Factors associated with colonization by carbapenem-resistant enterobacteria in oncological patients: a case-control study

Introduction: Colonization and infections caused by Carbapenemase Producing Enterobacteria (CPE) are a global problem, being associated with an increase in hospitalization time, costs for health services, and morbidity and mortality rates. Oncologic patients represent a group of special interest and there are few studies involving CPE colonization among these patients. Aim: to investigate factors associated with colonization in cancer patients. Outlining: Case-control study developed in a tertiary reference hospital in cancer treatment in Porto Alegre, Brazil, from January to December 2017. The population consisted of patients diagnosed with cancer in clinical or surgical hospitalization. Results: The univariate analysis showed that variables associated with colonization by CPE were age, male sex, tumors with bone type of surgical hospitalization, number of intra-hospital transfers since hospitalization, hospitalization time >30 days, ICU hospitalization in the last 30 days, ICU time more than 15 days, surgical procedure in the last 30 days, use of antibiotics in the last 30 days, presence of tumor wound, and KPC infection. After multivariate analysis, male sex, external hospital as origin, hospital stay longer than 30 days, antibiotic use in the last 30 days, and presence of tumor wound, remained associated with EPC colonization. Use of aminoglycosides, and linezolide were associated with CPE colonization. Implications: We identified variables associated with CPE colonization in oncologic patients. Our results may indicate actions to prevent CPE colonization and consequent development of infections.


INTRODUCTION
The emergence and spread of Carbapenemase-Producing-Carbapenemase (CRE)resistant Enterobacteriaceae is a major public health problem, emerging as a global threat in the last decade. 1hese micorganisms are associated with an increase in hospitalization length of stay, costs for health services, and morbidity and mortality rates.Organ transplants, intensive care unit (ICU) hospitalizations, complex surgical procedures, prolonged hospitalizations, and oncological and onco-hematological diseases are also impacted with infections due to CRE and increasing in antimicrobial resistance.For this reason, oncologic patients may represent a group of special interest and there are few studies involving CRE colonization among them. 1 Besides, oncological and onco-hematological diseases are also associated with microbial resistance.
Moreover, an early and rapid identification of colonized hospitalized patients is mandatory to avoid the spread of these highly resistant pathogens. 3search and knowledge about multidrug-resistant bacteria are relevant in order to limit their dissemination, encouraging the reduction of morbidity and mortality rates, accompanied by microbiological surveillance.Outbreaks in the hospital environment has been highlighted as a challenge to be considered, since it has become a difficult problem to solve.It is also necessary to consider the increasing number of patients with different conditions, as well as the frequency of immunosuppression conditions.
Considering that there are few studies involving CRE colonization among these oncological patients, the aim of this study was to investigate factors associated with CRE colonization in oncologic patients.For the descriptive analysis, the quantitative variables were presented by mean and standard deviation or median and interquartile range.

DISCUSSION
In this case-control study we identified an association of CRE colonization with variables such as male sex, type of hospitalization, time of hospitalization, hospitalization in ICU, previous use of antimicrobial agents, and presence of a tumor wound.CRE are increasingly prevalent in health institutions and their worldwide dissemination contributes to morbidity and mortality, leading to the need for active surveillance to identify and investigate associated factors. 8though there are studies on the prevalence and dissemination of ERC, there are relatively few studies that specifically address factors associated with colonization and infection by these microorganisms, especially among cancer patients.
We observed an association between male sex and CRE colonization.A recent study showed an association between male sex and the development of infection caused by carbapenemase-producing enterobacteria, when this group was compared to the group of patients infected with enterobacteria with non-susceptible carbapenems by mechanisms other than carbapenemase production.Such association had previously been described for the acquisition of other microorganisms. 9r study demonstrates the marked capacity of dissemination of carbapenemases among different bacterial species, which is consistent with previous studies. 10lthough most cases occurred due to colonization by KPC-producing K. pneumoniae, a wide variety of Enterobacterales was described.It is interesting to note that blaKPC production was mostly related to K. pneumoniae (in only three cases the enzyme was not detected in carbapenem-resistant isolates of this species), whereas the occurrence of NDM-producing isolates was more related to Enterobacter spp.This fact corroborates reports of worrisome dissemination at both national and international levels. 11e univariate analysis indicated different factors potentially associated with CRE colonization and allowed us to deeply investigate the influence of such factors. The treatment of cancer patients is challenging due to the exposure of intensive chemotherapy protocols, the use of monoclonal antibodies or other biological agents, the increasing age of patients with cancer disease and the frequent presence of multiple comorbidities. 14he patients with hematologic malignancies and recipients of hematopoietic stem cell transplants (HSCT) are at high risk of developing invasive infections due to enteric bacteria due to chemotherapy-induced neutropenia and gastrointestinal mucositis. 1us, our study strengthens this association with surgical procedures, and patients undergoing surgery are usually hospitalized in the long term, which could increase the risk of exposure.] Since our hospital is a national tertiary reference center for oncological care, most patient on oncological disease treatment undergo prolonged medical treatment and are exposed to multiple and prolonged hospitalizations, and to extensive use of broad-spectrum antimicrobials. 15his may explain the high CRE prevalence among these patients.
Hospitalization in ICU is also typically associated with resistance development, what was confirmed in our study.
Previous antibiotic use was significantly associated with CRE colonization; previous studies have shown that recent antimicrobial use may be a risk factor for colonization due to CRE. [16][17][18][19] The antimicrobials that stood out were piperacillin-tazobactam, carbapenems, linezolid and aminoglycosides.Treatment with carbapenems prior to colonization correlates with previous reports. 20rbapenem use is a well-defined factor associated with CRE, although this association was not present in some studies. 18,21hese discrepancies among the studies merit careful analysis but may be related to different definitions of antimicrobial exposure and the possibility of uncontrolled confounders in retrospective studies. 22iperacillin-tazobactam is widely used in our institution, what represents a cause for concern, since it has been associated with colonization by CRE.To the best of our knowledge, the association of colonization (or infection) by CRE and use of aminoglycosides or linezolid has not been previously described.Linezolid is selectively used in our hospital and its role as suppressor of gram-positive microbiota seems to confer advantages for survival of enterobacteria in the intestinal environment.Interestingly, the use of linezolid as a selective agent enhances the in vitro detection of CRE when applied to surveillance cultures. 23unds influence the outcome of clinical treatment.Oncological lesions merit a specific approach to each patient. 24his is the first study that identified oncological wounds as a factor associated with colonization with CRE, raising a specific discussion not yet explored in previous studies.
Measures to address patient-to-patient transmission include: hand hygiene, contact isolation precautions, environmental cleanliness, decolonization protocols, and surveillance programs to identify the asymptomatic carrier.In contrast, tackling the resistance effort requires the application of antimicrobial stewardship policies to avoid the necessary use of broad-spectrum agents, especially carbapenems. 25n Brazil, ANVISA 4 defines a series of specific measures, such as: importance of hand hygiene, use of personal protective equipment (PPE), isolation of colonized and infected patients, epidemiological surveillance system, Active Hospital Infection (CCIH), rational use of antimicrobials, among others.In this way, the early identification of asymptomatic carriers by active surveillance cultures is an ideal strategic approach to track the carriage of these bacteria, to explain transmission and control outbreaks.25 In our institution, the use of surveillance cultures is a well established practice. Ts approachhas become an essential tool in infection control programs, not only during outbreaks, but also as a routine measure in ERC-endemic settings.
Screening patients to identify asymptomatic colonization and instituting preventive contact isolation measures such as patient-to-patient screening and colonization pressure improve patient outcomes. 25r study has limitations because the data are collected in a single center and because it was a retrospective analysis, so we may not be able to

CONCLUSION
We identified factors associated with transmission (surgical hospitalization, hospital stay, including ICU, and presence of a tumor wound).We emphasize the association with previous antimicrobial use., which may serve as a basis for the adoption of strategies for better use of antimicrobials.Our results emphasize the need of a strict control at oncology unit in order to avoid the burden of CRE colonization.
Categorical variables were described by absolute and relative frequencies.Univariate and multivariate analyzes were performed.In the univariate analysis, t-student tests were applied for independent samples (quantitative variables of symmetrical distribution), Mann-Whitney (asymmetric quantitative variables) or chi-square supplemented by the analysis of the adjusted residuals (categorical variables).For multivariate analysis, in the control of potential confounding factors, a logistic regression model with backward extraction method, and odds ratio and confidence interval to measure the effect of each factor were used for the variables with a value of p <0.20 in the univariate analysis.A value of p <0.05 was considered to indicate statistical significance.All data were analyzed with SPSS version 21.0 (IBM-SPSS Inc, Armonk, NY).The research was approved by the Research Ethics Committee of the Santa Casa de Misericórdia Hospital of Porto Alegre, under protocol number 2.157.743.
p=0.001), surgical hospitalization (OR = 2,23, 95% CI: 1.08-4.59;p=0,030), hospital stay longer than 30 days (OR = 3,25, 95% CI: 2,01-5.25,p = <0.001),ICU in the last 3 months (OR=2,33,95% CI: 1,31-4,14); p=0.004, antibiotic use in the last 30 days (OR = 2,48, 95% CI: 1,41-4,35; p = 0.002), and presence of tumor wound (OR = 4,11, 95% CI:1.50-11.2,p=0.006), remained as factors significantly associated with CRE colonization.The following antibiotics were found to be statistically significant after adjustment for each periodicos.ufpi.brRev Pre Infec e Saúde.2023;9:3760 type of antibiotic: aminoglycosides (OR = 7.95, 95% control all possible confounding factors.Many studies on the subject have been identified in the literature, but few focused exclusively on CRE colonization, instead of on infection.External validation of oncological patients hospitalized in large tertiary hospitals with high levels of CRE may be useful to evaluate the reproducibility of our results.It is important to highlight that the microorganisms in some of the cases may have acquired the gene responsible for the resistance during the hospitalization period.Unfortunately, our study did not analyze the clonality of the isolates, which could have provided the conditions to identify this subpopulation.

Table 1 -
Distribution of microorganisms and carbapenemase-producing genes among 139 carbapenemase-resistant enterobacteria obtained from rectal cultures from patients at Hospital Santa Rita, January-December 2017.

Table 2 -
Univariate analysis and multivariate Logistic Regression Analysis of factors associated with carbapenem-resistant enterobacteria from patients at Santa Rita Hospital, January-December 2017.

Table 3 -
Case-control comparison of antibiotic use by class from patients at Santa Rita Hospital, January-December 2017.