NATIONAL INSTITUTES OF HEALTH

Researchers at WashU Medicine engineered a yeast probiotic to deliver immunotherapy directly to the gut, shrinking gastrointestinal tumors in mice, offering a novel strategy for hard-to-reach gut cancers.

COTA's study reveals DLBCL patients initiating treatment immediately post-diagnosis may experience worse outcomes compared to those starting later, with shorter median times to next-line therapy and overall survival rates. Presented at the 66th ASH 2024 Annual Meeting, the research also highlights higher ECOG scores and high-risk clinical features in these patients, emphasizing the need for informed future trials and practices.

NIH developed AI algorithm TrialGPT to match volunteers to relevant clinical trials on ClinicalTrials.gov, streamlining the process and improving enrollment. TrialGPT uses large language models to process patient summaries, identify eligible trials, and explain eligibility criteria, achieving nearly the same accuracy as human clinicians while saving time. The tool aims to enhance clinical trial recruitment and reduce barriers for underrepresented populations.

UND's Department of Psychology, with support from the College of Arts & Sciences, is conducting Pediatric Eye Disease Group (PEDIG) clinical trials for children's eye conditions like strabismus and amblyopia, using virtual reality headsets. The trials, led by David Biberdorf and coordinated by Dmitri Poltavski, are the first in North Dakota and funded by the National Institutes of Health and National Eye Institute.

NIH introduces TrialGPT, an AI algorithm matching patients to clinical trials on ClinicalTrials.gov, aiming to accelerate research and improve enrollment. Developed by NIH's NLM and NCI, TrialGPT uses LLMs to process patient data, exclude ineligible trials, and rank eligible studies. It achieved near-human accuracy in a benchmarking study and demonstrated potential to save clinicians 40% of screening time.

Illinois researchers discovered SARS-CoV-2 RNA triggers immune cells with anti-cancer properties, potentially treating cancers resistant to current therapies. The study, led by Dr. Ankit Bharat, found these 'inducible nonclassical monocytes' (I-NCMs) attack cancer cells, with observed responses in melanoma, lung, breast, and colon cancers. The research, published in 'The Journal of Clinical Investigation,' offers hope for advanced cancer patients and requires further development and clinical trials.

Cedars-Sinai researchers developed an AI method using pathology reports to automate cancer staging, speeding up clinical trial candidate selection and expanding AI's healthcare applications.