Purpose

Renal fibrosis, associated with the deposition of collagen in the cortical interstitial space, is considered an important predictor for allograft prognosis and can be used to adapt treatment ¹. Renal fibrosis is currently evaluated using histopathological analysis of biopsied specimens of the kidney. However, biopsies may have severe complications, including bleeding, and are subject to sampling bias. There is recent interest in advanced multiparametric MRI methods that may be used to noninvasively assess renal fibrosis ¹. T_1ρ mapping, which is sensitive to the interactions between water molecules and macromolecules including collagen, may be of specific interest for assessment of renal fibrosis. T_1ρ has shown to be sensitive to the degree of fibrosis in the liver ^2,3. However, there are no reports in which T_1ρ is assessed as potential biomarker for prediction of fibrosis in the kidney. The goal of this preliminary study was to investigate the utility of T_1ρ MRI for the assessment of fibrosis in renal allografts.

Methods

15 patients with stable functional allograft (M/F 9/6, mean age 56y, mean eGFR 65.9±12.6 ml/min/1.73 m²) and 6 patients with chronic dysfunctional allograft and biopsy-confirmed renal fibrosis (M/F 2/4, mean age 56y, mean eGFR 23.4±4.1 ml/min/1.73 m²) were prospectively enrolled in this IRB-approved study. All patients underwent multiparametric MRI at 1.5T. The protocol included T_1ρ imaging obtained during 4 x 10s breath holds in a single coronal slice throughout the renal allograft (spin-lock-prepared FLASH sequence, spin-lock strength 500 Hz, spin-lock times 4.8, 9.6, 19.2 and 38.4 ms, TE 3.5 ms, TR 20 ms, flip angle 15°, matrix 128x128, FOV 360x360 mm²). Repeatability of the T_1ρ measurement, as determined by coefficient of variation (CV) measurements, was tested in 4 patients (time between scans 17-45 days). Regions of interest (ROIs) were drawn in the cortex and medulla [three ROIs (upper pole, mid pole, lower pole) for both cortex and medulla)]. Average T_1ρ values in the cortex and medulla ROIs were quantified. Differences in T_1ρ values between the groups were assessed using Student t-tests. ROC analysis was employed to determine the diagnostic performance of T_1ρ for differentiation between functional and fibrotic allografts.

Results

T_1ρ measurements were more repeatable in the cortex than in the medulla (mean CV T_1ρ cortex 6.4%, medulla 14.6%). Representative T_1ρ maps in functional and fibrotic allografts are shown in Figure 1. While T_1ρ values were not significantly different between functional and fibrotic allografts in the medulla (T_1ρ medulla functional 132.2±25.5 ms, fibrotic 136.9±10.3 ms, P=0.677), significant differences were observed in the cortex (T_1ρ cortex functional 104.9±14.8 ms, fibrotic 119.5±8.3 ms, P=0.036; Figure 2). Preliminary ROC analysis showed an AUC of 0.856 (sensitivity 100%, specificity 73.3%, T_1ρ threshold 107.8 ms) for differentiation between functional and fibrotic allografts (Figure 3).

Discussion and Conclusions

The significant increase in T_1ρ in fibrotic renal transplants mirrors earlier findings in liver fibrosis ^2,3. It has been suggested that T_1ρ increase in fibrosis is associated with elevated collagen levels ³. In the near future, we are planning to recruit more patients with renal fibrosis and we will correlate the T_1ρ measurements with tissue collagen content measurements from histopathological analysis. In conclusion, our preliminary results suggest that T_1ρ may be a suitable biomarker for assessment of renal fibrosis, which will be verified in a larger study.

Acknowledgements

This research was supported by the Society of Abdominal Radiology (SAR) Morton Bosniak Research Award and Guerbet LLC Grant.