News
Curated oncology headlines from reputable sources, ranked by importance with Claude Opus 5.
A clinically-oriented foundation model for intraoperative pathology
CRISP, a vision-based foundation model trained exclusively on frozen section slides, outperformed existing foundation models in supporting intraoperative treatment decisions, according to its authors. Validation included a prospective patient cohort. Specialising in frozen sections is intended to address the artefacts and time constraints specific to intraoperative diagnostics. The paper describes…
Autoimmune Events From Checkpoint Inhibitors Often Demand Aggressive Treatment
Autoimmune reactions to immune checkpoint inhibitors (ICIs) in cancer patients referred to rheumatologists were generally more severe than previously reported in a German study, with corticosteroids alone frequently proving insufficient to control symptoms. The findings suggest that many ICI-induced rheumatic and autoimmune adverse events require escalation to second-line immunosuppressive agents.…
H. Pylori Exposure and Colorectal Cancer: The Evidence Grows
A risk-attribution modeling study based on estimates from 43 studies suggests that Helicobacter pylori exposure, long established as a cause of gastric cancer, may also contribute to colorectal cancer risk. The meta-analytic approach strengthens the epidemiological link between H. pylori infection and CRC, potentially expanding the cancer burden attributable to this bacterium. If confirmed in pros…
A guide to understanding tumour evolution through the lens of population genetics
A new Review in Nature Reviews Cancer synthesises how population-genetics models of mutation, selection, and genetic drift can be applied to cancer sequencing data to quantitatively reconstruct tumour evolutionary dynamics. The framework enables researchers to distinguish genuine selective forces from neutral drift and non-genetic processes shaping tumour heterogeneity. By moving beyond pattern-ba…
AI model predicts which breast-cancer drugs work best
An AI model trained on millions of protein measurements can predict which drugs will be most effective in tissue samples from people with triple-negative breast cancer. It gauges drug effectiveness directly from proteomic profiling of biopsy material, in a subtype with historically limited targeted treatment options.
Proximity-guided graph learning reveals tumour-associated proximity antigens
A proximity-mapping atlas of tumour cell surfaces has defined a new class of targets called tumour-associated proximity antigens — membrane proteins whose spatial co-localisation on cancer cells distinguishes them from healthy tissue. The study identified disease-associated membrane spatial communities and pinpointed the EGFR–CDCP1 pair as a co-target that enhances tumour killing when engaged simu…
Integrated signatures define mutational processes in prostate cancer
Researchers identified eight integrated mutational footprints that collectively explain the mutational landscape of 85% of primary prostate cancer genomes. Published in Nature, the study combines multiple layers of genomic data into unified signatures, providing a comprehensive framework for understanding how mutations accumulate in prostate tumors. The findings could improve patient stratificatio…
TRI-611, a selective, brain-penetrant molecular glue degrader of ALK
TRI-611 is a novel brain-penetrant molecular glue degrader that induces selective degradation of ALK fusion proteins through a previously undescribed CRBN recruitment motif. In preclinical models the compound demonstrated anti-tumour activity against ALK-positive non-small-cell lung cancer, including CNS disease, positioning it as a potential new therapeutic approach for patients who progress on c…
A serpin–myeloid axis in pancreatic cancer heterogeneity and immune evasion
SERPINE1 and SERPINB2 drive the formation of fibrin-rich niches in pancreatic ductal adenocarcinoma that locally reprogram macrophages toward an immunosuppressive phenotype and exclude cytotoxic T cells, establishing spatially organized immune evasion. The study, published in Nature, reveals a mechanistic serpin–myeloid axis that explains intra-tumoral heterogeneity in immune infiltration across d…
An operational perturbation proteomics-based virtual cell model
Researchers developed ProteinTalks, a virtual cell model built from temporal protein-abundance measurements in systematically perturbed breast cancer cell lines, enabling computational prediction across diverse drug discovery tasks. The model leverages large-scale perturbation proteomics to capture dynamic cellular responses and translate them into an operational tool for identifying drug targets …