Neutropenic fever (NF) is often the first sign of infection in patients with hematologic malignancies, but its cause is frequently unknown, leading to broad-spectrum antibiotic use without confirmed infections. Although research links gut microbiome disruptions to treatment-related infections, it typically examines NF as the outcome, leaving a gap in understanding how the microbiome and metabolic factors distinguish infectious from non-infectious cases. Stool samples from acute myeloid leukemia patients were analyzed to characterize gut microbiome composition and fecal metabolites at baseline and at fever onset. Machine learning models, network analyses, and functional profiling were used to differentiate infectious NF vs non-infectious NF at baseline and at fever onset. The baseline model (area under the receiver operating characteristic [AUROC] = 0.769) identified higher levels of Akkermansia , Enterobacter , Escherichia – Shigella , and Flavonifractor as predictors of infectious NF, while Collinsella , Lachnospiraceae , Coprococcus , and acetic acid were linked to non-infectious cases. At fever onset, Enterococcus was enriched in infectious NF, whereas Gemella , Butyrate, Lachnospiraceae , Ruminococcaceae , and Eisenbergiella abundances favored non-infectious NF outcomes (AUROC = 0.752). Network analyses also revealed greater functional diversity and microbiome–metabolome connectivity in non-infectious cases at fever onset. This study suggests that gut microbiota and metabolites may serve as biomarkers for distinguishing infectious from non-infectious neutropenic fever, warranting further validation in larger cohorts. IMPORTANCE: Our study tackles the challenge of managing neutropenic fever (NF) in immunocompromised patients whose numbers have increased due to various immunodeficiencies and treatments that suppress immune function. Fever is often the only sign of a serious infection in these patients, yet there are neither clear patterns linking risk factors to infection nor biomarkers reliable for ruling out non-infectious causes. As a result, febrile patients are typically empirically treated for major pathogens, even in the absence of confirmed infections, which propagates antimicrobial resistance and gut dysbiosis. Our research utilizes gut microbiome and targeted metabolomic profiling from two cohorts of patients with acute myeloid leukemia undergoing chemotherapy and employs a machine learning framework to distinguish between infectious and non-infectious NFs at baseline and upon fever onset.

Background: Broad-spectrum antibiotics are often used for suspected infections in patients with hematologic malignancies due to the risk of severe infections. Although antibiotic use can lead to antimicrobial resistance and microbiome dysbiosis, the effects of antibiotics on the microbiome and resistome in patients with acute myeloid leukemia (AML) undergoing remission induction chemotherapy (RIC) are not well understood. Methods: Various statistical models were utilized to examine the effects of antibiotic administration on the microbiome and resistome over time, as well as differences in AR-infection (ARI) and colonization (ARC) by important CDC-threats in 119 AML patients. Results: A greater number of unique antibiotic classes administered correlated with a loss of unique antibiotic resistance genes (ARGs) (R = −0.39, p = 0.008). Specifically, although a greater number of oxazolidinone administrations was correlated with a greater loss of diversity (R = −0.58, p < 0.001), each additional day of linezolid reduced the risk of ARC by ~30% (HR: 0.663, p = 0.047) and decreased the odds of acquiring genes predicted to confer macrolide (HR: 0.50, p = 0.026) resistance. Conclusions: The number of antibiotic administrations and the types of antibiotics used can influence the risk of antibiotic resistance gene (ARG) expansion and ARC events in AML patients undergoing RIC. While certain antibiotics may reduce microbial diversity, they are not always linked to an increase in ARGs or ARC events.

Background/Objectives: Chimeric antigen receptor (CAR) T cells are a powerful cancer therapy, but their function depends heavily on internal signaling domains and metabolic adaptability. Most studies evaluate CAR behavior upon antigen exposure, yet intrinsic signaling properties may pre-program CAR T cell states even in the absence of stimulation. This study investigates how CAR design and metabolic support shape baseline transcriptional programs, focusing on tonic signaling and NF-κB-related pathways. Methods: We engineered CAR T cells targeting HER2 or GPC3 antigens, incorporating either 4-1BB or CD28 co-stimulatory domains, respectively. A subset of cells was further modified with adenosine deaminase 1 (ADA1) and CD26 to degrade extracellular adenosine and supply inosine, a metabolic strategy termed metabolic refueling (MR). Bulk RNA-seq was performed on resting T cells without antigen stimulation. We analyzed differential gene expression, gene set enrichment (GO, KEGG, Hallmarks), and transcription factor activity (DoRothEA) to assess the impact of CAR design and MR on T cell programming. Results: All CAR T cells exhibited activation of NF-κB–centered inflammatory programs at baseline, indicating tonic signaling. GPC3 CAR T cells showed stronger baseline activation than HER2 CAR T cells. Metabolic refueling amplified these programs without altering their directionality, enhancing inflammatory, survival, and effector modules. Transcription factor activity scores mirrored these trends, highlighting RELA, FOS, and STATs as key regulatory nodes. Conclusions: CAR-intrinsic features, notably co-stimulatory domain choice, define the tonic NF-κB activation tone in resting CAR T cells. Metabolic refueling boosts these baseline states without overstimulation, suggesting it may be especially valuable for weaker CAR constructs. These findings provide a framework for tuning CAR T cell function through combinatorial design strategies targeting signaling and metabolism.

Bloodstream infections (BSIs) are a major cause of morbidity and mortality in acute myeloid leukemia (AML) patients undergoing induction chemotherapy. Staphylococcus epidermidis, typically a skin commensal, is increasingly recognized as a pathogen in these vulnerable individuals. This study investigated whether genomic differences exist between infectious and gastrointestinal colonizing S. epidermidis isolates from AML patients and how these compare to colonizing and infectious isolates from other patient groups and biogeographic sites. We analyzed 114 isolates—44 from AML patients (23 infections, 21 GI colonizers) and 70 from public datasets (34 infections, 36 colonizers). Stool samples underwent 16S rRNA sequencing and culture to identify colonization, while bloodstream isolates were sequenced and compared. Genomic profiling using Roary, Scoary, Phyre2, and InterProScan revealed that infectious and GI-colonizing AML isolates were phylogenetically close but genomically distinct. Infectious isolates from AML patients were significantly enriched for resistance genes (e.g., mecA, mecR1, mecI, ANT(4′)-Ib) and the biofilm-associated gene icaA. AML infectious isolates harbored more resistance genes and mobile elements than non-AML strains but lacked widespread classical virulence factors. These results suggest that S. epidermidis pathogenicity in immunocompromised hosts is driven by genomic adaptability and antibiotic tolerance rather than traditional virulence mechanisms.