Object: Among several non-invasive methods of liver fat analysis, the most important role is played by MR imaging and spectroscopy (MRS). This study describes the 1H MRS at 3T measurement of liver fat volume fraction Φ(fat) in a group of liver transplant patients, an at-risk group for the development of de novo steatosis.
Materials and methods: Seventy-seven liver transplant recipients who underwent routine protocolar posttransplant examination were divided into three groups: CON-PAT (control group for the cross validation test, 48 patients), PAT-PAT (patients test group for the cross validation test, 29 patients), and PAT (pooled data). Single voxel 1H MRS at 3T was used for the determination of Φ(fat) and histology results (His) were used as the reference standard.
Results: Linear and non-linear regression models were used to describe the relationship between Φ(fat) and His. Strong correlation was found for both models with r = 0.83-0.94 (P < 0.001); a higher r was found for non-linear regression in all tested groups. Areas under receiver operation curves were calculated for cut points His ≥ 5 and > 33% and were found in the range of 0.77-0.86. Fibrosis influences the calculation of Φ(fat) and different slopes were obtained for fibrosis stages F0-F1 and F2-F3, respectively.
Conclusion: Significant correlation was found between the results of histology and 1H MRS measurement of liver fat content. The method is suitable for non-invasive repetitive examination of liver fat in liver-transplants patients between protocol biopsies and for the screening of steatosis in other liver diseases.