Introduction: Several reports described Gd-EOB-DTPA-enhanced MRI (EOB-MRI) can be potentially used for evaluation of liver function. This modality has better versatility and spatial resolution compared with 99mTc-GSA scintigraphy. We assessed whether functional remnant liver volumetry (FRLV) using EOB-MRI predicted post hepatectomy liver failure (PHLF). Method: We retrospectively analyzed 114 cases performed hepatectomy of more than one segment (MOS). Signal intensity (SI) of remnant liver were measured by putting region of interest in T1WI image. SI of spinal erector muscle and spleen were calculated for standardization of SI value. Functional remnant liver score was derived in division of average SI of liver by average SI of muscle (or spleen) as liver-muscle ratio(LMR) and liver-spleen ratio(LSR), respectively. Future remnant volume was calculated by volume analyzer, then FRLV was calculated by multiplying LMR(or LSR) and remnant volume. Standardized FRLV(sFRLV) was calculated by dividing FRLV by body surface area. Definition of PHLF is used ISGLS classification. Results: LMR correlated with ICG-R15(p=0.004), but not LSR, FRLV or sFRLV. PHLF(≥GradeB) occurred in 5 patients (4.4%). In univariate analysis, predictive factors of PHLF were operative time(p=0.010), blood loss(p=0.013), FRLV(LMR) (p=0.021), sFRLV(LMR)(p=0.015), FRLV(LSR)(p=0.040), and sFRLV(LSR)(p=0.044). In multivariate analysis, only sFRLV(LMR) was independent predictive factor of PHLF(p=0.042). ROC analysis revealed that cut-off value of sFRLV(LMR) predicting PHLF was 599ml/m2(AUC: 0.913). In analysis of PTPE cases (n=17), 10 cases over cut-off of sFRLV did not occur PHLF, but 7 cases lower than cut-off occurred PHLF in 4 patients (57%) (p=0.019). Conclusion: sFRLV using EOB-MRI predicts PHLF in hepatectomy of MOS.
Read full abstract