Renal artery stenosis (RAS) may cause renal fibrosis, which is characterized by excessive and often irreversible collagen deposition.¹ Currently, the standard method to assess renal fibrosis is invasive tissue biopsy. Magnetization transfer (MT) is an MRI technique that can be used to evaluate the interaction between macromolecules and free water molecules. Compared to conventional MRI techniques, MT is more specific in the evaluation of macromolecules.² In this study, we hypothesize that MT can detect collagen deposition in murine kidneys with unilateral RAS.

Phantom Study MRI studies were performed on a vertical 16.4 T animal scanner (Bruker Biospin, Billerica, MA) equipped with a 38 mm inner diameter birdcage coil. To select the optimal offset frequency, MT study was performed on a phantom with collagen I&III concentrated at 20%, 10%, 5% and 0%. An MT-prepared fast low angle shot (FLASH) sequence was utilized. FLASH images without MT module (M₀) were first acquired with the following parameters: TR 400 ms; TE 2.9 ms; flip angle 20°; slice thickness 1 mm; slice number 5; FOV 3.0×3.0 cm²; matrix size 128×128; number of averages 4. Then MT-weighed FLASH images (M_t) were acquired by adding Gaussian MT pulses with the following parameters: offset frequency -6000 to 6000 Hz; pulse power 10 μT; pulse length 9.13 ms; pulse bandwidth 300 Hz; flip angle 585°; pulse number 2. Magnetization transfer ratio (MTR) was calculated as: (M₀ -M_t)/M₀.

In Vivo Study Fifteen male 129S1 mice were used in the study. At the age of 10 weeks, all mice underwent a baseline MRI scan, after which they were randomized for RAS (n=10) or sham (n=5) surgeries. Follow-up MRI scans were performed 2, 4, and 6 weeks post surgery. Both the stenotic (STK) and contralateral (CLK) kidneys were scanned. In the MT study, the FOV and acquisition matrix were prescribed at 2.56×2.56 cm² and 256×256 respectively, giving rise to an in-plane resolution of 100×100 μm². To maintain signal to noise ratio, 8 averages were used. MT pulses with optimized offset frequency were implemented. All other imaging parameters were the same as in the phantom study.

Histology and Hydroxyproline Assay All mice were euthanized within one day after the 6-week MR scan with kidneys harvested and fixed. Masson’s trichrome and Sirius red staining were performed on 5-μm axial slices of tissue, collected from locations corresponding to the imaging slices. Fibrosis was quantified as the fraction of fibrotic area over the total cross sectional area of the tissue. Hydroxyproline assay was performed on 20 mg of kidney tissue and the concentration of hydorxyproline was measured.

Phantom Study Representative M₀ and M_t images and the calculated MTR map at 1500 Hz are shown in Fig. 1a. The MTR values obtained at different offset frequencies are shown in Fig. 1b. To maintain high MT sensitivity with little saturation of free water signal, an offset frequency at 1500 Hz was chosen for in vivo MT study.

In Vivo Study Representative MTR maps of the STK, CLK, and a control kidney at baseline, 2, 4, and 6 weeks after surgery are shown in Fig. 2a-c. MTR in the post-stenotic cortex increased markedly from baseline to 2 weeks and then stabilized (Fig. 2d). Contrarily, medullary MTR decreased slightly from baseline to 2 weeks (white arrow in Fig. 2a), but subsequently increased dramatically (Fig. 2d). Both low and high MTR values were particularly observed at 6 weeks at the cortico-medullary junction (red and black arrows in Fig. 2a). In the CLK, a slight decrease in MTR was observed at 4 and 6 weeks, possibly due to pressure natriuresis. In the sham-operated controls, no change in MTR was observed throughout the study (Fig. 2c&d).

Histology and Hydroxyproline Assay Representative Masson’s trichrome and Sirius red stained slices and the MTR map of the corresponding slice are shown in Fig. 3a-c. A good agreement between ex vivo staining and in vivo MT imaging was observed in regions of fibrosis, necrosis and pelvis. A linear regression analysis also showed a good correlation between the fibrosis quantified from the stained slices and the MTR values (Fig. 3 d&e). A significant increase in the hydroxyproline concentration was observed in the STK compared to the sham controls, which correlated well with MTR (Fig. 3f).

The current study suggests that MTI can be used for measuring and monitoring renal fibrosis and longitudinal progression in mice with unilateral RAS. Therefore, MTI may be valuable for noninvasive diagnosis of renal pathology and evaluation of therapeutic interventions in patients with RAS.