched controls. Patients with S. maltophilia BSI were more likely to be dependent in their functional status upon admission, be hemodialysis dependent, with higher rates of recent hospitalizations, and mechanical ventilation (Table 1). Patients who developed S. maltophilia BSI had higher rates of recent exposure to antibiotics (overall), ie, specifically to cephalosporins and to tigecycline. The time from last antibiotic use, was also significantly shorter compared to controls. In multivariable analysis, the only independent predictors to S. maltophilia BSI remained hemodialysis (aOR = 3.6, CI-95 = 1.3–10.2), and recent exposure to (any) antibiotics (aOR = 4.1, CI-95 = 1.4–12.5).
Outcomes were significantly worse among patients with S. maltophilia BSI (Table 1). In separate multivariable models, S. maltophilia BSI remained independently associated with in-hospital mortality (aOR = 7.4, CI-95 = 3.2–18.4), 30- and 90-day mortality (aOR = 3.2, CI-95 = 1.5–6.9, and aOR = 2.7, CI-95 = 1.3–5.7, respectively), and with readmission in the following three months (aOR = 3.2, CI-95 = 1.3–9.0).
Eight A. xylosoxidans BSI were matched to 24 matched controls. Patients with A. xylosoxidans BSI had median length of stay from admission to infection of 7 days (IQR 5–14 days). Compared to controls, patients with A. xylosoxidans BSI had significant shorter duration since their last hospitalization (20 days (IQR 5–77) vs. 100 days (IQR 27–460), p = 0.04), had significant higher number of days on mechanical ventilation (146 days (IQR 16–276) vs. 1 day (IQR 1–7), p = 0.01), and had elevated rates of recent exposure to fluoroquinolones (63% vs 13%, p = 0.01). In multivariable model, the only independent modifiable predictor to A. xylosoxidans BSI, remained recent exposure to fluoroquinolones (aOR = 7.1, CI-95 = 1.01–59.1). Outcomes of patients with A. xylosoxidans BSI were not significantly worse compared to matched controls.
Discussion
The epidemiological burden and significance associated with nosocomial XDR Gram-negative infections is substantial, primarily due to lack of effective therapeutics, specifically infections resulting from non-glucose fermenting XDR isolates. Reference Jean, Hsueh and Lee3 The reported incidences of S. maltophilia and A. xylosoxidans infections from multiple hospitals are rising, Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4–Reference Lobo, Tulu, Aris and Noone6 with devastating outcomes reported from few case-series analyses. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4,Reference Brooke5 In this nested case-control trial, we explored the epidemiology of 66 S. maltophilia infections and of 8 A. xylosoxidans infections, for a 12-year period (2011–2022). In order to avoid classification bias and the inclusion of asymptomatic carriers as cases, Reference Alcaraz, Centron, Camicia, Quiroga, Di Conza and Passerini de Rossi8,Reference Dupont, Jumas-Bilak, Michon, Chiron and Marchandin9 only patients with BSI (per established definition Reference Singer, Deutschman and Seymour7,Reference Horan, Andrus and Dudeck10 ) were enrolled as cases. The incidences of S. maltophilia infections and of A. xylosoxidans infections, had not increased at SMC during the study.
Each BSI case was matched to three control patients, per strict matching criteria, determined a-priori, in accordance to established recommendations, Reference Kaye, Harris, Samore and Carmeli11 in order to power the small sample size calculations. per multivariable analyses, the independent predictors for S. maltophilia BSI were chronic hemodialysis and recent antibiotic usage (of any class) in the previous three months. Previous recent usage of fluoroquinolones was the only independent predictor for A. xylosoxidans BSI, with wide confidence intervals for the models’ adjusted odds ratios, due to the low sample sizes. Nearly all patients who developed S. maltophilia infections (85%), or A. xylosoxidans infections (88%), had recent exposure to antimicrobials, implying that selective pressure is the main promotor for emergence of these resistant strains in hospitals. This might also imply that patient-to-patient transmission was not the common mode of transmission or dissemination at SMC during 2011–2022 of these pathogens, but this needs to be trialed specifically.
There were various worse outcomes, which were independently associated with S. maltophilia infections. Mortality rates (ie, in-hospital mortality, 30-day mortality, 90-day mortality) and rates of readmissions in the following three months among those who survived the index hospitalization, were all independently associated with S. maltophilia BSI. In previous analyses, S. maltophilia “colonization” was not associated with worse clinical outcomes. Reference Lobo, Tulu, Aris and Noone6 This further highlights the importance of the controlled design executed herein, and the fact that only ‘true’ BSI cases were enrolled as cases. In additional analyses who tried to enroll as cases, patients with S. maltophilia or A. xylosoxidans “clinical infection” (various definitions applied), infections were also statistically associated with worse outcomes. Reference Perez Barragan, Sandino Perez, Corbella, Orellana and Fernandez-Ruiz4,Reference Brooke5 The associations of A. xylosoxidans BSI with increased mortality and readmission rates were evident in our univariable analyses, but due to the strict case definition criteria and the low sample size, it did not reach significance per multivariable models.
The main limitations of the study are the small sample of cases, specifically of the A. xylosoxidans cases. As previously mentioned, three control patients were matched to each case, in order to somewhat abate this limitation. In addition, the study has multiple inherent limitations associated with its retrospective chart review-based design, from a single Israeli center. The conclusions of this study could not be generalized nor extrapolated to other centers or to ambulatory settings (eg, patients with cystic fibrosis), without conducting additional detailed investigations. However, as far as we know, this is one of the largest analyses ever published in this evolving epidemiological field, which was executed with controlled design.
To conclude, S. maltophilia and A. xylosoxidans invasive infections were identified among 66 and 8 patients with prolonged hospitalization, during the years 2010–2020. The incidences are not rising at SMC over the years, but the infections are associated with serious and devastating outcomes. Selective pressure seems to promote infections, while in opposite to other XDR ICU non-glucose fermenting Gram-negative pathogens, Reference Marchaim, Navon-Venezia, Leavitt, Chmelnitsky, Schwaber and Carmeli12 this is the major suspected mode for emergence of these inherently resistant isolates. Independent predictors for infections were hemodialysis and prior antibiotic therapy in the previous three months for S. maltophilia BSI, and exposure to fluoroquinolones in the prior three months for A. xylosoxidans BSI. This should be considered in the management of nosocomial ICU patients, with prolonged hospitalizations, on mechanical ventilation, with extensive recent exposures to antimicrobials.
Acknowledgments
Author contribution: Conceptualization, S.Z.I. and D.M.; methodology, S.Z.I. and D.M.; formal analysis, S.Z.I., R.B. and D.M.; investigation, S.Z.I., E.C., L.A.P, A.Y., K.J., and D.M.; resources, D.M.; data curation, S.Z.I.; writing—original draft preparation, S.Z.I. and D.M.; writing—review and editing S.Z.I. and D.M.; supervision, D.M.; project administration, S.Z.I. and D.M. All authors have read and agreed to the submitted version of the manuscript.
Financial support
None.
Competing interests
All authors declare no conflict of interest.
Open access
