Premium
Elastogram quality assessment score in vibration‐controlled transient elastography: Diagnostic performance compared to digital morphometric analysis of liver biopsy in chronic hepatitis C
Author(s) -
Mendes L. C.,
Ferreira P. A.,
Miotto N.,
Zanaga L.,
Gonçales E. S. L.,
Pedro M. N.,
Lazarini M. S.,
Júnior F. L. G.,
Stucchi R. S. B.,
Vigani A. G.
Publication year - 2018
Publication title -
journal of viral hepatitis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.329
H-Index - 100
eISSN - 1365-2893
pISSN - 1352-0504
DOI - 10.1111/jvh.12822
Subject(s) - transient elastography , medicine , liver biopsy , biopsy , receiver operating characteristic , elastography , pathology , fibrosis , radiology , ultrasound
Summary Vibration‐controlled transient elastography ( VCTE ) is widely used for noninvasive fibrosis staging in chronic hepatitis C. However, internal validation is based solely on variability and success rate and lacks reproducible quality indicators. We analysed the graphic representation of shear wave propagation in comparison with morphometric results of liver biopsy, eliminating observer variability bias. Individual elastograms were classified according to two morphologic criteria: extension of wave propagation (length of the graphic representation) and shear wave dispersal (level of parallelism displayed in the elastogram). Then, a score based on these criteria stratified the elastogram in classes I through III (highest to lowest technical quality). Liver stiffness results of each measurement were compared with collagen contents in liver biopsy by morphometric analysis. A total of 3243 elastograms were studied (316 patients). Digital morphometry in liver biopsy showed significant fibrosis in 66% of samples and advanced fibrosis in 31%. Elastogram quality analysis resulted in 1438 class I measurements (44%), 1070 class II (34%) and 735 class III . Area under the receiver operating curve ( AUROC ) for severe fibrosis according to class (I, II and III ) was 0.941, 0.887 and 0.766, respectively. For advanced fibrosis, AUROC s were 0.977, 0.883 and 0.781, respectively. Spearman's correlation testing for all classes and levels of fibrosis demonstrated significant independent association ( r 2 = −.95, P < .01). Our study is the first to propose measurable quality criteria for VTCE and to validate them against objective assessment of liver biopsy through digital morphometric imaging analysis. We concluded that VCTE performance is significantly influenced by quality assessment of individual measurements. Considering these criteria in clinical practice may improve accuracy.