A deep learning system (DL-T1b) leveraging digital pathology achieved strong performance (AUC 0.910) for predicting lymph node metastasis in T1b gastric cancer. Integration with six clinical risk factors into a nomogram further boosted predictive accuracy, offering a precise tool for personalized treatment decisions. This study establishes proof-of-concept for deep learning-based pathomics in early gastric cancer risk stratification.

A study published in PhotoniX Life presents a label-free method for mapping subcellular dynamics using wide-field interferometric scattering (iSCAT) microscopy and power spectral density (PSD) analysis. By fitting the pixel-wise PSD of iSCAT signals to an inverse-power-law relationship over 30–1,250 Hz, the researchers generated spectral exponent maps that visualize the characteristics and strength of subcellular motion. The method distinguished between mitotic and interphase cells, between live and apoptotic cells, and among thyroid cancer cell subtypes with increasing malignancy, highlighting a potential intrinsic optical marker for mechanobiology, cancer research, and stem cell assessment.

A new mechanism based on a novel natural BH3 mimetic nobiletin combined with an HDAC inhibitor to synergistically regulate apoptosis and autophagy in non-small cell lung cancer (NSCLC) by down-regulating TRKC expression is provided, and a new strategy for NSCLC therapy is opened.

Salk Institute researchers investigated the impacts of entinostat, a drug that targets HDAC proteins, and found it inactivates DNA damage repair genes in pancreatic cancer cells. Their discovery led to new treatment strategies that pair entinostat with DNA-damaging therapies, as well as the development of a nanoparticle-based delivery approach that limits toxicity by selectively delivering entinostat to tumors. The findings could improve treatment outcomes and expand therapeutic options for pancreatic cancer, and similar strategies could be applied for treating other cancer types that resist DNA-damaging therapies.