Patients undergoing hematopoietic stem cell transplantation and high-dose chemotherapy for hematological diseases have impaired immune function, making them highly susceptible to febrile neutropenia and colonization or infection with resistant bacteria. Prompt and accurate initiation of empirical therapy is crucial to improve patient outcomes. At the 2024 European Hematology Association (EHA) Annual Meeting, Dr. Sizhou Feng's team from the Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences, reported a study showing a high proportion of bloodstream infections caused by carbapenem-resistant Acinetobacter spp. (CRA) in hematology patients, with homology between bloodstream isolates and colonizing bacteria. This indicates the need to strengthen hospital infection control to reduce resistant bacterial colonization and to promptly initiate appropriate empirical antimicrobial therapy for suspected infections in patients colonized with resistant bacteria. "Oncology Frontier - Hematology Frontier" invited Dr. Sizhou Feng to introduce and interpret this research at the EHA meeting.Oncology Frontier – Hematology Frontier: How common is the colonization of resistant bacteria like CRO in hematology patients? What are the potential impacts of inappropriate empirical antimicrobial therapy?
Dr. Sizhou Feng: With the increasing number of new antitumor therapies, particularly targeted therapy, immunotherapy, and hematopoietic stem cell transplantation, the prognosis of hematology patients has improved. However, they still face significant challenges with infections, especially those with severe immune impairment and febrile neutropenia, where resistant bacterial infections cannot be ignored. Enhancing the efficacy of antimicrobial therapy is crucial to reducing patient mortality.
Resistant bacterial infections are quite common in hematology patients, and colonization with resistant bacteria is also prevalent. Domestic studies show that the colonization rate of carbapenem-resistant Gram-negative bacteria (CR-GNB/CRO) in hematology patients is 10.8%[1], including carbapenem-resistant Pseudomonas aeruginosa (CRPA), Klebsiella pneumoniae (CRKP), Escherichia coli (CREC), and other Enterobacteriaceae (CRE), which are common pathogens causing infections or colonization in these patients. There is a correlation between colonizing bacteria and bloodstream infection pathogens, and accurately distinguishing between colonizing and infecting bacteria is key to determining the appropriateness of initial empirical therapy. During the initial empirical treatment phase, the causative pathogen and its drug sensitivity are unknown. Appropriate selection of antimicrobial drugs can significantly improve patient survival rates; conversely, inappropriate initial empirical therapy can result in a mortality rate of up to 50%[2]. Therefore, for patients undergoing hematopoietic stem cell transplantation or high-dose chemotherapy, necessary screening for resistant bacterial colonization before treatment is essential to guide the selection of empirical antimicrobial agents.
Oncology Frontier – Hematology Frontier: Your team’s study at this conference analyzed the genetic homology between CRA colonization and bloodstream infection and its impact on empirical therapy. Could you share these research findings and their clinical significance?
Dr. Sizhou Feng: As mentioned, it is essential to screen for resistant bacterial colonization, including oral and perianal swabs, before high-dose chemotherapy and hematopoietic stem cell transplantation. However, whether colonizing bacteria are homologous to pathogenic bacteria, i.e., how many pathogenic bacteria in immunocompromised, febrile neutropenic hematology patients originate from colonizing bacteria, is a clinical question that needs investigation.
In our study presented at the EHA conference[2], we analyzed 46 hematology patients, with 39.1% (n=18) having prior colonization with Acinetobacter spp. Among colonized patients, the proportion of CRA bloodstream infections (BSI) was significantly higher than in non-colonized patients (72.2% vs. 17.9%, P<0.001), with higher levels of inflammatory markers and a 30-day mortality rate (78.6% vs. 50%). Notably, we found that in eight patients with matched respiratory colonizing strains and bloodstream infection CRA respiratory isolates, next-generation sequencing (NGS) showed genomic homology. Based on multilocus sequence typing (MLST) and single nucleotide polymorphism (SNP) phylogenetic trees, the CRA bloodstream infections in these patients originated from respiratory colonizing strains. These results indicate that CRA bloodstream infections in hematology patients are highly homologous to colonizing bacteria. Clinically, once such colonizing bacteria are screened, and febrile neutropenia occurs, CRA infection should be suspected, and empirical antimicrobial therapy should be guided by the colonizing bacteria’s drug sensitivity results. Additionally, hospital infection control measures should be strengthened to reduce resistant bacterial colonization in hematology patients.
These findings have been published in Infect Drug Resist[3]. Our team also has an unpublished study further exploring whether resistant bacterial colonization in hematology patients is due to previous irrational antimicrobial use or host factors. In a large cohort of nearly 2000 hematology patients over 10 years, we found that the rate of resistant bacterial colonization was significantly higher in patients with relapsed/refractory acute leukemia compared to those in remission, indicating that immune function status is also related to resistant bacterial colonization. There were differences in gut microbiota and CRE colonization rates between these two groups.
Oncology Frontier – Hematology Frontier: In the absence of precise NGS sequencing, what clinical, infection markers, or pathogen characteristics do you use to analyze CRO colonization and infection and their treatment strategies?
Dr. Sizhou Feng: When NGS detects CRO colonization or infection, there is a high likelihood of bloodstream infection. In the absence of genetic testing methods, individualized assessments are based on various clinical characteristics of the patient. Our team previously established a prediction model for multidrug-resistant Pseudomonas aeruginosa (MDR-PA) infection[4]. We found that previous exposure to antimicrobial agents is an independent risk factor for MDR-PA infection, particularly prolonged use of quinolones (8.0 vs. 4.0 days, P=0.010) and carbapenems (12.5 vs. 8.0 days, P=0.006). We developed a rapid scoring prediction model based on several simple clinical factors, with a positive predictive rate of 86.7% for high-risk patients with a score ≥6.
In clinical practice, for patients with severe mucositis, high fever, chills, or prolonged antimicrobial exposure, if resistant colonizing bacteria are detected before hematopoietic stem cell transplantation or high-dose chemotherapy, there is a high risk of bloodstream infection caused by resistant colonizing bacteria. Before microbiological evidence is clear, empirical antimicrobial therapy can be guided by the drug sensitivity of colonizing bacteria to improve the accuracy of initial empirical therapy.
References:
[1] Hu H, Wang Y, Sun J, et al. Risk factors and molecular epidemiology of intestinal colonization by carbapenem-resistant Gram-negative bacteria in patients with hematological diseases: a multicenter case‒control study. Microbiol Spectr. Published online June 7, 2024. doi:10.1128/spectrum.04299-23
[2] Li J, et al. Strong Homology Between Colonizing and Bloodstream Carbapenem-Resistant Acinetobacter Spp.: Implications for Empiric Antibiotic Therapy in Hematological Patients. EHA 2024. Abstract P1577.
[3] Li J, Guo W, Wang J, et al. Strong Homology Between Colonizing and Bloodstream Carbapenem-Resistant Acinetobacter Spp.: Implications for Empiric Antibiotic Therapy in Hematological Patients. Infect Drug Resist. 2024;17:1827-1838. Published 2024 May 9. doi:10.2147/IDR.S458427
[4] Zhao Y, Lin Q, Zhang T, et al. Pseudomonas aeruginosa bloodstream infection in patients with hematological diseases: Clinical outcomes and prediction model of multidrug-resistant infections. J Infect. 2023;86(1):66-117. doi:10.1016/j.jinf.2022.08.037
