A feasibility study of computed tomography muscle phenotyping for physiotherapy selection in peripheral arterial...

Patients with peripheral arterial disease commonly undergo computed tomography (CT) angiography for anatomical assessment, yet body composition data from these scans remain unexploited for treatment planning. We evaluated the feasibility of extracting metabolic phenotype information from routine CT imaging and explored whether baseline body composition parameters might relate to physiotherapy outcomes in this hypothesis-generating pilot investigation.

This prospective feasibility study enrolled 10 patients with symptomatic peripheral arterial disease (Fontaine stages IIa-IIb); eight completed a 3-month supervised physiotherapy program consisting of progressive treadmill walking to moderate claudication and resistance training. We performed CT body composition analysis at the L3 vertebral level at baseline and follow-up, measuring the skeletal muscle index (SMI), muscle density in Hounsfield units (HU), visceral adipose tissue area, and subcutaneous adipose tissue area using automated segmentation. Functional assessments included the chair stand test, handgrip strength, and 2-minute walk test. The primary outcome was treatment success (continued conservative management) vs failure (revascularization required). We explored correlations between baseline body composition parameters and functional performance using Spearman rank correlation.

CT body composition analysis proved technically feasible in all patients, requiring minimal additional processing time. Five patients (62.5%) achieved treatment success. Baseline muscle density showed correlation with chair stand performance (ρ = 0.829; P = .021), whereas the SMI showed no functional relationships. Visceral adipose tissue was associated in five patients with walking distance (ρ= -0.900; P = .037) and handgrip strength (ρ = -0.775; P = .041). Patients achieving treatment success had a lower body mass index (24.9 kg/m² vs 29.6 kg/m²) and higher muscle density (41.1 HU vs 35.9 HU). Paradoxically, patients requiring revascularization had higher SMI (59.7 cm²/m² vs 48.6 cm²/m²) but lower muscle quality. Body composition changes during the intervention showed no relationship to outcomes.

This feasibility study demonstrates that metabolic phenotyping using existing CT imaging is technically viable in patients with peripheral arterial disease undergoing physiotherapy evaluation. Exploratory observations suggest testable hypotheses: that muscle quality may prove more functionally relevant than muscle quantity, that visceral adiposity may associate with impaired rehabilitation capacity, and that myosteatotic obesity could emerge as a high-risk phenotype. These preliminary findings support the rationale for adequately powered prospective studies to validate whether CT-based metabolic phenotyping could inform patient selection for the conservative management of claudication.