Prognostic Value of Psoas Major Muscle Volume in Assessing Sarcopenia in Elderly Patients With Rectal Cancer
Abstract
ABSTRACT Background This study aimed to evaluate the clinical utility and prognostic significance of sarcopenia classification based on the psoas major muscle volume index (PVI) in elderly patients with rectal cancer. Methods We retrospectively analyzed 428 patients aged ≥ 65 years who underwent laparoscopic or robot‐assisted radical rectal cancer resection between 2014 and 2018. PVI was calculated by dividing the semi‐automatically measured psoas major muscle volume on three‐dimensional computed tomography (CT) by the cube of height (m 3 ). Sex‐specific PVI cutoff values for 5‐year overall survival (OS) were determined using time‐dependent receiver operating characteristic analysis. Associations between PVI, nutritional and inflammatory markers, conventional sarcopenia indices, and long‐term outcomes were investigated. Results PVI cut‐off values were ≤ 58.9 cm 3 /m 3 for males and ≤ 39.9 cm 3 /m 3 for females. Patients with low PVI had significantly worse nutritional and inflammatory profiles. The 5‐year OS (86.6% vs. 94.6%, p = 0.001), cancer‐specific survival (93.5% vs. 97.4%, p = 0.003), and relapse‐free survival (76.2% vs. 85.4%, p = 0.021) were significantly lower in the low PVI group than in the normal‐high PVI group. Multivariate analysis identified low PVI (hazard ratio: 2.55; p = 0.012) and Prognostic Nutritional Index < 40 as independent predictors of poor OS. Conclusion PVI‐based sarcopenia classification reflects nutritional and inflammatory status and enables effective prognostic stratification in elderly patients with rectal cancer. It may serve as a simple and objective tool for preoperative risk assessment, offering improved prognostic stratification compared with conventional CT‐based indices.
Fields of Study & Subjects
Nutrition and Health in Aging
Frailty in Older Adults
Stoma care and complications
Publication details
- Journal / source
- Annals of Gastroenterological Surgery
- Volume
- 10
- Issue
- 3
- Pages
- 770–778
- Publication date
- 5 January 2026
- Language
- EN