First Automated Software
Quantitative Liver Function

Liver Surface Nodularity as an Estimate of Quantitative Liver Function and Portal Hypertension

John Hoefs, M.D
Depts of Medicine, Divs of GI & Hepatology, University of California Irvine, Irvine, California
Hepatology, Liver Specialty Center, Irvine, CA, United States


Background and Aims:

A nodular liver (NL) surface correlates with the presence of cirrhosis and, therefore, is an indirect estimate of hepatic function, portal hypertension and risk of clinical disease. The Perfused Hepatic Mass (PHM) using Hepatiq is a precise, direct measure of quantitative hepatic function and functional spleen volume (fSV) that correlates with the hepatic functional mass (AmJGastro;92:2054) and with clinical disease (Hepat;55:1019). We compared a nodular liver surface by ultrasound (US) with shear wave Elasticity (SWE), functional liver volume (fLV: N 7-12), fSV (N < 2.5 cc/lb IBW) and PHM (N 100-110). HYPOTHESIS:  NL will correlate with PHM and fSV.


Patients: We evaluated 190 consecutive patients with SWE at US and Hepatiq. Diagnosis of C was based on combinations of parameters including liver surface nodularity, abnormal blood tests (platelet count, INR, Bilirubin), CT/MRI showing collaterals, Varices by EGD and liver biopsy.  Patients had cryptogenic cirrhosis, HBV, HCV, PBC, PSC, ACAH and miscellaneous. Non-cirrhotic patients (NC) included 144 with underlying pre-cirrhotic CLD, non-cirrhotic portal hypertension and acute liver disease with jaundice. 46 cirrhotic patients (C) included those never having clinical complications (ascites, Variceal Bleed and encephalopathy) (C1=19 patients); those with these complications in past, but resolved not requiring further treatment (C2=13); active complications requiring ongoing treatment (C3=11) and those on the liver transplant list (C4=3).  C3 and C4 have decompensated cirrhosis (DC).

Method: All patients had a fasting US (GE LOGIQ E9) with SWE (Shear wave velocity=SWV) and post-prandial Hepatiq. The quantitative liver spleen scan was performed with Hepatiq as previously described (Hepat;55:1019). Clinical data was abstracted from the records and expressed as mean (+/-SD). Data were expressed as the cumulative percent (Cum%) from normal to abnormal in groups with NL or Smooth surface (SS).


Hepatiq and US were successful in all 190 patients and SWE in 184 (97%). 38 patients with NL were compared with 152 patients with SS. 34/38 (89.5%) patients with a NL had C compared to 10/152 (6.6 %) (p<.001) with a SS. Using the Cum% (smallest to largest) as a way of evaluating the distribution of parameters in NL and SS, there was no difference for fLV. The distribution curve of SWV was similar, but the curve shifted to the right for C (not shown). The distribution of PHM and fSV were significantly different in NL vs SS (p<.001) (figure). Quantitative dysfunction (PHM<95) was found in 62% with NL vs 4.7 % with SS (p<.001). All patients with ascites/encephalopathy had PHM < 95 whether NL or SS. Splenomegaly (fSV >2.5) was found in 79% NL vs 16% with SS (p<.001). Of the 23 patients with SS and fSV > 2.5, 6 had C (2 with decompensation), 2 severe acute liver diseases, 1 non-C portal hypertension, 1 post- LT and 13 NC (10 with fSV<4).


NL correlated with C, quantitative liver dysfunction and portal hypertension (splenomegaly). 2. SS did not rule out C, serious liver dysfunction or absence of portal hypertension, but made these less likely. 3. Although hepatic surface nodularity is helpful, PHM and fSV should be measured directly to determine prognosis and to follow patients.

Click here to return to list of research publications.