Associations of Ultrasound-Derived Fat Fraction and MRI PDFF Measurements: A Prospective Study in Pediatric Patients...

BACKGROUND. As pediatric metabolic dysfunction-associated steatotic liver disease (MASLD) increases in prevalence, ultrasound-based methods have emerged as accessible alternatives to MRI for hepatic fat quantification in children. OBJECTIVE. The purpose of the study was to evaluate associations of ultrasound-derived fat fraction (UDFF) measurements obtained under free-breathing and breath-hold conditions with proton-density fat fraction (PDFF) measurements in pediatric patients with suspected MASLD. METHODS. This prospective study enrolled pediatric patients with clinical suspicion for MASLD from November 2023 to October 2024. Participants underwent an investigational liver MRI examination with PDFF measurements and an investigational right upper quadrant ultrasound examination with UDFF measurements under both breath-hold and free-breathing conditions. A second operator repeated the ultrasound examinations in the first 14 patients to assess interobserver agreement. Concordance correlation coefficients (CCCs) and Spearman correlation coefficients were computed. Bland-Altman analyses were conducted, including linear regression fits. RESULTS. The study included 55 patients (33 male patients, 22 female patients; mean age, 13.8 ± 2.6 [SD] years; age range, 9-18 years). PDFF measurements ranged from 0.1% to 56.1%; UDFF measurements ranged from 2.0% to 36.2%. UDFF measurements showed good interobserver agreement under breath-hold (CCC = 0.79) and free-breathing (CCC = 0.76) conditions and excellent agreement between breath-hold and free-breathing conditions (CCC = 0.97). UDFF and PDFF measurements showed strong positive correlations under breath-hold (ρ = 0.84) and free-breathing (ρ = 0.83) conditions. In Bland-Altman analysis, UDFF showed no significant systematic bias versus PDFF under breath-hold condition (mean difference, -2.0%) but significant systematic underestimation bias under free-breathing condition (-2.6%); UDFF and PDFF showed a significant inverse proportional bias under both conditions (β = -0.5, p < .001). In patients with a PDFF of less than 5% (n = 28), UDFF showed a significant systematic overestimation bias versus PDFF under breath-hold (mean difference, 2.0%) and free-breathing (2.1%) conditions; in patients with a PDFF of 5% or greater (n = 27), UDFF showed a significant systematic underestimation bias versus PDFF under breath-hold (mean difference, -6.3%) and free-breathing (-7.4%) conditions. CONCLUSION. UDFF measurements in both breath-hold and free-breathing conditions showed strong positive correlations with PDFF measurements in children. However, UDFF overestimated and underestimated PDFF in patients with relatively low and high liver fat content, respectively. CLINICAL IMPACT. UDFF may have a role in screening for steatosis in children with suspected MASLD, although clinical integration must account for systematic biases between UDFF and PDFF measurements.