Evaluation of split renal function using MR-derived structure and arterial spin-labeled perfusion features in obstructive nephropathy

doi:10.21203/rs.3.rs-4670612/v1

Download PDF

Research Article

Evaluation of split renal function using MR-derived structure and arterial spin-labeled perfusion features in obstructive nephropathy

https://doi.org/10.21203/rs.3.rs-4670612/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objectives To explore the value of renal arterial spin-labeled (ASL) and structural MRI in evaluating split renal function (SRF) by renal scintigraphy (RS) in obstructive nephropathy.

Materials and methods Consecutive patients with obstructive hydronephrosis were recruited for a 3.0-T conventional and ASL examination from January 2022 to August 2023 prospectively. The 92 independent kidneys were classified into normal renal function, mild to moderate renal impairment, severe renal impairment, or no function groups. Renal blood flow (RBF), renal parenchymal thickness (RPT), and renal pelvic dilatation (RPD) were assessed based on ASL and T1WI maps. One-way analysis of variance (ANOVA) and Kruskal-Walli’s test were used to analyze the differences. Pearson and Spearman rank correlation analysis approaches were adopted to analyze the correlation between RBF, RPT, RPD, and glomerular filtration rate (GFR) by RS. The ROC curve was used to compare their performance in SRF evaluation.

Results Renal cortical RBF, RPT, and RPD could distinguish SRF pairwise subgroups with different renal function impairments (p<0.001), with correlation coefficients of 0.767, 0.635, and -0.589, respectively (all p<0.001). RBF combined with RPT and RPD achieved the best efficacy in distinguishing between normal and mild, mild and severe, normal and severe renal dysfunction, with AUCs of 0.869 (95% CI, (0.768-0.937)), 0.883 (95% CI, (0.756-0.959)) and 0. 994 (95% CI, (0.931-1.00)).

Conclusions MR structural features and ASL imaging enable quantitative measurements of RPT, RPD, and renal cortical RBF, which were correlated with RS-GFR. MR combination demonstrated clinical potential in noninvasive evaluation of SFR in obstructive nephropathy.

Urinary Tract Obstructions

Arterial Spin-Labeled Imaging

Renal Parenchymal Thicknesses

Magnetic Resonance Imaging

Obstructive nephropathy is an important cause of chronic kidney disease (CKD) characterized by renal dysfunction and renal interstitial fibrosis. It is a global health issue affecting a wide range of people. Obstructive nephropathy is caused by chronic urinary tract obstructions (UTOs). Renal calculi are the most common cause in young adults affecting ~ 10% of adults worldwide[1]. And benign prostatic hyperplasia, prostate cancer, and bladder tumors are common causes in older adults. To date, surgery remains the primary intervention for UTOs. Split renal function testing is a sensitive indicator for early detection of obstructive nephropathy. It is an important indicator for surgical time selection, prognosis evaluation, and the determination of kidney preservation[2]. The split renal function (SRF) assessment is of great significance in UTOs.

Split glomerular filtration rate (GFR) obtained through renal scintigraphy (RS) is the standard for assessing SRF currently [3]. Nevertheless, the radiotracers such as glomerular (99mTc-DTPA) or tubular (99mTc-MAG3) limit the clinical applications due to radiation exposure, operator dependence, and prolonged examination time, et al. [3, 4]. Even though the MR studies on renal interstitial fibrosis and estimation of glomerular filtration rate (eGFR) based on serum creatinine values have been recognized, the quantitative evaluation of SRF remains challenging. Unilateral ureteral obstruction (UUO) is the most important animal model for identifying novel mechanisms underlying renal fibrosis that date back centuries, however, the inability to accurately detect renal function is an important limitation. The research results of MRI based on UUO models have encountered bottlenecks in clinical translation, which leaves a gap of knowledge in obstructive nephropathy in humans.

Renal structural features, including parenchymal thickness (RPT), renal pelvis dilatation (RPD), and renal volume were indicators of split renal dysfunction. However, the morphological changes occur late with low sensitivity in the assessment of renal function. Arterial spin labeling (ASL) imaging uses water molecules as freely diffusing tracers to achieve quantitative measurement of blood flow perfusion without contrast agents[4, 5]. The value of ASL in evaluating cerebral blood flow perfusion has been recognized, and it has good reproducibility in RBF[6, 7]. The renal perfusion characteristics of obstructive nephropathy and its value in evaluating renal dysfunction are worth exploring.

This study aimed to explore the feasibility of RBF, RPT, and RPD in the assessment of renal dysfunction in UTOs, and to investigate their value in the quantitative evaluation of RS-SRF.

Human subjects

This prospective study protocol was approved by the Institutional Review Board for Clinical Research of our hospital. Written informed consent was obtained from all subjects. Eighty-five patients with UTOs were recruited and underwent a 3-T conventional and ASL-MR examination. Exclusion criteria included poor image quality, missing clinical data, or comorbidities with other diseases. The split kidneys were divided into 3 groups based on split GFR obtained by RS. (1) normal renal function: GFR≥40ml/min, renal parenchymal radioactivity uptake and clearance is good; (2) mild to moderate impairment: 20ml/min≤ GFR＜40ml/min, a mild or moderate reduction in renal parenchymal radioactivity uptake; (3) severe impairment or no function: GFR＜20ml/min, significant reduction in renal parenchymal radioactivity uptake. The patient's demographic and clinical information were derived from medical records, including age, sex, and history of diabetes and hypertension.

MRI Technique and Image Analysis

MRI was performed using a 3.0 Tesla MR system (GE Healthcare, Discovery 750w, Chicago, IL) with a 16-channel phased-array torso Coil. Conventional plain MRI and three-dimensional (3D)-ASL were conducted. For 3D-ASL, the post-labeling delay time (PLD), 2025ms; repetition time (TR), 4377ms; echo time (TE), 10.7ms; slice thickness, 4.0mm; field of view (FOV), 260×260mm²; NEX,3; bandwidths, 62.5 kHz/pixel. Axial T1-weighted imaging, TR, 5.7ms; TE, 2.5ms; slice thickness, 2.4mm; FOV, 260×260mm²; NEX,1; bandwidths, 142.86 kHz/pixel; flip angle, 12.

The 3D-ASL data was performed using function software in the GE AW 4.7 workstation. They were anonymously analyzed independently by two trained radiologists (with 2 and 15 years of experience in abdominal radiology respectively) without knowledge of clinical information. ASL measurements were determined by manually drawing small circular or elliptical regions of interest (ROI) on the renal cortex with an area of 20-120 mm² in the upper, middle, and lower poles of each kidney, respectively. The average value was taken as the renal blood flow.

On T1WI, a straight line parallel to the renal vein was used as the long-axis reference line, and a short-axis line perpendicular to the long-axis reference line was made at the maximum transverse diameter. On the layer with a clear boundary between renal parenchyma and renal sinus, the distances from the outer edge of renal sinus fat to the renal peritoneum were measured. In the upper, middle, and lower poles of the split kidney, a total of nine path lines were drawn. The average of them was taken as RPT, d (unit: mm). Measure the maximum diameter (mm) of renal pelvic dilatation at the renal pelvis-ureteral junction [10], which was taken as the degree of RPD (Fig. 1a-1f).

Statistical methods

Continuous variables were summarized as the mean ± standard deviation (SD) or median (interquartile range) as appropriate. We calculated the intraclass correlation coefficients (ICC) to assess the interobserver agreements (0.81-1.00, excellent agreement). One-way analysis of variance (ANOVA) and Kruskal-Walli’s test was used to analyze the differences among the three groups. Independent samples t-test and Mann-Whitney U-test were used to analyze the differences between every two groups. ROC analysis was performed to compare the capacity of MRI-derived parameters. AUC=0.50-0.70 was considered to be low diagnostic power, AUC=0.70-0.90 means moderate diagnostic power, and AUC=0.90-1.0 means high diagnostic power. Pearson correlations (normal distribution) or Spearman correlations (non-normal distribution) were used to assess the association between numerical variables. |r|≥0.7 was considered a high correlation, 0.4≤ |r|<0.7 was a moderate correlation, and |r|<0.4 was a low correlation. Statistical analysis was performed using SPSS 25.0, MedCalc 20, and GraphPad Prism 9.5.1. p<0.05 was considered as a statistically significant difference.

Clinical characteristics

46 patients (92 split kidneys) with UTOs were included (M: F=23:23), including 39 patients with unilateral hydronephrosis and 7 patients with bilateral hydronephrosis. The patients’ age was 55±11.2 years. There were 43, 29, and 20 split kidneys in the normal, mildly to moderate renal impairment, and severely renal impairment or no-function groups, respectively.

Interobserver agreement

All MR parameters demonstrated excellent repeatability. The ICC values of renal cortical RBF, RPT, and RPD were 0.982, 0.906, and 0.994, respectively.

Comparisons of renal cortical RBF, RPT, RPD

There were significant differences in renal cortical RBF, and RPT, RPD among the three groups (p<0.001), see Table S1. Two-by-two comparisons between groups were also significantly different (Fig. 1g-1i).

Correlation of renal cortical RBF, RPT, RPD, and GFR

Cortical RBF and RPT showed a significant positive correlation with RS-GFR (p<0.001; r=0.767 and 0.635, respectively), and RPD showed a significant negative correlation with RS-GFR (p<0.001, r=-0.589). RBF, RPT, and RPD were all associated with RS-GFR to varying degrees in the intra-group comparison within the mild to moderate renal impairment and severe impairment or no function (Fig. 2, Table S1).

Receiver operating curve (ROC) analysis

The discriminatory efficacy of RBF, RPT, and RPD was statistically different in two-by-two comparisons among the three groups (p<0.05). RPT efficacy was best when discriminating between groups with normal renal function and mild to moderate impairment (AUC=0.785). Renal cortical RBF efficacy was best when discriminating between groups with mild-to-moderate impaired renal function and those with severely impaired or non-function (AUC=0.878). RBF combined with MRI-derived structural features (RPT and RPD) in the renal cortex further enhances the diagnostic efficacy when comparing two-by-two between all three groups (Fig. 3, Table S2).

In the present study, we have applied the structure features（RPT and RPD）and ASL to UTOs patients and explored the value of MRI in assessing split GFR by RS. Our results demonstrate that RPT, RPD, and ASL-RBF contributed to distinguishing different degrees of renal dysfunction, and have a high correlation with split renal GFR. MRI offered an attractive approach for evaluating SRF without a contrast agent.

Renal structural imaging helps in renal function assessment. Previous studies have shown RPT ranges from 12.8 to 20.8mm in healthy adults[8]. Previous CT studies[9] showed that PRT was about 18.2mm in obstructive nephropathy. Our study distinguished RPT in different degrees of renal dysfunction furtherly, which showed RPT was approximately 17.71±2.67mm, 14.19±3.57mm, and 9.87±4.85mm in normal renal function, mild to moderate dysfunction, and severe dysfunction or no-function（p＜0.05）. RPT decreased with split GFR measured by RS（r=0.635）. Compared with previous CT studies, the correlation between RPT and GFR is lower than renal volume[10, 11]. However, RPT did not demonstrate value in distinguishing the renal function impairment in lupus nephritis from healthy volunteers[12], which may be related to its different etiological mechanisms. RPD is another structural indicator to assess renal function and hydronephrosis, which is defined as the width measured in the renal pelvis ureteral transition zone on MR images. In our study, RPD is significantly negatively correlated with split RS-GFR (r=-0.589). A previous study about obstructive hydronephrosis of UUO mice demonstrated that the volume of the cortex and inner stripe of the outer medulla increased while the inner medulla decreased[13], causing the decline of its renal function. RPD and RPT are objectively quantified imaging parameters helping for preliminary screening of renal dysfunction.

A unique aspect of renal circulation is the presence of separate cortical and medullary circulations. All blood flow to the kidney passes through the glomeruli of the renal cortices[6, 14], which provides the driving pressure for glomerular filtration. DCE-MRI showed that hemodynamic abnormalities could distinguish obstructive and non-obstructive conditions as early as 1 day after ureteral obstruction in animal models, before morphological abnormality[14]. Studies based on both children and adults have shown that DCE and contrast-enhanced CT are more reliable methods for evaluating renal function, and can achieve renal function evaluation results similar to RS-GFR compared to eGFR[15-18]. However, the impact of contrast agents on renal function is an unavoidable issue, which limits the application of DCE and enhanced CT in renal function evaluation. Marica Cutajar et al.[18] reported that ASL had better reproducibility than DCE with no significant difference in the perfusion parameters. In our study, the examination of renal cortical RBF also achieved high repeatability (ICC=0.982), RBF decreased as renal function damage worsened (r=0.767). In obstructive renal dysfunction, the mechanical stress caused by the accompanying hydronephrosis compresses the renal parenchyma, resulting in the loss of renal units and the decrease of renal blood perfusion. Long-term ischemia and hypoxia in renal tissue led to renal tubular dilation and atrophy, decreased GFR, and renal interstitial fibrosis, ultimately leading to renal insufficiency and even renal failure[19-21]. ASL could reflect the potential pathological damage of CKD and distinguish early CKD from those with normal renal function[22]. Prasad et al.[22] reported that RBF in the renal cortex is 109. 54 ± 25. 38 ml / (100 g·min) correlating with eGFR in CKD. However, the studies above did not conducted research on the RBF characteristics of different renal dysfunctions in-depth. In our study, renal cortical RBF distinguished between severe renal dysfunction, mild to moderate dysfunction, and normal renal function, with RBF of 57.17±24.24, 100.03±29.01, and 131.21±31.06 ml/(100 g·min) separately. Guoqing Fan et al.[13] reported that Cry-micro-optical sectioning tomography (cryo-MOST) autofluorescence imaging showed vascular damage in UUO mouse kidney, including segmental artery and vein atrophy and arcuate vein dilation, and the density of function capillaries in the cortex was reduced by 37.2%. RBF could reflect renal hemodynamics and help in early identification and grading diagnosis of renal dysfunction in obstructive nephropathy.

In terms of renal function assessment in obstructive nephropathy, RPT, RPD, and RBF exhibited different efficacy. RPT obtained better efficacy in the differentiation between normal from mild to moderate impairment of independent kidney (AUC=0.785), which is a key parameter parallel to RS-GFR. All these were consistent with the conclusions obtained from previous studies on determining the grading of renal hydronephrosis[23]. Besides these, our study showed that RBF had better diagnostic efficacy in the differentiation between mild-moderate and severe renal dysfunction groups (AUC=0.878), and all of RPT, RPD, and RBF achieved good diagnostic efficacy in distinguishing normal and severe renal dysfunction (all AUC＞0.9). Jiule Ding et al. evaluated renal dysfunction in CKD using texture analysis based on DWI, BOLD, and susceptibility-weighted imaging, and the quantitative MR parameters above achieved good efficacy in distinguishing renal function by eGFR (AUC=0.821~0.840)[24]. RBF combined with RPT and RPD further improved the diagnostic efficacy in each two groups in UTOs (AUC=0.869, 0.883, 0.994), which achieved good diagnostic performance.

This study is deficient in the following aspects: (1) due to the effect of radiation from the radiotracers of RS, we took the non-obstructive side of the kidney in patients with urinary tract obstruction as a control based on the GFR obtained by RS, ignoring the compensatory performance of renal impairment; (2) ASL examination of kidney was affected by abdominal breathing, the stability of image quality needs further improvement; (3) the sample size of this study is small and needs to be validated through a larger sample size.

In summary, ASL and structural features (RPT and RPD) hold promise as useful biomarkers for assessing renal dysfunction in UTOs, which is correlated with RS-GFR. The combination of RBF with the structural features may improve the diagnostic efficacy of split renal impairment furtherly, and provide a basis for non-invasive assessment of SRF.

Author's contribution XCL: protocol/project development. YD: protocol/project development、manuscript writing/editing. YQC: data collection or management, data analysis, manuscript writing/editing. LW: MRI image acquisition. WHG: data collection or management. JZY: data collection or management.

Funding This study has received funding from the “1+X” program for Clinical Competency enhancement–Clinical Technical Level Improvement Project, The Second Hospital of Dalian Medical University (2022LCJSYS06), and applied basic research project of Liaoning Province (2023JH2/101300124).

Data availability All data can be made available for further study.

Conflict of interest None declared.

Ethical approval All procedures performed in the study were consistent with the institution's ethical standards and the 1964 Ethical Standards of Helsinki and its subsequent amendments or similar ethical standards. In addition, the study was also approved by the Ethics Committee of Dalian Medical University.

Informed consent Informed consent was obtained from all participants in this study.

Klahr, S., Obstructive nephropathy. Intern Med, 2000. 39(5): p. 355-361.http://dx.doi.org/10.2169/internalmedicine.39.355
Grenier, N., P. Merville, and C. Combe, Radiologic imaging of the renal parenchyma structure and function. Nat Rev Nephrol, 2016. 12(6): p. 348-359.http://dx.doi.org/10.1038/nrneph.2016.44
Durand, E., P. Chaumet-Riffaud, and N. Grenier, Functional renal imaging: new trends in radiology and nuclear medicine. Semin Nucl Med, 2011. 41(1): p. 61-72.http://dx.doi.org/10.1053/j.semnuclmed.2010.08.003
Piepsz, A. and H.R. Ham, Pediatric applications of renal nuclear medicine. Semin Nucl Med, 2006. 36(1): p. 16-35.http://dx.doi.org/10.1053/j.semnuclmed.2005.08.002
Kim, D.W., et al., Measurement of arterial transit time and renal blood flow using pseudocontinuous ASL MRI with multiple post-labeling delays: Feasibility, reproducibility, and variation. J Magn Reson Imaging, 2017. 46(3): p. 813-819.http://dx.doi.org/10.1002/jmri.25634
Odudu, A., et al., Arterial spin labelling MRI to measure renal perfusion: a systematic review and statement paper. Nephrol Dial Transplant, 2018. 33(suppl_2): p. ii15-ii21.http://dx.doi.org/10.1093/ndt/gfy180
Zhu, D., et al., Evaluation of 3D stack-of-spiral turbo FLASH acquisitions for pseudo-continuous and velocity-selective ASL-derived brain perfusion mapping. Magn Reson Med, 2023. 90(3): p. 939-949.http://dx.doi.org/10.1002/mrm.29681
Ebrahimi, B., S.C. Textor, and L.O. Lerman, Renal relevant radiology: renal functional magnetic resonance imaging. Clin J Am Soc Nephrol, 2014. 9(2): p. 395-405.http://dx.doi.org/10.2215/cjn.02900313
Kaplon, D.M., et al., Renal parenchyma thickness: a rapid estimation of renal function on computed tomography. Int Braz J Urol, 2009. 35(1): p. 3-8.http://dx.doi.org/10.1590/s1677-55382009000100002
Li, J., et al., Estimation of Renal Function Using Unenhanced Computed Tomography in Upper Urinary Tract Stones Patients. Front Med (Lausanne), 2020. 7: p. 309.http://dx.doi.org/10.3389/fmed.2020.00309
Morrisroe, S.N., et al., Differential renal function estimation using computerized tomography based renal parenchymal volume measurement. J Urol, 2010. 183(6): p. 2289-2293.http://dx.doi.org/10.1016/j.juro.2010.02.024
Wang, Y., et al., Integrated Diffusion Tensor Imaging and Renal Parenchymal Volume for Early Detection and Grading of Split Renal Functional Impairment in Lupus Nephritis. Acad Radiol, 2024.http://dx.doi.org/10.1016/j.acra.2023.12.046
Fan, G., et al., 3D autofluorescence imaging of hydronephrosis and renal anatomical structure using cryo-micro-optical sectioning tomography. Theranostics, 2023. 13(14): p. 4885-4904.http://dx.doi.org/10.7150/thno.86695
Pallone, T.L., A. Edwards, and D.L. Mattson, Renal medullary circulation. Compr Physiol, 2012. 2(1): p. 97-140.http://dx.doi.org/10.1002/cphy.c100036
Kurugol, S., et al., Prospective pediatric study comparing glomerular filtration rate estimates based on motion-robust dynamic contrast-enhanced magnetic resonance imaging and serum creatinine (eGFR) to (99m)Tc DTPA. Pediatr Radiol, 2020. 50(5): p. 698-705.http://dx.doi.org/10.1007/s00247-020-04617-0
Wang, Y.C., et al., The Accuracy of Renal Function Measurements in Obstructive Hydronephrosis Using Dynamic Contrast-Enhanced MR Renography. AJR Am J Roentgenol, 2019. 213(4): p. 859-866.http://dx.doi.org/10.2214/ajr.19.21224
Shi, W., et al., Assessment of Split Renal Function Using a Combination of Contrast-Enhanced CT and Serum Creatinine Values for Glomerular Filtration Rate Estimation. AJR Am J Roentgenol, 2020. 215(1): p. 142-147.http://dx.doi.org/10.2214/ajr.19.22125
Strother, M.C., et al., The Delayed Nephrogram: Point-of-care Quantitative Measurement, Validation as an Indicator of Obstruction, and Novel Use as a Predictor of Renal Functional Impairment. Eur Urol Focus, 2022. 8(6): p. 1809-1815.http://dx.doi.org/10.1016/j.euf.2022.01.019
Hu, G., et al., Intravoxel Incoherent Motion and Arterial Spin Labeling MRI Analysis of Reversible Unilateral Ureteral Obstruction in Rats. J Magn Reson Imaging, 2019. 50(1): p. 288-296.http://dx.doi.org/10.1002/jmri.26536
Cai, X.R., et al., Use of intravoxel incoherent motion MRI to assess renal fibrosis in a rat model of unilateral ureteral obstruction. J Magn Reson Imaging, 2016. 44(3): p. 698-706.http://dx.doi.org/10.1002/jmri.25172
Xu, X., et al., Diffusion-weighted imaging and pathology of chronic kidney disease: initial study. Abdom Radiol (NY), 2018. 43(7): p. 1749-1755.http://dx.doi.org/10.1007/s00261-017-1362-6
Mao, W., et al., Capability of arterial spin labeling and intravoxel incoherent motion diffusion-weighted imaging to detect early kidney injury in chronic kidney disease. Eur Radiol, 2023. 33(5): p. 3286-3294.http://dx.doi.org/10.1007/s00330-022-09331-z
Onen, A., Grading of Hydronephrosis: An Ongoing Challenge. Front Pediatr, 2020. 8: p. 458.http://dx.doi.org/10.3389/fped.2020.00458
Ding, J., et al., Evaluation of renal dysfunction using texture analysis based on DWI, BOLD, and susceptibility-weighted imaging. Eur Radiol, 2019. 29(5): p. 2293-2301.http://dx.doi.org/10.1007/s00330-018-5911-3

No competing interests reported.

SupplementaryMaterial.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Evaluation of split renal function using MR-derived structure and arterial spin-labeled perfusion features in obstructive nephropathy

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1