The increase in infections caused by CRE is a great challenge for patients with hematologic malignancies. Owing to immunocompromised status, prolonged hospitalizations, frequent antimicrobial use and neutropenia, those patients have a higher risk for CRE infection and treatment failure. The GITMO performed a retrospective study based on data from 52 stem cell transplant centers and demonstrated a CRE-related mortality rate of 64.4% in allogeneic HSCT recipients [6]. Our present study described the clinical outcomes and microbiological characteristics of 94 hematological patients with BSI due to CRE from 2012 to 2021. The result was encouraging, with 30-day mortality rate was 28.7% for all 94 patients, and was only 7.1% for 28 patients treated with CAZ-AVI.

Studies in the literature have indicated that timely and appropriate administration of empirical therapy is essential for managing CRE BSI [7,8,9]. Falcone M et al. evaluated risk factors for mortality in 102 patients with KPC-producing K. pneumoniae bacteremia and found median time to appropriate antibiotic therapy was shorter in patients who survived (8.5 h) versus those who died (48 h). Receipt of an in vitro active therapy within 24 h was independently associated with lower 30-day mortality (HR = 0.36, p = 0.0021) [9]. In the present study, we also observed a lower percentage of mortality among cases who received appropriate empirical therapy within 24 h of BSI episode (16.3% vs. 42.2%, p = 0.006). Further analysis, patients with MDRO colonization or infection in the previous 3 months tended more to receive active therapy within 24 h of infection. Therefore, close monitoring of CRE colonization is essential for initiating timely therapy once upon the onset of CRE infection.

Prior to the introduction of CAZ-AVI, the treatment of CRE infections was based on limited last-resort agents, such as tigecycline, polymyxins, fosfomycin and aminoglycosides. Though the optimal treatment for CRE BSI has not been well established, combination therapy with at least two active agents has been recommended in most published studies [10,11,12]. In addition, some studies suggested that combination therapy rather than monotherapy may be more beneficial for patients with septic shock or a high mortality score [13, 14]. A multicenter retrospective study investigated the effect of appropriate therapy on mortality of 437 patients with BSIs due to carbapenemase-producing Enterobacteriaceae. Though overall mortality was not different between those receiving combination therapy or monotherapy, lower mortality had been associated with combination therapy among patients with high-mortality-score stratum (48% vs. 62%, p = 0.020) [13]. Analyzing clinical outcomes of 66 patients treated with OAAs in our study, we also observed a trend towards decreased 30-day mortality in patients receiving combination therapy compared with monotherapy (29.3% vs. 52%, p = 0.065). Therefore, combination therapy is recommended for CRE BSI, especially for critically ill patients.

Ceftazidime-avibactam was the first new antibiotic approved by the US Food and Drug Administration (FDA) in 2015 for the treatment of CRE infections. Avibactam, a non-β-lactam β-lactamase inhibitor, has activity against Ambler class A and certain class D carbapenemases but not against class B metallo-lactamases [15]. Most studies published [8, 16,17,18,19] have demonstrated the survival benefit of CAZ-AVI in treatment of CRE infection, including higher clinical success, decreased mortality, and lower toxicity. Tumbarello M et al. [17] conducted the largest studies to date evaluating the efficacy of CAZ-AVI for KPC-K. pneumoniae infections. A total of 577 patients were included, and they received treatment with CAZ-AVI alone or with ≥ 1 other active antimicrobials. The all-cause mortality rate 30 days after infection onset was 25%, which was significantly lower than rates achieved with earlier non-CAZ-AVI-based drug regimens. Shields RK et al. [19] also indicated superiority of ceftazidime-avibactam to other treatment regimens against CRKP bacteremia. Moreover, nephrotoxicity was significantly less likely with ceftazidime-avibactam than with colistin- or aminoglycoside-containing regimens. Against MBLs, the combination of CAZ-AVI and ATM represents a promising treatment option. ATM remains hydrolytic activity against MBLs, however, it cannot be used alone due to frequent co-production of other enzymes (e.g., ESBLs, OXA-48) by MBL-producing Enterobacteriaceae [20]. Emerging data [21,22,23] has supported the combination of CAZ-AVI and ATM for MBL producers. A recent prospective study [21] evaluated the efficacy of CAZ-AVI plus ATM in patients with BSIs due to MBL-producing Enterobacteriaceae. Overall, 52 patients received a combination therapy of CAZ-AVI + ATM, whereas 50 were treated with OAAs. The 30-day mortality rate was significantly lower in the CAZ-AVI + ATM group than in the OAAs group (19.2% vs. 44%, P = 0.007). In our study, a total of 28 patients received an antimicrobial treatment containing CAZ-AVI. Corroborating previous studies, we also observed a remarkable decreased mortality for patients treated with CAZ-AVI than with OAAs (7.1% vs. 37.9%, p = 0.003). In multivariate analysis, the administration of CAZ-AVI was the independent predictor for favorable outcome. With regard to the obvious survival benefit, we considered CAZ-AVI as the preferred treatment option for CRE BSI.

As the distribution of carbapenemases varies in nations, regions and even centers, knowledge of the prevalence and molecular characteristics of CRE is of vital importance. KPC-producing Enterobacteriaceae are widespread in the United States, Latin America, Italy and Greece [24]. NDM-producing Enterobacteriaceae are mainly detected in Indian subcontinent [25], while OXA-48-producing Enterobacteriaceae are endemic in Turkey [26]. In China, blaKPC and blaNDM are the most common carbapenemase genes among CRE strains. Data from a longitudinal large-scale CRE study in China (2012–2016) proved that KPC and NDM are the major carbapenemases produced by CRE, while KPC was predominant in K. pneumoniae (77%), NDM was predominant in E. coli (75%) and E. cloacae (53%) [27]. Another study [28] collected 935 non-duplicate CRE strains from 36 hospitals across China, and reached similar conclusions. In our study, NDM (66.7%, 36/54) was the most common carbapenemase gene, followed by KPC (29.6%, 16/54). Different from other centers in China, all 25 carbapenemase-producing E. coli in our study were identified NDM gene. Besides, NDM also accounted for a large proportion (37.5%, 9/24) in carbapenemase-producing K. pneumoniae. Considering the predominance of NDM gene in our center, the majority (75%, 21/28) of patients in CAZ-AVI group received the combination with ATM, which contributed to the decreased mortality and higher clinical success. Therefore, detection of Carbapenemase genes was crucial as it could guide the rational choices of antibiotics.

The present study had several limitations. First, it was a single-center retrospective study, designed exclusively for hematological patients. The results from our study may not be representative of the experience at other centers. Second, considering the heterogeneous treatment regimens in the study, we did not perform further comparison for mortality between different antibiotics or combinations. Third, though we observed a distinct advantage of CAZ-AVI therapy for CRE BSI, the number of patients in the CAZ-AVI group were relatively small. More large-scale studies are warranted.

In conclusion, this study indicates the severity and poor prognosis of CRE BSI in patients with hematologic malignancies. The initiation of appropriate empirical therapy within 24 h of BSI onset is crucial for managing CRE BSI. We confirm the superiority of CAZ-AVI to other treatment regimens against CRE BSI. Detailed knowledge of the prevalence trends and distribution of carbapenemase-producers is quite essential. Unlike previous reports in China, we found a larger proportion of blaNDM in carbapenemase-producing Enterobacteriaceae in our center. Taking this into account, we recommend the combination of CAZ-AVI with ATM when choose CAZ-AVI-containing regimen.