Perfusion computed tomography of the liver

Background: Perfusion CT (P-CT) is a relatively new imaging technique that permits the visual and quantitative assessment of the micro- and macrocirculation of a target organ and focal lesions. P-CT has shown promising results in the evaluation of hyper-vascular tumors such as hepatocellular carcinoma (HCC). HCC is the sixth most common cancer globally and it has a poor prognosis when discovered at a late tumor stage. Any improvement in HCC detection would be directly beneficial for patient care. This thesis aims to investigate the strengths and limitations of whole liver P-CT and to evaluate if PCT can improve the detection of hyper-vascular liver lesions in patients with chronic liver disease. Methods: Study I: Twenty-four patients, who underwent dynamic P-CT for detection of HCC were retrospectively divided into three groups: (1) without portal-venous hypertension (PVH) (n = 8), (2) with PVH (n = 8), (3) with PVH and thrombosis (n = 8). Time to peak splenic- and peak renal enhancement (PSE resp. PRE), as well as arterial liver perfusion (ALP), portal- venous liver perfusion (PLP) and hepatic perfusion-index (HPI) of the liver and HCC derived from PSE- versus PRE- based modelling were compared between the groups. Study II: Group A (n=15) and Group B (n= 38) underwent P-CT using 50 ml contrast medium (CM). Group B underwent an additional standard multiphasic liver CT using 120ml (70-143 ml). Triple-arterial CT image sets were reconstructed from P-CT by fusing three dedicated arterial time points. Triple-arterial CT and single-arterial CT were compared by two readers (R1, R2), who assessed subjective image quality (IQ) and HCC detection rate. A third reader assessed objective IQ.Study III: Fifty study participants (Group A) were scanned with P-CT, a high CM volume protocol and bolus-tracking technique to depict ten arterial phases. Time attenuation curves were created for hyper-vascular liver lesions, liver parenchyma and hepatic vascular structures. 16 participants of Group A with lesions were further analyzed and radiation dose-neutral quadruple arterial phase image sets were created (Group A1). Group A1 was then compared to a Control Group (Group B) consisting of 16 consecutive patients undergoing standard single arterial phase scans. Lesion depiction and quantitative IQ were compared. Results: Study I: Time to PSE was significantly delayed in PVH groups 2 and 3 (P = 0.02), whereas PRE was similar in groups 1, 2 and 3 (P > 0.05). In group 1, liver and HCC perfusion parameters were similar for PSE- and PRE-based modelling (all P > 0.05), while significant differences were seen for PLP and HPI (liver only) in group 2 and ALP in group 3 (all P < 0.05). Study II: The mean CTDIvol of triple-arterial CT and single-arterial CT was equivalent (P=0.73). Triple-arterial CT showed lower image noise and better contrast-to-noise-ratio (P<0.001, P=0.032, respectively), but no significant difference in lesion-to-liver-contrast-ratio (P=0.31). Subjective IQ was good for both protocols. The detection rate of the 65 HCC lesions was 82%/83% (R1/R2) at triple-arterial CT and 80%/77% (R1/R2) at single-arterial CT (P=0.4). Study III: Both Group A1 and B had 33 hyper-enhancing liver lesions each. The mean CTDIvol of quadruple-arterial CT and single-arterial CT was equivalent (P=0.16). The mean time to reach peak lesion-to-liver contrast (LLC) was 20.1s (±4.2s) with a range of 12.5s to 29.1s. Quadruple arterial CT performed significantly better than the Control Group in regards to LLC (P= .009), CNR (P= .002), Image Noise (P<0.001) and hepatic artery enhancement(P