Utilization of muscle area in an accurate prediction formula for renal function for patients with hepatocellular carcinoma.

ABSTRACT

OBJECTIVE: Accurate assessment of renal function prior to surgery for hepatocellular carcinoma is important for patient outcome, but current methods such as the estimated glomerular filtration rate (eGFR) are inadequate. We developed a new prediction formula that incorporates preoperative computed tomography (CT) imaging data to determine renal function. METHODS: We retrospectively analyzed 400 patients who underwent hepatectomy for hepatocellular carcinoma between January 2010 and December 2021. Predictors associated with renal function were identified by multivariate analysis. RESULTS: Age, sex, body height, body weight, body surface area, body mass index, serum creatinine, and muscle areas including third lumbar vertebra total muscle area (L3 TMA) determined by preoperative CT were identified as independent predictors likely to be associated with renal function. These were used to construct a new prediction formula using multiple regression analysis performed with a stepwise method: 232.2 + (-1.17 × age) + (-89.0 × serum creatinine) + (0.28 × L3 TMA). The median difference between conventional eGFR and CCr was 47.6 ml/min (range, 1.7-137.9 ml/min), while that between the new eGFR and CCr was 14.3 ml/min (range, 0.02-64.7 ml/min). Spearman rank correlation analysis revealed that the new eGFR was more positively correlated with CCr than conventional eGFR (ρ = 0.623, P < 0.05; ρ = 0.700, P < 0.05, respectively), and hence more accurately reflected renal function. CONCLUSION: A new prediction formula based on L3 TMA determined by CT is more accurate than conventional eGFR for evaluating renal function.