Frontiers in Cardiovascular Medicine · Nov 12, 2026

Preoperative computed tomography-derived psoas muscle index is associated with mortality after heart transplantation

By Carl-Thaddäus Braun, Angelika Costard-Jäckle, Nils Rodenburger et al.

Background: Frailty assessment is recommended when evaluating heart transplant candidates, but conventional physical measures are confounded by congestion and are often impossible in hospitalized or mechanically supported patients. Computed tomography (CT)-derived psoas muscle metrics offer an objective alternative, yet data specific to heart transplantation are scarce. Methods: In a single-center retrospective cohort study, 683 consecutive adult recipients (2013-2023) with a CT performed before transplantation were analyzed. At L4/L5 the summed cross-sectional area of both psoas muscles was divided by the square of body height to give the psoas muscle index (PMI; muscle quantity), and mean psoas attenuation (Hounsfield units) was recorded as a measure of muscle quality. Contrast administration was verified from the DICOM record for every examination. The primary outcome was all-cause mortality; secondary outcomes were in-hospital mortality and postoperative morbidity. Cox and logistic models were adjusted for established recipient risk factors and transplant year, with multiple imputation and bootstrap internal validation. Results: Over a median 5.2 years (3,714 patient-years), 251 patients (36.7%) died, 86 (12.6%) during the index admission. PMI was sex-dimorphic (7.80 vs 5.77 cm²/m²; p<0.001). Lower PMI independently predicted all-cause mortality (adjusted hazard ratio [HR] 1.18 per 1-standard-deviation decrease, 95% confidence interval [CI] 1.01–1.37; p=0.032), unchanged in the 578 unenhanced examinations (1.20, 1.03–1.40; p=0.022) and with contrast status added as a covariate (1.18, 1.02–1.37; p=0.030). Hazards were non-proportional; with follow-up split at one year, the adjusted PMI HR was 1.27 (1.00–1.62) in the first year and 1.05 (0.85–1.31) thereafter (interaction p=0.24). Lower attenuation was associated with mortality overall (HR 1.14, 1.01–1.29; p=0.040) but was not confirmed once contrast was accounted for (1.12, 0.99–1.28 in unenhanced examinations; contrast raised attenuation by 9.6 HU). Neither metric independently predicted in-hospital death, and incremental discrimination was small and non-significant (C-index 0.600–0.604). Associations were unaffected by the CT-to-transplant interval (PMI p=0.63). Conclusion: Low preoperative psoas muscle quantity is independently associated with mortality after heart transplantation. An association for muscle quality was not confirmed once contrast administration was accounted for. PMI may serve as an objective marker of physiological reserve and a target for prehabilitation rather than an exclusion criterion.

The paperJournal siteBiomarkers

AgingResearch.News