Background
MR elastography is a noninvasive technique that provides high diagnostic accuracy for the staging of liver fibrosis; however, it requires external hardware and mainly assesses the right lobe.
Purpose
To evaluate the diagnostic performance of MRI cine‐tagging for staging fibrosis in the left liver lobe, using biopsy as the reference standard.
Study Type
Institutional Review Board (IRB)‐approved two‐center prospective study.
Population
Seventy‐six patients with chronic liver disease who underwent an MRI cine‐tagging examination and a liver biopsy within a 6‐week interval.
Field Strength/Sequence
2D‐GRE multislice sequence at 3.0T with spatial modulation of the magnetization preparation sequence and peripheral pulse‐wave triggering on two coronal slices chosen underneath the heart apex to capture maximal deformation with consecutive breath‐holds adapted to patient cardiac frequency.
Assessment
A region of interest was selected in the liver close to the heart apex. Maximal strain was evaluated with the harmonic phase (HARP) technique.
Statistical Tests
Spearman's correlation, Kruskal–Wallis test, Mann–Whitney U‐test, and receiver operating characteristic (ROC) analysis were performed.
Results
Liver strain measured on tagged images decreased with higher histological fibrosis stage (ρ = –0.68, P < 0.0001). Strain values were significantly different between all fibrosis stages (P < 0.0001), and between groups of fibrosis stages ≤F3 vs. F4 (P < 0.05). Areas under the ROC curves were 0.95 (95% confidence interval: 0.89–1.00) to distinguish fibrosis stages F0 vs. F4, 0.81 (0.70–0.92) for stages F0 vs. ≥F1, 0.84 (0.76–0.93) for stages ≤F1 vs. ≥F2, 0.86 (0.78–0.94) for stages ≤F2 vs. ≥F3, and 0.87 (0.77–0.96) for stages ≤F3 vs. F4.
Data Conclusion
MRI cine‐tagging is a promising technique for measuring liver strain without additional elastography hardware. It could be used to assess the left liver lobe as a complement to current techniques assessing the right lobe.
Level of Evidence: 1
Technical Efficacy: 3
J. Magn. Reson. Imaging 2020;51:1570–1580.