Abstract
Objectives: Peripheral arterial disease (PAD) is a common vascular condition with amputation as a major complication. While muscle volume and vascular severity is often considered in risk prediction, the prognostic value of muscle density remains underexplored. Methods: In this exploratory study, we retrospectively analyzed 134 patients undergoing lower-limb computed tomography angiography between January 2018 and December 2023. Muscle density (MD), muscle volume, intermuscular adipose tissue (IMAT), and vascular severity scores were quantified using deep learning software. Patients were stratified into non-PAD, mild PAD, and critical limb ischemia (CLI) groups. Multivariate Cox regression assessed associations with amputation risk. Results: PAD patients, especially those with CLI, had lower muscle volumes (e.g., total leg: 7945.3 ± 2012.5 cm(3) in CLI vs. 11,161.6 ± 4670.4 cm(3) in non-PAD), lower muscle densities (e.g., lower leg: 34.0 ± 10.5 HU in CLI vs. 44.1 ± 6.9 HU in non-PAD), higher intermuscular adipose tissue volume percentage (e.g., total leg: 15.6 ± 5.4% in CLI vs. 10.5 ± 3.6% in non-PAD), and higher vascular severity scores (e.g., total leg: 12.6 ± 5.0 in CLI vs. 0.1 ± 0.3 in non-PAD), compared to non-PAD individuals. Only muscle density (MD) including that of abdominal muscle, thigh muscle, and lower leg muscle remained significant predictors of amputation risk after adjusting for confounders. Multivariate Cox regression models, adjusted for demographics and comorbidities, revealed that lower MD of abdomen (<18.7 HU; HR, 6.50, 95% CI, 1.95-21.77), thigh (<27.8 HU; HR, 5.00, 95% CI, 1.60-15.66), and lower leg (<32.4 HU; HR, 6.89, 95% CI, 2.17-21.93) muscles were independently associated with increased amputation risk. Conclusions: Muscle density, reflecting muscle quality rather than quantity, was an independent predictor of amputation risk in PAD. These findings highlight the prognostic value of muscle quality and support the integration of advanced imaging techniques, such as automated CTA-based body composition analysis, for risk stratification in PAD patients.