Are computed tomography-based measures of specific abdominal muscle groups predictive of adverse outcomes in older cancer patients?

PURPOSE: It is unknown whether computed tomography (CT)-based total abdominal muscle measures are representative of specific abdominal muscle groups and whether analysis of specific abdominal muscle groups are predictive of the risk of adverse outcomes in older cancer patients. METHODS: Retrospective single-center cohort study in elective colon cancer patients aged ≥65 years. CT-based skeletal muscle (SM) surface area, muscle density and intermuscular adipose tissue (IMAT) surface area were determined for rectus abdominis; external- and internal oblique and transversus abdominis (lateral muscles); psoas; and erector spinae and quadratus lumborum (back muscles). Outcomes were defined as severe postoperative complications (Clavien-Dindo score >2) and long-term survival (median follow-up 5.2 years). RESULTS: 254 older colon cancer patients were included (median 73.6 years, 62.2% males). Rectus abdominis showed the lowest SM surface area and muscle density and the back muscles showed the highest IMAT surface area. Psoas muscle density, and lateral muscle density and percentage IMAT were associated with severe postoperative complications independent of gender, age and cancer stage. CONCLUSIONS: CT-based total abdominal muscle quantity and quality do not represent the heterogeneity that exists between specific muscle groups. The potential added value of analysis of specific muscle groups in predicting adverse outcomes in older (colon) cancer patients should be further addressed in prospective studies.