BMI-dependent discordance between transient and MR elastography: risk of fibrosis overestimation in severe obesity

Elastography is central to fibrosis risk stratification in metabolic dysfunction-associated steatotic liver disease (MASLD). While vibration-controlled transient elastography (VCTE) is most widely used, magnetic resonance elastography (MRE) is more accurate. A discordance between them has been reported, and body mass index (BMI) has been proposed as a determinant. We investigated the impact of BMI on VCTE-MRE discordance in a real-world cohort.

This retrospective study included 268 adults with MASLD who underwent VCTE and MRE within 30 days. The primary endpoint was high-grade cross-threshold discordance, defined as VCTE ≥ F3 with MRE < F2 or MRE ≥ F3 with VCTE < F2. Secondary endpoints included discordance at ≥F2,≥F3, and ≥ 2-stage disagreement. Predictors of high-grade discordance were evaluated using logistic regression adjusted for age, sex, and diabetes. BMI-stratified calibration models were developed to harmonize TE values into MRE-equivalent scores.

VCTE and MRE showed moderate correlation (Spearman ρ = 0.54; log-Pearson r = 0.70) and moderate staged agreement (κ = 0.50). High-grade discordance occurred in 19.0% overall and increased sharply with BMI: 7.7% (< 25), 10.3% (25-29.9), 21.4% (30-34.9), and 64.7% (≥ 35). Discordance was predominantly TE-high/MRE-low. Multivariable analysis showed that every 5 kg/m² increase in BMI nearly tripled the odds of high-grade discordance (OR 2.70; 95% CI 1.93-3.93). Calibration modelling revealed that advanced VCTE values in patients with BMI ≥ 35 corresponded to MRE values well below the threshold for advanced fibrosis.

In MASLD, VCTE-MRE discordance is strongly BMI-dependent and clinically consequential in severe obesity, supporting BMI-aware fibrosis triage and modality harmonization.