Noninvasive imaging of renal urea handling by CEST-MRI

Abstract

Purpose: Renal function is characterized by concentration of urea for removal in urine. We tested urea as a CEST-MRI contrast agent for measurement of the concentrating capacity of distinct renal anatomical regions.

Methods: The CEST contrast of urea was examined using phantoms with different concentrations and pH levels. Ten C57BL/6J mice were scanned twice at 7 T, once following intraperitoneal injection of 2M 150 µL urea and separately following an identical volume of saline. Kidneys were segmented into regions encompassing the cortex, outer medulla, and inner medulla and papilla to monitor spatially varying urea concentration. Z-spectra were acquired before and 20 minutes after injection, with dynamic scanning of urea handling performed in between via serial acquisition of CEST images acquired following saturation at +1 ppm.

Results: Phantom experiments revealed concentration and pH-dependent CEST contrast of urea that was both acid- and base-catalyzed. Z-spectra acquired before injection showed significantly higher CEST contrast in the inner medulla and papilla (2.3% ± 1.9%) compared with the cortex (0.15% ± 0.75%, P = .011) and outer medulla (0.12% ± 0.58%, P = .008). Urea infusion increased CEST contrast in the inner medulla and papilla by 2.1% ± 1.9% (absolute), whereas saline infusion decreased CEST contrast by -0.5% ± 2.0% (absolute, P = .028 versus urea). Dynamic scanning revealed that thermal drift and diuretic status are confounding factors.

Conclusion: Urea CEST has a potential of monitoring renal function by capturing the spatially varying urea concentrating ability of the kidneys.

Keywords: CEST; MRI; kidney; urea.

Figure 1.

Concentration and pH-dependent responses of urea phantoms. (a) MTR asymmetry map of a phantom with different concentrations (#1: 100, #2: 200, #3: 400, #4: 600 and #5: 800 mM) of urea at pH 6. (b) Z-spectra from urea phantoms with different concentrations shown in (a). (c) Concentration-dependent urea MTR asymmetry. (d) MTR asymmetry map of a phantom with different 800mM urea at varying pH (#1: 6.25, #2: 6.5, #3: 6.75, #4: 7.0, #5: 7.25, #6: 7.5). (e) Lorentzian-fitted z-spectra from urea phantoms shown in (d). (f) pH-dependent urea MTR asymmetry at different concentrations.

Figure 2.

Magnitude-reconstructed images of urea-infused mouse. (a) High resolution T2-weighted anatomical image delineating cortex (C), outer medulla (OM) and inner medulla and papilla (IM+P) of the kidneys. (b) Unsaturated S0 image. (c) Raw CEST image with saturation at 1 ppm before urea infusion. (d) Raw CEST image with saturation at 1ppm after urea infusion.

Figure 3.

In vivo MTR asymmetry maps acquired before and after infusion of either urea or saline. Pseudo color-coded +1 ppm MTR asymmetry maps are overlaid on anatomical T2-weighted images. The inner medulla and papilla of each kidney is delineated with a blue line. MTRasym values increase significantly in the inner medulla and papilla following urea infusion, but not following saline infusion.

Figure 4.

Representative z-spectra with 4-pool Lorentzian fitting acquired from each part of the kidney before and after urea infusion. Each Lorentzian function shows the contribution of each pool to the measured z-spectra. Red arrows indicate increased urea CEST contrast at +1 ppm after infusion of urea. Other parts of the kidney do not show changes in saturation transfer at +1ppm.

Figure 5.

Representative z-spectra with 4-pool Lorentzian fitting acquired from each part of the kidney before and after saline infusion. Saline adminsitration did not generate any difference in z-spectra, including the inner medulla. The red arrow in the inset in pre-infusion inner medullary z-spectra (top right) indicates the urea contrast at 1 ppm.

Figure 6.

(a) MTR asymmetry (MTRasym) measured from each region of the kidney before and after the infusion of urea. (b) MTRasym measured before and after the infusion of saline. (c) The change in MTRasym following infusion of urea was greatest in the inner medulla and papilla. No changes in MTRasym were observed following saline infusion. *P < 0.05, ^#P < 0.05 vs. cortex (C), ^##P < 0.01 vs. C, ⁺P < 0.05 vs. outer outer medulla (OM), ⁺⁺P < 0.01 vs. OM.

Figure 7.

Dynamic time curves of MTR asymmetry change and S₀ image signal intensity in each part of the kidney. (a) Dynamic measurement of MTRasym without thermal drift correction reveals a steady increase in MTRasym in each part of the kidney following urea infusion. (b) Normalized signal intensities of S_o images acquired during dynamic scanning reveal a time dependent decrease in S_o values, likely due to thermal drift. (c) When dynamic MTRasym measurements from (a) are corrected for thermal drift by accounting for the decrease in S_o image signal intensities over time (b), changes in MTRasym are only observed in the inner medulla and papilla. (d) While such correction enabled dynamic measurement in some mice, others still demonstrated highly noisy dynamic measurements with early peaks in inner medulla and papilla MTRasym and rapid declines. Black arrows indicate the time point when the urea infusion was made.