A lexicon for hepatocellular carcinoma surveillance ultrasonography: benign versus malignant lesions

Abstract

Background/AimsTo suggest a lexicon for liver ultrasonography and to identify radiologic features indicative of benign or malignant lesions on surveillance ultrasonography.MethodsThis retrospective study included 188 nodules (benign, 101; malignant, 87) from 175 at-risk patients identified during surveillance ultrasonography for hepatocellular carcinoma. We created a lexicon for liver ultrasonography by reviewing relevant literature regarding the ultrasonographic features of hepatic lesions. Using this lexicon, two abdominal radiologists determined the presence or absence of each ultrasonographic feature for the included hepatic lesions. Independent factors associated with malignancy and interobserver agreement were determined by logistic regression analysis and kappa statistics, respectively.ResultsLarger tumor size (odds ratio [OR], 1.12; 95% confidence interval [CI], 1.06-1.183; P<0.001), multinodular confluent morphology (OR, 7.712; 95% CI, 1.053-56.465; P=0.044), thick hypoechoic rim (OR, 5.878; 95% CI, 2.681-12.888; P<0.001), and posterior acoustic enhancement (OR, 3.077; 95% CI, 1.237-7.655; P=0.016) were independently associated with malignant lesions. In a subgroup analysis of lesions <2 cm, none of the ultrasonographic features were significantly associated with malignancy or benignity. Interobserver agreement for morphology was fair (κ=0.36), while those for rim (κ=0.427), echogenicity (κ=0.549), and posterior acoustic enhancement (κ=0.543) were moderate.ConclusionsFor hepatic lesions larger than 2 cm, some ultrasonography (US) features might be suggestive of malignancy. We propose a lexicon that may be useful for surveillance US.

Sonographic appearance of focal liver lesions and likelihood of hepatocellular carcinoma in adult Thais with chronic hepatitis B virus infection

“…[14][15][16] Nonetheless, nodule echogenicity does not distinguish HCC and benign nodules. 17 While uncommon in our study, a hypoechoic rim was found more frequently with HCC (20%) than benign (0.2%) nodules. In 2008, Forner et al reported that a halo or hypoechoic rim is found in 76.47% of HCCs, whereas in only 11.54% in non-HCC nodules.…”

“…All US surveillance was performed by or under the supervision of board-certified radiologists. The US features assessed included nodule size, echogenicity, rim, and posterior acoustic enhancement (adopted from An et al 17 ).…”

“…[14][15][16] A hypoechoic rim is correlated with a fibrous capsule of the HCC, 25 especially when it has a thick hypoechoic rim (≥2 mm). 17 Posterior acoustic enhancement is a US artifact caused by different attenuation properties of tissues found in fluid-filled lesions, for example, simple cysts or simple internal architecture like the homogenous cell population in HCC. 26 US technique also affects the presence of this artifact, and US surveillance using the spatial compounding mode will decrease this artifact.…”

“…Posterior acoustic enhancement is found in 46%-59% of HCCs, mostly those larger than 3 cm, 15,26 and is independently associated with HCC. 17…”

Direct comparison of biopsy techniques for hepatic malignancies

“…In this study, the results by directly comparing the sensitivity of the three methods suggest that CNB may be the most sensitive one (93.6%) for both primary liver cancer and metastatic cancers, and the TIC has comparable sensitivity in metastatic lesions (93.4%). In the clinical practice, prior to receiving liver biopsies, those patients usually undergo ultrasonographic and radiologic examinations, revealing certain degree of suspicion for malignancy [ 9 , 10 ]. The results of the current study, including the sensitivity of each method as well as the factors associated with the sensitivity, may help the clinical physicians select the appropriate techniques with satisfactory sensitivity and safety for patients with suspected hepatic malignancies on images.…”