Duan et al. benchmark six large language models using multi-center real-world LDCT reports for lung cancer screening. GPT-4 leads in accuracy, while Claude 3 Opus excels in readability. Exploratory analysis shows that Open-source DeepSeek-R1 and GPT-4o perform comparably to GPT-4 and outperform GPT-3.5, underscoring the growing clinical promise of open-source models.

Yao et al. identify that IL-17C in peritoneal fluid drives endometriosis-associated carcinogenesis via IL-17RE/PI3K-mediated EMT. They demonstrate that blocking IL-17C-IL-17RE signaling suppresses malignant progression in vivo and in organoids, identifying a potential therapeutic target for preventing cancer development in endometriosis patients.

Wang et al. demonstrate that nonsense-mediated mRNA decay (NMD) is abnormally activated in CRC. Inhibiting NMD remodels the tumor microenvironment by activating the TRAF6-TBK1 axis and boosting CD8+ T cell function—ultimately improving the efficacy of ICB therapy, offering a clinically actionable strategy to improve CRC immunotherapy outcomes.

Zhou et al. show that activation of Ngn2 and Isl1 enabled reactive astrocytes to convert into neurons, which promotes functional recovery after spinal cord injury, with new neurons forming synaptic connections with ascending and descending neural pathways, and promoted propriospinal axon regeneration and improved the neuromuscular junction morphology and function of muscle.

Jin et al. integrate drug sensitivity testing, transcriptomics, and functional assays to uncover key determinants of venetoclax-azacitidine (VEN/AZA) response in AML. They develop an eight-gene model (RF8) that links drug response with mutations and survival, offering a precision medicine framework to guide VEN/AZA therapy beyond conventional genetic markers.

The tumor microenvironment (TME) is a highly complex and dynamic ecosystem composed of diverse cellular and extracellular components that critically influence cancer initiation, progression, and resistance to therapy. Over the past few decades, extensive research has elucidated the interactions between tumor cells, the surrounding stroma, and the immune system, underscoring the TME’s pivotal role in oncogenesis and responses to therapy.

What do you envision as the most promising future directions for therapeutic strategies aimed at modulating the tumor microenvironment?

Karin de Visser is senior group leader at the Division of Tumor Biology & Immunology at the Netherlands Cancer Institute. In this interview with Cell Reports Medicine, Karin de Visser talks about her journey in science and shares her insights about tumor microenvironment modulation to improve patient therapies.

In vivo CAR engineering is an emerging therapeutic platform for cancer and autoimmune diseases. Enabled by advanced delivery systems such as immune-evasive lentiviral vectors and targeted lipid nanoparticles, this strategy has progressed to early clinical testing, showing initial safety and efficacy. Here, key opportunities and translational hurdles are critically discussed.

The discriminative dimensionality reduction tree (DDRTree) framework provides a powerful approach to unravel metabolic heterogeneity and delineate clinically relevant subtypes. In this issue, Qiu et al. and Jia et al. apply it to type 2 diabetes and obesity, identifying subtypes with distinct risk profiles.

Everolimus is approved for treating advanced ER+ breast cancers, yet many patients develop resistance. Medina et al. identify SMARCD3 regulons linked to activation of alternate growth-factor signaling and MEK1/2 inhibitor sensitivity in resistant cells, revealing an effective strategy to overcome everolimus resistance in ER+ breast cancer.

Wang et al. develop a clinical-imaging-multi-omics prognostic index for diffuse large B cell lymphoma by integrating quantitative PET/CT biomarkers, genetic subtypes, and clinical risk factors. By analyzing large cohorts of patients, key biomarkers—total metabolic tumor volume and lactate dehydrogenase—that reflect immunosuppressive microenvironments and tumor proliferation were identified.

Peng et al. reveal that the impaired cytotoxicity and exhausted state of T cells are significantly associated with lethal outcomes in SFTS patients and demonstrate that Gn-targeted bispecific T cell engager antibodies significantly protect mice against lethal SFTSV infection by redirecting T cells to kill infected cells.

Zhao et al. translationally demonstrate how maternal diabetes induces early-onset endothelial dysfunction in sons, but not daughters, even when offspring are metabolically healthy. This challenges prior findings that metabolic dysfunction is the primary driver of cardiovascular disease in this population, guiding future clinical practice and risk management of these individuals.

Monroe et al. analyze genome sequencing from individuals with early-onset ventricular arrhythmias. Combining the effect of common genetic variants with rare coding and noncoding regulatory variants in cardiomyopathy and arrhythmia genes predicts outcomes. Each component additively predicts arrhythmia with the cumulative GAPS performing best.

Zhang et al. utilize large-scale plasma proteomic data to investigate the association between 2,920 plasma proteins and the risk of incident epilepsy. They characterize the temporal trajectories of epilepsy-associated protein levels preceding disease onset and perform clustering analysis. Ultimately, they develop a protein-based predictive model for assessing future epilepsy risk.

In this review, Aprile et al. highlight the clinical application of gene therapy by gene addition and gene editing, and the novel findings about hematopoietic stem cell and bone marrow niche features and function in beta-thalassemia and sickle cell disease.

Monroe et al. analyze genome sequencing from individuals with early-onset ventricular arrhythmias. Combining the effect of common genetic variants with rare coding and noncoding regulatory variants in cardiomyopathy and arrhythmia genes predicts outcomes. Each component additively predicts arrhythmia with the cumulative GAPS performing best.

Zhao et al. translationally demonstrate how maternal diabetes induces early-onset endothelial dysfunction in sons, but not daughters, even when offspring are metabolically healthy. This challenges prior findings that metabolic dysfunction is the primary driver of cardiovascular disease in this population, guiding future clinical practice and risk management of these individuals.

Tu et al. present a systematic analysis of the tumor microenvironment across 14,837 breast cancers, identifying distinct cellular patterns that predict disease-specific survival independently of genomic and intrinsic subtypes. Characterizing microenvironment composition may provide critical prognostic markers that can complement current molecular classification for treatment selection and preventing recurrence.

Lyu et al. show that short term oral or intramuscular glucocorticoid interventions trigger a gut dysbiosis in randomized controlled trials of healthy men. The dysbiosis shows both administration-route specific and shared features. The authors identify specific bacterial species with abundance that associate with measures of host metabolism.

(Cell Reports Medicine 5, 101350; January 16, 2024)

Tailor et al. identify enolase 1 (ENO1) as the target of SU212, a promising therapeutic candidate for triple-negative breast cancer (TNBC). By inducing ENO1 degradation, SU212 reduces tumor growth with minimal toxicity, maintains efficacy in diabetic models, and shows favorable pharmacology, underscoring its strong potential for clinical translation.