An artificial intelligence algorithm characterizes the deterioration of body composition from systemic therapy for patients with advanced prostate cancer
Menée à partir de 2 342 clichés de TEP-TDM réalisés sur 468 patients atteints d'un cancer de la prostate de stade avancé, cette étude met en évidence l'intérêt d'un algorithme d'intelligence artificielle pour caractériser la fonte musculaire associée aux traitements systémiques
Skeletal muscle is a major determinant of functional status. The impact of advanced prostate cancer (aPC) systemic therapies on body composition is unknown. We identified aPC patients with serial PET-CTs receiving androgen deprivation therapy (ADT), androgen receptor pathway inhibitors (ARPI), chemotherapy, and/or Lutetium-177. An artificial intelligence algorithm quantified body composition changes from PET-CT images. Linear mixed effects models examined body composition changes over time by treatment type. The algorithm examined 2,342 PET-CTs from 468 patients. On multivariable analysis, ADT was associated with a 9% reduction in muscle density (-1.91 Hounsfield units), ARPIs an additional 6% (-1.4 Hounsfield units), and Lutetium-177 a subsequent 6% (-1.3 Hounsfield units), demonstrating progressive reduction in muscle density across later stages of systemic therapy. This large longitudinal analysis of body composition in aPC patients supports the use of an automated PET-CT algorithm to identify aPC patients on systemic therapy at risk for muscle wasting.
Journal of the National Cancer Institute , résumé, 2026