Published January 13, 2026
UB’s Clinical and Translational Science Institute (CTSI) Translational Pilot Studies Program annually provides seed money to investigators to assist them in developing promising technologies and therapeutics from the conceptual stage to clinical studies.
These include dental implant failure, infections caused by antibiotic resistant bacteria, diabetes and obesity treatment, opioid use disorder, pancreatic cancer, and periodontitis. The 10 studies incorporate cutting-edge technology, including artificial intelligence, metagenomic sequencing, and novel wearable technology.
“The projects awarded pilot studies grants for 2026 address important clinical and translational science questions,” says CTSI Director Sanjay Sethi, MD, SUNY Distinguished Professor of medicine and senior associate dean for clinical and translational research in the Jacobs School of Medicine and Biomedical Sciences.
“These innovative projects will contribute to our efforts to enhance health and health care in Western New York and beyond. Congratulations to the investigators involved in this group of studies.”
This year’s awarded projects are summarized below. In addition, four studies featuring translational science elements are noted.
Drugs that target the protein KRAS, which drives pancreatic cancer, hold immense promise for treating the disease, but for unclear reasons the cancer often becomes resistant to these drugs. This study will examine how MET-signaling, which is controlled both by tumor cell-intrinsic and extrinsic mechanisms, contributes to KRAS inhibitor resistance and determine the efficacy concomitant KRAS/MET targeting in pancreatic cancer.
Translational science element: Understanding how MET-signaling contributes to resistance to KRAS targeting drugs, as well as understanding the mechanisms that drive MET-signaling, will lead to strategies that will enhance both the clinical efficacy and durability of KRAS targeting therapies in the treatment of pancreatic cancer.
Periodontitis (gum disease) is a life-long, chronic inflammatory disease that requires continuous clinical management to avoid recurrence, suggesting the participation of maladaptive immune memory. This project will investigate how memory T cells and their clonal features contribute to disease persistence by combining spatial immune profiling and single-cell transcriptomics of human gingival tissue.
Older adults face high risk of harm after being discharged from the hospital as the instructions they receive are often hard to understand. This can lead to serious communication errors, resulting in preventable readmissions and patient harm. This project proposes to develop and evaluate patient-friendly discharge instructions created by generative artificial intelligence with physician supervision to improve communication at the time of discharge and ultimately patient safety.
Translational science element: Developing novel methods to improve patient communication at the time of discharge could improve patient safety.
Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are used in diabetes and obesity treatment. GLP-1Ras are known for having side effects including dehydration, which can pose a problem on its own and exacerbate gastrointestinal symptoms of GLP-1RAs. Investigators will evaluate whether an intervention aimed at increasing fluid intake in patients who are newly taking GLP-1RAs can improve biomarkers of hydration and reduce the experience of other side effects of the medication.
This study investigates the temporal dynamics of microbial communities surrounding dental implants prior to the clinical onset of peri-implantitis, a major contributor to implant failure. Using metagenomic sequencing, investigators aim to identify taxonomic and functional shifts in the submucosal biofilm that may serve as early indicators of disease development, supporting future strategies for early diagnosis and targeted intervention.
Phages are viruses that only infect bacteria and are increasingly used to treat infections caused by antibiotic resistant bacteria. This project seeks to characterize and define interactions between phages, bacteria, and mammalian cells and biomolecules within ex vivo human fluids. Characterization of these interactions will directly inform phage dosing in upcoming compassionate use treatments and clinical trials addressing antibiotic resistant and difficult-to-treat infections.
This project develops a machine learning framework to improve Alzheimer’s disease diagnosis using MRI data collected from multiple institutions. By enabling pre-trained models to adapt to new data in a parameter-efficient fashion, the project addresses key challenges in integrating diverse imaging datasets. Such a continual learning framework helps diagnostic agentic AI systems learn from new data without starting over each time, making them more practical for real-world healthcare settings.
Translational science element: By developing scalable continual learning methods that enable AI models to adapt across diverse clinical sites without full retraining, this pilot removes operational obstacles to implementing early Alzheimer’s disease detection in real-world healthcare settings throughout Western New York.
Investigators are developing small molecule therapeutics that target the neurobiological circuits underpinning reward, craving, and withdrawal in people suffering from opioid addiction. Using a combination of receptor binding assays, functional testing, in vitro pharmacokinetic profiling, and AI-guided structural optimization, the team aims to refine lead compounds that reduce fentanyl intake in vivo and exhibit favorable drug-like properties.
Adolescent and young adult sarcoma survivors often struggle with long-term, chronic pain that is poorly understood and difficult to treat. This longitudinal observational study will evaluate the feasibility and acceptability of a three-month, daily ecological momentary assessment protocol to assess chronic pain, opioid use, and associated risk factors in these individuals. Investigators aim to better understand what makes pain worse to help design better treatments.
Translational science element: Limited treatments exist for chronic pain in cancer survivors. Understanding factors that contribute to long term chronic pain and opioid use in adolescent and young adult sarcoma survivors would help inform the development and adaptation of behavioral pain treatments that are low burden, personalized, cost effective, and efficacious.
This project aims to develop a wearable patch that uses ultrasound and photoacoustic imaging technologies to continuously and automatically monitor a baby’s head and oxygen levels during labor to determine when a baby needs a C-section. The patch is designed to work without needing a skilled operator and can provide accurate imaging as a novel wearable technology in future health care applications.
This program is supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award number UM1TR005296 to the University at Buffalo, as well as: UB’s Office of the Provost, Office of the Vice President for Research and Economic Development, and Office of the Vice President for Health Sciences; Roswell Park Comprehensive Cancer Center; and the deans of UB’s Jacobs School of Medicine and Biomedical Sciences, School of Dental Medicine, School of Pharmacy and Pharmaceutical Sciences, School of Engineering and Applied Sciences, School of Public Health and Health Professions, and School of Nursing.
