SP224
TRI−EXPONENTIAL APPROACH FOR INTRAVOXEL INCOHERENT MOTION ANALYSISOF MULTI B−VALUE DIFFUSION WHEIGTED MRI DATA FOLLOWS GFR CHANGES IN HEALTHY HUMANS

Introduction and Aims: The currently available methods to assess kidney perfusion and tubular flow are costly, time-consuming and do not differentiate between kidneys. Potentially, a tri-exponential intravoxel incoherent motion (IVIM) model for diffusion weighted MR-images (DWI) may provide a cheaper, faster and non-invasive method. In this study we hypothesized that changes in glomerular perfusion and tubular urine flow can be reliably detected by diffusion parameters, as derived from a tri-exponential IVIM model.

Methods: Tri-exponential IVIM modeling was assessed in DW-scans performed in three groups of young healthy subjects. First, in two groups of 6 participants (19-28 yrs) intra and inter-session variability was assessed. Secondly, in 8 subjects (18-24 yrs), we assessed the response of the IVIM model to changes in measured GFR and effective renal plasma flow (ERPF, by ¹²⁵I-thalamate and ¹³¹I-hippuran clearing tests) induced by continuous angiotensin-II (Ang-II) infusion at 0, 0.3, 0.9 and 3.0 ng/kg/min. DW MRI consisted of 2D coronal slices acquired during free breathing, with b-values set at 0, 2, 4, 8, 12, 18, 24, 32, 40, 50, 75, 110, 200, 300, 450, 600 s/mm², 9 gradient directions and a voxel size of 3.5x3.5x3.5 mm. Respiratory motion was compensated using 2D rigid registration of each slice to the corresponding mean b=0 s/mm² image. For each voxel, a tri-exponential model was fitted returning the diffusion coefficient (D), pseudo-diffusion coefficients (D*¹, D*²) and signal fractions for D*¹, D*² and D (f₁, f₂, f_D). We delineated the cortex and medulla and took the mean parameter values of all delineated voxels.

Results: Based on simulations that we ran on the pooled baseline data, both pseudo diffusion coefficients (D*^,1 and D*^,2) were fixed to values obtained from a fit to all delineated baseline data. For D, f₁, f₂, f₃ the inter-session CVs were 1.7%, 10.4%, 13.8% and 2.2% and the intra-session CVs were 3.5%, 14.4%, 6.5% and 1.9%, respectively. The mean values at baseline were 1.93±0.08×10^-3 mm²/s, 11.3±2.4%, 7.8±1.9% and 80.8±2.2% for D, f₁, f₂ and f₃, respectively. The perfusion fractions in the kidneys changed as function of Ang-II dose. The slopes of linear fits to f₁ and f₂ as function of Ang-II were -4.7±0.7% (p=0.02) and 6.4±0.2% (p=0.001) per ng/kg/min Ang-II respectively. Changes in f₁ (fraction of signal from blood in larger vessels) correlated to ERPF was R=0.42 (p=0.01), f₂ (fraction of signal from blood in the capillaries/glomerulus and pre-urine in the renal tubuli) correlated to GFR R=-0.62 (p<0.001).

Conclusions: Our data indicate that the tri-exponential model for DWI-data has potential as a surrogate marker for kidney function; the method is highly repeatable and responsive to changes in GFR and ERPF induced by Ang-II. Changes in IVIM parameters correlate to measured ERPF and GFR ¹²⁵I-thalamate and ¹³¹I-hippuran clearing tests.