Comparison of different methods to quantify future liver remnants after preoperative portal vein embolization to predict postoperative liver failure.

Background/aims: Even though preoperative portal vein embolization (PVE) can lead to hypertrophy of the future liver remnant (FLR) in candidates with small remnant liver prior to anticipated major hepatic resection, quantification of FLR after PVE is important to ensure adequate hepatic reserve in order to avoid postoperative hepatic failure. This study aims to determine the accuracy and reliability of three commonly used FLR quantification methods by correlating their ability to detect postoperative hepatic failure. The role of the degree of liver hypertrophy (DLH) in this context was also explored.

Methodology: The records of 98 consecutive patients considered for PVE prior to major hepatic resections were reviewed retrospectively. Out of these 98 patients, 66 patients who underwent major liver resection after PVE were included in the present study. Pre- and post-PVE FLR volume and volumes of other liver segments were studied using MRI scans. The three FLR quantification methods employed were: (A) the estimation of the Total Liver Volume (TLV) directly by MRI; (B) by indirect estimation of TLV from patients' body surface area; and (C) by indirect estimation of TLV by patients' body weight. The difference of pre- and post-FLR% determined the DLH by all three methods. Receiver-operator characteristic (ROC) curve analysis was used to demonstrate the efficacy of the three methods in predicting postoperative hepatic failure (HF).

Results: The assessment of FLR% by method B was significantly better (p < 0.005) than by method A. However, there was no significant difference among these two methods in determining the DLH. Ten out of 66 patients developed postoperative HF and 8 patients recovered. From the ROC curves plotted to demonstrate efficacy in predicting postoperative HF, it was evident that all three methods were comparable due to small FLR%, all methods having a significant predictability. The DLH also had significant predictability for postoperative HF but there was significant inverse correlation between the DLH and the pre-FLR%.

Conclusions: All 3 methods were equally efficient in predicting postoperative hepatic failure. The DLH assay alone should not be used to predict postoperative hepatic failure but should be used in conjunction with FLR%.