A Novel Model Incorporating Pectoralis Muscle Measures to Predict Mortality After Ventricular Assist Device Implantation

Background: We have previously demonstrated that pectoralis muscle mass and tissue attenuation obtained on preoperative CT scans were powerful predictors of mortality after left ventricular assist device implantation. In this analysis, we confirm our findings in a separate left ventricular assist device implantation cohort, and we present a novel, user-friendly mortality-prediction model incorporating these measures.

Methods and results: Patients with chest CTs performed ≤ 3 months prior to left ventricular assist device implantation at University of Minnesota (n = 143) and Houston Methodist Hospital (n = 133) were identified. Unilateral pectoralis muscle mass indexed to body surface area (PMI) and attenuation (approximated by mean Hounsfield units) (PHU_m) were measured on preoperative chest CT scans. To develop a prediction model incorporating pectoralis muscle measures, we implemented a cross-validated model-selection approach using Cox proportional hazards regression models. The final model included PHU_m, PMI, African American race, creatinine, total bilirubin, body mass index, bridge to transplant, and presence or absence of contrast. Receiver-operating characteristic curves for 30-, 90- and 365-day survival were generated. The area under the curve for the model at 30, 90 and 365 days was 0.78, 0.76 and 0.76, respectively.

Conclusions: The Minnesota Pectoralis Risk Score had favorable discrimination in this multicenter dataset. These skeletal-muscle measures appear to add important information to preoperative risk assessment.