Magnetic resonance diffusion tensor imaging for evaluation of histopathological changes in a rat model of diabetic nephropathy

Abstract

Objectives: The aim of this study was to investigate whether magnetic resonance (MR) diffusion tensor imaging (DTI) allows assessment of renal pathologies in a rat model of diabetic nephropathy.

Materials and methods: Twenty-one male Sprague-Dawley rats were divided into 3 groups: (1) untreated controls, (2) diabetes (DM), (3) diabetes with uninephrectomy (DM UNX) to accelerate renal impairment. Eight weeks after diabetes induction with streptozotocin, MR imaging was performed in a 1.5-T scanner using an 8-channel wrist coil. Morphological proton density images and echoplanar DTI were obtained (b = 0 and 300 s/mm, 6 diffusion directions). Renal apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values were calculated for each of the different anatomical layers of the kidney. Imaging results, laboratory parameters of diabetic state and kidney function, and renal histopathological changes (glomerulosclerosis, tubular dilatation, and renal fibrosis) were compared between groups. Correlations between FA and histopathological changes were evaluated.

Results: All diabetic animals developed hyperglycemia and hypoinsulinemia. Uremia, albuminuria, and histopathological changes were most pronounced in DM UNX animals. Fractional anisotropy was significantly reduced in DM UNX animals in the cortex (CO) (0.167; confidence interval [CI], 0.151-0.184; P < 0.001), outer stripe of the outer medulla (OS) (0.254; CI, 0.225-0.283; P = 0.038), and inner medulla (IM) (0.459; CI, 0.395-0.523; P = 0.008) compared with control animals (CO, 0.251; CI, 0.224-0.277; OS, 0.309; CI, 0.267-0.350; IM, 0.559; CI, 0.515-0.603). In DM-without-UNX animals, only cortical FA was significantly lower than in controls (P < 0.001). Between groups, ADC values were not different, except for cortical ADC, which was higher in DM UNX animals than in controls. Significant negative correlations were observed between the FA of different anatomical layers and the extent of glomerulosclerosis (CO, P = 0.003, r = -0.65; and OS, P = 0.022, r = -0.52), tubulointerstitial fibrosis (IM, P = 0.028, r = -0.50), and tubular dilatation (CO, P = 0.015, r = -0.55; and IM, P = 0.006, r = -0.61), respectively.

Conclusions: Magnetic resonance DTI by reduction of FA identified renal pathologies of diabetic nephropathy such as glomerulosclerosis, interstitial fibrosis, and tubular damage. Representing different stages of disease, DM and DM UNX animals could be differentiated. Thus, MR DTI may be valuable for noninvasive detection and monitoring of renal pathology in patients with diabetes.