Centre for Translational Medicine, Semmelweis University Budapest, Budapest, Hungary
Ruben Zsolt Borbély, MD1, Vivien Vass, 1, Katalin Márta, MD1, Brigitta Teutsch, MD1, Balint Eross, MD, PhD1, Péter Jenő Hegyi, MD1, Áron Vincze, MD, PhD2, Eszter Ágnes Szalai, MD1, Andrea Szentesi, 1, Nándor Faluhelyi, MD, PhD2, Peter Hegyi, MD, PhD1 1Centre for Translational Medicine, Semmelweis University, Budapest, Budapest, Hungary; 2University of Pécs, Pécs, Baranya, Hungary
Introduction: Accurately predicting the severity of acute pancreatitis (AP) is a challenge in clinical practice. Recent advancements in neural network computing enable the use of CT scans for body composition analysis in addition to their diagnostic role. However, the prognostic application of various CT metrics has been inconsistent, resulting in conflicting data regarding the most reliable indicators of disease severity. The purpose of this study is to clarify these discrepancies.
Methods: We conducted a retrospective analysis of 279 AP patients from the University of Pécs, Hungary, all of whom underwent CT scans within 24 hours of hospital admission. Utilizing the TotalSegmentator tool in 3D Slicer, we quantified areas of visceral and subcutaneous adipose tissues (VAT and SAT) and skeletal muscle area (SMA) at the third lumbar vertebra (L3), with measurements expressed in square centimeters. These areas were normalized for height to generate indices analogous to BMI, including the Visceral Adipose Tissue Index (VATI), Subcutaneous Adipose Tissue Index (SATI), and Skeletal Muscle Index (SMI). Additionally, muscle density was measured in Hounsfield Units (HU) to evaluate fatty infiltration. We assessed the predictive accuracy of these metrics for severe AP using ROC (receiver operator characteristic) curves and AUC (area under the curve) analysis, differentiated by gender. The outcomes were correlated with actual AP progression. Preliminary results are provided below.
Results: The dataset comprised 174 male and 105 female patients, categorized by AP severity into 139 mild, 116 moderately severe, and 24 severe cases. Refer to Table 1 for the AUC values of various CT metrics, differentiated by gender.
Discussion: CT-derived body composition metrics, particularly index-based measurements and muscle radiodensity, demonstrate significant potential for predicting the severity of acute pancreatitis. Integrating these metrics into clinical practice could facilitate early risk stratification and personalized patient management, potentially improving outcomes. Our future research objective is to refine these metrics and establish precise cutoff values to enhance their clinical utility.
Note: The table for this abstract can be viewed in the ePoster Gallery section of the ACG 2024 ePoster Site or in The American Journal of Gastroenterology's abstract supplement issue, both of which will be available starting October 27, 2024.
Disclosures:
Ruben Zsolt Borbély indicated no relevant financial relationships.
Vivien Vass indicated no relevant financial relationships.
Katalin Márta indicated no relevant financial relationships.
Brigitta Teutsch indicated no relevant financial relationships.
Balint Eross indicated no relevant financial relationships.
Péter Jenő Hegyi indicated no relevant financial relationships.
Áron Vincze indicated no relevant financial relationships.
Eszter Ágnes Szalai indicated no relevant financial relationships.
Andrea Szentesi indicated no relevant financial relationships.
Nándor Faluhelyi indicated no relevant financial relationships.
Peter Hegyi indicated no relevant financial relationships.
Ruben Zsolt Borbély, MD1, Vivien Vass, 1, Katalin Márta, MD1, Brigitta Teutsch, MD1, Balint Eross, MD, PhD1, Péter Jenő Hegyi, MD1, Áron Vincze, MD, PhD2, Eszter Ágnes Szalai, MD1, Andrea Szentesi, 1, Nándor Faluhelyi, MD, PhD2, Peter Hegyi, MD, PhD1. P3441 - Predicting Acute Pancreatitis Severity With CT-Derived Body Composition: A Retrospective Study, ACG 2024 Annual Scientific Meeting Abstracts. Philadelphia, PA: American College of Gastroenterology.
