Purpose
To assess the diagnostic value of multiparametric magnetic resonance imaging (MRI) including dynamic Gd-EOB-DTPA-enhanced (DCE) and diffusion-weighted (DW) imaging for diagnosis and staging ...of hepatic fibrosis in primary sclerosing cholangitis (PSC) using transient elastography as a standard reference.
Material and methods
Multiparametric MRI was prospectively performed on a 3.0-Tesla scanner in 47 patients (age 43.9±14.3 years). Transient elastography derived liver stiffness measurements (LSM), DCE-MRI derived parameters (hepatocellular uptake rate (Ki), arterial (Fa), portal venous (Fv) and total (Ft) blood flow, mean transit time (MTT), and extracellular volume (Ve)) and the apparent diffusion coefficient (ADC) were calculated. Correlation and univariate analysis of variance with post hoc pairwise comparison were applied to test for differences between LSM derived fibrosis stages (F0/F1, F2/3, F4). ROC curve analysis was used as a performance measure.
Results
Both ADC and Ki correlated significantly with LSM (r= -0.614;
p
<0.001 and r= -0.368;
p
=0.01). The ADC significantly discriminated fibrosis stages F0/1 from F2/3 and F4 (
p
<0.001). Discrimination of F0/1 from F2/3 and F4 reached a sensitivity/specificity of 0.917/0.821 and 0.8/0.929, respectively. Despite significant inter-subject effect for classification of fibrosis stages, post hoc pairwise comparison was not significant for Ki (
p
>0.096 for F0/1 from F2/3 and F4). LSM, ADC and Ki were significantly associated with serum-based liver functional tests, disease duration and spleen volume.
Conclusion
DW-MRI provides a higher diagnostic performance for detection of hepatic fibrosis and cirrhosis in PSC patients in comparison to Gd-EOB-DTPA-enhanced DCE-MRI.
Key Points
• Both ADC and hepatocellular uptake rate (Ki) correlate significantly with liver stiffness (r= -0.614; p<0.001 and r= -0.368; p=0.01).
• The DCE-imaging derived quantitative parameter hepatocellular uptake rate (Ki) fails to discriminate pairwise intergroup differences of hepatic fibrosis (p>0.09).
• DWI is preferable to DCE-imaging for discrimination of fibrosis stages F0/1 to F2/3 (p<0.001) and F4 (p<0.001).
Summary
Background
In patients with primary sclerosing cholangitis follow‐up magnetic resonance imaging (MRI) with magnetic resonance cholangiopancreatography (MRCP) is performed by many centres, ...particularly for the early detection of biliary malignancies and strictures. Clinically meaningful MRI‐based definitions of primary sclerosing cholangitis related complications are, however, lacking.
Aim
To investigate how primary sclerosing cholangitis experts interpret follow‐up MRI/MRCP with a focus on conclusions that may impact clinical decision‐making in primary sclerosing cholangitis.
Methods
Within the International Primary Sclerosing Cholangitis Study Group, an online survey on 16 real‐life primary sclerosing cholangitis cases including clinical and biochemical information as well as a T2‐weighted liver MRI/3D‐MRCP was conducted. The interpretation of images and subsequent recommendations were assessed using a multiple‐choice questionnaire. An inter‐rater reliability calculation (Fleiss′ kappa) was performed and factors potentially affecting the interpretation of magnetic resonance images were analysed using generalised linear mixed‐effect models.
Results
Forty‐four members/associates of the International Primary Sclerosing Cholangitis Study Group (median experience in the care of primary sclerosing cholangitis patients: 14 years) completed the survey. The MRI interpretation significantly varied among the participants. The lowest agreement was found with respect to the indication to perform subsequent endoscopic retrograde cholangiopancreatography (ERCP; Κ = 0.12, 95%CI 0.11‐0.14). Elevated total bilirubin was the variable with the strongest effect on the rate of suspected dominant strictures, cholangiocarcinoma or ERCP recommendations. Liver cirrhosis did not prevent participants from recommending ERCP. Overall, the survey participants′ recommendations contrasted the real‐life management and outcome.
Conclusions
In primary sclerosing cholangitis, the interpretation of follow‐up MRI/3D‐MRCP significantly varies even among experts and seems to be primarily affected by bilirubin levels. Generally accepted MRI‐based definitions of primary sclerosing cholangitis‐related complications are urgently needed.
Linked ContentThis article is linked to Stratton and Williams paper. To view this article visit https://doi.org/10.1111/apt.14898.
To evaluate the benefit of extracellular gadolinium-based contrast agent (GBCA) enhanced magnetic resonance imaging (MRI) in addition to conventional non-enhanced T2-weighted imaging (WI) for ...quantification of inflammatory or fibrotic alterations in the liver parenchyma of patients with primary sclerosing cholangitis (PSC).
MRI (3 T) examinations were reviewed retrospectively by two radiologists in 27 PSC patients (age 42.9±15.6 years), and 19 controls. Regions of interest (ROIs) were drawn onto T2 hyperintense and T2 isointense areas and copied to section position matched non-enhanced and delayed-phase contrast-enhanced T1WI. Signal intensities (SI) obtained from ROIs of the multiphase T1WI were used to calculate relative liver enhancement (RLE). The interobserver agreement of RLE and quantified T2 signal was calculated using Bland–Altman analysis. RLE assessed for both T2 hyperintense (RLEhyper) and T2 isointense (RLEiso) areas were compared in patients and controls (RLEhealthy).
The interobserver agreement of RLE in affected hyperintense areas (bias −0.77, limits of agreement −51.7 to 50.1) was superior to the quantification of T2 signal only in these areas (bias −3.35, limits of agreement −162.4 to 155.7). The RLEhyper (86.2±9.7%) was higher than the RLEiso (59.8±6.2%, p=0.03) and the RLEhealthy (53.2±2.7%, p=0.002). The mean RLEiso was not significantly different from the RLEhealthy (p=0.3).
The extracellular gadolinium-based RLE of T2 hyperintense areas could be a useful add-on for routine follow up MRI in the detection of early inflammatory changes, possibly preceding formation of fibrotic scarring in PSC patients, if validated in larger cohorts.
•The RLE and T2 signal intensity are typically shifted in primary sclerosing cholangitis.•Hyperintensity on T2 weighted imaging is connected to gadolinium-based contrast enhancement.•The relative liver enhancement (RLE) provides a quantitative measure of the liver parenchyma.
PDF
