Editorial Note: Neutropenia, immunosuppressive treatments, and various other factors make patients with hematological disorders particularly susceptible to infections from diverse pathogens. These infections often progress rapidly, with high mortality rates, posing a severe threat to patient survival. Consequently, effectively preventing and managing infections in this patient population remains a critical challenge in hematology. Recently, Dr. Sizhou Feng's team from the Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences, published two significant studies. One examines the clinical characteristics and 28-day mortality risk factors of Stenotrophomonas maltophilia (SM) bloodstream infections in adult hematological patients. The other study presents a clinical and imaging-based model to improve the differential diagnosis between central nervous system leukemia (CNSL) and central nervous system infections (CNSI). Both studies offer valuable insights for clinical management and prognosis.1. Clinical Characteristics and Risk Factors for 28-Day Mortality in Adult Hematological Patients with Stenotrophomonas maltophilia Bloodstream Infections
Stenotrophomonas maltophilia (SM) is a Gram-negative, non-fermentative bacterium known for its ability to adhere to respiratory epithelial cells and invasive medical devices. This adhesion promotes biofilm formation, enhancing colonization and resistance in the host. SM harbors multiple antimicrobial resistance genes, exhibiting intrinsic resistance to most β-lactam antibiotics, including carbapenems and cephalosporins. Moreover, it employs enzymatic mechanisms and efflux pumps to resist aminoglycosides and fluoroquinolones. Hematological patients, often hospitalized long-term, with central venous catheters, compromised immune systems, and altered microbiomes, are particularly vulnerable to SM infections, which are associated with high severity and mortality. However, research on SM bloodstream infections in hematological patients is limited, and optimal treatment strategies remain undefined.
This retrospective study analyzed 140 adult hematological patients diagnosed with SM bloodstream infections from January 2012 to July 2023. The overall 28-day mortality rate was 31.43% (44/140), with a median patient age of 44 years. The median hospital stay before the onset of SM bloodstream infections was 25 days, and 69.29% of patients were neutropenic at the time of infection.
All patients had received broad-spectrum antibiotics in the previous month, with 69.29% developing breakthrough bloodstream infections after more than 48 hours of carbapenem treatment. Compared to survivors, patients who died had longer durations of broad-spectrum antibiotic use (22 vs. 16 days, P < 0.001). The deceased group also had a higher frequency of fluoroquinolone (27.27% vs. 9.38%, P = 0.013) and tigecycline (31.82% vs. 6.25%, P < 0.001) exposure.
Independent risk factors for mortality included a Sequential Organ Failure Assessment (SOFA) score ≥ 5, tigecycline exposure, age ≥ 60, and concurrent pulmonary infections.
High-risk patients with ≥2 risk factors had significantly higher 28-day mortality rates than low-risk patients (56.52% vs. 7.04%, P < 0.001). Further evaluation showed that high-risk patients treated with trimethoprim-sulfamethoxazole (TMP-SMX) alone (P = 0.008) or in combination with cefoperazone-sulbactam (CSL) (P = 0.005) had better outcomes. In contrast, CSL monotherapy did not affect survival (P = 0.372).
Among tested antibiotics, minocycline exhibited the strongest antimicrobial activity against SM isolates, followed by levofloxacin, TMP-SMX, and CSL. Survivors demonstrated greater sensitivity to these antimicrobials.
In conclusion, SM bloodstream infections are frequently observed in hematological patients with prolonged hospital stays, persistent neutropenia, and extensive broad-spectrum antibiotic use, particularly carbapenems. Mortality risk is higher in patients with elevated SOFA scores, older age, pulmonary infections, or recent tigecycline exposure. TMP-SMX is the preferred treatment, with TMP-SMX plus CSL offering a viable alternative.
2. A Clinical and MRI-Based Nomogram for Differentiating CNS Infections and CNS Leukemia in Hematological Malignancies
Central nervous system leukemia (CNSL) and central nervous system infections (CNSI) are severe complications in acute leukemia (AL) patients. Their overlapping clinical and imaging characteristics make differentiation challenging. Accurate diagnosis is critical for prognosis, necessitating prompt and precise diagnostic and therapeutic interventions. CNSI diagnosis often relies on neurological symptoms, imaging, cerebrospinal fluid (CSF) microbiological analysis, and biopsies. Meanwhile, CNSL diagnosis depends on neurological symptoms, imaging, and CSF cytology. This retrospective study aimed to develop a clinical and MRI-based model to enhance the differentiation between CNSL and CNSI.
The study included 149 suspected CNSL or CNSI cases from January 2020 to December 2023, analyzing clinical data, laboratory findings, CSF cytology, and microbiology results. MRI scans for all patients were independently reviewed by two neuroradiology experts, with a senior radiologist resolving discrepancies. Univariate and multivariate regression analyses identified independent predictors for distinguishing CNSL from CNSI.
Among the 108 patients included in the final analysis (65 CNSL and 43 CNSI cases), significant differences were noted between the groups in hyperleukocytosis, treatment methods, bone marrow remission status, fever, seizures, altered consciousness, asymptomatic presentations, lymphopenia, and co-infections (P < 0.05 for all).
MRI findings were classified into five types: parenchymal, meningeal, cranial nerve, mixed, and normal. Abnormal MRI findings were observed in 40 CNSL and 30 CNSI cases, with significant differences in MRI patterns and diffusion-weighted imaging (DWI) hyperintensity between the groups.
Univariate analysis identified non-remission bone marrow status, allogeneic hematopoietic stem cell transplantation (allo-HSCT), fever, nausea/vomiting, co-infections, parenchymal MRI patterns, and DWI hyperintensity as independent predictors. Multivariate analysis confirmed hyperleukocytosis and non-remission bone marrow as independent predictors for CNSL, while fever, altered consciousness, co-infections, and parenchymal MRI patterns were independent predictors for CNSI.
Based on these predictors, the team developed a nomogram to estimate the likelihood of CNSL or CNSI. The model’s calibration was assessed using the Hosmer-Lemeshow test, and its discriminative ability was evaluated via the area under the ROC curve (AUC = 0.947, 95% CI: 0.9105–0.984), indicating excellent performance. Calibration results showed strong agreement between predicted and actual probabilities.
These studies represent significant advancements in understanding infection risks and improving differential diagnosis strategies for hematological patients. Dr. Sizhou Feng’s team continues to provide invaluable contributions to the field, improving patient care and outcomes.
