MULTI-PARAMETRIC RENAL MAGNETIC RESONANCE IMAGING IN EARLY KIDNEY TRANSPLANTATION

Abstract Background and Aims Existing methods of investigating renal transplant dysfunction do not provide reliable information regarding diagnosis nor prognosis. Multi-parametric magnetic resonance imaging (MRI) may provide novel biomarkers for evaluation of transplant dysfunction. We aim to determine how MRI parameters change over the first year of transplantation, and how these relate to future renal function. Method Patients receiving a kidney transplant attended for study visits at 6, 26 and 52 weeks post operatively, comprising measurement of clinical and biochemical parameters, together with research multi-parametric MRI. Imaging measurements comprised kidney volume, arterial spin labelling (ASL) perfusion, T1 relaxation time, T2*, apparent diffusion coefficient (ADC) and fractional anisotropy (FA). Imaging was performed at 3.0 Tesla using a Siemens MAGNETOM Prisma system. Regions of interest were drawn in whole kidney (WK), cortex (Cx) and medulla (Md) (figure 1). Results 20 patients were included: 16 were male, with age 55.5±12.8 years, baseline eGFR 54.0±23.6 ml/min/1.73m2, and blood pressure 146/80 ± 15/15 mmHg. 14 received deceased, and 6 received live, donor transplants. Patients were all managed with tacrolimus, mycophenolate and low dose prednisolone, following induction therapy with either basiliximab or anti-thymocyte globulin. 6 week ADC was 1.69±0.14 in WK, 1.65±0.08 in Cx, and 1.67±0.10 ×10-3 mm2/s in Md. FA was 0.19±0.04 in WK, 0.14 ± 0.04 in Cx and 0.22 ± 0.10 in Md. T2* was 57.6±9.4 in WK, 63.9±8.7 in Cx and 45.0±8.0 ms in Md. Over the 3 visits there was reduction in FA (p=0.008) and medullary T2* (p<0.001). There was no significant change over 12 months in any other MRI parameter. Over 1 year the median change in eGFR was -2ml/min/1.73m2. There was correlation between baseline eGFR and the following variables: volume (r=0.29, p=0.04), whole kidney ADC (r=0.36, p=0.01), cortical ADC (r=0.46, p=0.001), representative cortex ADC (r=0.48, p<0.001) and medullary ADC (r=0.29, p=0.04). There was no correlation between eGFR and other imaging biomarkers. There was correlation between baseline whole kidney ADC (r=0.54, p=0.02), cortical ADC (r=0.49, p=0.03), and renal function at 12 months. Conclusion Diffusion weighted MRI measurements correlate with eGFR and may allow improved prognostication regarding future renal function. Certain MRI parameters including FA and R2* vary depending on time point of transplantation which may reflect changes in transplant microstructure in the early postoperative period. Multi-parametric MRI provides a novel method of evaluating renal transplants non-invasively and may allow more accurate prediction of future transplant function than existing biochemical measures.