We propose a physics-based framework in which cancer cell state is defined by position and velocity in a continuous space of directly measurable physical variables-cell surface area (S) and volume (V)-and motion through S-V space as an interpretable proxy for plasticity. Therapy generates S-V vector fields that govern trajectories, enabling the design of drug combinations to steer heterogeneous cell populations toward nonviable states, offering a predictive and physically interpretable alternative to therapies directed against oncogenic mutations and/or predefined cell subpopulations.

Our systematic and comprehensive genomic and spatially resolved profiling of Hodgkin lymphoma revealed a paradigmatic link between a genetic disease subtype and TME composition and function. Insights into mutation-driven mechanisms of deregulated cytokine signaling will pave the way for therapeutic interventions that interfere with oncogenic signaling and cellular ecosystems in cancer.

Hepatocellular carcinoma (HCC) is a highly fatal tumor, for which risk stratification is crucial, yet remains challenging. Here, we develop an interpretable machine-learning framework for HCC risk stratification based on routinely collected clinical data. We utilize prospectively collected multimodal data from over 900,000 individuals and 983 cases of HCC across two population-scale cohorts: the "UK Biobank study" (development) and the "All of Us Research Program" (external testing). We assess individual and cumulative contributions of data modalities including demographics, lifestyle, health records, blood, genomics, and metabolomics. Our final, random-forest-based models significantly outperform all publicly available state-of-the-art risk-scores on both internal and external test sets. We demonstrate robustness across ethnic subgroups, provide comprehensive interpretability and release all code, model weights and a web-calculator for external validation and agentic integration. Our study presents PRE-Screen-HCC, a robust and interpretable machine-learning framework for HCC risk stratification and early detection.

Five-year data from the phase IIb KEYNOTE-942 study show that the mRNA-based neoantigen vaccine intismeran autogene continues to reduce the risk of melanoma recurrence or death by roughly 49% when combined with pembrolizumab. This durability strengthens confidence in the personalized vaccine strategy, but a larger phase III trial will provide the decisive findings.

CAR therapy has transformed the treatment landscape for hematological malignancies but its efficacy in solid tumors is limited, owing in part to insufficient functional persistence of the engineered T cells. To elucidate the basis for their functional decline, we conducted integrated chronic in vivo and in vitro screens of 400 transcription factors, which revealed NFIL3 as a driver of CAR T cell dysfunction. Genetic disruption of NFIL3 in CAR T cells sustains their expansion, increases cytokine production, overall restraining terminal differentiation. Loss of NFIL3 enhances CAR T cell efficacy, improving tumor control and prolonging survival in xenograft and syngeneic mouse tumor models across different CAR designs. Under chronic stimulation, disruption of NFIL3 establishes a transcriptional state predictive of favorable clinical outcomes. Our findings underscore the power of comprehensive in vivo genetic screens integrated with multi-parameter in vitro assessment and identify NFIL3 as a novel therapeutic target to enhance cancer immunotherapy.

Childhood and adolescent cancer rates increased during 2001 to 2016 and then declined significantly. This temporal pattern was largely driven by rates of leukemias, lymphomas, CNS tumors, and other epithelial tumors. Changes in cancer classification, screening, and diagnostic technology likely influenced cancer-specific patterns, in addition to potential changes in risk factor prevalence.