Could MRI Be Used To Image Kidney Fibrosis? A Review of Recent Advances and Remaining Barriers

Abstract

A key contributor to the progression of nearly all forms of CKD is fibrosis, a largely irreversible process that drives further kidney injury. Despite its importance, clinicians currently have no means of noninvasively assessing renal scar, and thus have historically relied on percutaneous renal biopsy to assess fibrotic burden. Although helpful in the initial diagnostic assessment, renal biopsy remains an imperfect test for fibrosis measurement, limited not only by its invasiveness, but also, because of the small amounts of tissue analyzed, its susceptibility to sampling bias. These concerns have limited not only the prognostic utility of biopsy analysis and its ability to guide therapeutic decisions, but also the clinical translation of experimental antifibrotic agents. Recent advances in imaging technology have raised the exciting possibility of magnetic resonance imaging (MRI)-based renal scar analysis, by capitalizing on the differing physical features of fibrotic and nonfibrotic tissue. In this review, we describe two key fibrosis-induced pathologic changes (capillary loss and kidney stiffening) that can be imaged by MRI techniques, and the potential for these new MRI-based technologies to noninvasively image renal scar.

Keywords: MRI; biopsy; chronic; chronic kidney disease; cicatrix; elastography; fibrosis; kidney; magnetic resonance imaging; microvascular injury; renal insufficiency; selection bias; stiffness.

Figure 1.

Multiple physical changes occur in the kidney as it scars. As the tubulo-interstitium scars, the deposited extracellular matrix fibrils compress and obliterate surrounding capillaries, resulting in reduced blood flow and oxygen delivery, diminished tissue water mobility, and an increased diffusion distance for oxygen to reach tubular epithelial cells. Progressive ischemic tubular injury leads to tubular cell apoptosis, and tubular atrophy/dilation. These processes reduce oxygen consumption, which can normalize blood oxygen levels in the chronically injured kidney, and induce the production of profibrotic stimuli that drive further fibrosis in a positive feedback loop. Replacement of compliant kidney cells with stiff, crosslinked extracellular matrix increases the stiffness of the scarring kidney, a phenomenon that also induces further fibrogenesis. Novel magnetic resonance imaging modalities can image a variety of these physical changes to provide estimates of whole-kidney scar burden.

Figure 2.

Diffusion-weighted magnetic resonance imaging detects reduced microvascular perfusion in fibrotic kidneys. (A) Representative pseudo-colorized apparent diffusion coefficient (ADC_perfusion) map from a kidney allograft with minimal fibrosis, and (B) representative ADC_perfusion map showing reduced microvascular perfusion in a kidney allograft with severe fibrosis. These images demonstrate the potential ability of diffusion-weighted magnetic resonance imaging to visualize the progressive reduction in microvascular blood flow that occurs as a result of kidney scarring. The color scale bar reflects the percentage of signal in a voxel that exhibits perfusion-like behavior at small b value diffusion gradient. Echo planar spin echo images were acquired with a water excitation pulse and a repetition time/echo time = 3500 ms/68 ms over a 380 × 380 mm field of view. Thirty-five-millimeter slices were acquired with a resolution of 1.5 × 1.5 mm, with a 20% gap between each slice. This acquisition was repeated with multiple diffusion gradients with increasing weighting of b = 0, 50, 100, 300, 600, 800, 1000 s/mm².

Figure 3.

Blood oxygenation level–dependent magnetic resonance imaging (BOLD MRI) detects basal differences in renal blood oxygenation only in the setting of severe fibrosis. Representative pseudo-colorized BOLD MRI–derived T2* images of (A) a transplant kidney with minimal fibrosis, (B) a transplant kidney with moderate fibrosis, and (C) a transplant kidney with severe fibrosis. T2* is the observed transverse relaxation time, which takes into account magnetic field inhomogeneity/susceptibility and is shorter than the inherent T2 of tissue. Note that T2* does not differ between kidneys with minimal versus moderate fibrosis, but is significantly reduced when fibrosis is severe. These images suggest that in the absence of a pharmacologic stress, BOLD MRI can detect only severe kidney scarring, when renal blood flow (and hence blood oxygenation) is significantly compromised. The color scale bar denotes T2* signal intensity in milliseconds. T2* imaging was acquired using a two dimensional gradient echo sequence (repetition time/echo time = 100 ms/20 echoes with first echo 0.67 ms, and equally spaced echos every 1.2 ms thereafter). Quantitative T2* maps were reconstructed from an exponential fit to the acquired data.

Figure 4.

Magnetic resonance elastography detects increases in kidney stiffness as a surrogate measure of renal fibrosis. (A) In magnetic resonance elastography, a gentle acoustic vibrational wave is applied on the skin overlying the kidney. The magnetic resonance imaging scanner acquires images at a frequency matched to that of the generated vibrational wave, capturing the small displacements in the vibrating organ. Taking advantage of the fact that these waves are longer and travel more rapidly in stiff organs than in soft organs, kidney stiffness can be calculated on the basis of the measured displacements and the amount of vibrational force applied to create “stiffness maps” (elastograms) of the imaged kidney. Sample pseudo-colorized displacement images are shown to the left and right of the magnetic resonance scanner in (A), with blue and red depicting alternating phases of the displacement waves. Below each displacement image is the corresponding pseudo-colorized elastogram (stiffness map), with blue depicting soft, and red depicting stiff tissue. (B) In these representative images of healthy (left panels) and fibrotic (right panels) transplant kidneys, standard transverse relaxation time (T2)-weighted images are shown in the top row. Note that no obvious differences can be identified. In contrast, in the bottom row, pseudo-colorized elastogram stiffness maps are shown of the same two kidneys, with red color reflecting stiffer tissue. The color bar is scaled from 0 kPa (blue) to red (8 kPa). Note the significant increases in stiffness in the fibrotic kidney. Phase images were acquired with a gradient echo sequence (repetition time/echo time = 50/20.79 ms) using a 12° flip angle. Acquired resolution of 1.5 × 1.5 × 5 mm over a 380 mm² field of view. Four separate phase delays were acquired over the 60.1 Hz cycle of the acoustic driver and reconstructed to yield stiffness images. The T2-weighted image was acquired using a two dimensional Half-Fourier-Acquired Single-shot Turbo spin Echo (2D HASTE) sequence using a repetition time/echo tine of 1400/90 ms for over 40 slices with a reconstructed resolution of 1.6 × 1.6 × 4.5 mm. Partial Fourier reconstruction was used over a 256-squared encoding matrix.