Evaluation of early stage diabetic nephropathy with bold MRG, diffusion tensor imaging and diffusion weighted imaging

Purpose: The aim of this prospective study is to investigate the effectiveness of BOLD (Blood Oxygenation Level Dependent) MRI, Diffusion tensor imaging (DTI) and Diffusion weighted imaging (DWI) sequences, which are non-contrast MRI (Magnetic Resonance Imaging) methods, in the evaluation of early-stage diabetic nephropathy. Materials and Methods: 53 patients (28 female, 25 male) diagnosed with type 2 Diabetes Mellitus (DM) and 31 healthy volunteers (19 female, 12 male) were prospectively included in the study. The patients were divided in to two groups according to albuminuria levels as 32 (17 female, 15 male) normoalbuminuria and 21 (11 female, 10 male) microalbuminuria. With a 3 Tesla MRI device, the kidneys were evaluated with BOLD MRI, DTI and DWI sequences in the coronal plane. By placing ''region of interest'' (ROI) in the upper pole, middle section and lower pole cortex and medulla of both kidneys, R2* in BOLD MRI, fractional anisotropy (FA) in DTI, the apperent diffusion coefficient (ADC) in DWI value was measured. Independent-Samples Mann-Whitney U Test was used for comparisons of two groups, Independent-Samples Kruskal-Wallis Test and Dunn's Post Hoc test with Bonferronni correction were used for comparisons between more than two groups. ROC (receiver operating characretistics) analysis was used to predict the presence of normoalbuminuric and microalbuminuric diabetic nephropathy (DN), and cut-off values were found. P<0.05 was accepted as the cut-off value of statistical significance. Findings: Renal cortex FA value was significantly higher in the normoalbuminuria group compared to the healthy control group (p=0.009) and significantly lower than in the microalbuminuria group (p=0.033). Renal cortex ADC value was significantly lower in the normoalbuminuria group compared to the healthy control group (p=0.018). In the estimation of the normoalbuminuria group, the specificity of renal cortex FA value was found to be 74.2%, sensitivity 59.4% (AUC 0.712, cut-off value >0.320) (p=0.001), the specificity of renal cortex ADC value 96.8%, sensitivity 34.4% (AUC) 0.706, cut-off value ≤1.693 (p=0.002); In the estimation of the microalbuminuria group, the specificity of the renal cortex ADC value was found to be 61.3%, the sensitivity 71.4% (AUC 0.673, and the cut-off value ≤1.775) (p=0.025). Conclusion: It is thought that the renal cortex FA and ADC values measured in DTI and DWI sequences, which is a noninvasive, non-contrast method, may play a role in the evaluation of early stage DN. Key Words: Diabetic nephropathy, BOLD MRI, Diffusion Tensor Imaging, Diffusion-Weighted Imaging, FA, ADC, R2*.