A Clinical Feasible Method for Computed Tomography-Based Assessment of Sarcopenia in Patients With Chronic Pancreatitis.

Abstract

Sarcopenia is a serious but often overlooked complication of chronic pancreatitis (CP). We investigated the prevalence and risk factors for sarcopenia in patients with CP and determined the utility of a computed tomography (CT)-based method, based on psoas muscle measurements, for easy and clinical feasible diagnosis of sarcopenia. This was a retrospective multicenter study of 265 patients with CP. We used segmentation of CT images to quantify skeletal muscle mass and diagnose sarcopenia. On the same CT image as used for muscle segmentation, psoas muscle thickness and cross-sectional area were measured and receiver operating characteristic analyses defined age and sex-specific cutoffs for diagnosing sarcopenia. The prevalence of sarcopenia was 20.4%. The optimal height-adjusted psoas muscle cross-sectional area cutoff for diagnosing sarcopenia was 3.3 cm/m in males and 2.5 cm/m in females. The corresponding area under the receiver operating characteristic curves were 0.8 and 0.9, with sensitivities of 84% and 81% and specificities of 62% and 81%, respectively. Comparable diagnostic performance characteristics were observed for psoas muscle thickness. Sarcopenia is present in 1 of 5 patients with CP. Assessment of psoas muscle parameters provides a clinical feasible method to diagnose sarcopenia.