Premium
Can point shear wave elastography differentiate focal nodular hyperplasia from hepatocellular adenoma
Author(s) -
Taimr Pavel,
Klompenhouwer Anne Julia,
Thomeer Maarten G. J.,
Hansen Bettina E.,
Ijzermans Jan N. M.,
de Man Robert A.,
de Knegt Robert J.
Publication year - 2018
Publication title -
journal of clinical ultrasound
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.272
H-Index - 61
eISSN - 1097-0096
pISSN - 0091-2751
DOI - 10.1002/jcu.22603
Subject(s) - medicine , focal nodular hyperplasia , intraclass correlation , elastography , radiology , hyperplasia , nuclear medicine , ultrasound , hepatocellular carcinoma , pathology , clinical psychology , psychometrics
Purpose Focal nodular hyperplasia (FNH) and hepatocellular adenoma (HCA) are liver tumors that require different management. We assessed the potential of point shear wave elastography (pSWE) to differentiate FNH from HCA and the interobserver and intraobserver reliability of pSWE in the examination of these lesions and of native liver tissue (NLT). Methods The study included 88 patients (65 FNH, 23 HCA). pSWE was performed by two experienced liver sonographers (observers 1 [O1] and 2 [O2]) and acquired within the lesion of interest and NLT. Group differences, optimal cutoff for characterization and interobserver reliability was assessed with Mann‐Whitney‐U, area under the ROC curce (AUROC) and intraclass correlation coefficient (ICC). Intraobserver reliability in NLT was assessed in 20 healthy subjects using ICC. Results Median stiffness was significantly higher in FNH than in HCA (7.01 kPa vs 4.98 kPa for O1 ( P = 0.017) and 7.68 kPa vs 6.00 kPa for O2 ( P = 0.031)). A cutoff point for differentiation between the two entities could not be determined with an AUROC of 0.67 (O1) and 0.69 (O2). Interobserver reliability was good for lesion‐ stiffness (ICC = 0.86) and poor for NLT stiffness (ICC = 0.09). In healthy subjects, intraobserver reliability for NLT‐stiffness was poor for O1 (ICC = 0.23) and moderate for O2 (ICC = 0.62). Conclusion This study shows that pSWE cannot reliably differentiate FNH from HCA. Interobserver and intraobserver reliability for pSWE in NLT were insufficient. Interpretation of results gained with this method should be done with great caution.