In vivo evaluation of renal function using diffusion weighted imaging and diffusion tensor imaging in type 2 diabetics with normoalbuminuria versus microalbuminuria

Abstract

This work aims to estimate the value of diffusion weighted imaging (DWI) and diffusion tensor imaging (DTI) in detecting early-stage kidney injury in type 2 diabetic patients with normoalbuminuria (NAU) versus microalbuminuria (MAU) prospectively. A total of 30 T2DM patients with normal kidney function were recruited and assigned to the NAU group (n = 14) or MAU group (n = 16) according to 8 h overnight urinary albuminuria excretion rate (AER) results. A contemporary cohort of health check-up recipients were included as controls (n = 12). DWI and DTI scans were performed on bilateral kidney using SE single-shot EPI, and apparent diffusion coefficient (ADC) and fractional anisotropy (FA) of the renal parenchyma was determined from ADC and FA maps of the three groups. ADC and FA values were compared among the three groups. According to DWI with a b value of 400 s/mm², the MAU and NAU groups showed significantly lowered mean ADC values compared with the healthy controls (P<0.01). The mean ADC in the MAU group [(2.22±0.07)×10^–3 mm²/s] was slightly lower than that of the NAU group [(2.31±0.22)×10^–3 mm²/s], but this difference was not statistically significant (P>0.05). The FA value in the MAU group was higher than that in the control group (0.45±0.07 vs. 0.39±0.03, P = 0.004) but did not differ from that in the NAU group (0.42±0.03) (P>0.05). ADC and FA values may be more sensitive than urine AER in reflecting early-stage kidney injury and, hence, may facilitate earlier detection and quantitative evaluation of kidney injury in T2DM patients. Combined evaluation of ADC and FA values may provide a better quantitative approach for identifying diabetic nephropathy at early disease stages.

Keywords

type 2 diabetes mellitus / microalbuminuria / diffusion weighted imaging / diffusion tensor imaging / early-stage kidney injury

Introduction

Although the pathogenesis and progression of renal damage in patients with type 2 diabetes mellitus (T2DM) have yet to be fully understood, histological changes induced by chronic hypoxia and hemodynamic abnormalities are considered the major contributing factors to the disease. Diabetic nephropathy (DN), a common complication of T2DM, is usually manifested by renal enlargement, glomerular hyperfiltration, and hyperperfusion in its early stages. Along with the progression of T2DM, DN patients may present albuminuria that gradually evolves from intermittent to persistent and from mild to severe. Renal lesions of DN are mostly intractable when diagnosed in clinical settings. Therefore, early detection and intervention for renal lesions of DN is necessary before the lesions advance to irreversible levels. Microalbuminuria (MAU) level, the current diagnostic gold standard for DN [ 1], also represents an index of diabetic renal damage severity [ 2– 4]. Persistently high MAU may suggest the advent of stage III of DN, which is the result of progressive glomerulosclerosis due to long-term hyperfiltration. Using the Mogensen classification, stage III DN is labeled microalbuminuria or incipient nephropathy with urine albumin excretion rates (UAERs) ranging between 20 and 200 μg/min. Many studies have targeted MAU for intervention and as a specific biomarker for progression among DN patients. However, the mechanism underlying MAU implies that detection of MAU may not correlate well with or may even lag behind DN progression because MAU is hardly detectable until the occurrence of clinically significant renal damage. In addition, MAU in early-stage DN is characterized by intermittent and uneven daily excretion, as demonstrated by our previous study [ 5] and others [ 6]. Moreover, test results are prone to influence by a variety of pathological and physiological factors. Given that short-term hyperglycemia, exercise, urinary tract infection, significant hypertension, heart failure, and acute febrile disease may all result in a transient increase in MAU, comparing the severity of early renal impairment between individuals based on MAU alone or evaluating the progression of early renal impairment in a given individual simply according to dynamic changes in MAU would be difficult.

Diffusion weighted imaging (DWI) and diffusion tensor imaging (DTI) are non-invasive imaging options for assessing renal function [ 7]. DWI is often used to determine the severity of ischemia [ 8], which is characterized by high sensitivity to cell edema and atrophy resulting from various causes, particularly to acute and chronic cell damage induced by hypoxia. The apparent diffusion coefficient (ADC) of renal parenchyma is detected to acquire comprehensive information on diffusion of water molecules within the kidney tissue, thus facilitating early detection of the severity of ischemic and hypoxic renal injury. As fractional anisotropy (FA) values calculated from DTI manifest distinctively varied FA of the renal parenchyma, FA values aid in further identifying kidney lesions, especially abnormalities in the microstructure of the kidney. Given that hypoxia inevitably occurs in the progress of DN, we hypothesize that parameters of DWI and DTI may help in the detection and evaluation of early renal damage in diabetic patients.

To explore the diagnostic value of ADC and FA values for early renal damage in T2DM patients, the present study performed routine MRI, DWI, and DTI of the kidney in T2DM patients with normal renal function but different degrees of albuminuria. The renal contour, dimension, anatomical structure, ADC values, and FA values of the renal parenchyma were compared with those of healthy controls for analysis of the correlation between these parameters and levels of albuminuria.

Subjects and methods

Ethics statement

The study protocol was approved by the Ethics Committee of the First Affiliated Hospital, Guangzhou Medical University (Approval Number: GYFYY-2010-18). All subjects provided written informed consent prior to joining the study.

Study subjects and group allocation

Between August 2010 and December 2010, we enrolled 30 T2DM patients who had been positively diagnosed in the Outpatient Department of Endocrinology of the First Affiliated Hospital, Guangzhou Medical University according to the diagnostic criteria for diabetes proposed by the World Health Organization (WHO) in 1999. The patients were grouped according to their MAU results as follows: (1) the normoalbuminuria (NAU) group (8 h overnight albuminuria<20 μg/min), which consisted of 14 patients, 6 males and 8 females, aged 38 to 67 years, with a median age of 56±7 years, disease duration of 5.8±6.4 years, and body mass index (BMI) of 24.7±2.8 kg/m²; (2) the microalbuminuria (MAU) group (8 h overnight albuminuria= 20 μg/min to 200 μg/min), which consisted of 16 patients, 7 males and 9 females, aged 43 to 73 years, with a median age of 58±8 years, disease duration of 9.2±6.1 years, and BMI of 23.1±2.6 kg/m². The exclusion criteria are as follows: (1) patients who had primary hypertension before or when diagnosed with DM; (2) patients who had a history of kidney disease resulting from causes other than T2DM; (3) patients who had concomitant coronary heart disease and a history of stenting; and (4) patients who had abnormal renal anatomy as shown by MRI. The normal control (NC) group consisted of 12 healthy adults who had physical check-ups in our hospital and whose renal functions were normal; this group included 7 males and 5 females aged 51 to 68 years, with a median age of 57±6 years and BMI of 24.0±2.4 kg/m². Control group subjects were free from diabetes, hypertension, coronary heart disease, and primary and secondary renal disease. All three groups were matched for age and gender.

Laboratory investigations and imaging studies

MAU measurement

Urine samples were collected when subjects were free of fever and female subjects were not on their menstrual period. The subjects were required to avoid high-fat diets and strenuous activity the night before urine was collected. All urine was collected from 22:00 to 6:00 the following day, and urine output was recorded. A urine sample of 5 ml was taken after uniform mixing for detection of MAU via the immune transmission turbidimetric method (Hitachi 7170 Automatic Biochemical Analyzer, Hitachi, Japan).

Blood sample tests

A 5 ml sample of fasting cubital venous blood was obtained in the early morning from each patient after fasting overnight for more than 8 h. The serum was separated after static standing for 30 min and stored at -80 °C. A Beckman UniCel DxC800 Synchron automatic biochemical analysis system was used to measure serum cystatin C (Cys C) by turbidimetric immunoassay and serum creatinine (sCr) by the picric acid method.

MRI protocol

MRI scanning of both kidneys was performed in three groups using a PHILIPS Achieva 1.5T superconducting MR scanner (Intera Nova Dual, Netherlands) and an 8-channel abdominal TORSOL coil. The scan sequences included conventional axial T1WI, T2WI, DWI, and DTI. The parameters were as follows: (1) T1WI (T1WI /FFE): TR: 8.7 ms; TE: 4.6 ms; matrix: 220×180; field of view (FOV): 345 mm×350 mm; and number of signal averaged (NSA) = 2. T2WI (T2WI /TSE): TR: 1800 ms; TE: 90 ms; matrix: 288×144; FOV: 405 mm×287 mm; and NSA= 3. (2) DWI Scan: DWI was performed with single shot spin-echo echo-planar imaging (SE-EPI) sequence according to the following parameters: b values: 400, 500, 600, and 800 s/mm², respectively, with diffusion gradient in three directions. The parameters of axial DWI were as follows: shortest TR; TE: 60 ms; slice thickness/gap: 3 cm/1 cm; FOV= 350 mm × 270 mm; NSA= 2; and scan time= 156.5 s. The parameters of coronal DWI were as follows: TR: 3500 ms; TE: 60 ms; slice thickness gap: 5 cm/1cm; matrix: 180 × 121; FOV: 360 mm × 330 mm; NSA= 2; and scan time= 157 s. (3) DTI scan: Coronal DTI was performed using single-shot spin-echo echo-planar sequence (SS-SEEPI) with b values of 0 and 500 s/mm²; 6–8 directions of diffusion gradient:16; TR: 730 ms; TE: 72 ms; slice thickness/gap: 5 mm/1 mm, FOV: 380 mm×380 mm; and NSA= 3.

Post processing of DWI and DTI images

The same operator processed the initial DTI images to acquire axial and coronal DWI images, ADC maps, and FA maps of kidneys on the Philips workstation. Each b value of 400, 500, 600, and 800 s/mm² was matched with b₀ images to obtain their corresponding ADC images on the Philips workstation. The region of interest (ROI) of 100 mm² voxels was located at the level of the renal hilum for axial and coronal ADC images. Given that the renal cortex could not be accurately distinguished from the real medulla on DWI images, the ROI was localized within the renal parenchyma, bypassing the renal pelvis and calyces. Three sites were selected from the anterior, medial, and posterior renal parenchyma on both sides for axial images, and another three sites were selected from the upper, middle, and lower poles of the renal parenchyma on both sides for coronal images. ADC values were measured and averaged in the above six sites on each kidney. When the renal cortex was clearly differentiated from the renal medulla in the coronal FA maps, FA values were calculated for the coronal plane images via the renal hilum. The ROIs of the renal cortex were placed in the upper, middle, and lower poles with a size of 40 mm². Similarly, the ROIs of the renal medulla were placed in the upper, middle, and lower poles with a size of 100 mm². Results were continuously measured and averaged by the same operator for three times. The FA values ranged from 0 to 1. The closer an FA value is to 1, the higher the anisotropic water diffusion, and vice versa.

Statistical analysis

Statistical analysis was performed using SPSS 11.5 statistical software. Measurement data were indicated using \overline{x} ±s. ADC values of kidney at different b values and ADC values between the left and right kidney were tested and compared for the same individual using analysis of variance for repeated measure design. ADC values were compared between groups using single-factor analysis of variance. The correlations of ADC and FA values with MAU, CysC, and sCr were evaluated using linear correlation analysis in all groups. A level of P<0.05 was considered to be statistically significant.

Results

In the NAU and MAU groups, levels of sCr [(64.8±22.4) and (60.5±15.1) μmol/L, respectively] and Cys C [(0.9±0.2) and (0.8±0.2) mmol/L, respectively] were within the normal range and not significantly different from those in the NC group (P>0.05). A negative correlation was observed between urine albumin level and ADC value at a b value of 400 s/mm² (Spearman correlation coefficient= -0.265, P<0.01). No correlation was found between ADC values and levels of sCr and Cys C in the NAU and MAU groups (all P>0.05).

Routine MRI appearance

A clear demarcation was found between renal medulla and renal cortex, as shown in axial T1WI and T2WI images. Renal contour, dimension, and anatomical structure in the two observation groups showed no significant difference from those in the NC group (Fig. 1 A–1 C). No differences were observed among the three groups in terms of dimension and structure of the two kidneys (Fig. 2 A– 2 C).

The ADC values of the renal parenchyma in the NAU and MAU groups were significantly lower than those in the NC group at different b values (P<0.01). ADC values in the MAU group tended to decrease at all b values compared with those in the NAU group; these differences, however, were not statistically significant (P>0.05). The ADC values of the three groups are summarized in Table 1.

FA values of the renal medulla gradually increased from the NC group to the NAU group, and then to the MAU group; here, significant differences between the MAU and NC groups (P = 0.004) were observed. No significant differences were observed among the three groups with respect to FA values of the renal cortex (P>0.05). Analyses of the respective correlations of FA values of the renal cortex and medulla with albuminuria levels indicated no statistically significant difference (Spearman correlation coefficients= -0.196 and 0.037, respectively; P = 0.408 and 0.878, respectively). The FA values of the three groups are summarized in Table 2.

Discussion

Numerous studies have demonstrated that, because renal failure is reversible in the early stages of DN [ 9], early diagnosis and objective assessment of the severity of renal impairment are critical in developing a proper treatment protocol and improving patient outcomes. At present, although quantitative detection of MAU is widely used as the gold standard for DN diagnosis, the severity and progress of diabetic kidney impairments remain difficult to judge accurately and objectively based solely on the quantitative results of MAU, which are variable and irreproducible in many cases. In the present study, we performed routine sequence MRI, DWI, and DTI on the kidney of T2DM patients with normal renal functions but varying degrees of albuminuria (from NAU to MAU). Results showed that ADC and FA values may be more sensitive than urine AER in reflecting early-stage kidney injury.

DWI can detect the diffusion of water molecules in human tissue. When a diffusion-sensitive gradient is applied, diffusion of water molecules results in attenuation of MR signals, which depends on the ADC and b values of water molecules. Normal renal parenchyma with a rich renal blood supply acquires high ADC values. Besides the random motion of water molecules, the water content of the kidney, microcirculation blood flow perfusion, glomerular filtration, tubular resorption, and tubular secretion affect the diffusion characteristics of the kidney. Diffuse lesions of the kidney often result in various degrees of impairment in the above-mentioned functions, thereby impeding the free motion of water within the extravascular-extracellular space. Consequently, the diffusion of water molecules is limited in the direction along the gradient field, which leads to a decline in renal ADC values. Therefore, decreased ADC values are often a comprehensive manifestation of multiple factors and may serve as a basis for evaluating renal function by DWI [ 10]. Ries et al. [ 11] reported that renal ADC values are significantly lower in diabetic mice than in normal mice. The present study demonstrated that T2DM patients with normal albuminuria levels presented with a significant decrease in ADC values of the renal parenchyma compared with normal cases, though no differences in renal contour, dimension, and anatomical structure in conventional MRI or in biochemical indicators of renal function, such as sCr and Cys C, were observed between groups. This finding suggests that renal hemodynamic abnormalities and local tissue ischemia and hypoxia, which are renal impairments that may be detected by DWI, have already occurred in T2DM patients before persistent MAU is detected. Given that persistent renal hemodynamic abnormalities and local tissue ischemia and hypoxia will continuously affect the free diffusion of water molecules in the kidney tissue in this population, these abnormalities may be the pathological basis for subsequent MAU, and such functional changes may be a prelude to subsequent structural changes of the kidney.

Distinct abnormalities detected by DWI, along with gradually increasing levels of albuminuria in T2DM patients at the MAU stage, as manifested by a corresponding downward trend in ADC values of the renal parenchyma, were also revealed in the present study. Thus, ADC values of renal functional impairments display incomparable advantages over albuminuria quantitative methods, blood tests of kidney function and conventional magnetic resonance imaging [ 12, 13]. ADC values can detect kidney lesions, which cannot be easily detected by traditional methods, such as albuminuria quantitation, in T2DM patients at the NAU stage. Quantitative assessment of renal impairments can be performed based on ADC values of the renal parenchyma acquired in T2DM patients at the NAU or MAU stage so that earlier intervention and efficacy observation can be facilitated by effective and objective indicators.

In this study, DTI detection revealed a progressively increasing trend in FA values of the renal medulla from the control group to the NAU and MAU groups; significantly higher FA values in the renal medulla of the MAU group than those in the control group were observed. This finding indicates that impairment of the renal medulla, besides functional impairment of the renal parenchyma, also occurs during renal injury development in early DN. This finding slightly differs from that of Lu et al. [ 14], who stratified renal function according to estimates of glomerular filtration rate in patients with renal impairment. This group proposed that FA values of renal medulla should be important in judging kidney injury because of the positive correlation of changes in FA values of the renal medulla with decreases in glomerular filtration rate during disease progression in DN patients. The present study, however, stratified albuminuria quantitative results to observe changes in FA values of the renal medulla in patients with normal renal function. Given that the biochemical characteristics and structure of kidney tissue can basically affect the diffusing capacity of water, the pathological structure of tissue fibers affects the diffusibility and anisotropic diffusion of water. In early DN, although swelling of renal tubular cells, vacuolar degeneration of epithelial cells, increased thickness of basement membrane, and tubulointerstitial lesions mainly manifested by interstitial fibrosis occur, the structure of most renal tubules remains intact. Changes in the microstructure of the tubular wall described above may strengthen their anisotropy. By contrast, in the middle and late stages of the disease, FA values of the renal medulla may exhibit a downward trend when tubular atrophy, thickening of basement membrane, lumen expansion, and even tubular necrosis develop. Findings by Lu et al. [ 14] and our group suggest that, at early stages of renal impairment (e.g., stage III and earlier), elevated FA values of the renal medulla reflect, to a certain extent, incessant compensatory state of renal impairment, whereas a progressive decline in FA values occurs with advancing development of impairment (e.g., abnormal rates of creatinine and glomerular filtration).

In summary, T2DM patients at the NAU stage generally show significantly lower average ADC values of the renal parenchyma than normal cases despite presenting normal kidney contours, dimensions, and anatomical structures. This finding indicates that DWI scanning may be helpful in early detection and quantitative assessment of renal lesions in T2DM patients, given its higher sensitivity to early kidney impairment in T2DM than albuminuria detection. Moreover, ADC values showed a downward trend while FA values of the renal medulla exhibited an upward trend as levels of albuminuria increased, which indicates that combined evaluation of ADC and FA values may provide an excellent quantitative approach for identifying DN at early disease stages.

Finally, we acknowledge that the limitations of this study, which include the small sample size, may prohibit drawing of solid conclusions. Nevertheless, the present study provides new non-invasive tools for the diagnosis and evaluation of renal injury in early DN. As such, the findings are valuable for future studies.

References

[1] Mogensen CE, Keane WF, Bennett PH, Jerums G, Parving HH, Passa P, Steffes MW, Striker GE, Viberti GC. Prevention of diabetic renal disease with special reference to microalbuminuria. Lancet1995; 346(8982): 1080-1084

[2] Al-Agha AE, Ocheltree A, Hakeem A. Occurrence of microalbuminuria among children and adolescents with insulin-dependent diabetes mellitus. Saudi J Kidney Dis Transpl2013; 24(6): 1180-1188

[3] Weir MR. Microalbuminuria in type2 diabetics: an important, overlooked cardiovascular risk factor. Clin Hypertens (Greenwich)2004; 6(3): 142-143

[4] Lane JT. Microalbuminuria as a marker of cardiovascular and renal risk in type 2 diabetes mellitus: a temporal perspective. Am J Physiol Renal Physiol2004; 286(3): F442-F450

[5] Chen XY. Prediction and multifactorial intervention for subclinical arteriosclerosis in patients with newly diagnosed type 2 diabetes mellitus. Central South China University. 2004. (in Chinese)

[6] Cheng SQ. NIU MC. Value of microalbuminuria in early diagnosis of diabetic renopathy. Chin J Lab Med (Zhonghua Jian Yan Yi Xue Za Zhi)2005; 28(7): 740-741 (in Chinese)

[7] Yang W, Lu J, Weng J, Jia W, Ji L, Xiao J, Shan Z, Liu J, Tian H, Ji Q, Zhu D, Ge J, Lin L, Chen L, Guo X, Zhao Z, Li Q, Zhou Z, Shan G, He J; China National Diabetes and Metabolic Disorders Study Group.. Prevalence of diabetes among men and women in China. N Engl J Med2010; 362(12): 1090-1101

[8] Ries M, Basseau F, Tyndal B, Jones R, Deminière C, Catargi B, Combe C, Moonen CWT, Grenier N. Renal diffusion and BOLD MR I in experimental diabetic nephropathy. Magn Reson Imaging2003; 17(1): 104-113

[9] Atkins RC, Polkinghorne KR, Briganti EM, Shaw JE, Zimmet PZ, Chadban SJ. Prevalence of albuminuria in Australia: the Ausdiab kidney study. Kidney Int Suppl2004; 66(92): S22-S24

[10] Goyal A, Sharma R, Bhalla AS, Gamanagatti S, Seth A. Diffusion-weighted MRI in assessment of renal dysfunction. Indian J Radiol Imaging2012; 22(3): 155-159

[11] Ries M, Basseau F, Tyndal B, Jones R, Deminière C, Catargi B, Combe C, Moonen CW, Grenier N. Renal diffusion and BOLD MRI in experimental diabetic nephropathy. Blood oxygen level-dependent. J Magn Reson Imaging2003; 17(1): 104-113

[12] Thoeny HC, De Keyzer F, Oyen RH, Peeters RR. Diffusion-weighted MR imaging of kidneys in healthy volunteers and patients with parenchymal diseases: initial experience. Radiology2005; 235(3): 911-917

[13] Michaely HJ, Schoenberg SO, Ittrich C, Dikow R, Bock M, Guenther M. Renal disease: value of functional magnetic resonance imaging with flow and perfusion measurements. Invest Radiol2004; 39(11): 698-705

[14] Lu L, Sedor JR, Gulani V, Schelling JR, O’Brien A, Flask CA, MacRae Dell K. Use of diffusion tensor MRI to identify early changes in diabetic nephropathy. Am J Nephrol2011; 34(5): 476-482

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg