Digital Biopsy & Network Analysis of dynamic 68Ga-FAPI data in pancreatic cancer patients

Ziel/Aim: With static ⁶⁸Ga-FAPI-PET/CT, distinguishing pathologies like pancreatic ductal adenocarcinomas (PDAC), inflammatory lesions of the pancreas (ILP), post- pancreatectomy reactive tissue (PRT) and recurrent-PDAC (RPDAC) is a challenge due to their marked increase in signal intensity. Dynamic imaging allows ⁶⁸Ga-FAPI kinetic profile analysis, highlighting differences between these pathologies. Heretofore, analysis of such dynamic PET data is challenging. The use of a voxel-level “digital biopsy” approach combined with network analysis and clustering could overcome this challenge. We hypothesise this approach will allow the identification of healthy, non- malignant pathological and malignant pathological kinetic signatures which could aid diagnosis.

Methodik/Methods: 47 Patients with unclear pancreatic lesions (30 primaries, 17 recurrent settings) underwent dynamic ⁶⁸Ga-FAPI-PET/CT. Primary cases underwent surgical resection/biopsy after PET imaging. Possible recurrences were classified according to CT and clinical course (minimum 18 months). A digital biopsy was sampled in each volume of interest (VOI) before being blinded and imported into Graphia V.3.1. Muscle, fat, kidneys, liver and blood were sampled as controls.

Ergebnisse/Results: A total of 47 individual-networks, and two combined-networks, were created. Within all individual scan networks, voxels tended to arrange and cluster within the sampled VOI. Networks typically arranged into healthy control, elimination organs and pathological regions. Pathologies tended to cluster with high purity (> 95% from the same VOI), with multiple clusters per VOI indicating heterogeneity within the pathological classification. Importantly, our analysis approach was able to differentiate between malignant and non-malignant pathologies in the case of PDAC vs. ILP, as well as RPDAC vs. PRT. This differentiation was driven by a slower clearance of FAPI within the malignant voxels.

Schlussfolgerungen/Conclusions: The unique kinetics of 68Ga-FAPI across the different regions, coupled with an easily feasible sampling and interpreter-independent analysis approach, allowed the discrimination and identification of healthy, non-malignant pathological and malignant pathological clusters and kinetic features, thus warranting further investigation.

Publikationsverlauf

Artikel online veröffentlicht:
12. März 2025