Do you obtain liver biopsy to confirm the diagnosis of cirrhosis if cirrhotic liver morphology is noted on imaging?

This question touches upon two interesting trends:

1) There is an increasing trend in Radiology to report "cirrhotic liver morphology" in the "Impressions" section. When you then review the Body of the report, often these cases are noted to only have a heterogeneous appearing liver with surface nodularity, but no other comments to support a diagnosis of cirrhosis (no splenomegaly/collaterals, no left lobe hypertrophy, nor ascites). These patients are typically older (>60 years old), underwent a scan for a non-liver indication, and otherwise have normal labs without evidence of advanced liver disease (normal platelet count >150K, normal albumin, normal INR, low MELD<10, and/or relatively normal liver enzymes <2xULN; if their FIB-4 score is e...

In general, if the radiographic imaging is compatible with cirrhosis and other signs of portal hypertension, I normally do not do a biopsy. There is also FibroScan that one could do to determine the liver stiffness- this is a noninvasive method to determine the degree of fibrosis/cirrhosis without doing biopsy.