Key words

1 Introduction

Kidneys play an important role in fluid regulation, waste products elimination, and homeostasis of salts and pH. Therefore, accurate measurement of kidney function is essential for monitoring early renal injuries, disease progression, renal transplantation and in assessing therapeutic efficacy after renal failure [1, 2]. Magnetic resonance imaging (MRI) has emerged as a powerful technique that combines anatomical details with functional information. In recent years, various MRI techniques have been investigated for assessing renal function, including dynamic contrast-enhanced (DCE) MRI, arterial spin labelling (ASL ), blood oxygenation level dependent (BOLD ), diffusion weighted imaging (DWI ), sodium imaging, chemical exchange saturation transfer (CEST), and hyperpolarized MRI [3,4,5,6,7,8,9]. These techniques have shown promising results for the noninvasive evaluation of several renal function, including filtration, perfusion, oxygenation, pH, homeostasis, and others [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27]. Among them, DCE-MRI or MR renography has emerged as a reliable method to reveal renal pathophysiology, functional and hemodynamic parameters. DCE-MRI relies on bolus administration of a gadolinium-based contrast agent (CA) [16], which changes the intrinsic longitudinal (T1) and transversal (T2) relaxation rates of protons in tissue. This means that administration of CA leads to an increased signal intensity on T1-weighted images or decreased signal intensity on T2-weighted images. DCE-MRI usually refers to T1-weighted MR imaging techniques combined with a gadolinium-based CA using an acquisition scheme with a high temporal resolution. In the kidney the acquired signal over time leads to an MRI renogram, where the absolute signal intensity, in complex manner, depends on the native T1 value, MRI sequence parameters, the specific magnetic relaxivity constant of the CA (the contrast agent efficiency, dependent also on its chemical structure [28,29,30,31,32,33,34]) and the concentration of the CA residing in the tissue [35, 36].

The main purpose of DCE-MRI is to reveal information about extravascular permeability, blood perfusion and compartmental kinetics. In practice, this method provides images prior and following the administration of the CA using a continuous image acquisition protocol. In kidneys, DCE-MRI is suitable for assessing kidney function in terms of filtration and perfusion [17]. Different postprocessing methods allow estimations of kidney function and perfusion, either from simple semiquantitative approaches to more elaborated model-based analyses [37,38,39,40,41,42].

Here we describe step-by-step an experimental dynamic contrast enhanced (DCE ) MRI protocol for monitoring the passage of a Gd-based tracer in the kidneys of rodents. The rationale for the choosing acquisition parameters is given in generic terms, together with specific parameter examples.

This experimental protocol chapter is complemented by two separate chapters describing the basic concepts and data analysis, which are part of this book.

This chapter is part of the book Pohlmann A, Niendorf T (eds) (2020) Preclinical MRI of the Kidney—Methods and Protocols. Springer, New York.

2 Materials

2.1 Animals

This experimental protocol was tailored for immune-competent (typically C57BL/6 and Balb/C) or immune-compromised (nude, NOD-SCID) mice with a body mass of 15–35 g. Advice for adaptation to rats is given as Notes where necessary. All animal studies have to be carried out in compliance with specific legislation covering the use of animals for scientific purposes. Therefore, all experiments must be authorized under national regulations.

2.2 Lab Equipment and Chemicals

  1. 1.

    Mouse tail illuminator/restrainer for catheterizing the tail vein.

  2. 2.

    Saline solution or heparin solution, 1-ml syringe, catheter (e.g., PE 20 polyethylene tubing) of known inner diameter to hold the saline flush and for contrast agent bolus is required.

  3. 3.

    Anesthesia: please refer to the chapter by Kaucsar T et al. “Preparation and Monitoring of Small Animals in Renal MRI” for an in-depth description and discussion of the anesthesia. Typically, 0.5–1.5% isoflurane is used for anesthesia administered to the mice using an anesthetic gas vaporizer (Leica Biosystems, Maryland, USA). For nonrecovery experiments, urethane solution (Sigma–Aldrich, Steinheim, Germany; 20% in distilled water) can provide anesthesia for several hours with comparatively fewer side effects on renal physiology, which is an important issue. For intramuscular anesthesia, please refer to Note 1.

  4. 4.

    Gases: O2, N2, and compressed air, as well as a gas-mixing system or general inhalation anesthesia equipment, including an anesthetic vaporizer, a flow meter, and an induction chamber.

  5. 5.

    Mouse cradle.

  6. 6.

    Gd-based contrast agents: typically, small molecular weight FDA-approved gadolinium-based contrast agents are employed, such as gadoteridol (Prohance®, Bracco Imaging SpA, Milano, Italy), gadobutrol (Gadovist®, Bayer Healthcare Pharmaceuticals, Germany), or gadoteric acid (Dotarem®, Guerbert, France). Such clinically available contrast agents have stock concentrations of 0.5–1.0 mol/l that need to be diluted for handling volumes with less error (see Note 2).

  7. 7.

    Small animal ventilator: animals (rats) needs to be ventilated so the DCE-MRI acquisition can be gated to minimize the respiratory motion artifacts.

  8. 8.

    Syringe infusion pump: an MRI-compatible single syringe pump (Harvard Bioscience, Holliston, USA) for contrast agent administration for allowing precise timing infusion.

2.3 MRI Hardware

The general hardware requirements for renal 1H MRI on mice and rats are described in the chapter by Ramos Delgado P et al. “Hardware Considerations for Preclinical Magnetic Resonance of the Kidney.” The technique described in this chapter was tailored for MRI preclinical systems at a magnetic field strength of 7 T but advice for adaptation to other magnetic field strengths is given where necessary (see Note 3). No special or additional hardware is required, except for the following:

  1. 1.

    A physiological monitoring system that can track the respiration and which is connected to the MRI system such that it can be used to synchronize DCE-MRI with respiration. Typically, we use an MRI-compatible rodent monitoring and gating system equipped with an air-pillow to monitor breath rate (Small Animal Instruments, New York, USA).

2.4 MRI Techniques

DCE-MRI experiments are performed by dynamic acquisitions of a strongly T1-weighted imaging technique. The acquired signal following the CA administration is converted to relative concentrations of the gadolinium agent using either phantom calibration scales or by mathematical calculation of the sequence equation that describes the relationship between signal and MRI sequence parameters. The MRI sequence parameters, required volume coverage, spatial resolution, and SNR will determine the final temporal resolution. Several T1 mapping approaches can be exploited for measuring the longitudinal relation times, a detailed description of different methods is provided in the chapter by Garteiser P et al. “Experimental Protocols for MRI Mapping of Renal T1.”

3 Methods

3.1 MR Protocol Setup

3.1.1 Variable Repetition Time (VTR) Acquisition for T1 Mapping

  1. 1.

    Use a 2D RARE (rapid acquisition with refocused echoes) sequence with variable repetition times (RAREVTR on a Bruker MR system).

  2. 2.

    Set the acceleration factor (RARE-factor or echo train length, ETL) to at least 4 in order to shorten the acquisition time; an acceleration factor of 8–16 can be used to further shorten the acquisition time by sampling more k-space lines within a single repetition period (TR). Centric encoding can be exploited to minimize T2 effects with increasing doses, but at the expenses or “blurred” images.

  3. 3.

    Set the echo time (TE) as short as possible for pure T1-weight; in our setup an effective TE of 25 ms (defined by the acquisition of the central k-space line) was used.

  4. 4.

    Choose enough averages in order to have a sufficient signal to noise ratio (SNR) of above 30.

  5. 5.

    Repeat the RARE acquisition a number of times by changing the value of the repetition time (TR) to sample changes in signal intensities according to the expected T1 values of the kidneys.

  6. 6.

    Because of the variability of the T1 values inside the kidney regions, choose the TRs values to span a range large enough to permit an accurate calculation of T1 by selecting 3–7 TR values in the range 100–10,000 ms (see Note 4).

  7. 7.

    For in vivo experiments, keep the overall acquisition time as short as possible by considering a trade-off between number of TRs, averages and matrix size (or spatial resolution).

  8. 8.

    For an example of a specific parameter set please see Note 5.

3.1.2 Variable Flip Angle (VFA) Acquisition for T1 Mapping

  1. 1.

    Use a 2D gradient-echo sequence (FLASH on a Bruker system).

  2. 2.

    Set the echo time (TE) as short as possible for heavily T1 weighting; in our setup a TE of 1.5 ms was used.

  3. 3.

    In order to have a sufficient signal to noise ratio, at least two averages have been acquired or enough averages for SNR > 30.

  4. 4.

    Repeat the FLASH acquisition a number of times by changing the value of the flip angle (FA). The changes in signal intensities will follow a parabolic curve, thus low flip angles and higher flip angles are needed to sample both side of this curve. Acquire a minimum of 2–3 flip angles, considering that experiments with more flip angles will improve the goodness of the T1 estimate.

  5. 5.

    Set the TR in the range 10–100 ms for providing good T1 weighting. Shorter TR increases T1 weighting, but with detrimental SNR values; thus, a trade-off must be chosen, even considering number of averages and overall acquisition time.

  6. 6.

    For an example of a specific parameter set please see Note 6.

3.1.3 Fast Low Angle Shot (FLASH) T1-Weighted Sequence for DCE-MRI Experiments

  1. 1.

    Use a T1-weighted sequence as a fast low angle shot (FLASH) sequence (in Bruker systems, this sequence refers to a dubbed FLASH sequence ).

  2. 2.

    Use short TR to increase T1-weighting, but this will decrease SNR. Therefore, a good balance should be found between strongly T1-weighted images and image quality with sufficient SNR. Besides, TE and FA affect T1-weighting and SNR. A good TR range could be between 5 and 100 ms (usually shorter values for 3D imaging and longer for 2D multislice acquisition).

  3. 3.

    Use the shortest echo time (TE) possible for pure T1 weight.

  4. 4.

    Flip Angle (FA): high FA corresponds to a strong T1 weighting, but at expenses of reduced SNR, therefore a good balance may result in FAs in the range 10–60°.

  5. 5.

    Use Partial Fourier acceleration to reduce the size of the acquired matrix and improve the time efficiency of data acquisition.

  6. 6.

    Use a high acquisition bandwidth (BW) to shorten TE, while keeping an eye on the SNR, which decreases with the square root of the increasing BW. For the image with the shortest TE an SNR of at least 30 is recommended.

  7. 7.

    Employ Fat Saturation modules. It is important to avoid fat signal overlaying with the kidney due to chemical shift.

  8. 8.

    Respiration trigger: there is no need to use respiratory triggering (per slice).

  9. 9.

    Adapt the geometry (or slice position) so that animal fits into FOV in Left-Right direction (approximately 30–40 mm according to the inner coil diameter) and use frequency encoding in Head-Feet direction. Use a slice with the lowest thickness the SNR allows, typically around 1–2 mm. The thinner the slices the better, as this improves the resolution which can be readily found on the raw images.

  10. 10.

    Use the highest in-plane resolution that the SNR allows, typically around 96 × 96 to 128 × 128 for FOV of 30 × 30 mm2 or 40 × 40 mm2. This resolution should be balanced by the need for sufficient SNR in the images. Zero-filling or use of partial Fourier techniques in the phase encoding direction can be helpful to speed up acquisition to minimize motion artifacts and contrast agent concentration changes during the acquisition. One may use half-Fourier acquisition schemes in read direction (asymmetric echo) to further shorten the TE. Reducing the excitation and refocusing pulse lengths to below 1.5 ms can then also help to shorten TE.

  11. 11.

    For an example of specific parameter sets for a 7 T MRI system, please see Notes 7 and 8, whereas for a scanner operating at 3 T please see Note 9.

3.2 Preparations of the Contrast Agent Injection and of the Catheter

  1. 1.

    Fill the 1-ml syringe with the contrast agent solution (see Note 2).

  2. 2.

    Prepare the catheter by filling the PE20 tubing with the contrast agent solution, and with 20–40 μl of a heparin solution (or a saline solution), between the needle (the animal) and the contrast agent solution to prevent clotting before the contrast agent will be injected.

  3. 3.

    According to the length of the catheter (usually 60–100 cm to exit from the magnet bore) the contrast agent solution can be injected directly from the syringe containing the contrast agent or by changing the syringe with one filled with a (heparin-free) saline solution for the postinjection flush.

3.3 Preparations of the Mouse

  1. 1.

    Place the mouse on a mouse tail illuminator restrainer for catheterizing the tail vein. The tail vein can be catheterized by placing a 27/29-G needle in the animal.

  2. 2.

    Induce anesthesia in animals via isoflurane in an induction box. Then transfer the animal to the MRI system (see Note 10).

  3. 3.

    Set up the temperature monitoring (rectal probe) and respiratory monitoring (balloon on chest) unit. During imaging, keep the animal’s respiration rate at 35–80 breaths per min by adjusting the dose of anesthesia (approximately 1–1.5% isoflurane with air and oxygen mixed at a 3:1 ratio administered through a nose cone attached to the animal bed and ophthalmic ointment will be applied to the eyes) (see Note 11).

  4. 4.

    Perform anatomical imaging as described in the chapter by Pohlmann A et al. “Essential Practical Steps for MRI of the Kidney in Experimental Research.”

  5. 5.

    Since the AIF should be sampled with the same temporal resolution of the DCE-MRI experiment, selection of anatomical slices including large vessels near the kidneys should be considered.

  6. 6.

    Perform localized shimming on the kidney imaging as described in the chapter by Pohlmann A et al. “Essential Practical Steps for MRI of the Kidney in Experimental Research.”

3.4 In Vivo T1 Mapping

Typically, a T1 map is acquired before the injection of Gd-based agent.

  1. 1.

    Select the same geometry of the anatomical images, with the same number of slices and position.

  2. 2.

    Load or select one of the several T1 mapping sequences previously described (3.1). Adapt the same slice orientation as for the anatomical image with coronal or axial view with respect to the kidneys.

  3. 3.

    Run the T1-mapping scan. Examples images are shown in Fig. 1 and Fig. 2.

Fig. 1
figure 1

Representative T1-weighted images in kidneys obtained with the VTR scheme at different repetition times (TR) values: 75 ms (a), 534 ms (b), 1076 ms (c), 1738 ms (d), 2589 ms (e), 3782 ms (f), 5791 ms (g), and 15000 ms (h). Acquisition details are provided in Note 5

Full size image
Fig. 2
figure 2

Representative T1-weighted images in kidneys obtained with VFA T1 mapping approach by varying the excitation pulse flip angle values: 5° (a), 15° (b), 30° (c), 45° (d), 60° (e), and 75° (f) with a 7 T MRI system. Acquisition details are provided in Note 6

Full size image

3.4.1 DCE-MRI Experiment

  1. 1.

    Load the T1-weighted sequence , adapt the slice orientation to provide a coronal or axial view with respect to the kidney (in scanner coordinates this is double-oblique) by importing the same geometry as for the anatomical (and T1 mapping sequence ).

  2. 2.

    Set the required number of repetitions accordingly to the sampling time and to the duration of the dynamic acquisition. Typically, 50–100 repetitions provide enough points to be exploited by any of the analysis procedures.

  3. 3.

    Initiate the DCE-MRI scan.

  4. 4.

    After some baseline image acquisition (5–10 images before the injection are enough to derive average baseline values and signal oscillations) perform an intravenous bolus injection of the Gd-based CA through the tail vein at a dose of 0.02–0.1 mmol Gd /kg b.w. by using a power injector to maintain the exact administration rate for each imaged mouse .

  5. 5.

    A demonstration of the signal intensity changes that can be expected for the DCE-MRI experiment across the kidneys is given in Fig. 3 (axial orientation) and Fig. 4 (coronal orientation).

Fig. 3
figure 3

Representative DCE-MRI images in kidneys (transversal orientation) at different time points following Gd-based contrast agent injection with a 7 T MRI scanner. Anatomical T2-weighted (a) and GRE-FLASH T1-weighted images before (b) and 4 s (c), 30 s (d), 5 min (e), and 10 min (f) after gadoteridol injection (dose 0.1 mmol Gd/kg b.w.). Acquisition details are provided in Note 7

Full size image
Fig. 4
figure 4

Representative DCE-MRI images in kidneys (coronal orientation) at different time points following Gd-based contrast agent injection with a 7 T MRI scanner. Anatomical T2-weighted (a) and GRE-FLASH T1-weighted images before (b) and 4 s (bc), 30 s (d), 5 min (e), and 10 min (f) after gadoteridol injection (dose 0.1 mmol Gd/kg b.w.). Acquisition details are provided in Note 8

Full size image

4 Notes

  1. 1.

    Intramuscular anesthesia can be administered as a mixture of xylazine 5 mg/kg (Rompun, Bayer, Italy) and tiletamine/zolepan 20 mg/kg (Zoletil 100, Virbac, Italy) with a dose dependent on the length of the MRI session. Usually, a 40–80 μl volume allows for a sleeping time of 30–60 min for a 25 g mouse .

  2. 2.

    A required dose for DCE-MRI in the kidneys is in the range 0.01–0.1 mmol Gd/kg b.w. that would require too small volumes of the stock solution. Therefore, it is advisable to dilute stock solutions to ca. 50 mmol/l. For instance, a 25 g mouse would receive 50 μl of 50 mmol/l Gd-complex to achieve a dose of 0.1 mmol Gd/kg. Please avoid the inclusion of air bubbles in the catheter filled with the saline solution or with the contrast agent solution that will be injected into the animals.

  3. 3.

    DCE-MRI experiments are described for Bruker MRI systems, such as Avance 300 (7 T) and Biospec (3 T), (Bruker, Ettlingen, Germany) or for a Magnetom Skyra 3 T (Siemens Healthcare, Erlangen, Germany). Standard 1H radiofrequency-coils (birdcage, quadrature) or phased-array 1H coils can be used for data reception.

  4. 4.

    Within the Bruker method, specific TR values can be suggested according to the estimated T1 values of the tissue.

  5. 5.

    For estimated T1 of 1–3 s, 10 TR values can be used (580, 810, 1075, 1380, 1740, 2180, 2740, 3525, 4830, and 10,000 ms) to correctly follow signal intensities recovery until plateau. With 10 TR values complete acquisition lasted 8:40 min (for 96 × 96 matrix and 8 slices).

  6. 6.

    For estimated T1 values of 1–3 s, we used six different FA values (5°, 15°, 30°, 45°, 60°, and 75°). With a TR of 80 ms and matrix size of 96 × 96, the single image acquisition lasted more than 15 s (8 slices). The complete acquisition time of the VFA T1 map with six FA values is approximately 2 min.

  7. 7.

    Example for DCE-MRI using a FLASH readout at 7 T: TR: 16 ms; TE: 1.6 ms; FA: 30°; scan averages: 2; slice thickness: 1.5 mm; slice number: 1; slice orientation: axial; frequency encoding: head-feet; FOV = 30 × 30 mm2; matrix size = 128 × 128; fat suppression = on; acquisition time per image: 4 s; number of repetitions: 150.

  8. 8.

    Example for DCE-MRI using a FLASH readout at 7 T: TR: 15 ms; TE: 1.7 ms; FA: 30°; scan averages: 2; slice thickness: 1.5 mm; slice number: 1; slice orientation: coronal; frequency encoding: head-feet; FOV = 30 × 30 mm2; matrix size = 128 × 128; fat suppression = on; acquisition time per image: 4 s; number of repetitions: 150.

  9. 9.

    Example for DCE-MRI using a 3D time-resolved angiography with stochastic trajectories (TWIST) sequence at 3 T: TR   =   3.4 ms, TE   =   1.4 ms, FA: 20°, FOV = 114 × 50 mm2, generalized autocalibrating partially parallel acquisition (GRAPPA) of factor 2, matrix = 192 × 84; slice number: 2, voxel resolution = 0.6 × 0.6 × 1.2 mm3. TWIST view sharing: 15% central region, 20% sampling density in the outer region, nominal temporal resolution 0.9 s per volume . Images were continuously acquired for 6 min.

  10. 10.

    Mice can be alternatively anesthetized with intramuscular anesthesia, tail veins heated with hot water/heater equipment at 37 °C and catheter placed in the animal.

  11. 11.

    It is important to monitor the respiration continuously throughout the entire experiment.