Intelligent 4D ultra-high-sensitive PET Imaging for early detection of pancreatic ductal adenocarcinoma
Description: The number of cancer deaths caused by pancreatic ductal adenocarcinoma (PDAC) will be responsible for the most overall cancer deaths after lung cancer over the next years. The reasons for this dismal outcome are multifactorial, but is mainly caused by detection at an advanced tumor stage due to long subclinical progression. Therefore, it is a compelling need to detect PDAC precursor lesions at an early, still curable tumor stage. However, to date there is no blood test or non-invasive imaging method available for early detection of PDAC. Recent development of cancer early diagnosis suggested a package of combining blood tests and PET/CT imaging, where PET may capture early molecular abnormalities before anatomical changes. Pathological investigations revealed the potential of several PET tracers including FDG (glycolysis), FLT (proliferation), Pentixafor (CXCR4), Trivehexin (avß6), and FAPi (Desmoplasia) in capturing the molecular alterations of early PDAC lesions. However, in contrast to PDAC lesions at late stages, early PDAC lesions are mainly small in size and their molecular abnormalities may be marginal leading to weak imaging signal. The opposing effect of reduced tracer delivery due to desmoplasia further hampers the contrast in imaging early PDAC lesions. Therefore, it is challenging for conventional PET scanners with limited sensitivity, conventional static imaging protocol and conventional data analysis methods to detect early PDAC lesions on PET imaging. This Sinergia project aims to develop a clinically transferrable intelligent 4D imaging approach on an ultra-high-sensitive PET scanner for the early detection of non-cystic neoplastic precursors or low-grade lesions of PDAC in combination with a novel blood test based on TKTL1 measurement within monocytes. For the first time, we introduce the concept of combining cutting-edge instrument, PDAC-specific PET tracers and 4D signature using advanced data analysis particularly artificial intelligence to detect marginal physiological abnormality of early PDAC lesions.A critical challenge for the development of computer-aided early diagnosis of pancreatic cancer is the limited availability of clinical data. It is very challenging to recruit a considerable number of subjects with identifiable very early PDAC lesions as no other effective methods exist and high-risk patients are not part of regular screening protocols. This leads to major challenges for the development of new imaging methods by exploring different tracers and scanning protocols. This dilemma hampers the technological breakthrough despite the revealed potential of ultra-high-sensitive PET imaging, artificial intelligence, and blood-test screening. This Sinergia project also proposes to develop physiologically and physically realistic in silico simulations including histology-driven lesion simulations and anthropomorphic PET imaging simulation to overcome the above-mentioned difficulties in tracer selection, protocol optimization and data augmentation. The in-silico simulations will be first validated and optimized on mice with orthotopic transplantation of patient-derived organoids. The complementary iterations between computational simulation and clinical measurements on PDAC patients with co-existing early PDAC lesions provide the unique opportunity to assist and accelerate the development of intelligent 4D imaging for early PDAC diagnosis. This project will employ a blood test to enhance the potential clinical translation of the developed methodology. The developed methodology will be finally tested on a relatively large group of high-risk individuals of early PDAC lesions from multiple clinical centers, which will further enhance the translational potential.
Amount: CHF 2,192,000
Funding type: research
Funder scheme: Sinergia
Start date: 2024-09-01
End date: 2028-08-31
Investigator: Rominger, Axel
Institution: University of Berne – BE
Subfield: Oncology
Field: Medicine
Domain: Health Sciences
Funded outputs count: 0
Funder award ID: 10000115
OpenAlex ID: G361436504