Archive
1250 articles, sorted by publication time.
The FDA has cleared a 250-patient randomized controlled study, moving CRB-701 from an early tumor-shrinkage signal toward pivotal validation; whether it can prolong survival and control toxicity must still be answered by TEMPO-1, which has yet to begin enrollment.
A 46-patient phase 1 trial found that relatlimab alone or with nivolumab could bring more CD8 T cells into tumors; the combination group had a one-year overall survival rate of 52.2%, but the nonrandomized design is not yet sufficient to demonstrate efficacy.
From a Phase 3 Alzheimer’s disease trial to FDA reviews of new drugs, Kalshi is attempting to use trading prices to represent market expectations; the experiment also brings risks involving inside information, liquidity, and patient perceptions to the fore.
The approval brings precision radiation delivery into the hormone-sensitive setting; while the risk of disease progression was reduced, hematologic toxicity and still-immature survival data also define the limits of the current evidence.
Researchers combined a protein inverse-folding model with evolutionary constraints to design RNA-guided nucleases whose sequences diverge substantially from natural versions yet still function in bacteria, plants, and human cells. The findings expand the design space for gene-editing tools, but multiple hurdles remain before therapeutic applications.
This Transformer model attempts to infer one molecular modality from more readily obtainable spatial measurements while preserving cellular neighborhood relationships. Data from Alzheimer’s disease brain tissue and breast cancer demonstrate its research potential, but inferred results still cannot replace direct experiments.
BW-41012 has begun a Phase I dose-escalation study in Australia, initially assessing safety, pharmacokinetics, and whether it can sustainably reduce Factor XI in healthy adults; its clinical benefit in preventing thrombosis and its bleeding risk remain to be answered by subsequent trials.
The Belgian startup does not begin with computer-generated molecules, but first observes how patient-derived organoids respond to drugs. The concept addresses the translational challenge in AI drug development, but the evidence remains at an early preclinical stage.
A research team combined computational design with yeast screening to create an antibody that recognizes a complex formed by a KRAS mutant peptide and a specific HLA; cell experiments have demonstrated selective killing, but the eligible patient population, in vivo safety, and efficacy remain to be confirmed.
After the first participants completed 30 days of safety observation, the REMIND trial was cleared to expand from up to 5 participants to 15; the procedure seeks to reconstruct cervical lymphatic drainage, but for now can address only questions of safety and feasibility.
Three de novo-designed miniature proteins precisely bind distinct fluorescent dyes and support super-resolution and fluorescence lifetime imaging in living human cells; the toolkit also advances the study of protein interactions into a new stage in which they can be activated and controlled by dyes.
Digital twins, AI-enhanced organoids, and virtual scientists are moving to the front lines of drug development. What will ultimately determine whether they can enter the regulatory evidence chain is not how novel the models are, but whether they can reliably answer a clearly defined biomedical question.