Sorry, this entry is only available in Italian.
Manetta Rosa, Daffinà Julia, Martinese Andrea, Gazzerro Alessia, De Simone Stefano, Ciccone Vincenzo, Masciocchi Carlo, Di Cesare Ernesto
Objective: The primary objective of this study is to identify and characterize a pattern of volatile compounds capable of discriminating, through appropriate statistical treatment, patients with prostate cancer (PCa) from healthy reference individuals.
Materials and Methods: This prospective pilot trial was conducted in the RadiodiagnosticsDepartment of L’Aquila in collaboration with the research and development company Predict srl(Bari, Italy), between May 2022 and June 2023. Alveolar breath was collected from 147 volunteers aged between 45 and 75 years on the day of multiparametric MRI (MRmp) execution. The sample was collected before MRmp execution using Mistral technology (Predict). Molecules were analyzed using gas chromatography and mass spectrometry (TD-GC-MS). Artificial intelligence algorithms were utilized to evaluate the selection of significant variables and the precision of the breath test in identifying the group of each sample.
Results: After a qualitative selection of 257 molecules, the breath test’s ability to accurately distinguish the group of each sample was tested using AI algorithms, demonstrating a predictive accuracy exceeding 74%.
Conclusions: The results obtained highlight that information about metabolic alterations induced by the presence of prostate tumor pathology is contained in the breath. The goal is to develop a new non-invasive diagnostic monitoring approach that supports and complements the gold standard reference techniques for this pathology.
