Feasibility of vertebral bone Diffusion Weighted MR Imaging (DWI) for early diagnosis of hepatic osteodystrophy in Primary Sclerosing Cholangitis
Sarah Keller1, Fabian Kording1, Hendrick Kooijman2, Christoph Schramm3, Roland Fischer4, Adam Gerhard1, Ansgar Lohse3, Harald Ittrich1, and Jin Yamamura1

1Diagnostic and Interventional Radiology, University Medical Center Hamburg Eppendorf (UKE), Hamburg, Germany, 2MRI, Philips GmbH, Hamburg, Germany, 3Department of Internal Medicine, University Medical Center Hamburg Eppendorf (UKE), Hamburg, Germany, 4Biochemistry, University Medical Center Hamburg Eppendorf (UKE), Hamburg, Germany


Hepatic osteodystrophy is a frequent complication in patients with chronic cholestatic and non-choleatatic liver disease, affecting up to 20-30% of patients [1, 2]. Besides conventional dual x-ray absorbtiometry (DXA) scans, diffusion-weighted MRI (DWI) has been performed for the evaluation of osteoporosis and osteopenia in risk patients. Follow up MRI examinations of PSC patients, to exclude malignancy and identify bile duct stenosis, are frequently performed in clinical routine. The aim of this study was to test the feasibility of add-on DWI sequences during routine examination for the detection of early changes of the bone marrow density (BMD) in comparison to DXA T-score values of the vertebral bone and healthy controls.


The aim of this study was to test the feasibility of add-on DWI sequences during routine examination for the detection of early changes of the bone marrow density (BMD) in comparison to DXA T-score values of the vertebral bone and healthy controls.


22 patients (age 49.5y;7-78y) with histopathological diagnosed PSC and 11 controls (age 48.0y;28-48y) were included. MR-imaging was performed at a Philips Ingenia 3.0 Tesla scanner (Ingenia, Philips Medical Systems, The Netherlands) with a eight-channel SENSE body coil. Axial T2w imaging (T2wI TSE Sense) for anatomical orientation (TR 1250ms; TE 80ms; FA 90; Fov 400x400mm) was performed over the epigastric region prior to the DWI. A diffusion weighted (DWI) spin echo echo-planar sequence (ssEPI) was generated in transversal orientation including the whole vertebral body (coil combination: TR 1985ms; TE 69ms; FOV 400x400mm; voxel size 1.79x1.79x3.0mm; slice thickness 5mm; intersection gap 0mm; b-factors 0, 50, 100, 200, 400, 800s/mm2; average 35 slices, NEX 2). SPIR technique was used for fat-suppression. A starting b-value of 50 s/mm² was applicated to suppress vascular signal in the initial T2w EPI image.

IMAGE ANALYSIS To avoid the susceptibility artefact of surrounding air in the basal lung, an upper lumbar vertebral body (L1 or L2) was selected. The grey value of the pixel corresponded to the ADC value [mm2/sx10-3] with a pixel-to-pixel ADC map calculated for each slice. The value itself was calculated with the equation Si(I) = Si(0) exp(-bi•ADC); where S(I) is the signal intensity measured on the ith b-factor image and b1 is the corresponding b-factor. S(0) estimates the signal intensity for a b-factor of 0s/mm2 i.e. without the noise induced by the MR measurement [3]. The ROIs (mean size 1.8mm2 (SD±0.56)) were localized onto the ADC-maps in the vertebral body excluding the intravertebral space (Fig. 1).

STATISTICAL ANALYSIS was performed using GraphPad Prism 6.0f (GraphPad Software Inc., USA). Correlation and significance was tested by the Student’s T-test and parametric Pearson correlation analyses. The difference was considered statistically significant if the significance level α = 0.05 was reached.


Patients mean ADC was slightly, but not significantly decreased compared to controls (02951±0.0648 mm2/s x10-3 vs. 0.3163±0.02221 mm2/sx10-3) (Fig. 2). No significant bias of the ROI location in the vertebral body (upper, mid and base portion) on the ADC-value was found. Mean T-score of the vertebral bone was -0.245±1.669, corresponding to a Z-score of -0.233±1.517. Using non-parametric Pearson test, a slight correlation between the patients ADC and T-score (r=0.3334; 95%CI-0.1280-0.6762; p=0.15) was evident. No difference between the mean ADC in patients with normal, osteopenic and osteoporotic BMD, according to T-score, was observed. Laboratory data resulted as followed: AP 231.0±180.16(U/l); IgG 13.1±1.4(g/l); Bilirubin 1.05±1.03(mg/dl); Calcium 2.3±0.09(mmol/l); bAP 29.04±26.93(μg/l); Osteocalcin 13.1±1.4(μg/l); 25OH-D3 22.67±10.03(μg/l); Parathormone 13±7.6(ng/l). The duration of PSC disease since diagnosis was 35.5 month (range 2-166.0 month) (Tabl. 1). Neither a correlation of the ADC with PTH (r=0.01143; 95%CI -0.4451-0.4632), 25OH VitD3 (r=0.4522; 95%CI -0.1643-0.8147), Osteocalcin (r=0.03783, 95%CI -0.4368 to 0.4959) or bAP (r=-0.1093; 95%CI -0.6429 to 0.4957), nor the liver parameters AP (r= -0.1825; 95%CI -0.5693-0.2705), or bilirubin (r=-0.1914; 95%CI -0.5939-0.2879) was observed. Analysing the T-score in further detail, a significant correlation with the patients duration of PSC disease was found (r=-0.4984; 95%CI -0.7768- -0.05718; p=0.0298).


OSTEOPOROSIS is a frequent comorbidity in chronic liver diseases [4].Particularly to date, no reports evaluating the ADC using diffusion weighted MR imaging (DWI) in patients with PSC, have been published. The aim of this study was to test the feasibility of an additional added vertebral bone DWI for the detection of hepatic osteodystophy during the standard follow up liver- MRI examination of PSC patients. The obtained ADC values of patients and controls were in the normal range of previous studies, reporting a vast range of 0.2-0.6 mm2/sx10-3 [5]. Nevertheless, ADC values obtained in our study did only show a tendency of correlation to the T-score. This might be explained by a lower number of study patients, especially with a T-score <-2.5. A pathological decrease of the T-score (<-1.0) was found in 8 of 22 patients proving the higher risk of PSC patients for low BMD. Furthermore, the T-score values correlated significantly to the disease duration upon diagnosis, underlining the hypothesis of disease progression on the osseous affection.


The results of this study show that the DWI add-on is applicable in the routine MRI follow up examinations of PSC patients and might serve to identify changes of the BMD during disease course. Larger studies cohorts, using optimized DWI sequences are needed for further proving.


No acknowledgement found.


1. Guanabens, N., et al., Low bone mass and severity of cholestasis affect fracture risk in patients with primary biliary cirrhosis. Gastroenterology, 2010. 138(7): p. 2348-56.

2. Menon, K.V., et al., Bone disease in primary biliary cirrhosis: independent indicators and rate of progression. J Hepatol, 2001. 35(3): p. 316-23.

3. Le Bihan, D., [From Brownian motion to mind imaging: diffusion MRI]. Bull Acad Natl Med, 2006. 190(8): p. 1605-27; discussion 1627.

4. Hirschfield, G.M., et al., Primary sclerosing cholangitis. Lancet, 2013. 382(9904): p. 1587-99.

5. Dietrich, O., et al., Diffusion imaging of the vertebral bone marrow. NMR Biomed, 2015.


Fig. 1: Selected area of ROI localization in the lumbar vertebral bone (L1) using an average size of 1.8mm2 (SD±0.56). A) ADC map B) T2w anatomical sequence.

Table 1: Patients (n=22) clinical and paraclinical details including ADC and T-score.

Fig. 2: a) No significant difference is observed in the vertebral bone mean±SD ADC in PSC patients and controls (0.2951±0.01380 vs. 0.3163±0.02221; p=0.4093). b) Tendency of correlation between ADC and patients T-score (r=0.3326, 95%CI-0.1289 to 0.6757, p=0.15). c) No significant difference is found in the mean ADC±SD of patients with normal (>-1.0), low (<-1.5 >-2.5) and osteoporosis (<-2.5) T-score.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)