Evaluation of blood perfusion in liver cirrhosis by dynamic contrast enhanced computed tomography
Background Hepatic blood perfusion has been suggested to change according to the severity of hepatic fibrosis and cirrhosis. This clinical study aimed to examine changes in hepatic arterial perfusion (AP) and portal perfusion (PP), by dynamic contrast enhanced computed tomography (DCE-CT), in patients with varying degrees of cirrhosis categorised by Child Pugh score (CP). We compared the perfusion values and the hepatic perfusion index (HPI) ((AP/AP+PP) x 100) in between the three respective CP groups, to evaluate if AP, PP and HPI correlated with the CP level and hepatic venous pressure gradient (HVPG). Methods In a prospective non-randomised study approved by the local ethics committee (H-15000815), we included patients clinical diagnosed with cirrhosis. We performed a standard categorisation of their liver disease by: calculating CP, measurement of the HVPG, gastroduodenoscopy, blood sampling and a DCE-CT. Patients were consecutively recruited from Department of Hepatology, Rigshospitalet from Dec 2015 to June 2017. We excluded patients with severely impaired kidney function (eGFR<30ml/min), cancer and patients with a TIPS. Results 31 patients with cirrhosis were included (75,8 % (22/29) male) with a mean of age 57 years (range 33-78). The most common ethiology was alcohol 62% (18/29). 74 % of the patients (20/27) had varices at the time of gastroduodenoscopy, and 27,5% (8/29) were treated with propranolol, two with CP A, three B and three C. The CP was A in eight patients B in 17 patients, and C in six patients. The mean values of AP were significantly increased in patients with CP C vs. A+B, (70,4 ml/ vs. 43.1) ml/min/100ml tissue (p=0.0003). Patients with Child Pugh A tended to have higher PP values, 117,7 and Child B+C 105.3 ml/min/100ml tissue, respectively (p=0,44). The HPI increased significant with the degree of cirrhosis (mean A 29.6/B 30,96/C 44.9 ml/min/100ml tissue ANOVA p=0,046). Testing for linear regression with PP and HVPG values, the r-values raised with CP score (CP A r=0,137 p=0,049, CP B r=0,314 p=0,215 and CP C r=0,427 p=0,12). We found no correlation between the HVPG and PP. Patients treated with propranolol had non-significant lower PP than patients without betablockade (mean 88,1/115,6 ml/min/100ml tissue p=0,078). Conclusion In this study we found an increase in the arterial hepatic perfusion (AP) comparing patients with CP A+B cirrhosis to patients with CP C and the resultant hepatic perfusion index. The portal flow flow (PP) was not significantly reduced. This shows that DCE-CT scanning is capable of detecting changes in the liver perfusion in patients with cirrhosis with various clinical degrees of liver incompensation.