Association Between Obesity and Discordance in Fibrosis Stage Determination by Magnetic Resonance vs Transient Elastography in Patients With Nonalcoholic Liver Disease

Background & Aims: Magnetic resonance elastography (MRE) and transient elastography (TE) are noninvasive techniques used to detect liver fibrosis in nonalcoholic fatty liver disease. MRE detects fibrosis more accurately than TE, but MRE is more expensive, and the concordance between MRE and TE have not been optimally assessed in obese patients. It is important to determine under which conditions TE and MRE produce the same readings, so that some patients can simply undergo TE evaluation to detect fibrosis. We aimed to assess the association between body mass index (BMI) and discordancy between MRE and TE findings, using liver biopsy as the reference, and validated our findings in a separate cohort. Methods: We performed a cross-sectional study of 119 adults with nonalcoholic fatty liver disease who underwent MRE, TE with M and XL probe, and liver biopsy analysis from October 2011 through January 2017 (training cohort). MRE and TE results were considered to be concordant if they found patients to have the same stage fibrosis as liver biopsy analysis. We validated our findings in 75 adults with nonalcoholic fatty liver disease who underwent contemporaneous MRE, TE, and liver biopsy at a separate institution from March 2010 through May 2013. The primary outcome was rate of discordance between MRE and TE in determining stage of fibrosis (stage 2–4 vs 0–1). Secondary outcomes were the rate of discordance between MRE and TE in determining dichotomized stage of fibrosis (1–4 vs 0, 3–4 vs 0–2, and 4 vs 0–3). Results: In the training cohort, there was 43.7% discordance in findings from MRE versus TE. BMI associated significantly with discordance in findings from MRE versus TE (odds ratio, 1.69; 95% confidence interval, 1.15–2.51; P =.008) after multivariable adjustment by age and sex. The findings were confirmed in the validation cohort: there was 45.3% discordance in findings from MRE versus TE. BMI again associated significantly with discordance in findings from MRE versus TE (odds ratio, 1.52; 95% confidence interval, 1.04–2.21; P =.029) after multivariable adjustment by age and sex. Conclusions: We identified and validated BMI as a factor significantly associated with discordance of findings from MRE versus TE in assessment of fibrosis stage. The degree of discordancy increases with BMI.