Editor’s Note: At the recently held Fourth China Hematology Development Conference, the “Chinese Expert Consensus on the Application of Metagenomic Next-Generation Sequencing (mNGS) in the Diagnosis of Infectious Pathogens in Hematological Patients” was officially released. Traditional microbiological tests have low positive rates and are time-consuming, making them insufficient for the diagnostic needs of patients with hematological infections. The emergence of metagenomic next-generation sequencing (mNGS) has optimized pathogen diagnosis and anti-infection treatment. “Infectious Disease Frontline” invited Professor SiZhou Feng , Deputy Director of the Stem Cell Transplantation Center at the Hematology Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, CAMS), to share the advantages of mNGS in the etiological diagnosis of hematological infection patients and the difficulties in initial empirical anti-infection treatment of neutropenia-associated fever, also discussing how to optimize such treatment plans based on multiple studies.
“Infectious Disease Frontline“: What advantages does mNGS have over traditional microbiological testing in diagnosing pathogens in patients with hematological infections? What research achievements have you and your team made in this area?
Professor SiZhou Feng : Hematological patients have high rates of infection and mortality, primarily due to compromised immune function from the disease and treatments like high-dose chemotherapy and hematopoietic stem cell transplantation (HSCT), leading to neutropenia and mucosal barrier damage. These patients are prone to bloodstream infections and pneumonia, posing serious threats to their lives. Traditional microbiological tests often have poor sensitivity, leaving many neutropenia-associated fever cases undiagnosed. Hence, there is an urgent clinical need for more sensitive diagnostic methods.
In recent years, mNGS has been widely used in clinical infections. Its advantages are mainly threefold: First, mNGS is far more sensitive than traditional microbiological tests, including blood and sputum cultures. Second, mNGS is less affected by antimicrobial drugs compared to traditional tests. Third, mNGS can identify rare pathogens that traditional tests often miss, especially in patients with strong pathogenic infections like toxoplasmosis and tuberculosis post-HSCT. Our team, under the leadership of Professor Wang Jianxiang in 2023, conducted a national multicenter prospective study [1], evaluating the diagnostic performance of plasma mNGS in acute leukemia patients with neutropenia-associated fever and its impact on clinical anti-infection treatment. The study showed: (1) mNGS had a significantly higher positive detection rate of clinically relevant infectious pathogens than traditional blood culture (50.9% vs. 21.3%), confirming mNGS’s higher sensitivity; (2) Antimicrobial exposure one day before blood sample collection significantly reduced the positive rate of blood cultures, but had a smaller impact on mNGS; (3) Hematological patients are prone to pneumonia post-chemotherapy and HSCT, making pathogen detection difficult via bronchoalveolar lavage fluid. Plasma mNGS can detect DNA fragments of pathogens released from infection foci, like Aspergillus and Mucorales, serving as a non-invasive method for diagnosing deep infections. These findings are highly instructive for future clinical practice.
Another study by our team assessed the application value of mNGS in infections following allogeneic hematopoietic stem cell transplantation (allo-HSCT). A retrospective analysis of 404 mNGS tests from 264 patients showed: (1) The positive detection rate of mNGS was significantly higher than traditional microbiological tests (82.7% vs. 39.4%); (2) Common infection sites and pathogen types varied at different stages of transplantation (pre-transplant, early post-transplant, late post-transplant). Especially in the late post-transplant stage, mNGS had a 20% higher detection rate for fungi compared to traditional tests (P＜0.01), playing a key role in detecting Aspergillus and Mucorales. This study also confirmed that mNGS is an indispensable diagnostic tool for identifying pathogens and optimizing anti-infection treatment in patients receiving allo-HSCT.
To optimize the indications and reporting interpretation of hematological infection mNGS testing, under the leadership of Professor Liu Qifa, head of the Anti-infection Group of the Hematology Branch of the Chinese Medical Association, we formulated the “Chinese Expert Consensus on the Application of Metagenomic Next-Generation Sequencing in the Diagnosis of Infectious Pathogens in Hematological Patients” [3], published in the Chinese Journal of Hematology, August 2023. The release of this consensus will further standardize the clinical use of mNGS, playing a significant role in the anti-infection treatment of hematological patients.
“Infectious Disease Frontline“: At the conference, you presented an excellent report titled “Choice of Appropriate Antibiotics for Initial Empirical Treatment in Patients with Neutropenia and Fever.” Could you please discuss the initial empirical anti-infection treatment strategies for these patients?
Professor SiZhou Feng : This is a classic topic in the anti-infection treatment field of hematological diseases and a skill that hematologists must master. When hematological patients present with neutropenia and fever, we first need to rule out non-infectious diseases. After exclusion, antibiotic treatment should be considered. Before starting antibiotic treatment, patients should undergo etiological testing, primarily traditional microbiological tests. For patients likely to progress rapidly to severe infections or suspected of central nervous system infections, traditional tests should be performed simultaneously with mNGS, which is vital for guiding subsequent anti-infection treatment. Additionally, patients should undergo risk stratification and assessment for resistant bacteria. Carbapenem-resistant Pseudomonas aeruginosa (CRPA) and carbapenem-resistant Enterobacterales (CRE) are major pathogens causing death in hematological infection patients, along with less common fungal infections like Aspergillus, Mucor, and Candida. Gram-negative bacteria are the main clinical isolates, and infections with resistant strains are a challenging issue. Therefore, before initiating anti-infection treatment, it’s essential to thoroughly assess the patient’s risk factors for resistant bacterial infections, including a history of colonization or infection with resistant pathogens, exposure to broad-spectrum antimicrobials, severe illness, hospital-acquired infection, prolonged and/or repeated hospitalization, indwelling catheters, elderly patients, and ICU patients.
In terms of initial empirical antimicrobial therapy, treatment strategies are primarily divided into step-up and step-down. For patients without complex clinical presentations and at lower risk of resistant bacterial infections, monotherapy with third or fourth-generation cephalosporins or β-lactamase inhibitor combinations is usually adopted, without the need to combine drugs against Gram-positive bacteria, following a step-up strategy. For patients in critical condition, with complex infections, or previous resistant bacterial infections, monotherapy with carbapenems or β-lactamase inhibitor combinations combined with aminoglycosides is typically chosen, following a step-down strategy.
Professor SiZhou Feng : When there is a risk of resistant bacterial infection (e.g., colonization or past infection with resistant bacteria), it is crucial to consider covering these pathogens during empirical treatment. Clinicians can use predictive models for resistant bacterial infections to aid decision-making. For instance, our research team previously developed a simple model [4] to predict bloodstream infections (BSI) caused by multidrug-resistant Pseudomonas aeruginosa (MDR-PA). In this model, previous use of carbapenems (1 point), piperacillin/tazobactam (1 point), quinolones (1 point), and BSI during antibiotic treatment (4 points) were independent risk factors for MDR-PA BSI. Patients with a score ≥6 were defined as high-risk, with a positive predictive value (PPV) of 86.7% and a negative predictive value (NPV) of 89.1%. These patients often show resistance to carbapenems, and treatment with β-lactamase inhibitor combinations or fourth-generation cephalosporins combined with aminoglycosides can be effective.
Additionally, Professor Jiang Erlie’s team at our hospital conducted a study [5] on a predictive model for BSI risk after colonization with carbapenem-resistant Enterobacteriaceae (CRE) in hematological patients. The study identified age, rectal swab strain minimum inhibitory concentration (MIC) for meropenem and imipenem >8 μg/mL, gastrointestinal symptoms, absolute neutrophil count nadir <0.025×109/L, and chills at peak body temperature as independently related to CRE BSI within 30 days. By stratifying patients based on the best cutoff value (0.215) from the ROC curve, high-risk patients had a significantly higher incidence of CRE BSI than low-risk patients (47.5% vs. 3.7%, P＜0.001). By stratifying risk, high-risk patients can be treated with ceftazidime/avibactam or polymyxin to reduce CRE BSI mortality, improve prognosis, and avoid overuse of antibiotics.
“Infectious Disease Frontline“: Guidelines emphasize that for high-risk neutropenia-associated fever patients, initial empirical treatment should cover Pseudomonas aeruginosa and other severe Gram-negative bacteria. Can you discuss the challenges in diagnosing and treating Pseudomonas aeruginosa infections?
Professor SiZhou Feng : This is a crucial issue for hematologists. The need to cover Pseudomonas aeruginosa (PA) in initial empirical treatment arises from its high infection mortality rate, especially in bloodstream infections. Studies show that neutropenic patients with PA infections have a mortality rate of about 20%-40%. Our research on PA BSI patients in hematological hospitals found that quinolones, carbapenems, and piperacillin/tazobactam contribute to PA resistance. Thus, it’s important to avoid creating resistance through irrational antimicrobial use. In a survey [6] on acute leukemia patients with PA BSI, we found a total 30-day mortality rate of 8.5%. Our low mortality rate is partly due to 81.2% of patients receiving appropriate empirical treatment within 48 hours of BSI onset, and our policy of not using quinolones for prophylaxis in neutropenic hematological patients. It’s advised not to use quinolones for bacterial infection prophylaxis in leukemia chemotherapy and HSCT patients, as prophylactic antibiotics may reduce bacteremia incidence but not mortality and may lead to resistant bacteria, especially resistant PA. Due to PA’s ability to form biofilms, increasing treatment difficulty and patient mortality risk, stratified treatment is necessary. Based on clinical experience, some patients may be treated with monotherapy for PA infections, but critical patients require combination therapy to overcome the high mortality of PA infections, particularly resistant PA.
“Infectious Disease Frontline“: How should carbapenems be used appropriately in neutropenia-associated fever patients, given their importance in antimicrobial stewardship (AMS)?
Professor SiZhou Feng : Due to the high incidence and mortality of infections in hematological neutropenic patients, carbapenems are a common choice for hematologists. Especially for neutropenia-associated fever patients, carbapenems often outperform third or fourth-generation cephalosporins and β-lactamase inhibitor combinations. However, as mentioned earlier, for patients with less severe conditions and lower risk of resistant bacterial infection, third or fourth-generation cephalosporins and β-lactamase inhibitor combinations should be the first choice for initial empirical treatment, reserving carbapenems as second-line drugs to control their use rationally.
Moreover, our team published a multicenter retrospective study [7] in Clinical Infectious Diseases in 2023, analyzing the rational use of antimicrobials, including carbapenems, and how to reduce resistance in neutropenia-associated fever patients with PA BSI. Previously, it was commonly believed that PA BSI required at least 14 days of treatment. However, our study found no significant difference between the short-course (7-11 days) and long-course (12-21 days) groups in terms of fever recurrence, relapse of PA infection, or death within 30 days, or PA reinfection within 90 days after stopping antimicrobials. This indicates that short-course patients have similar clinical outcomes as those in the long-course group. Therefore, for PA BSI, if clinical symptoms and signs disappear and blood cultures turn negative, there is no need to extend treatment to 14 days. Treating for 7-11 days and deciding whether to stop based on culture results not only shortens antimicrobial use and hospital stay but also helps reduce infections with resistant bacteria, particularly CRPA.
Professor SiZhou Feng is a distinguished medical professional associated with the Hematology Hospital of the Chinese Academy of Medical Sciences (Institute of Hematology, CAMS). Holding a Doctor of Medicine degree, he serves as a chief physician, second-grade professor, and a mentor for doctoral and postdoctoral research students. Recognized as one of the foremost medical experts in Tianjin, he has served as a delegate in the 17th and 18th Tianjin People’s Congress.
From July 1999 to July 2001, Professor Feng conducted postdoctoral research at the University of Tokyo, Japan. He currently holds several notable positions including the Deputy Director of the Stem Cell Transplantation Center at the Hematology Hospital of the Chinese Academy of Medical Sciences, Vice Chairman of the Pharmacovigilance Committee of the Chinese Pharmacological Society, and Vice Chairman of the Blood Disease Precision Diagnosis and Treatment Professional Committee of the China Research Hospital Society.
Additionally, Professor Feng is the Vice Chairman of the Hematopoietic Stem Cell Transplantation Committee of the China Primary Health Care Foundation, the Deputy Head of the Anti-infection Group of the Hematology Branch of the Chinese Medical Association, a member of the Bacterial Infection and Resistance Prevention and Control Branch of the Chinese Medical Association, a standing committee member of the Hematology Branch of the Chinese Geriatrics Society, and a standing committee member of the Anti-Leukemia Alliance Expert Committee of the Chinese Society of Clinical Oncology.
He also serves as a committee member of the Tumor Hematology Professional Committee of the Chinese Anti-Cancer Association and the Chairman of the Geriatric Oncology Professional Committee of the Tianjin Anti-Cancer Association. As an editorial board member of several journals, including Chinese Organ Transplantation, Chinese Journal of Infection and Chemotherapy, Clinical Hematology, Clinical Internal Medicine, and Internal Medicine Emergency and Critical Care, Professor Feng has made significant contributions to medical literature.
He has published over 40 SCI papers as a corresponding and first author in esteemed journals such as J Infection, Clinical Infectious Diseases, Clinical Microbiology and Infection, and Br J Haematol. His achievements include receiving 5 provincial and ministerial-level scientific and technological progress awards, the Tianjin May Day Labor Medal, and being honored as a doctor who has earned the satisfaction and respect of the people of Tianjin.
