Diagnostic sensitivity of hepatocellular carcinoma imaging and its application to non-cirrhotic patients

Background and Aims: The American Association for the Study of Liver Disease issued guidelines that proposed that hepatocellular carcinoma (HCC) can be diagnosed if a mass is larger than 2 cm in a cirrhotic liver and shows typical features of HCC at triphasic liver computed tomography (CT) or dynamic magnetic resonance imaging (MRI). In non-cirrhotic livers, the criteria were not applicable. The aim of the present study was to retrospectively analyze the sensitivity of imaging by samples of definite HCC postoperatively and test their application to diagnose HCC in non-cirrhotic livers. Methods: From January 2006 to November 2008, a total of 343 pathologically-diagnosed HCC patients via surgical resection were reviewed. Among the 343 patients, 204 patients had undergone liver CT examination, and 80 patients underwent MRI examination; serum alpha-fetoprotein had been checked for all 343 patients prior to operation. The diagnostic sensitivity of HCC by imaging was evaluated and compared in patients with/without cirrhosis by ultrasound and histology. Results: The diagnostic sensitivity of HCC by single imaging was approximately 65-80% (liver CT or MRI). A higher sensitivity of HCC diagnosis was found in patients with ultrasound-diagnosed cirrhosis than non-cirrhosis, but the difference in sensitivity disappeared after histologically-cirrhotic validation. The results indicated that regardless of the presence or absence of cirrhosis (histology), a typical vascular pattern could diagnose HCC with equally high sensitivity. Conclusions: We provide evidence that the sensitivity of HCC diagnosis by imaging is not influenced by the cirrhotic background. Further study is needed to validate the specificity and accuracy.