Volume 08 · Issue 03 · 2026

In patients with hematological disorders, the high risk of complex infections caused by immune dysfunction and intensive therapies poses a major challenge to the use of conventional microbiological tests (CMTs). Plasma cell-free DNA (cfDNA) metagenomic next-generation sequencing (mNGS) has emerged as a revolutionary noninvasive tool that enables unbiased, broad-spectrum, and rapid pathogen identification directly from blood samples. This review summarizes the core applications of plasma cfDNA mNGS in patients with hematological disorders, including the diagnosis of febrile neutropenia, bloodstream infections, focal infections, and infections caused by uncommon/fastidious pathogens. It highlights the advantages of this technology in overcoming antibiotic interference, enabling early detection, and providing diagnostic value in cases without clear infection foci or when invasive sampling is not feasible. This review further discusses how China has facilitated the widespread adoption of this technology through a localized application model, cost reduction, and the development of clinically relevant interpretation models. Nevertheless, challenges remain, such as lower sensitivity than site-specific specimens in focal infections, and the difficulty in predicting antimicrobial resistance (AMR) on the basis of cfDNA mNGS. Future developmental directions should focus on technical optimization (eg, combined plasma cell-fraction testing), quality assurance and quality control management, multidimensional data integration (eg, host immune response analysis), artificial intelligence (AI)-assisted interpretation, and cost reduction through technology popularization and insurance coverage. These efforts will advance cfDNA mNGS from a pathogen detection tool toward an intelligent clinical decision-support platform, ultimately improving the diagnostic accuracy and clinical outcomes of hematological patients with infections.

Platelets are traditionally considered a homogeneous population of anucleate cells derived from megakaryocytes, primarily responsible for hemostasis and thrombus formation. However, advances in platelet phenotyping and functional assays have revealed substantial heterogeneity within the platelet population, identifying distinct subpopulations with specialized functions. These subpopulations, including reticulated, procoagulant, aggregatory, and coated platelets, are characterized by distinct molecular signatures, activation thresholds, and functional roles that differentially influence thrombin generation, clot stability, and vascular integrity. Their variable reactivity and interactions with leukocytes and coagulation pathways position them as critical mediators of thromboinflammation. These subpopulations also display varied responses to commonly used antiplatelet medications. Reticulated platelets (RPs), for instance, are younger and more reactive, and have been associated with relative resistance to aspirin and P2Y12 inhibitors. Despite significant advances, the precise contributions of platelet subpopulations to hemostasis and disease progression remain poorly understood, largely due to the lack of standardized markers and methodological variability across studies. This review critically summarizes current insights into platelet heterogeneity, emphasizing the functional relevance of specific subpopulations and their pharmacologic profiles while addressing existing limitations and unresolved controversies. Furthermore, it discusses emerging research strategies aimed at refining platelet classification and explores how a deeper understanding of platelet heterogeneity may facilitate the development of precision-based diagnostic tools and targeted antiplatelet therapies. Such approaches may ultimately improve risk stratification and therapeutic outcomes in thrombotic and inflammatory disease.

Perturb-seq enables high-throughput linkage of CRISPR perturbations to single-cell transcriptomic phenotypes; however, inference quality depends on accurate single-guide RNA (sgRNA) assignment. In 10x Genomics-based single-cell workflows, assignment can be distorted by ambient RNA, overloaded droplets, and amplification artifacts, including cross-library polymerase chain reaction (PCR) chimeras. We demonstrate that standard Cell Ranger processing—with independent correction of gene expression and CRISPR libraries—does not explicitly resolve cross-library molecular collisions, in which a single-cell barcode-unique molecular identifier (CBC-UMI) pair is assigned to discordant features. To address this limitation, we developed Perturb-Audit, a diagnostic and denoising framework that integrates molecule-level collision auditing with statistical background suppression. Across Perturb-seq datasets of T-cell exhaustion, targeted collision removal provides high-specificity cleanup, whereas global denoising with CellBender yields broader improvements in assignment quality and phenotypic separation. Improved assignment fidelity increases detectable perturbation effect sizes and enables the recovery of biologically relevant immune cell signals. Applying this approach, we recapitulated the known Klf2-deficient phenotype in antiviral CD8+ T-cell Perturb-seq data. Furthermore, we found that suppression of Eomes triggers an exhaustion-biased shift, whereas Tox deficiency promotes effector-like differentiation. Collectively, these findings support an audit-first strategy to improve assignment fidelity and biological interpretability in single-cell CRISPR screens.

Transforming growth factor-β (TGF-β) signaling plays a crucial role in maintaining the quiescence and self-renewal potential of hematopoietic stem cells (HSCs). Despite decades of research, the underlying mechanisms of TGF-β signaling in HSCs remain elusive due to conflicting phenotypes of various knockout (KO) mouse models. Here, we show that HSCs co-express Tgfbr2, Tgfbr3, and Endoglin (Eng) but rarely express Tgfbr1, whereas lymphocytes co-express Tgfbr1 and Tgfbr2 but rarely express Tgfbr3 or Eng. We also demonstrate that Tgfbr3 is dispensable for the maintenance of immune homeostasis, in contrast to Tgfbr1 and Tgfbr2, either of which is essential for lymphocyte homeostasis. Serial transplantation assays revealed that deletion of Tgfbr3 in HSCs had little effect on short-term reconstitution but impaired long-term self-renewal potential, a similar phenotype observed in Eng conditional KO mice. Therefore, we propose that lymphocytes rely on the Tgfbr1/Tgfbr2 complex as suggested by the classical model, whereas HSCs require the unique Tgfbr2/Tgfbr3/Eng complex to orchestrate TGF-β signaling. Collectively, this study reveals Tgfbr3 as a critical regulator in the maintenance of HSC self-renewal potential and suggests a novel TGF-β receptor complex specific to HSCs.

Acute myeloid leukemia (AML) is a group of genetically and clinically heterogeneous malignancies characterized by clonal expansion of immature myeloid progenitors and profound disruption of normal hematopoiesis. Emerging evidence suggests that alterations in cellular homeostasis shape cancer progression. However, the mechanisms underlying AML progression remain largely unclear. Here, we identify lysophosphatidylcholine acyltransferase 3 (LPCAT3), a key enzyme of the Lands’ cycle, as a critical regulator of AML progression. Analysis of public transcriptomic datasets and patient-derived CD34+ cells revealed robust LPCAT3 overexpression in AML and an association between high LPCAT3 levels and inferior overall AML patient survival. Suppression of LPCAT3 by shRNA or CRISPR-Cas9 in MOLM-13 and THP-1 cells markedly impaired proliferation, induced apoptosis, and caused G0/G1 cell-cycle arrest. Conversely, enforced overexpression promoted cell survival. In xenograft murine models, LPCAT3 depletion reduced leukemic burden. RNA-seq following LPCAT3 loss showed that genes differentially expressed were significantly enriched in granulocyte chemotaxis-related pathways, suggesting a role of LPCAT3 in modulation of leukemic differentiation programs and microenvironmental interactions. Collectively, these data established that LPCAT3 as a previously unrecognized mediator of AML cell fitness and as a potential therapeutic target.

Signal transducer and activator of transcription 3 (STAT3) is a pivotal oncogenic driver in multiple myeloma (MM), and its constitutive activation promotes malignant plasma cell proliferation, survival, and drug resistance in the bone marrow microenvironment. Despite therapeutic advances, MM remains incurable due to persistent STAT3-driven tumorigenesis and the resilience of MM stem cells. We investigated the therapeutic potential of napabucasin (BBI608), a novel STAT3 inhibitor, in MM. Our data demonstrated that BBI608 potently suppressed MM cell proliferation in vitro and in vivo, while significantly impairing the clonogenic potential and inducing robust apoptosis. Mechanistically, BBI608 exhibited dual efficacy by targeting bulk tumor cells and eradicating the stem-like compartment of MM cells, thereby addressing a critical therapeutic challenge. Moreover, we revealed that BBI608 triggered immunogenic cell death (ICD) via the activation of endoplasmic reticulum (ER) stress and the unfolded protein response (UPR), which subsequently enhanced T-cell-mediated anti-tumor immunity. Our findings highlight STAT3 inhibition as a promising strategy to simultaneously eradicate MM cells, target stem cell reservoirs, and harness anti-tumor immunity, providing a robust rationale for the clinical translation of BBI608 in MM therapy.

Continuous intravenous heparin infusion is widely used for deep vein thrombosis (DVT) in the intensive care unit (ICU), but accurate prediction of activated partial thromboplastin time (APTT) remains challenging due to patient heterogeneity and complex drug responses. A deep learning model was developed using heparin administration data, laboratory tests, and patient history to predict future APTT values in ICU patients with DVT. Data from 796 patients receiving continuous heparin infusion were collected, including demographics, comorbidities, treatment, and laboratory information, with external validation performed on 514 patients from the Medical Information Mart for Intensive Care (MIMIC) database. A 2-layer Long Short-Term Memory model with dropout was trained and evaluated. Optimal performance was achieved with the [t - 12, t - 2] window, yielding a mean absolute error of 1.885, root mean square error of 4.732, R2 of 0.945, and mean absolute percentage error of 4.433. Subgroup analyses demonstrated robust accuracy across critical APTT zones and therapeutic categories (subtherapeutic, therapeutic, supratherapeutic), with area under the curve values of 0.92, 0.81, and 0.92. The model predicted abnormal APTT values a mean of 6 hours in advance and could improve guidance over physician-led management in approximately 70% of cases. External validation confirmed good generalizability. Feature importance identified the difference between 45 s and APTT, cumulative heparin dose, and infusion rate as the leading predictors. Collectively, this deep learning model accurately forecasts future APTT values in critically ill patients with DVT receiving intravenous heparin, supporting a shift from reactive to proactive, data-driven heparin management in the ICU with potential benefits for patient safety and treatment efficacy.

The precise impact of febrile neutropenia (FN) during chemomobilization on the collection of CD34+ peripheral blood stem cells and post-transplant infection(s) in patients with multiple myeloma (MM) and lymphoma undergoing autologous stem cell transplantation (ASCT) remains insufficiently characterized. Real-world records of 148 patients diagnosed with MM and 130 with lymphoma, who underwent ASCT between October 2010 and January 2024, were retrospectively analyzed. All patients underwent chemotherapy with granulocyte colony-stimulating factors (G-CSF) for stem cell mobilization. Statistical analysis was performed using GraphPad Prism 9 (p < 0.050). FN was associated with poorer estimated 5-year overall survival in patients with MM (61.0% vs 84.4%; p = 0.002). Patients with MM and FN required greater G-CSF stimulation (p < .001), longer apheresis (p = 0.007), and had lower CD34+ cell yield (p < .001). They also exhibited lower optimal-mobilization rates (p < .001), and delayed neutrophil (p = 0.006) and platelet (p = 0.028) engraftment. In patients with lymphoma, FN was associated with prolonged apheresis (p = .001), delayed platelet engraftment (p = 0.017), and an increased risk for pre-engraftment infection (p = 0.034). FN during chemomobilization was associated with worse long-term survival and impaired stem cell mobilization in patients with MM, along with a heightened risk for pre-engraftment infection(s) in those with lymphoma.