Spleen-corrected MRI signal intensity parameters for predicting posthepatectomy liver failure in hepatocellular carcinoma

To evaluate two spleen-corrected signal intensity parameters, the signal intensity-based corrected hepatocellular enhancement index (SI-CHEI) and signal intensity-based hepatocellular enhancement fraction (SI-HEF), for predicting posthepatectomy liver failure (PHLF) in hepatocellular carcinoma (HCC) using gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI).

This retrospective study included 233 patients with HCC who underwent Gd-EOB-DTPA-enhanced MRI before hepatectomy. Patients were divided into training (n = 163) and test (n = 70) groups. Signal intensity was measured in liver and spleen on precontrast and hepatobiliary-phase images. A two-compartment model was used to calculate SI-CHEI and SI-HEF. Predictive performance was assessed with receiver operating characteristic analysis and multivariable logistic regression.

In total, 65 patients (27.9%) developed PHLF. The area under the curve (AUC) values for SI-HEF were 0.759 (95% confidence interval [CI]: 0.673-0.837) and 0.760 (95% CI: 0.634-0.870) in the training and test groups, respectively. For SI-CHEI, AUCs were 0.721 (95% CI: 0.628-0.811) and 0.721 (95% CI: 0.599-0.842). The liver-to-spleen ratio was 0.737 (95% CI; 0.637-0.826). SI-HEF had the highest sensitivity (0.810) and negative predictive value (0.889). SI-HEF values were lower in patients with grades B and C PHLF than those with grade A PHLF (P = 0.041); SI-CHEI showed no difference. A multivariable nomogram integrating clinical and volumetric variables with SI-HEF showed AUCs of 0.891 (95% CI: 0.841-0.940) and 0.881 (95% CI: 0.784-0.979); SI-HEF was the largest contributor to model predictions.

Spleen-corrected parameters provide robust PHLF prediction. SI-HEF offers high sensitivity and complements conventional indices, and its integration with clinical variables improves preoperative risk stratification.