Viszeralmedizin NRW 2026. 192. Jahrestagung der Niederrheinisch-Westfälischen Gesellschaft für Chirurgie, 34. Jahrestagung der Gesellschaft für Gastroenterologie
Viszeralmedizin NRW 2026. 192. Jahrestagung der Niederrheinisch-Westfälischen Gesellschaft für Chirurgie, 34. Jahrestagung der Gesellschaft für Gastroenterologie
AI-based metabolic phenotyping reveals preferential adipose tissue depletion after esophagectomy
Text
Background and objective: Weight loss after esophagectomy is associated with impaired recovery and reduced long-term survival, yet the underlying postoperative metabolic phenotype remains poorly characterized. While sarcopenia is commonly assumed, the relative contribution of adipose tissue versus skeletal muscle loss has not been comprehensively quantified. We hypothesized that esophagectomy induces a distinct metabolic phenotype characterized by preferential adipose tissue depletion.
Method: Patients undergoing curative esophagectomy with contrast-enhanced CT imaging available at diagnosis and 6 months postoperatively were retrospectively analyzed. A deep learning–based Body and Organ Analysis (BOA) algorithm enabled fully automated three-dimensional segmentation and volumetric quantification of skeletal muscle and adipose tissue compartments (total, visceral, subcutaneous, intramuscular) across thoracic and abdominal regions. Longitudinal changes and sex-specific differences were assessed using paired analyses and two-way ANOVA.
Result: Seventy-five patients (24% female) with paired imaging were included. Across all compartments, significant tissue loss occurred at 6 months (all p<0.001). Adipose tissue depletion markedly exceeded muscle loss. Total adipose tissue declined by −32.7% in the thorax and −42.0% in the abdomen, whereas skeletal muscle volume decreased by −11.1% and −10.2%, respectively. Visceral adipose tissue exhibited the greatest reduction (up to −50%). Men demonstrated greater losses of both adipose and muscle compartments than women. Despite substantial weight loss, skeletal muscle volume remained relatively preserved compared with adipose tissue.
Summary: Esophagectomy induces a postoperative metabolic phenotype dominated by preferential adipose tissue depletion rather than isolated sarcopenia. AI-based volumetric imaging enables non-invasive, compartment-specific metabolic phenotyping and challenges the prevailing focus on muscle wasting alone, with important implications for metabolically tailored postoperative nutritional strategies.



