Hepatic function imaging using dynamic Gd‐EOB‐DTPA enhanced MRI and pharmacokinetic modeling

    loading  Checking for direct PDF access through Ovid


INTRODUCTIONHepatic function evaluation is important in preoperative planning for liver surgery 1. Serum bilirubin (total bilirubin (TBil) and direct bilirubin (DBil)) and certain proteins (albumin (Alb), prealbumin (PA) concentration, and prothrombin time (PT)) are used widely in clinical practice to quantify the liver function, because they reflect the function of liver 2. However, quantification of regional liver function is more important in surgery planning 4. Thus, medical imaging methods with localized function evaluation have been proposed. For example, technetium‐99m diethylenetriaminepenta acid galactosyl human serum albumin (99mTc‐GSA) clearance, which reflects the liver function, quantified by single photon emission computed tomography (SPECT) has been found to be correlated with indocyanine green (ICG) retention rate–related liver function 5. However, it suffers from low clinical availability, ionizing radiation exposure, and low spatial resolution.Recently, dynamic contrast enhanced (DCE) MR imaging with a hepatobiliary contrast agent, Gd‐EOB‐DTPA, has been proposed to quantify hepatic functions 6. Gd‐EOB‐DTPA can not only permeate into the extravascular extracellular space (EES), but also be transported into hepatocytes through polypeptides organic anion transporter OATP1B1 and OATP1B3, and then be excreted by multidrug resistance protein MRP2 through bile 8. The pharmacokinetics of Gd‐EOB‐DTPA provides a good opportunity to quantify regional hepatic function, because the transfer and excretion rate directly reflect the hepatic function. First, descriptive parameters were proposed using a simple imaging protocol with limited dynamic phases 10. However, the descriptive parameters do not have physical meanings and cannot model the real hepatic pharmacokinetics of Gd‐EOB‐DTPA. Recently, a dual‐input two‐compartment model was proposed and investigated 11. However, the hepatic function parameters generated from pharmacokinetic modeling of DCE‐MRI with Gd‐EOB‐DTPA have not been validated with clinical serum‐based liver function tests. Moreover, the most suitable assumptions of this model are still needed to be determined with a clinical available imaging protocol, including the output assumptions of contrast excretion through bile and veins.The purpose of this study is to use the 6‐min dynamic Gd‐EOB‐DTPA–enhanced MRI protocol to fit four dual‐input two‐compartment models with different output assumptions and a one‐compartment model known as the Van Beers model 12 to determine whether hepatic function parameters quantified by pharmacokinetic modeling correlate with serum‐based liver function tests, and to compare goodness of fit using the Akaike information criterion (AIC) fitting error.THEORYBased on the previously developed dual‐input two‐compartment model 11, we further discussed the outputs of models based on the pharmacokinetics of the contrast and imaging protocol. The models investigated in this study are composed of two compartments: the EES and intracellular space. The whole pharmacokinetic process of Gd‐EOB‐DTPA 13 is show in Figure 1. For an EES compartment with concentration , there are two inputs: the inflow from the hepatic artery with concentration , and the portal venous inflow with concentration , in which and are the arterial and portal venous delay, respectively. Usually, can be neglected 11, as the region of interest (ROI) of portal venous input function (PIF) we selected was much nearer than that of arterial input function (AIF). Considering the contrast outputs from EES, one output connects the intracellular space with a fixed intracellular uptake rate ; another output flows out through the hepatic vein with concentration . is used as a substitute of 11, because it is difficult to directly measure from images because of the complicated structure of the hepatic vein.

    loading  Loading Related Articles