Novel non-invasive and quantitative assessment of the renal function of transplanted kidneys using Doppler ultrasonography with the vascular index of superb microvascular imaging
Article information
Abstract
Purpose
This study assessed the reproducibility and clinical value of the vascular index (VI), derived from superb microvascular imaging (SMI) using Doppler ultrasonography, for evaluating renal function in transplanted kidneys.
Methods
This retrospective study included 63 renal transplant patients who underwent grayscale and Doppler ultrasonography with SMI from January 2022 to February 2023. The VI of the transplanted kidneys was measured using three methods (VIbox, VIF1, VIF2). The VI was compared across chronic kidney disease (CKD) groups categorized by estimated glomerular filtration rate (eGFR) and Kidney Disease: Improving Global Outcomes (KDIGO) CKD risk groups based on eGFR and albuminuria. The correlation between VI and renal function was evaluated. Univariate and multivariate linear regression analyses were used to identify predictors of eGFR.
Results
Significant differences in VI were observed among CKD groups based on eGFR (VIbox, P=0.001; VIF1, P<0.001; VIF2, P<0.001) and KDIGO CKD groups based on eGFR and albuminuria (VIbox, P=0.039; VIF1, P=0.001; VIF2, P<0.001). VIF1 and VIF2 demonstrated moderate/high correlations with eGFR (r=0.627, P<0.001 and r=0.657, P<0.001, respectively) and serum creatinine (r=-0.626, P<0.001 and r=-0.649, P<0.001, respectively). VIbox moderately correlated with eGFR (r=0.445, P<0.001). Multivariate regression identified the urine albumincreatinine ratio (ACR) (adjusted odds ratio [aOR], 1.122; 95% confidence interval [CI], -0.007 to, 0.000; P=0.030) and VIF2 (aOR, 1.114; 95% CI, 0.466 to 1.235; P<0.001) were independently associated with eGFR.
Conclusion
The VI measured by drawing a region of interest along the border of the transplanted kidney in SMI (VIF2) is highly reproducible and correlates well with eGFR. Both VIF2 and ACR independently predict eGFR.
Introduction
As the prevalence of chronic kidney disease (CKD) increases globally, the number of kidney transplantations has risen dramatically over the years [1,2]. In 2021, the total number of kidney transplantations in the United States reached 25,487 [2]. Although kidney transplantation is the most effective treatment for end-stage renal disease, the 5-year graft loss rate is reported to be approximately 30%, and nearly 50% of patients lose their graft within 10 years, despite advances in immunosuppression [3].
One of the key factors in the development of CKD in the kidney allograft is microvascular rarefaction [4]. The loss of peritubular capillaries is associated with lower long-term estimated glomerular filtration rate (eGFR) in the transplanted kidney [4]. Steegh et al. [5] reported that early peritubular capillary loss during the first 3 months after transplantation predicts lower renal function at 1 year. Although renal biopsy is the gold standard for assessing graft status, including microvasculature, percutaneous core needle biopsy is invasive and carries a risk of bleeding [6,7].
Ultrasonography (US) is a completely non-invasive method for the gross evaluation of the renal allograft [8]. Doppler US can depict the vascularity of the renal allograft and assess its vascular complications [8,9]. Conventional color Doppler US is effective for evaluating medium to large vessels; however, its ability to assess small and low-flow vessels is limited [10]. Various advanced techniques, including power Doppler imaging (PDI) and contrastenhanced US (CEUS), have been explored for assessing renal microvasculature [10-14]. However, PDI is vulnerable to respiratory motion and probe movement due to its low frame rate [10], and while CEUS is relatively safe for evaluating renal vascularity, it is limited by longer examination times and high cost.
Recently, superb microvascular imaging (SMI) has been introduced as a novel technique that visualizes microvasculature and lowvelocity flow using an adaptive algorithm [15-17]. SMI offers advantages over conventional color Doppler and PDI, providing high resolution, high frame rate, increased sensitivity, and fewer motion artifacts [15-17]. However, data regarding renal allograft evaluation using SMI remain scarce, and few studies have employed SMI for renal allografts [14,18]. Moreover, an optimal quantification method for renal allograft vascularity using SMI has not yet been established. Therefore, this study aimed to investigate the reproducibility and value of the vascular index (VI) derived from SMI in Doppler US for assessing renal function in the transplanted kidney.
Materials and Methods
Compliance with Ethical Standards
This single-center retrospective study was approved by the authors’ affiliated Institutional Review Board (IRB No. 2023-05-010), and the requirement for informed consent was waived due to its retrospective design.
Study Design and Patients
From January 2022 to February 2023, 84 consecutive patients with renal transplantation underwent renal US with Doppler imaging, including the SMI technique, for clinical purposes. Patients were excluded if the SMI was obtained at a low frame rate, if raw SMI data for VI measurement were unavailable, if marked flash artifacts (defined as a definite flash artifact in perirenal fat adjacent to the renal cortex or parenchyma) were present, or if the imaging did not include the entire longitudinal diameter of the transplanted kidney.
Renal US Examination
All ultrasound examinations of the transplanted kidney were performed using one of four identical US machines (Aplio i800 System, Canon Medical Systems Corporation, Tokyo, Japan) with a multi-frequency convex transducer (i8CX1) operating at a center frequency of 4 MHz. Patients were scanned in the supine position. A board-certified radiologist with 10 years of experience in renal US and 5 years of experience in SMI conducted all examinations while remaining blinded to the patients’ clinical information, including serum creatinine, eGFR, and albuminuria. The US examination included grayscale imaging, a color Doppler study, a spectral Doppler study with measurement of the resistive index (RI), and SMI.
On grayscale imaging, the kidney size, the distance from the skin to the transplanted kidney, renal parenchymal echogenicity, and the presence of hydronephrosis or focal lesions were evaluated. The size of the transplanted kidney was measured in three planes: the transverse and anteroposterior lengths on the transverse scan and the longitudinal length on the longitudinal scan. For RI measurement, both color and spectral Doppler modes were used. RI values were measured at least three times in the upper, mid, and lower poles at the level of the interlobar or arcuate arteries, and the average value was recorded.
SMI was obtained using the following scan parameters: time gain of 40; color velocity scale of 2.3 cm/s; frame rate of 27 fps; and color frequency of 3.0 MHz. From the stored SMI still image of the transplanted kidney, one radiologist retrospectively drew a region of interest (ROI) to measure the VI, defined as the percentage ratio between the Doppler signal pixels and the grayscale pixels within the ROI. Three methods were used to measure the VI of the transplanted kidney. VIbox was defined as the average of three vascular indices measured from 1.5×0.5-cm box ROIs placed in the upper, mid, and lower pole cortices. VIF1 was defined as the VI obtained from a freehand ROI manually drawn along the outer margin of the transplanted kidney. VIF2 was defined as the VI derived from a freehand ROI manually drawn along the outer margin of the transplanted kidney, excluding the renal sinus fat (Fig. 1).
Three methods of measuring the vascular index (VI) of the transplanted kidney.
VIbox is defined as the average of three VIbox values measured in the upper (A), middle (B), and lower (C) pole of the transplanted kidney. VIF1 is defined as the VI derived from a region of interest (ROI) drawn along the outer margin of the transplanted kidney (D). VIF2 is defined as the VI derived from an ROI drawn along the outer margin of the transplanted kidney, excluding renal sinus fat (E). SMI, superb microvascular imaging.
To assess interobserver agreement, the VI of 30 patients was measured independently by a second radiologist (with 6 years of experience in renal US and 1 year of experience in SMI) using all three methods (VIbox, VIF1, and VIF2).
Data and Statistical Analysis
Patients were categorized into three groups based on eGFR (G1, eGFR >60 mL/min/1.73m2; G2, eGFR 45-60 mL/min/1.73m2; G3, eGFR <45 mL/min/1.73m2) and into three groups according to the Kidney Disease: Improving Global Outcomes (KDIGO) CKD risk classification based on eGFR and albuminuria (low, moderate, and high risk) [19]. In this study, the "high risk" and "very high risk" categories were combined into a single high-risk group (Fig. 2). Clinical data—including age, sex, kidney donor type, transplantation vintage, etiology of CKD, and renal function test results (serum creatinine, eGFR, and urine albumin-creatinine ratio [ACR])—were collected from electronic medical records. Laboratory results were obtained from tests performed within one week before or after the ultrasound examination.
Kidney Disease: Improving Global Outcomes (KDIGO) risk groups.
CKD, chronic kidney disease; GFR, glomerular filtration rate; eGFR, estimated glomerular filtration rate.
VI values measured by the three methods were compared among CKD groups based on eGFR and among KDIGO risk groups using one-way analysis of variance (ANOVA) with Dunnett’s post hoc test for pairwise comparisons. Comparisons of clinical data among CKD groups were performed using one-way ANOVA or the Kruskal-Wallis test, as appropriate, with the Shapiro-Wilk test assessing normality for continuous variables. Pearson’s correlation analysis was used to evaluate the relationship between VI and renal function, as indicated by serum creatinine or eGFR. The intraclass correlation coefficient (ICC) was calculated to assess interobserver agreement for the three VI measurement methods [20]. Reliability was classified as excellent (ICC >0.75), fair to good (ICC=0.40-0.75), or poor (ICC ≤0.40) [21]. Univariate and multivariate linear regression analyses were used to identify factors predictive of eGFR. Statistical analyses were performed using Rex 3.1.2 (https://rexsoft.org/), with statistical significance defined as P<0.05.
Results
Patients
Of the 84 eligible patients, 21 were excluded for the following reasons: SMI obtained with a low frame rate (n=9), unavailable raw SMI data for VI measurement (n=6), a marked flash artifact on SMI (n=4), and imaging that did not include the entire longitudinal diameter of the transplanted kidney (n=2). Ultimately, 63 patients (36 men and 27 women) who underwent renal US with Doppler imaging including SMI were included (Fig. 3). The mean age was 56.8 years (standard deviation, 9.62; range, 38 to 77 years). Patient demographics are detailed in Table 1. Among the KDIGO risk groups, the low-risk group contained a significantly higher proportion of women (P=0.019). No other significant differences in patient characteristics were observed among the eGFR groups or KDIGO risk groups.
Flowchart of the study.
SMI, superb microvascular imaging; USG, ultrasonography.
Patient characteristics by eGFR groups and KDIGO CKD risk groups
Comparison of VIs Measured by Three Methods According to Renal Function
Statistically significant differences in VI were observed among CKD groups based on eGFR (VIbox, P=0.001; VIF1, P<0.001; VIF2, P<0.001) and KDIGO CKD risk groups based on eGFR and albuminuria (VIbox, P=0.039; VIF1, P=0.001; and VIF2, P<0.001) (Table 2). VIBox and VIF1 demonstrated a statistically significant difference between the eGFR group 1 (G1) and group 3 (G3) (VIBox, P=0.005 and VIF1, P<0.001) and between group 2 (G2) and G3 (VIbox, P=0.020 and VIF1, P=0.000). However, no statistically significant difference was found between G1 and G2 (VIbox, P=0.543 and VIF1, P=0.163). VIF2 showed statistically significant differences in all between-eGFR group comparisons (G1 vs. G2, P=0.016; G1 vs. G3, P<0.001; and G2 vs. G3, P=0.005, respectively) (Fig. 4). All three VI measurement methods revealed statistically significant differences between KDIGO CKD groups 1 (low-risk group) and 2 (moderate-risk group) and between groups 1 and 3 (high-risk group), but with no statistically significant difference between groups 2 and 3 (Fig. 5).
Comparison of VI measured by three methods according to CKD groups
Comparison of the vascular index (VI) values measured using three methods in groups based on the estimated glomerular filtration rate (eGFR).
VIF2 shows a statistically significant difference in all between-group comparisons. VIbox and VIF1 exhibit statistically significant differences in G1 vs. G3 and G2 vs. G3, and with no significant difference between G1 vs. G2. *P<0.05.
Comparison of vascular index (VI) values measured by three methods in the Kidney Disease: Improving Global Outcomes (KDIGO) chronic kidney disease (CKD) groups.
The VI measured by all three methods showed statistically significant differences in G1 vs. G2 and G1 vs. G3. *P<0.05.
Correlations between VI of the Transplanted Kidney, eGFR, and Serum Creatinine
Among the three methods of measuring the VI of the transplanted kidney, VIF1 and VIF2 showed moderately high correlations with eGFR (r=0.627, P<0.001 and r=0.657, P<0.001, respectively) and serum creatinine (r=-0.626, P<0.001 and r=-0.649, P<0.001, respectively). VIBox exhibited a moderate correlation with eGFR (r=0.445, P<0.001) and serum creatinine (r=-0.447, P<0.001) (Fig. 6).
Correlation between the vascular index (VI) of the transplanted kidney, estimated glomerular filtration rate (eGFR), and serum creatinine (Cr).
A strong correlation was defined as a correlation coefficient of > 0.5; moderate correlation, correlation coefficient ≥0.4 to ≤0.75; and poor correlation, a correlation coefficient of <0.4.
Clinical and Doppler US Factors Associated with eGFR
Univariate linear regression analysis revealed that the urine ACR (P=0.003) and VIF2 (P<0.001) were associated with eGFR. Other parameters (age, sex, post–kidney transplantation period, and donor type) exhibited no associations with eGFR.
Multivariate linear regression analysis showed that the ACR (adjusted odds ratio [aOR], 1.122; 95% confidence interval [CI], -0.0007 to -0.000; P=0.030), and VIF2 (aOR, 1.114; 95% CI, 0.466 to 1.235; P<0.001) was independently associated with eGFR (Table 3).
Clinical and Doppler ultrasound factors associated with eGFR
Inerobserver Agreement of the Three VI Measurement Methods
The interobserver agreement of VI measurements was excellent between the two radiologists for all three methods (ICC: VIBox, 0.953; VIF1, 0.988; VIF2, 0.932, respectively) (Table 4).
Inter-observer agreement of the VI measurement methods
Discussion
This study demonstrated that the VI measured by SMI is significantly lower in patients with reduced eGFR and that both ACR and VIF2 are independently associated with eGFR. Moreover, the interobserver agreement was excellent for all three VI measurement methods. These findings suggest that the VI derived from SMI can serve as a reliable indicator of renal function in transplanted kidneys, with VIF2 emerging as the most robust measurement method. In contrast, VIbox and VIF1 were not independently associated with eGFR. The VIbox method may be limited by its evaluation of only a small region of the transplanted kidney, whereas the VIF1 method may be confounded by the inclusion of renal sinus fat and large vessels within the ROI.
Assessing the cortical microvasculature is crucial for predicting microvascular rarefaction in kidney allografts. Although renal biopsy remains the most definitive method for analyzing the kidney microvasculature, it is an invasive procedure. A previous study reported that quantifying relative renal blood volume using enhanced computed tomography (CT) correlated with capillary rarefaction [22]; however, CT exposes patients to ionizing radiation, and iodine contrast may induce nephrotoxicity.
SMI is a state-of-the-art Doppler technique for detecting slow flow and small vessels using a novel clutter suppression algorithm, which separates flow signals from tissue motion artifacts with a high frame rate [15,23,24]. The kidney is an ideal organ for evaluation with Doppler US to obtain functional and vascular information [25]. A recent study demonstrated that SMI is more sensitive than color Doppler or power Doppler US in depicting cortical microvasculature in native kidneys [24].
However, the utility of SMI for evaluating both native and transplanted kidneys remains uncertain. Few studies have investigated the use of SMI in transplanted kidneys [14,18]. Kim et al. [14] reported that VI measured by the VIbox method on SMI was not significantly different between rejection and non-rejection groups, which contrasts with the findings of the present study; however, their study relied solely on the VIbox method, which may have limited the evaluation of overall kidney vascularity. In contrast, Gurbuz et al. [18] demonstrated that the distance between the kidney capsule and SMI-detected vessels could predict the chronic allograft damage index.
This study found that VIF2 had a strong correlation with eGFR and was independently associated with eGFR. The authors believe that the VIF2 method minimized selection bias by capturing the ratio of color to grayscale pixels across the entire renal parenchyma. However, unlike the eGFR groups, VIF2 did not reflect a stepwise difference across KDIGO risk groups. This discrepancy may stem from the complexity of the KDIGO grading system, which incorporates both eGFR and albuminuria. Since albuminuria reflects both glomerular and tubular damage, it may account for the observed differences between KDIGO risk groups and VI [26]. To the authors’ knowledge, this study is the first to use the entire renal parenchyma of an image for acquiring the VI of the transplanted kidney. The excellent interobserver agreement supports that VI measurement on SMI is highly reproducible and suitable as a quantitative marker for renal function evaluation.
This study has several limitations. First, its retrospective design and relatively small sample size may introduce selection bias. Second, renal biopsy results were not analyzed because not all kidney transplant patients undergo allograft biopsy regardless of renal function, and biopsy is an invasive procedure with bleeding risk. Third, only single-vendor US machines were used, and the quantitative measurement of microvascular imaging (VI) is a unique feature of these machines. Fourth, the study provided only an indirect assessment of allograft vascularity using a two-dimensional VI from SMI, which does not quantify blood flow velocity because it is based on a pixel-counting method within the ROI. Fifth, the VIF2 method did not exclusively capture small vessels but also included some larger vessels within the ROI. Owing to the nature of manual ROI delineation, isolating only small vessels is challenging, and VIF2 was selected as a practical alternative with a clearly defined contour. Finally, renal function was assessed at a single time point, so whether VI can predict future allograft function remains unknown. Future longitudinal studies are warranted to evaluate the relationship between VI on SMI and the prognosis of transplanted kidneys.
In conclusion, the VI measured by drawing a freehand ROI along the border of the transplanted kidney in SMI (VIF2) is a highly reproducible method that correlates well with eGFR. Both VIF2 and ACR are independent predictors of eGFR. The VI derived from SMI in Doppler US provides a non-invasive, quantitative, and reproducible assessment of renal function in transplanted kidneys.
Notes
Author Contributions
Conceptualization: Bae SH, Lee EJ. Data acquisition: Bae SH, Lee EJ. Data analysis or interpretation: Lee EJ, Hwang J, Hong SS, Chang YW, Nam B. Drafting of the manuscript: Bae SH. Critical revision of the manuscript: Lee EJ, Hwang J, Hong SS, Chang YW, Nam B. Approval of the final version of the manuscript: all authors.
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
This work was supported by the Soonchunhyang University Research Fund. The concepts and information presented in this paper are based on research results that are not commercially available.
References
Article information Continued
Notes
Key point
The vascular index measured on superb microvascular imaging with a freehand region of interest along the outer margin of the transplanted kidney, excluding renal sinus fat, correlates well with estimated glomerular filtration rate (eGFR) and is an independent predictor of eGFR. Vascular index measurement on superb microvascular imaging is a non-invasive and quantitative method for assessing renal function in the transplanted kidney.