Magnetization transfer imaging predicts porcine kidney recovery following revascularization of renal artery stenosis

Abstract

Objectives:

Magnetization transfer imaging (MTI) is a novel non-invasive tool for detection of kidney fibrosis, but its association with kidney function and hemodynamics is unclear. Renovascular disease (RVD) associated with metabolic derangements elicits renal fibrosis and is often unresponsive to percutaneous transluminal angioplasty (PTRA), but tools to predict therapeutic success are unavailable. We hypothesized that MTI predicts kidney recovery following PTRA in swine with unilateral RVD.

Materials and Methods:

Stenotic-kidney (STK) and contralateral kidney magnetization transfer-ratio (MTR, M_t/M₀) were measured at 3.0T MRI, at offset-frequencies of 600 and 1000Hz, before and 1-month post-PTRA in 7 RVD pigs. STK MTR was correlated to renal perfusion, renal blood flow (RBF), and glomerular filtration rate (GFR), determined using multidetector computed-tomography, and with ex-vivo renal fibrosis (trichrome staining). Untreated RVD (n=6) and normal pigs (n=7) served as controls.

Results:

RVD induced hypertension and renal dysfunction. Blood pressure and renal perfusion were unchanged post-PTRA, but GFR and RBF increased. Baseline cortical STK MTR predicted post-PTRA renal perfusion and RBF, and MTR changes associated inversely with changes in perfusion and normalized-GFR. STK MTR at 600 Hz showed closer association with renal parameters, but both frequencies predicted post-PTRA cortical fibrosis.

Conclusion:

Renal STK MTR, particularly at 600Hz offset, is sensitive to hemodynamic changes following PTRA in swine RVD, and capable of non-invasively predicting post-PTRA kidney perfusion, RBF, and fibrosis. Therefore, STK MTR may be a valuable tool to predict renal hemodynamic and functional recovery, as well as residual kidney fibrosis after revascularization in RVD.

Introduction

Atherosclerotic renovascular disease (RVD) accounts for almost 90% of the cases of renal artery stenosis (RAS) (1), a frequent cause of progression to chronic kidney disease (CKD) and end-stage renal disease. The affected stenotic kidneys (STKs) in RVD typically undergo progressive and chronic fibrosis involving deposition and accumulation of extracellular matrix components, which is mainly composed of tissue macromolecules fibronectin and collagens (2).

Based on existing literature, it is evident that RVD is associated with inconsistent response to revascularization techniques, such as percutaneous transluminal angioplasty (PTRA), in terms of blood pressure control and renal function recovery (3–7). Various factors have been suggested to affect differential response to PTRA in patients with RVD. For example, patients with high albuminuria (more severe disease) may fare worse after renal artery stenting than those with low albuminuria (8). However, current predictive tools based on conventional risk factors fail to estimate degree of response to revascularization in RVD (9). Thus, novel biomarkers are urgently required to evaluate the severity of renal fibrosis and recovery potential with therapeutic interventions, such as revascularization, and thereby guide patient management strategies.

Magnetization transfer imaging (MTI) represents a powerful molecular magnetic resonance (MR) technique based on the presumed co-existence of two tissue proton pools, an observable unbound free water pool and a restricted water pool bound to local tissue macromolecules (10). Taking advantage of this phenomenon, magnetization transfer ratio (MTR) is a MTI-calculated measure that probes tissue macromolecule content, and has been used to study fibrosis in the kidney (11), nervous system (12, 13), heart (14), musculoskeletal system (15, 16), gastrointestinal tract (17, 18), genitourinary system (19), and cancers (20, 21). Higher MTR values indicate greater availability of bound tissue macromolecules (e.g., collagen) to exchange magnetization with mobile water macromolecules.

Renal fibrosis is a significant predictor of kidney disease progression (22, 23), which is characterized by deteriorating renal hemodynamics and function. Increased collagen content is the hallmark of renal tissue fibrosis. We previously validated the capability of MTI-measured MTR to noninvasively detect renal fibrosis in mice (16.4T) (24, 25) and swine (3.0T) experimental models of unilateral RAS (26). Furthermore, we also found in experimental pig models of acute graded ischemia that several imaging techniques are sensitive to alterations in renal hemodynamics, whereas a sudden decline in renal perfusion does not impact MTR (27), underscoring its ability to probe the renal microstructure. This is particularly relevant in RAS, in which a fall in renal blood flow (RBF) might decrease renal fluid content and potentially confound assessments of renal fibrosis.

However, whether MTR-derived fibrosis can predict recovery of the ischemic kidney in RVD remains unknown. Therefore, this study was designed to test the hypothesis that MTR is a correlate of renal functional and hemodynamic measures in pigs with RVD, and predicts renal recovery in response to PTRA.

Materials and Methods

Study protocol

All animal experiments were approved by the Institutional Animal Care and Use Committee. Seven 3-month-old female domestic pigs (Sus scrofa; Manthei Hog Farm LLC, Elk River, MN) initiated a high-cholesterol/high-carbohydrate diet (to induce early atherosclerosis) containing (in % kcal) 17% protein, 20% complex carbohydrates, 20% fructose, and 43% fat and supplemented with 2% cholesterol and 0.7% sodium cholate by weight) (Figure S1). All animals had free access to water ad libitum.

Six weeks later, animals were anesthetized with IM tiletamine hydrochloride/zolazepam hydrochloride (0.25g, Telazol®, Fort Dodge Animal Health, New York) and 0.5g of xylazine, and maintained with intravenous ketamine (0.2mg/kg/min) and xylazine (0.03mg/kg/min). RAS was induced in 7 pigs (Figure S1) by placing an irritant coil in the main renal artery using fluoroscopy, a procedure that gradually develops unilateral RAS within 1–2 weeks. After 12 weeks of diet animals underwent PTRA and stenting, achieved by inflation of a balloon in the stenotic renal artery and implantation of a stent (28, 29). All animals underwent MTI and multi-detector computed-tomography (MDCT) imaging studies before PTRA and 4 weeks later, at 12 and 16 weeks after initiation of diet.

Untreated RVD pigs (n=6) and 7 healthy pigs fed a standard pig chow (13% protein, 2% fat, 6% fiber, Purina Animal Nutrition LLC, MN, USA) underwent similar MTI and MDCT studies after 16 weeks of diet, to serve as controls.

In-vivo MTI and image analysis

MRI studies were performed on a GE Signa HDxt 3.0 T scanner (GE Healthcare, Waukesha, WI). The body coil was used as the transmitter and an 8-channel surface coil as the receiver. To avoid respiratory motion artifacts, scans were performed with suspended respiration under isoflurane anesthesia (1–2%). A complete set of MRI scans took approximately 50 minutes per animal.

Kidney MTR was measured in each STK, contralateral kidney (CLK), and the right normal kidney using an MT-prepared gradient echo (GRE) sequence. Images without MT preparation (M_o) (Figure 1a) were initially acquired in the coronal plane with the following parameters: TR=300ms, TE=5.3ms, flip angle=30 degrees, slice thickness=4mm, slice number=5, FOV=15×15cm², matrix size=128×128, and number of averages=1. Subsequently, MT-weighed images (M_t) (Figure 1b) were acquired by adding Fermi pulses prior to GRE image acquisition. The MT pulse parameters were: offset frequency at 600, pulse width=16ms, and flip angle=800°. To select the best parameter predictive of recovery, also we tested a 1000Hz offset frequency. This was based on a previous phantom study, showing high MT sensitivity with minimal direct saturation of the free water pool at offset frequency range of 600–1,000Hz (26).

Figure 1.

Representative M₀ (a) and M_t (b) coronal images of both kidneys in a pig with renovascular disease (RVD). The magnetization transfer ratio (MTR) map was calculated at 600Hz and 1,000 offset-frequencies, and the normalized MTR map overlaid with M₀ image. Renal MTR was normalized by the MTR of the ipsilateral psoas muscle major to correct for B1 variations (c). Representative MTR maps (d) pre- and post-percutaneous transluminal angioplasty (PTRA) in stenotic (STK) and contralateral (CLK) kidneys at the offset frequency of 600Hz. STK: stenotic kidney, PTRA: percutaneous transluminal angioplasty.

MTI images were analyzed using MatLab (Mathworks, Natick, MA) modules developed in-house. The MTR (Figure 1c) map was calculated in the STKs and CLK (pixel-wise) based on the (M₀-M_t)/M₀ formula. We then normalized kidney MTR by the MTR of dorsal muscle to adjust for intra- and inter-subjectB₁ inhomogeneity. Considering the modest contrast in MTI images, T₁/T₂*-weighted anatomical images of the same slices were used to select cortical and medullary regions of interest (ROIs). Greater T₁ in renal medulla yields reduced signal intensity across T₁-weighed image (30), while greater hypoxia makes the medulla appear darker than the cortex in T₂*-weighted images (31). Accordingly, a combination of T₁/T₂* weighting produces satisfactory corticomedullary contrast (Figure 1d). We implemented a MatLab-based module to semi-automatically select cortical and medullary ROIs (Figure S2) in the following steps: 1) Exclusion of the collecting system, 2) thresholding for image segmentation (32), 3) correction of misclassified pixels in renal cortex by image opening, and 4) edge detection of the borders of cortex and medulla. We manually edited the threshold for image segmentation and the kernel size for image opening to correct ROI selection. Subsequently, the selected ROIs were propagated and applied for quantification of cortical and medullary MTRs. The mean values from all ROIs in different slices were averaged.

In-vivo MDCT and image analysis

MDCT studies evaluated single-kidney hemodynamics and function 2 days after completion of MTI studies under continuous ketamine (0.2mg/kg per minute, Hospria Inc., Lake Forest, IL) and xylazine (0.03mg/kg per minute) anesthesia. Selective angiography was used to estimate the degree of RAS. The inferior vena cava was accessed to collect a blood sample, and a carotid artery catheter measured mean arterial pressure (MAP). A total of 140 consecutive scans were continuously acquired in 3 contiguous 5-mm slices following a bolus injection of iopamidol (0.5mL/kg in 2 seconds, Omnipaque®, Novalplus, USA) through a catheter placed in the right atrium for quantification of renal perfusion and glomerular filtration rate (GFR) (33). Renal volume was measured 15 min after the flow study to allow contrast washout. A second bolus of iopamidol (0.5mL/kg over 5 sec) was injected and 30–34 contiguous 5mm-thick slices covering the whole kidney were imaged.

The Analyze™ software package (Biomedical Imaging Resource, Mayo Clinic, Rochester, MN) was used to analyze MDCT images. The sum of cortical and medullary volumes, which were manually assessed on all frames where kidney was observed, was used to quantify kidney volume. Tissue attenuation curves were generated from aorta, renal cortex, and medulla, and fitted using a Matlab module developed in-house to define renal perfusion and normalized GFR (GFRn, per unit volume) as estimates of renal hemodynamics and function (34, 35). Renal blood flow (RBF) was calculated as the sum of the multiplication of cortical and medullary perfusions by their corresponding volumes, and GFR (ml/min) by multiplying total kidney volume (cc) and GFRn (ml/min/cc).

Ex-vivo fibrosis assessment

Pigs were euthanized 3–5 days after the final MDCT study, and kidneys harvested, fixed in 10% formalin, and processed for histology. Trichrome staining was performed on 5-μm axial slices of tissue. Fibrosis was quantified as the fraction of fibrotic over the total cross-sectional area of the tissue using AxioVision4.8 (Carl Zeiss SMT, Oberkochen, Germany).

Statistical Analysis

Statistical analysis was performed using JMP 10.0 (SAS Institute, Cary, NC). Normality of the data was assessed using the Shapiro-Wilk test. Results are expressed as means±standard deviations. As appropriate, the unpaired Student’s t-test/Wilcoxon test and paired t-test/Wilcoxon Signed-Rank test were performed to compare cortical/medullary and baseline/post-PTRA measures, respectively. The Pearson’s correlation was used to correlate baseline and post-PTRA MTR with study parameters measured in-vivo and ex-vivo. A p≤0.05 was considered statistically significant.

Results

Baseline characteristics of pigs and effect of PTRA

The pre- and post-PTRA characteristics of RVD+PTRA pigs (n=7) are summarized in Figure 2. The average body weight of RVD+PTRA pigs increased from baseline to post-PTRA (p<0.001), due to growth and the continuous high-fat diet (Figure S1). PTRA decreased the degree of RAS from 70% to ~zero% (p<0.001), while MAP remained unchanged (p=0.2). There were no significant changes in mean STK-GFRn (p=0.4) or perfusion (p=0.3), although 3/7 and 4/7 of RVD+PTRA pigs had improved perfusion and GFRn post-PTRA, respectively. Furthermore, PTRA increased mean STK-GFR (p<0.001), RBF (p=0.010), and renal volume (p<0.001) in RVD+PTRA pigs compared to baseline (Figure 2).

Figure 2.

Changes from baseline in renal characteristics of pigs with renovascular disease undergoing percutaneous transluminal angioplasty (RVD+PTRA) after PTRA. RVD+PTRA pigs had increased body weight, glomerular filtration rate (GFR), and renal blood flow (RBF), but reduced degrees of renal artery stenosis (RAS) compared to baseline. MAP: mean arterial pressure, GFRn: normalized GFR, MTR: magnetization transfer ratio.

Comparison of RVD+PTRA pigs 4 weeks after revascularization to non-revascularized RVD and controls is shown in Table 1. Both groups had higher weight and MAP, and lower GFRn, GFR, and perfusion than Normal control pigs (Table 1). Yet, RVD+PTRA had lower degrees of RAS, and higher GFR and RBF compared to untreated RVD pigs.

Table 1.

Systemic and Single kidney (stenotic or right) characteristics of pigs with renovascular disease (RVD) with and without undergoing percutaneous transluminal angioplasty (RVD+PTRA) 4 weeks earlier, and Normal control pigs.

Variables	Normal Control (n=7)		RVD Control (n=6)		RVD+PTRA (n=7)
Body weight (kg)	55.1±5.6		86.2±9.0*		85.6±10.7*
Degree of RAS (%)	0		89.2±11.6*		0†
MAP (mmHg)	96.1±7.9		118.1±12.7*		111.1±8.5*
GFR (ml/min)	86.3±10.7		48.9±15.8*		69.7±9.1*†
GFRn (ml/min/cc)	68.5±10.6		31.8±5.0*		36.9±3.6*†
Perfusion (ml/100ml/min)	482.7±139.1		272.0±70.6*		367.2±65.8*†
RBF (ml/min)	603.5±140.9		509.4±214.0*		640.6±95.7†
Cortical MTR (600Hz)	0.73±0.02		0.74±0.02		0.79±0.05*†
Medullary MTR (600Hz)	0.69±0.02		0.70+0.03		0.73±0.05*†

^†p≤0.05 RVD+PTRA vs. RVD control

^*p≤0.05 Normal control

RVDrenovascular disease, PTRApercutaneous transluminal angioplasty, MAPmean arterial pressure, GFRglomerular filtration rate, GFRnnormalized GFR, RBFblood flow, MTRmagnetization transfer ratio.

In-vivo MTR at 600Hz-offset frequency

Representative MTR maps at offset frequencies of 600 and 1000Hz acquired from the STKs and CLKs pre- and post-PTRA are shown in Figure 1c–d. In both STKs (p=0.020) and CLKs (p<0.001) of RVD+PTRA pigs, MTR values were higher in the cortex compared to the medulla (Figure 2). Prior to PTRA, STKs of RVD+PTRA pigs also had higher cortical (0.77±0.03 vs. 0.72±0.03, p=0.005) and medullary (0.73±0.04 vs. 0.67±0.04, p=0.018) 600Hz-MTR compared to the corresponding values in CLKs.

By 4 weeks post-PTRA, mean cortical (p=0.161) and medullary (p=0.381) STK MTRs were unchanged from baseline (Figure 2) (although they fell in 3/7 and 4/7 of RVD+PTRA pigs, respectively), but were higher than both RVD and Normal controls (Table 1). In contrast, the RVD+PTRA CLK had cortical (0.73±0.02 vs. 0.74±0.02, p=0.119) and medullary (0.68±0.03 vs. 0.70±0.03, p=0.241) MTRs comparable to the RVD CLK and to Normal controls (both p=0.241), but similar to STKs, their MTRs remained unchanged after revascularization (Figure 2), likely due to the ongoing atherogenic diet and hypertension.

Baseline medullary (albeit not cortical) 600Hz-MTR inversely correlated with baseline perfusion (Figure 3a), whereas post-PTRA MTR in both regions inversely correlated with their perfusion (Figure 3a). Cortical 600Hz-MTR measured at baseline correlated inversely with post-PTRA perfusion and RBF (Figure 3a). Furthermore, changes in cortical and medullary 600Hz-MTR from baseline to 4 weeks after PTRA inversely correlated with alterations in kidney perfusion and GFRn (Figure 3b). Contrarily, pre-PTRA cortical and medullary MTR did not correlate with post-PTRA GFRn (r²=0.099 and r²=0.0002, respectively) or the change in GFRn (r²=0.010 and r²=0.234, respectively).

Figure 3.

a) Correlation of magnetization transfer ratio (MTR) with perfusion pre- and post- percutaneous transluminal angioplasty (PTRA), and pre-PTRA cortical MTR with post-PTRA perfusion and renal blood flow (RBF), at offset frequency of 600Hz. Basal cortical 600Hz-MTR correlated inversely with post-PTRA perfusion and RBF, whereas basal cortical 1000Hz-MTR correlated with RBF, but not perfusion. b) Association of changes in magnetization transfer ratio (MTR) with changes in perfusion and normalized glomerular filtration rate (GFRn) at offset frequency of 600Hz. The change (Δ) in cortical and medullary MTR correlated inversely with Δperfusion at 600Hz. By comparison, Δcortical, but not Δmedullary, MTR correlated inversely with ΔGFRn at 600Hz.

In-vivo MTR at 1,000Hz-offset frequency

See Supplementary Materials.

Ex-vivo renal fibrosis

Figure 4 shows trichrome staining-derived fibrosis in STKs and CLKs from RVD+PTRA and RVD as well as normal kidneys in control pigs, harvested at protocol completion. All STKs showed extensive fibrosis in both the cortex and medulla, which was greater than in the CLKs. Both the cortex and medulla of RVD+PTRA STKs had greater fibrosis than normal pig kidneys (Figure 4). Baseline cortical MTR at 600Hz correlated directly with post-PTRA ex-vivo cortical fibrosis (but not medullary) on trichrome staining (Figure 5), whereas post-PTRA MTR correlated directly with renal fibrosis in both the cortex and medulla (Figure 5).

Figure 4.

Representative ex-vivo assessment and quantification of renal fibrosis using trichrome staining in pigs with renovascular disease and undergoing percutaneous transluminal angioplasty (RVD+PTRA), RVD control pigs, and Normal pigs in the cortex and medulla. Stenotic kidneys (STKs) of both RVD and RVD+PTRA pigs had higher percentage of fibrosis compared to their contralateral kidneys (CLKs), and also kidneys of Normal pigs. *p≤0.05 Normal vs. STK or CLK fibrosis. †p≤0.05 STK vs. CLK fibrosis.

Figure 5.

Correlation of ex-vivo renal fibrosis measured by trichrome staining with pre- and post-PTRA MTR values. Baseline cortical MTR and post-PTRA cortical and medullary MTR values correlated with ex-vivo renal fibrosis at the 600-offset frequency.

Discussion

This study shows that baseline MTI-measured STK MTR associates with MDCT-derived renal perfusion and RBF, and is sensitive to subtle changes in renal function and hemodynamics (indicated by perfusion and GFR) following PTRA in experimental pigs with co-existing metabolic syndrome and unilateral RAS, a model recapitulating the features of human RVD. In addition, STK cortical and medullary MTR values correlated inversely with the concurrent corresponding perfusion. Furthermore, baseline STK cortical MTR also correlated directly with ultimate renal fibrosis assessed ex-vivo by trichrome, and inversely with post-PTRA perfusion and RBF. Although PTRA abolished RAS and increased GFR and RBF, it did not affect MAP, kidney perfusion, or GFRn values, and both cortical and medullary STK and CLK-MTR (at 600Hz) remained similarly unchanged. The observation that STK-MTR tracks with renal fibrosis and perfusion supports MTR (particularly at 600Hz offset) as a useful tool to predict renal outcomes following PTRA.

RVD is responsible for nearly 15% of cases of CKD (36), carries a worse prognosis than other causes of CKD (37), and is highly prevalent among those with atheromatous disease of other organs (38). Randomized controlled trials have largely failed to detect clear-cut benefits for renal revascularization over medical therapy in most RVD patients (3, 4). Indeed, PTRA did not restore MAP in our study. These findings are also consistent with our previous observations that PTRA increased both RBF and GFR in RAS swine, but failed to alter blood pressure or perfusion (28).

Overall, clinical trials show that following revascularization only ~30% of RVD patients demonstrate improvement in renal function (5, 39), which might in fact worsen (6), possibly linked to preexisting intrarenal damage and ischemia-reperfusion injury. Furthermore, 7% of RVD patients assigned to revascularization in the ASTRAL trial experienced a major procedure-related complication (40), underscoring the need to identify subjects likely to benefit from this procedure. However, indices that determine a meaningful restoration of renal function in RVD revascularization are ambiguous, and sensitive and non-invasive tools to predict response to therapy in RVD are urgently required (9).

The present study addressed two important hypotheses. Firstly, we tested the hypothesis that MTR is a correlate of renal functional and hemodynamic measures in pigs with RVD. Early detection of renal fibrosis is imperative, as accumulating evidence suggest that the primary event for progression of CKD is tubulointerstitial fibrosis (22). MTR is considered to be an index of the tissue macromolecule content (e.g., proteins, including collagen). We have previously validated the use of MTI for non-invasive assessment of fibrosis in murine (at 16.4T) (24, 25) and swine (at 3.0T) (26) models of unilateral RAS. Furthermore, we demonstrated that during acute graded ischemia in healthy pigs without renal fibrosis, MTI-measured MTR remained relatively stable despite decreases in renal perfusion (27), suggesting that MTR is independent of renal hemodynamics. Extending our prior studies, the current study shows that baseline MTI-measured STK cortical MTR correlates well with the degree of residual kidney fibrosis observed after revascularization in RVD, probably because fibrosis did not regress.

The experimental swine RVD model captures distinct characteristics of human RVD, noted by kidney tissue hypoxia, excessive release of inflammatory cytokines, and infiltration of inflammatory cells (41–47). Interstitial fibrosis in RVD likely underlies the association of MTR with kidney function and hemodynamics. Thus, MTR may provide a reliable early, non-invasive, and sensitive assessment of residual renal fibrosis, and detection of subtle changes in kidney function and hemodynamics in patients with RVD.

Secondly, we tested the hypothesis that MTR predicts renal recovery in response to PTRA. We observed that PTRA improved perfusion and GFRn in about half the pigs, and increased mean STK-GFR and RBF. Interestingly, similar to clinical observations, mean STK cortical MTR was slightly higher following PTRA than in control RVD pigs, possibly due to progression of renal fibrosis and microvascular loss (48), given the continuous high-fat diet after revascularization, as well as possibly PTRA-inducted damage. Nevertheless, changes cortical and medullary 600Hz-STK-MTRs following PTRA correlated well with changes in their perfusion and in GFR. Importantly, baseline cortical 600Hz-MTR inversely correlated with post-PTRA perfusion and RBF. This relationship might be driven through fibrosis-induced microvascular rarefaction, a mediating or non-observable factor in our study (22, 23). Therefore, baseline MTR potentially provides an indirect measure of renal microvascular reserve, an important factor in PTRA outcomes (49, 50), and predicts the ability of cortical perfusion to improve. Contrarily, baseline cortical and medullary MTRs did not correlate with post-PTRA GFRn, and thus do not necessarily predict the response of filtration function to PTRA.

Lastly, 600Hz-MTR was found to be slightly superior to 1000Hz-MTR for detection of renal fibrosis, as we observed before (26), as well as for prediction of hemodynamic changes.

Our study bears some limitations, most importantly a limited sample size. MTI is semi-quantitative, yet future development of quantitative magnetization transfer techniques may provide robust estimates of renal fibrosis (51, 52). In our young pigs, PTRA was successful in restoring RBF and GFR, but not perfusion and GFRn, which (unlike whole-kidney RBF and GFR) do not account for changes in kidney volume that precede regional functional changes per unit tissue. Interestingly, STK-MTR was slightly higher following PTRA compared to untreated RVD, whereas trichrome staining of collagenous connective tissue fibers was not. Possibly, MTR may detect a wider range of macromolecules than staining. Furthermore, while STK-MTR correlates with renal fibrosis on trichrome staining, it may have more limited value in certain models, particularly in association with tissue edema, such as unilateral ureteral obstruction (25). The precise nature of the macromolecules captured by MTI also warrants further studies.

In summary, STKs of swine RVD have elevated cortical and medullary MTI-measured MTR values at both 600 and 1000Hz, which PTRA did not reduce. Baseline STK-MTR, particularly at the 600Hz-offset, is a useful tool to predict perfusion, RBF, and residual kidney fibrosis in experimental swine RVD undergoing revascularization. Additionally, changes in STK-MTR in response to PTRA correlate well with concurrent changes in perfusion and GFR. Strong linear correlation of changes in MTR with changes in perfusion and GFRn suggest a promise for quantitative applications in predicting renal outcomes. These findings indicate the capacity to predict renal hemodynamic recovery potential, and help direct management of subjects with RVD in clinical settings, given the wide availability of imaging sequence in clinical scanners, non-contrast scans, and relatively short acquisition time (~1.5 minutes for 1 M0 and 2 Mt scans). Future experimental and clinical studies are required to the association of MTR with measures of renal hemodynamics in acute non-fibrotic as well as chronic fibrotic kidney disease settings.