2703. Suboptimal Hepatobiliary Phase Image in Gd-EOB-DTPA-Enhanced Liver MRI For the Surveillance of HCC: Predictive Factors
Authors* Denotes Presenting Author
Hunkyu Ryeom *;
Kyungpook National University Hospital
Objective:
To determine the influencing laboratory values leading to suboptimal hepatobiliary phase (HBP) image in Gd-EOB-DTPA enhanced liver MRI for the surveillance of HCC in patients with chronic liver disease.
Materials and Methods:
A total of 307 patients (236 men, 71 women, mean 64 years old) with chronic liver disease (CLD) who underwent liver MRI using Gd-EOB-DTPA for surveillance of HCC were included in this study. Seventy-nine patients with higher signal intensity (SI) of the portal vein than adjacent liver parenchyma in HBP images during the recent five years and 228 consecutive patients who underwent liver MRI during recent 3 months were included. Correlation analysis between the liver function tests (LFT) and estimated glomerular filtration rate (eGFR) and several parameters calculated from the SI measurements on the HBP images were performed: SI ratio between the portal vein and adjacent liver parenchyma, L/PV; SI ratio between liver parenchyma and spleen, L/S; presence of biliary tree excretion, B-excretion. Comparison analysis was done between the groups of suboptimal HPB images (L/PV = 1, L/S < 1.5, no B-excretion) and optimal image groups were done. The receiver operating characteristics (ROC) analysis was performed to determine optimal cutoff values.
Results:
In correlation analysis, L/PV showed moderate correlation with serum albumin level (r = .562, p < .01). L/S revealed mild correlation with serum albumin level (r = .446, p < .01). The eGFR showed no correlation with any HPB image parameters tested. In comparison between B-excretion and no B-excretion groups ALP, total bilirubin, direct bilirubin, total protein, albumin values were different. Optimal cutoff value for no B-excretion was a total bilirubin level of 1.49 mg/dL (AUC 0.749, sensitivity 60%, specificity 77%). Bright portal vein (L/PV = 1) was associated with direct bilirubin, total protein, and albumin values. The optimal cutoff value for L/PV = 1 was albumin level of 4.05 g/dL (AUC 0.971, sensitivity 65%, specificity 82%). Comparison between L/S =1.5 and L/S < 1.5 groups showed different AST, ALT, ALP, total bilirubin, direct bilirubin, total protein, and albumin levels. The optimal cutoff value for L/S <1.5 was direct bilirubin level of 0.83 mg/dL (AUC 0.830, sensitivity 69%, specificity 84%). The optimal cut off value of the inadequate HBP image (L/S < 1.5 and no B-excretion) was serum direct bilirubin level of 0.76 mg/dL (AUC 0.762, sensitivity 76%, specificity 77%). Suboptimal HBP image for surveillance of HCC in patients with CLD is associated with serum direct bilirubin value. The renal function is not related to suboptimal HBP images.
Conclusion:
In liver MRI using Gd-EOB-DTPA for surveillance of HCC, suboptimal HBP image quality can be predicted by liver function tests before performing MRI. In patients with CLD and high serum bilirubin level use of Gd-EOB-DTPA contrast material can be avoided.
