Objectives—The aim of this study was to identify the main influencing factor of the shear wave velocity (SWV) of the kidneys measured by acoustic radiation force impulse elastography.
Methods—The SWV was measured in the kidneys of 14 healthy volunteers and 319 patients with chronic kidney disease. The estimated glomerular filtration rate was calculated by the serum creatinine concentration and age. As an indicator of arteriosclerosis of large vessels, the brachial-ankle pulse wave velocity was measured in 183 patients.
Results—Compared to the degree of interobserver and intraobserver deviation, a large variance of SWV values was observed in the kidneys of the patients with chronic kidney disease. Shear wave velocity values in the right and left kidneys of each patient correlated well, with high correlation coefficients (r = 0.580–0.732). The SWV decreased concurrently with a decline in the estimated glomerular filtration rate. A low SWV was obtained in patients with a high brachial-ankle pulse wave velocity. Despite progression of renal fibrosis in the advanced stages of chronic kidney disease, these results were in contrast to findings for chronic liver disease, in which progression of hepatic fibrosis results in an increase in the SWV. Considering that a high brachial-ankle pulse wave velocity represents the progression of arteriosclerosis in the large vessels, the reduction of elasticity succeeding diminution of blood flow was suspected to be the main influencing factor of the SWV in the kidneys.
Conclusions—This study indicates that diminution of blood flow may affect SWV values in the kidneys more than the progression of tissue fibrosis. Future studies for reducing data variance are needed for effective use of acoustic radiation force impulse elastography in patients with chronic kidney disease.
Key Words—acoustic radiation force impulse; brachial-ankle pulse wave velocity; chronic kidney disease; genitourinary ultrasound; renal blood flow; shear wave velocity.
Quantification of Kidney Stiffness by ARFI Elastography
The SWV was measured with an Acuson S2000 ultrasound system (Siemens Medical Solutions) using a 3.5-MHz convex probe. Kidney images were obtained in the prone position so that the longitudinal section was visible on the monitor. An ROI of 10 × 5 mm was set adjacent to the inferior pole of the cortex in the dorsal area of the renal parenchyma to exclude vessels depicted by the color Doppler mode. Placement of the ROI was accurate in almost all participants. To prevent respiratory motion the SWV was measured on inhalation breath holding 5 to 6 times consecutively by a sonographer. The mean SWV values were calcul ated in the right and left kidneys, respectively. While being blinded to the clinical data, 2 experienced sonographers (J.T. and Y.T.) performed ARF elastography.
In ARFI elastography for chronic liver disease, the SWV increases in more advanced stages because progressing interstitial fibrosis predominantly affects tissue elasticity, as observed in liver cirrhosis.
However, the main affecting factor of ARFI elastography in the kidneys has not been elucidated presumably for two reasons. Namely, a large variance of SWV values in the kidneys, as demonstrated by Goertz et al, has yielded results with low reliability, and the degree of interstitial fibrosis in the kidneys of patients with chronic kidney disease is not as marked as that in chronic liver disease. Since approximately 20% of cardiac output flows into the kidneys, which constitute less than 1% of body mass, we suspected that renal blood flow might be the main influencing factor of the SWV in the kidneys instead of interstitial fibrosis.
In the feasibility study, interobserver and intraobserver deviation was proven to be small when the SWV was measured in the kidney of a healthy volunteer. Although no significant correlation was obtained between SWV and estimated GFR values in 14 healthy volunteers in the first trial, a significant correlation was found between the SWV and estimated GFR when the ROI setting and measurement timing during arterial pulsation were reviewed. Interestingly, the SWV decreased concurrently with a decline in the estimated GFR in all cohorts of patients with chronic kidney disease despite the large variance noted. This finding means that kidney tissue stiffness decreases at advanced stages of chronic kidney disease despite the increasing prominence of interstitial fibrosis. Low SWV values were obtained in patients with a high brachial-ankle PWV. Considering that the brachial-ankle PWV represents arteriosclerosis of large vessels, we hypothesized that diminution of renal blood flow succeeding atherosclerosis of renal arteries may cause decreased elasticity of renal parenchyma in advanced chronic kidney disease. We also assumed that the large variance of SWV values in the kidneys of patients with chronic kidney disease derived from the structural heterogeneity of renal parenchyma and pressure fluctuation resulting from pulsating blood flow, rather than technical variance in measurement.
Renal parenchyma is grossly divided into the cortex and medulla. The cortex consists of proximal and distal tubules and renal glomeruli. The medulla mostly consists of the loop of Henle and the lower part of the collecting tubule. To meet the large oxygen consumption for massive reabsorption, the renal tubules are surrounded by a dense vascular plexus in both the cortex and medulla. In kidneys with low estimated GFRs, the number of glomeruli with global sclerotic changes increases. The renal tubule, located downstream of the sclerosing glomerulus, becomes atrophic, and peritubular fibrosis subsequently progresses.
Blood flow in the peritubular vascular plexus decreases according to sclerotic changes of the glomeruli, since blood flows from the glomeruli to the vascular plexus. Considering the remarkable damage of microcirculation in advanced chronic kidney disease, it is conceivable that blood flow rather than interstitial fibrosis dominantly affects the elasticity of kidney tissue in chronic kidney disease. In addition, it is widely known that the incidence of cardiovascular events increases concurrently with a decline in the estimated GFR. As shown in Figure 6, patients with chronic kidney disease who had a high brachial-ankle PWV tended to have a low SWV in the kidney. From these results, we strongly suspected that a combination of microcirculatory damage in the renal tissue and arteriosclerosis of renal arteries diminished renal blood flow and reduced kidney stiffness in patients with chronic kidney disease.
This study had several limitations. Since our hypothesis was based on hemodynamic changes in the kidneys, parameters directly related to renal blood flow and renal vessel resistance should be demonstrated. We measured the peak systolic velocity (Vmax) and end-diastolic velocity (Vmin) by using Doppler sonography and calculated the resistive index by the following equation: resistive index =(Vmax– Vmin)/Vmax. However, the data fit between the estimated GFR and Vmax and the estimated GFR and resistive index was poor (data not shown). For the ARFI measurement, ROI setting was frequently difficult when measuring the SWV in the kidneys of patients with advanced chronic kidney disease. In patients with high estimated GFRs, the SWV could be measured at the cortex because the thickness of the renal parenchyma was still sufficient.
However, for patients with advanced chronic kidney disease, the renal parenchyma was atrophic, and distinction between the cortex and medulla was often difficult.
Renal atrophy prevented accurate SWV measurement in 4 patients and Vmax and Vmin measurement in 16 patients. Several potential future studies are proposed. Instead of the brachial-ankle PWV, the cardio-ankle vascular index could be used as a novel indicator of arterial stiffness. Since the cardio-ankle vascular index is less affected by systemic blood pressure, it can be used as an alternative method to evaluate large-vessel arteriosclerosis. Shear wave velocity values can be normalized by systemic blood pressure and synchronization with electrocardiography, which ensures consistent aortic pressure during the cardiac cycle. Early-stage SWV assessment of the kidneys in diabetic patients may be feasible because of the hyperdynamic blood flow of the diabetic kidney in the early stages. If our hypothesis is correct, SWV values should be high in those kidneys. A comparison of ARFI assessment with surrogate markers of tissue fibrosis could be performed.
In conclusion, our study suggests that the SWV measured by ARFI elastography in patients with chronic kidney disease may represent the diminution of blood flow that succeeds arteriosclerosis, as opposed to the development of renal fibrosis. In our results, the standard deviations of SWV values were considerably high. However, further improvement of this method may result in obtainment of more consistent SWV values in the kidneys.