Controlled attenuation parameter for steatosis grading in chronic hepatitis C compared with digital morphometric analysis of liver biopsy: impact of individual elastography measurement quality.

Background and objective Controlled attenuation parameter (CAP) diagnostic performance for steatosis grading has been controversial and considerable observer-related variability in liver biopsy has been reported. This is a subanalysis of a larger chronic hepatitis C study on noninvasive fibrosis staging. Materials and methods Patients were prospectively enrolled for paired liver biopsy and transient elastography. Biopsy fragments were subjected to digital morphometric steatosis quantification. Associated patient and technical factors, including a newly described elastogram quality score, were evaluated. Results A total of 312 patients were included in the final analysis. The mean liver stiffness was 8.7 +/- 2.1 kPa. Morphometry showed S0 in 19.2% of patients, S1 in 28.5%, S2 in 31.1%, and S3 in 21.2%. CAP showed S0 in 11.2% of patients, S1 in 26.6%, S2 in 56.7%, and S3 in 5.4%. Spearman coefficient showed a positive and independent correlation between CAP and morphometric analysis (r = 0.48, P < 0.05), except for distinguishing S1 and S2 (P = 0.11). Area under the receiver operating characteristic curves for the presence or absence of steatosis was 0.944; differentiation between levels I, II, and III were 0.776, 0.812, and 0.879. Elastogram quality independently predicted accuracy [odds ratio (OR): 6.95, 95% confidence interval (95% CI): 4.45-9.06 as well as CAP interquartile range OR: 2.81, 95% CI: 1.67-3.99] and liver stiffness (OR: 0.78, 95% CI: 0.51-0.80). Conclusion We present an external validation for CAP against the objective steatosis quantification provided by digital morphometry. Fairly good performance indicators were found, except for S1 versus S2 differentiation. Variability and higher liver stiffness were associated with lower performance. Achieving higher quality measurements, however, overcame such limitations with excellent accuracy. Copyright (C) 2018 Wolters Kluwer Health, Inc. All rights reserved.