Analytic Morphomics Research Articles

Quantifying the eyeball test: Sarcopenia, analytic morphomics, and liver transplantation

The physiological implications of end-stage liver disease (ESLD) are so pervasive that measuring the severity of the disease is challenging and imperfect. Although the Model for End-Stage Liver Disease (MELD) score is convenient and reliable, clinicians appreciate the broad range of disease severity among patients with any given MELD score. Within this context, most MELD-based allocation systems allow exception points for patients whose risk of wait-list mortality is poorly reflected by their calculated MELD scores. In the United States, the majority of MELD exceptions are granted for objective diagnoses such as hepatocellular carcinoma and hepatopulmonary syndrome, and patients who are subjectively assessed to be ‘‘sicker than their MELD score’’ likely have increased wait-list mortality. Importantly, allocation policy and treatment algorithms work best with objective clinical assessments (eg, the MELD score), which are well suited for standardization and the rigorous modeling of predicted risk. In all, better objective measures of disease severity in patients with liver disease are needed. One such objective measure of disease severity, sarcopenia, is described by Tandon et al. 1 in this issue of Liver Transplantation. This good work has several key take-home points. First, the authors demonstrate that sarcopenia is a strong and independent predictor of mortality among patients waiting for liver transplanta

Analytic Morphomics, Core Muscle Size, and Surgical Outcomes

Assess the relationship between lean core muscle size, measured on preoperative cross-sectional images, and surgical outcomes. Novel measures of preoperative risk are needed. Analytic morphomic analysis of cross-sectional diagnostic images may elucidate vast amounts of patient-specific data, which are never assessed by clinicians. The study population included all patients within the Michigan Surgical Quality Collaborative database with a computerized tomography(CT) scan before major, elective general or vascular surgery (N = 1453). The lean core muscle size was calculated using analytic morphomic techniques. The primary outcome measure was survival, whereas secondary outcomes included surgical complications and costs. Covariate adjusted outcomes were assessed using Kaplan-Meier analysis, multivariate cox regression, multivariate logistic regression, and generalized estimating equation methods. The mean follow-up was 2.3 years and 214 patients died during the observation period. The covariate-adjusted hazard ratio for lean core muscle area was 1.45 (P = 0.028), indicating that mortality increased by 45% per 1000 mm(2) decrease in lean core muscle area. When stratified into tertiles of core muscle size, the 1-year survival was 87% versus 95% for the smallest versus largest tertile, whereas the 3-year survival was 75% versus 91%, respectively (P < 0.003 for both comparisons). The estimated average risk of complications significantly differed and was 20.9%, 15.0%, and 12.3% in the lower, middle, and upper tertiles of lean core muscle area, respectively. Covariate-adjusted cost increased significantly by an estimated $10,110 per 1000 mm(2) decrease in core muscle size (P = 0.003). Core muscle size is an independent and potentially important preoperative risk factor. The techniques used to assess preoperative CT scans, namely analytic morphomics, may represent a novel approach to better understanding patient risk.