Background: Hepatic macrosteatosis (HMS) is prevalent among high BMI patients, but a lack of validation of non-invasive measures of liver fat hampers non-alcoholic liver disease (NAFLD) investigation in general. Recent work suggests BMI adjusted, non-contrasted computed tomography (nc-CT) attenuation data (Hounsfield units) reflects liver fat accumulation in a normal weight population. However, this and other CT-based HMS studies have only approximated macrosteatosis (%) histologically, but have not validated findings with chemical liver triglyceride (TG) concentrations (mg/gram protein). Also, all previous CT based steatosis studies excluded high BMI subjects, whose habitus may affect properties of the scan. We hypothesized that in high BMI patients nc-CT attenuation measurements expressed in Hounsfield units (HU) accurately estimate liver triglyceride concentrations as well as histological macrosteatosis.
Methods: With informed consent, 15 patients underwent nc-CT scan of the abdomen prior to weight loss surgery with intraoperative wedge and core needle liver biopsy. Mean left lobe nc-CT Hounsfield units (CT(L)), liver TG (mg/g Pr), HMS (%), BMI (kg/m(2)), liver-spleen index (CT(L/S) = hepatic HU/splenic HU), and liver-spleen difference (CT(L-S) = hepatic HU - splenic HU) were a priori outcomes.
Results: In 15 patients (11 female) with a BMI of 44.4 ± 1.1 (mean ± SEM), CT(L/S), CT(L-S), and CT(L) measures were significantly associated with liver TG concentrations (r = -0.80, P < 0.001; r = -0.80, P < 0.001; and r = -0.71, P < 0.01, respectively; Table 1). Macrosteatosis (%) and liver triglyceride concentration were positively associated (r = 0.83; P < 0.0001). BMI did not correlate strongly to liver triglyceride (r = 0.44, P = NS).
Conclusion: Estimates of liver fat obtained by nc- CT scans (esp. CT(L/S), CT(L-S)) correlate to chemical measurement of liver triglyceride concentrations, suggesting non-contrasted CT may be a suitable non-invasive "gold standard" for hepatic steatosis quantification in these patients.