MP368
MULTIPARAMETRIC MAGNETIC RESONANCE IMAGING ASSESSMENT OF CHRONIC KIDNEY DISEASE

INTRODUCTION AND AIMS: Progression of Chronic Kidney Disease (CKD) occurs as a result of a common pathway of mechanisms of inflammation and fibrosis that may be independent of the underlying aetiology. In current clinical practice, assessment of renal parenchymal damage is confined to renal biopsy, which is limited by sampling error and is associated with patient risk. Recent advances in Magnetic Resonance Imaging (MRI) allow assessment of structural and functional changes relevant to kidney disease. We performed a multiparametric MRI study to assess its utility and reproducibility in patients with CKD.

METHODS: 21 people were studied, 14 patients with CKD Stage 3-4 who had renal biopsies performed as part of routine clinical care, seven healthy volunteers (HVs) as a comparator group. Patients and HVs had two multiparametric renal MRI scans performed 7-14 days apart. Biochemical and clinical parameters were collected at the first scan. MRI scans were performed on a 3T Philips Ingenia scanner. Structural assessment included kidney volume, longitudinal relaxation time (T₁)mapping and diffusion weight imaging (DWI) to compute apparent diffusion coefficient (ADC) as markers of fibrosis and/or inflammation. Functional assessments were Arterial Spin Labelling (ASL) to measure renal perfusion, phase contrast to measure blood flow in the renal artery and Blood Oxygenation Level Dependent (BOLD) Imaging as a measure of renal oxygenation. Coefficient of variance (CV) was calculated for each MRI measure between the two scans.

RESULTS: CKD patients had a mean age of 57±18yrs, ten were male, mean baseline estimated Glomerular Filtration Rate (eGFR) was 39±13mls/minute/1.73m³ and mean urine Protein Creatinine Ratio (PCR) was 59±64mg/mmol. Biopsy results showed six patients had ischaemic nephropathy, four had tubulointerstial disease and four had IgA nephropathy. HVs had a mean age 35±15yrs, six were male and all subjects had an eGFR>60mls/minute/1.73m³ and urine PCR <15mg/mmol. CKD patients had higher T₁ and lower ADC values than HVs indicating the presence of more fibrosis/inflammation. CKD cortical and medullary T₁ values were 1587±79ms and 1765±70ms respectively compared to HVs of 1410±85 and 1677±78 (p<0.001) respectively. ADC values were lower in CKD patients at 2.05±0.34(µm²/ms) versus 2.29±0.09(µm²/ms) in HVs (p=0.01). Renal perfusion (ASL) was lower in CKD patients, 146±36ml/100g/min versus 222±79ml/100g/min (p=0.003). There were no differences in renal volumes and T₂* (BOLD) between groups. No single MR measure correlated with CKD patient eGFR. Reproducibility was excellent for cortical and medullary T₁ (CV=2.65% and 1.98% respectively), T₂*(CV=2.58%), ADC (CV=8.16%) and renal volume (CV=2.98%).

CONCLUSIONS: This study demonstrates that mutiparameteric MRI differentiates between HVs and CKD patients, and is reproducible. The MR results reflect the known pathophysiological changes expected in CKD; reduced perfusion and the presence of fibrosis and/or inflammation, but interestingly did not demonstrate a difference in T₂* a potential marker of renal hypoxia. In future work, we plan to correlate histology findings on renal biopsy with MR measures and investigate methods of differentiating between fibrosis and inflammation using MRI. Further studies are required to build on this initial work to determine how best multiparametric MRI can be used to assess whole kidney pathology and prognosis.