Clinical and epidemiological characteristics of carbapenem-resistant Klebsiella pneumoniae infections in a tertiary hospital in China

Background(cid:0)The infections due to carbapenem-resistant Klebsiella pneumonia (CR-KP) have become an important problem. The aim of the study is to evaluate the clinical and epidemiological characteristics of CR-KP. Results: The CR-KP infections overall mortality was 37.3%, and bloodstream infections mortality was 66.2%. Survival analysis revealed that there were statistically signicant differences between bloodstream infection and pulmonary and drainage uid infection. Hemopathy, age (>60 years), tumors, diabetes, septic shock, acute kidney injury and stroke were independent predictors associated with the 30-day mortality. Multivariate linear regression showed that survival time was negatively correlated with APACHE II score and SOFA score, while positively correlated with LYM. Chi-square test showed that antimicrobial regimen combined carbapenems, tigecycline with polymyxin B was superior the one combined carbapenems with polymyxin B. But there was not statistically signicant difference between carbapenems plus tigecycline and carbapenems plus polymyxin B. Ceftazidime avibactam-based antimicrobial regimens also had no advantage over other therapeutic regimens. Conclusions: Our study conrmed there is a high mortality rate in CR-KP infections, especially in the bloodstream infections. The outcome is greatly inuenced by the patients’ clinical conditions. Antimicrobial regimen combined carbapenems, tigecycline with polymyxin B might be a better choice.

The rst CR-KP was identi ed in 1993, and since then the CR-KP strains are endemic in many countries [8]. The Annual Report of the European Antibiotic Surveillance Network was published in 2016, reporting a mean percentage of carbapenems resistance equal to 6.1%. In 2018 in China, Klebsiella pneumonia had a resistance rate of 15.4% to imipenem and 17.9% to meropenem [9,10].
Although several studies have demonstrated the e cacy of combination regimens in terms of decreased mortality, an effective treatment is still a challenge for clinicians [11][12][13][14]. The outcome measured was death within 30 days since the rst positive culture. Survivor and nonsurvivor subgroups have been compared to identify predictors of mortality.

Antimicrobial therapies
The attending physician determined the appropriate initial antimicrobial therapy. Antimicrobial therapy was administered within 5days after infection onset was de ned as empirical therapy. Appropriate de nitive antibiotic therapy was de ned as that matching the in vitro susceptibility results according to the Clinical and laboratory Standards Institute (CLSI) criteria.

De nitions
Septic shock is de ned as sepsis plus persistent hypotension, requiring vasopressor to maintain mean arterial pressure (MAP) ≥65 mmHg, and serum lactate levels > 2 mmol/L despite adequate uid resuscitation.
Samples of microbial cultures were routinely collected when patients got fever and there was evidence of clinical suspicion or infection. Bloodstream infection was de ned as hospital-acquired if the index blood culture had been collected > 48 h after hospital admission.

Statistical analysis
Statistical analysis was performed by using SPSS 22.0 (IBM, Armonk, USA). Categorical variables were recorded as absolute numbers and their relative frequencies, and they were compared by the χ 2 or Fisher exact test. Categorical variables were analyzed by multivariate logistic regression analysis to identify independent risk factors for mortality. The empirical and de nitive therapies for patients with CR-KP in the survivor and death group are summarized in Table 1. The most frequently used antibiotics for both empirical and de nitive therapy were carbapenems in two groups. Then, we did a statistical analysis for the different antibiotic therapeutic regimens and the results were listed in Table 2. Chi-square test showed that both antibiotic regimen B (carbapenems plus tigecycline and polymyxin B, χ 2 =3.96, p=0.04) and antibiotic regimen A (ceftazidime avibactam plus tigecycline, χ 2 =4.52, p=0.033) were superior to antibiotic regimen C (carbapenems combined with polymyxin B) and the differences were of statistical signi cance. However, there was no statistically signi cant difference among antibiotic regimen D (carbapenems plus tigecycline), antibiotic regimen C (carbapenems plus polymyxin B) and regimen E (carbapenems plus tigecycline plus fosfomycin). It seemed that polymyxin B and tigecycline have synergistic effect and carbapenems plus polymyxin B was not a recommended therapeutic regimen ( Figure 1). Ceftazidime avibactam-based antimicrobial regimens also had no advantage over other therapeutic regimens.
In addition, we investigated the e cacy of tigecycline combination with other different antibiotics in the treatment of CR-KP.
Chi-square test showed that there was no statistically signi cant difference among different antibiotic regimens. Considering the price of antibiotics, we recommend biapenem combined with tigecycline as a therapeutic regimen.

Prognostic factors
In the multivariate logistic regression analysis, hemopathy, age (>60 years), solid tumors, diabetes, septic shock, acute kidney injury and stroke were independent predictors associated with the 30-day mortality ( Table 3). Correlation analysis and ROC curves Multivariate linear regression was performed in APACHE II score, SOFA score, lymphocyte absolute value (LYM) and survival time. Survival time was negatively correlated with APACHE II score and SOFA score, and positively correlated with LYM (Figure 2 and 3). In addition, ROC curves were also drawn for APACHE II score, SOFA score and LYM, with AUC of 0.825, 0.876 and 0.797, and with cut-off value of 17, 6 and 0.775 respectively.

Survival analysis
Survival analysis showed that pulmonary and drainage uid infections were statistically signi cance differences from bloodstream infection with CR-KP (Figure 4).

Discussion
In this study we evaluated 30-day mortality in patients with CR-KP infection. The total mortality of patients with CR-KP infection was 37.3%, while with bloodstream infection mortality was 66.2% [2,3,5]. It was patients' characteristics, antibiotic therapeutic regimens as well as immune status of the patients made mortality so high.
The relatively higher mortality compared with previous studies might be attributable to several factors. First, our hospital is a tertiary referral hospital, and many patients had been very serious when they were admitted to our hospital. The states of the patients were directly related to mortality and had repeatedly reported in previous studies [4,5]. Second, a previous study has reported that administering the appropriated empirical antibiotic therapy and de nitive therapy was an important predictor of patients' outcome [6, 17,18]. Therefore, it is very important to take the antibiotics reasonably before patients transferred to ICU. Clinical departments, especially surgery, should pay more attention to the rational use of antibiotics. Finally, it is important to detect pathogen and adjust the therapeutic regimen to treat CR-KP infection [19,24]. But the positive rate of bacterial culture was low. The positive rate of rectal swabs for the carbapenem-resistant enteric bacteria was only 30.2% in our study. In recent years, Next Generation Sequencing, which had emerged to help to diagnose CR-KP infection, demand a further study [20].
Independent risk factors associated with mortality were hemopathy, age (>60 years), solid tumors, diabetes, septic shock, acute kidney injury and stroke. Correlation analysis showed that survival time was positively correlated with LYM. Lymphocytes were involved in the patient's innate immunity and adaptive immunity, and low immune status was directly related to patient prognosis [21][22][23].
We investigated the e cacy of tigecycline combination with different antibiotics (meropenem, imipenem, biapenem, ceftazidime avibactam, and 3rd or 4th ephalosporins) in the treatment of CR-KP, and the differences had not statistically signi cant. Considering the price of antibiotics, we recommend biapenem combined with tigecycline as a therapeutic regimen. Tigecycline combination with meropenem or ceftazidime avibactam had a higher survival rate. Because there were more neurosurgical patients, meropenem was used more frequently.
Chi-square test showed that both antibiotic regimen 1 (carbapenems plus tigecycline and polymyxin B) and antibiotic regimen 2 (ceftazidime avibactam plus tigecycline) were superior to antibiotic regimen 3 (carbapenems combined with polymyxin B). The reasons may be that polymyxin B was a narrow-spectrum antibiotic, and some patients often had multiple drug-resistant bacterial infections. Beside, polymyxin B and tigecycline may have synergistic effect in treating CR-KP infection. Finally, ceftazidime avibactam-based antimicrobial regimens also had no advantage over other therapeutic regimens, but the sample size is too small and it needs further study.
There are some limits in our study. The size of some samples is too small to allow us to detect subtle differences in treatment outcome. Since our data come from a single center, a multi-centers prospective study with more extensive collection of potential confounders is required.

Conclusion
Our study con rmed a high mortality rate of CR-KP, especially in the bloodstream infections. The outcome is heavily in uenced by the patients' clinical conditions. Carbapenems, tigecycline and polymyxin B combined antimicrobial regimen might be a better choice. Ceftazidime avibactam-based antimicrobial regimens had no advantage over other therapeutic regimens. Additional data needs to further elucidate this nding especially in light of the introduction of the new agents.

Declarations
Ethics approval and consent to participate: The study was approved by the Human Ethics Committee of the First A liated Hospital of Zhengzhou University. Written informed consents were obtained from all patients.
Consent for publication: All authors consent for publication.
Availability of data and materials: The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: Not applicable.
Authors' contributions: ZW C collected data and wrote the initial draft of the manuscript. LR W searched the literatures. W C also collected data and expanded the discussion. MH L was responsible for design and revision of the manuscript. YX L and M F revised the manuscript critically for intellectual content. All authors read and approved the nal manuscript.