Imaging Fibrosis in Pediatric Kidney Transplantation: A Pilot Study

Abstract

Background:

Non-invasive alternatives to biopsy for assessment of interstitial fibrosis and tubular atrophy (IFTA), the major determinant of kidney transplant failure, remain profoundly limited. Elastography is a non-invasive technique that propagates shear waves across tissues to measure their stiffness. We aimed to test utility of elastography for early detection of IFTA in pediatric kidney allografts.

Methods:

We compared ultrasound (USE) and MR elastography (MRE) stiffness measurements, performed on pediatric transplant recipients referred for clinically indicated biopsies, and healthy controls.

Results:

Ten transplant recipients (median age 16 years) and eight controls (median age 16.5 years) were enrolled. Three transplant recipients had “stable” allografts and seven had Banff Grade 1 IFTA. Median time from transplantation to biopsy was 12 months. Mean estimated glomerular filtration rate was 61.5mL/min/1.73m² by creatinine-cystatin-C CKiD equation at time of biopsy. Mean stiffness, calculated through one-way ANOVA, was higher for IFTA allografts (23.4kPa USE /5.6kPa MRE) than stable allografts (13.7 kPa USE /4.4kPa MRE) and controls (9.1kPa USE /3.6kPa MRE). Pearson’s coefficient between USE and MRE stiffness values was strong (r=0.97). AUC for fibrosis prediction in transplanted kidneys was high for both modalities (0.91 USE and 0.89 MRE), although statistically non-significant (p>0.05). Stiffness cut-off values for USE and MRE were 13.8kPa and 4.6kPa, respectively. Both values yielded a sensitivity of 100% but USE specificity (72%) was slightly higher than MRE (67%).

Conclusion:

Elastography shows potential for detection of low grade IFTA in allografts although a larger sample is imperative for clinical validation.

Graphical Abstract

Means of elastography stiffness markers were significantly higher for interstitial fibrosis and tubular atrophy (IFTA) allografts than stable allografts and controls. Therefore, elastography is showing a potential for early detection of IFTA in renal allografts but a larger sample is needed for clinical validation.

Introduction:

One of the major histopathologic determinants for progression to kidney allograft failure is the development of interstitial fibrosis and tubular atrophy (IFTA)[1]. The pathological course of IFTA progresses in two phases; the early phase, which typically occurs in the first-year post-transplantation is characterized by new-onset fibrogenesis and tubulointerstitial injury as a result of ischemic or immunological insult; and the late phase, which typically occurs beyond the first year involves a more severe pattern of allograft injury (interstitial fibrosis, tubular atrophy, glomerulosclerosis, and arteriolar hyalinosis) leading to irreversible graft loss[2, 3].

To date, percutaneous biopsy remains the gold standard and the only reliable method for evaluation and diagnosis of IFTA. However, biopsies present risks of complications, which include hematuria, pain at the biopsy site, arteriovenous fistulas, hematomas, infections, or, in rare cases, graft loss, despite ultrasound guidance[4–7]. Previous studies found that serum creatinine rises only in the late stages of graft failure, when lesions become irreversible, therefore it is not a reliable surrogate for early progression of IFTA [8, 9]. The risks of complications associated with biopsies and the diagnostic limitations of serum creatinine highlight the critical need to develop a non-invasive alternative method to achieve accurate and reproducible identification of allograft fibrosis.

Elastography is a non-invasive technique that propagates shear waves across tissues to provide quantitative measurement of their stiffness relative to the imparted force. Each imaging modality uses a specific energy level to trigger tissue excitement and requires modality and technique-specific cut-off values[10]. In magnetic resonance elastography (MRE), a measurable phase shift is induced by the motion-sensitizing gradient, which allows us to calculate the displacement at each voxel and directly image the acoustic waves. In contrast, ultrasound elastography (USE) uses combined push and time-aligned sequential tracking for the generation of large elasticity maps superimposed on the grayscale image obtained by using conventional ultrasound. Elastography cut off values vary by technique and vendor and as such, the direct comparison between USE and MRE is difficult in routine clinical practice. Both USE and MRE techniques have been approved by the U.S. Food and Drug Administration for use in liver disease and have high sensitivity and specificity in the differentiation between normal and cirrhotic liver [11]. USE advantages include wider availability, shorter duration easier-to-perform study, patient comfort and lower cost; while MRE allows for the evaluation of the entire organ, accounting for heterogeneous sampling of other techniques and it is not operator dependent. Currently, correlations between elasticity and fibrosis for both native and transplanted kidneys have been published in adults, but their results are not yet conclusive for either MR or US elastography [12–15]. In children, there is even less evidence [16]. We aim to compare ultrasound and MR elastography in order to determine normative values for pediatric kidney transplant recipients and healthy controls and evaluate their performance as independent but equivalent noninvasive imaging biomarkers.

Methods:

Study design and selection criteria:

This pilot study is part of an ongoing single-center, HIPAA-compliant, IRB-approved and NIH funded prospective observational cohort study: “Imaging Modalities in Pediatric Assessment of Pediatric Kidney Transplants’ (IMPAKT). IMPAKT eligible cases are pediatric transplant recipients older than 6 years of age scheduled to undergo a clinically indicated (either surveillance or “for-cause”) kidney allograft biopsy who are followed at a tertiary pediatric center that performs ~ 16–20 transplant cases per year and care for ~130 children with transplants to the present time. Surveillance protocol biopsy program has been established at the center since November 2019. IMPAKT healthy controls were randomly selected from pediatric center’s population database (see below) and defined as children > 6 years of age who do not have known history of kidney disease, urinary tract malformations, or febrile urinary tract infection (pyelonephritis) and have a body mass index between 5^th and 85^th percentile. IMPAKT controls were not matched to cases with kidney transplant since the transplanted kidneys come from donors of all ages. Subjects that met study criteria between February 2020 and February 2022 were enrolled in this pilot study. Informed written consent was obtained in compliance with regulatory guidelines from IMPAKT eligible cases and healthy controls.

During the screening process, subjects were excluded from the study if there was any contraindication for MRI (e.g., cochlear implants, pacemakers, or spinal stimulation devices), inability to tolerate MRI (claustrophobia, age), or if the subject was pregnant. Healthy controls were randomly identified through the Recruitment Enhancement Core of the Clinical Research Support Office at Children’s Hospital of Philadelphia. Transplant recipients had all imaging studies performed prior to allograft biopsy; no biopsy was performed on healthy controls.

USE imaging technique:

USE measurements were obtained by a clinically approved Philips ultrasound scanner (Philips Medical Systems, Bensalem, PA) using age and size-appropriate linear high-frequency (15–4 MHz) or convex low-frequency (2–5 MHz) transducers. In the case of transplant recipients, given the proximity to skin surface, all US elastography studies were performed with a curve low frequency (C1–6) probe. The equipment comes with a dedicated kidney preset that directly converts the velocity measurement into a quantitative value of elasticity, expressed in kilopascals (kPa) and/or shear wave velocity expressed in meters per second. Elasticity values are displayed as a real-time, color-coded, 2D quantitative map of tissue stiffness over a conventional grayscale B-mode image. The positions of the circular ROI used for stiffness/velocity measurement are operator adjustable. The research scans were conducted by two qualified sonographers with over 20 years of experience. The quality of the images and elastographic measurement were subsequently evaluated by a pediatric radiologist with 9 years of postgraduate experience and research interest in advanced renal imaging.

Native kidney stiffness measurements were obtained from mid, upper, and lower poles of the kidney depending on patient habitus and field of view, avoiding the renal sinus, large crossing vessels, the renal capsule, and perinephric fat. Kidney allograft measurements were obtained from the lower pole (i.e.; on the region where biopsy sample was to be obtained). We acquired a minimum of 10 measurements per kidney unit using a 10-mm diameter circular ROIs. Depth of ROI placement depended on the size of the participant. Subjects were lying in a prone position for native kidneys and in a supine position for kidney transplants.

Since reliability thresholds for USE have not yet been established, we applied liver accepted methodology where the average of the 10 valid measurements are used to calculate the mean. In the liver literature, an IQR/median of less than 30% of all included measurements is clinically accepted as “good quality image”[17]. We similarly report the IQR/median ratio for each case and performed a secondary review of all images with IQR/median >30%. After review, all images were found to be adequate as obtained and all values were included in the analysis.

MRE imaging technique:

MRE was performed on a 3 Tesla MR scanner (Siemens Healthineers, Malvern, PA) equipped with MRE hardware consisting of an active-passive driver system. As previously described, the active driver is similar to an audio subwoofer and generates low-amplitude 60 Hz vibrations which are passed via pneumatic pressure through a ¾-inch diameter, 24-ft long hollow plastic tube to the passive driver, which is placed on the subject’s lower abdomen (for kidney transplant recipients) or under the back (for native kidneys in healthy controls) (Figure 1)[18, 19].

Figure 1:

Schematic of kidney MRE paddle positioning. The passive driver is positioned posteriorly to be relatively closer to the kidney from the dorsal side. A Y-shaped splitter is used to position 2 paddles in parallel under both kidneys. For patients with single allograft kidney, the Y-shaped paddle is replaced with a single conventional paddle.

Shear wave images were generated by obtaining multiple slices through the kidney(s). A 2D spin-echo planar MRE (SE-EPI) sequence was used to acquire coronal wave images with the following parameters: repetition time ms/echo time ms, 1000/30; continuous sinusoidal vibration, 60 Hz; field of view, 32–42 cm; matrix size, 100 × 100; flip angle, 30°; section thickness, 6 mm; 4 evenly spaced phase offsets; acceleration factor of 2, and 4 pairs of 60-Hz trapezoidal motion encoding gradients with zeroth and first moment nulling along the through-plane direction. All processing steps were applied automatically without manual intervention to yield quantitative images of tissue shear stiffness in kilopascals. Stiffness maps (elastograms) were produced automatically by the scanner software, including 95% confidence maps indicating areas of good wave propagation. On each section of the image on MRE, ROIs were manually drawn on the elastograms within regions bound by the confidence maps, which included the maximal amount of renal parenchyma in each slice, while avoiding the edges and large vessels. The mean of measurements on 4 slices was used. The mean stiffness (in kPa) and area (in cm²) of each slice were calculated and recorded. Overall kidney’s mean stiffness was calculated as the average of stiffness measurements from each slice, weighted by the ROI area of each slice. MRE measurements were obtained by a single reader with three years of experience who was blinded to the USE measurements recorded by sonographers.

Pathology:

Biopsy procedures were performed by the pediatric nephrology service under ultrasound guidance assisted by a pediatric ultrasound technician. One to three cores of the lower pole were obtained using 16 or 18-gauge needles (Bard Magnum, Covington, GA, USA). Specimens were embedded in paraffin. Paraffin blocks were sliced in 3 μm samples and then stained with hematoxylin & eosin, periodic acid-Schiff (PAS), trichrome Masson stain, and methenamine silver stains. Immunohistochemistry staining was performed for C4d and BK virus (SV40 T antigen), and additional IHC stains such as EBER (in situ hybridization) was performed as requested based on clinical scenario. The pathology report was formulated according to the updated 2018 Banff classification which stratified pathology by stages of severity of interstitial fibrosis and tubular atrophy; Grade 0 = 0 to 5% (also referred to as “stable allografts”), Grade 1 = 6–24%, Grade 2=25–49%, and Grade 3 = ≥50% [1]. The report was generated by a single pathologist with 13 years of post-fellowship experience blinded to the results of elastography scans.

Data collection and Statistical Analysis:

Data derived from MR, US elastography and pathology reports were recorded in RedCap^® (Vanderbilt University, Nashville, TN). Three subject groups were defined: healthy controls with eGFR >90mL/min/1.73m²; stable allografts with IFTA Grade 0; and IFTA+ allografts with IFTA Grade ≥ 1 on histologic review.

Descriptive statistics were used to calculate counts and frequencies for categorical variables and means, medians, standard deviations, and interquartile ranges (IQR) for continuous variables. The stiffness data met parametric distribution on histograms. Accordingly, comparison of the mean stiffness values for MRE and USE between the three subject groups (healthy controls, stable allografts, and IFTA+ allografts) was done through one-way ANOVA test while comparison of the mean stiffness values between each pair of subject groups was done using independent t-test.

In the subset of IFTA+ allografts, the t-test was also used to compare the stiffness values between rejected and non-rejected grafts. Box and Whisker plots were used to compare the median and IQR of the stiffness values between the three patient groups. Pearson correlation coefficient was used to determine the degree of correlation between MRE and USE stiffness measurements. Receiver operating characteristic (ROC) curve was used to determine the area under the curve (AUC) for both MRE and USE and the cut-off stiffness values that have the highest sensitivity and specificity in differentiating stable from IFTA+ allografts. P-value ≤ 0.05 was considered statistically significant. Statistical computation was performed using IBM^® SPSS^® V. 25 (SPSS Inc., Chicago, IL) and JMP Pro 15^® (SAS Institute, Cary, NC).

Results:

Ten transplant recipients (8 males and 2 females) and 8 controls (2 males and 6 females) were enrolled. Median age was 16 years, IQR (12.5–18) for transplant recipients and 16.5 years (13.5–19 years) for controls. Five of the 10 transplant recipients (50%) received allografts from deceased donors. Three transplant recipients had “stable” allografts with no IFTA on pathology and 7 had Banff Grade 1 (G1) IFTA. Among these 7 IFTA allografts, one had acute antibody mediated rejection and one had borderline T-cell mediated rejection; the rest did not show evidence of rejection on histology.

Congenital anomalies of the kidneys and urinary tract (CAKUT) was the most common cause of renal failure (n=5), Table 1, which are representing of similar cohorts with kidney disease[20, 21]. Time from transplantation to allograft biopsy ranged from 2 weeks to 6 years. No biopsy complications were reported. Details of participants’ demographics are illustrated in Table 1.

Table 1:

Participants’ characteristics

	Total (n=18)	Recipients with “stable” allografts (n=3)	Recipients with “IFTA+” allografts (n=7)	Controls (n=8)
Gender
− Male, n (%)	10 (56)	3 (100)	5 (71)	2 (25)
Race
− Asian, n (%)	1 (6)	0 (0)	0 (0)	1 (13)
− Black/African American, n (%)	7 (39)	2 (67)	3 (43)	2 (25)
− White, n (%)	7 (39)	1 (33)	3 (43)	3 (37)
− Other, n (%)	3 (17)	0 (0)	1 (14)	2 (25)
Median age (IQR) (y)	16 (12.5–18)	14 (13–16)	16 (14–18)	16.5 (13.5 – 18)
Number of Kidneys analyzed	26	3	7	16
Number of MRE exams	17	3	6	8
Number of USE exams	11	3	4	4
Mean BMI (Kg/m2)	22	22	24	21
Mean serum creatinine (mg/dl)	1	1.2	1.4	0.7
Mean eGFR mL/min/1.73m2 (Creatinine-Cystatin C CKiD equation 2012) [45]	77	67	56	109
Mean eGFR mL/min/1.73m2 (average CKiD U25 equation) [46]	78	69	56	108
Transplant characteristics*:  Median age at time of transplant (y)	13.5	11.9	15	---
Median time from transplant to biopsy (m)	12	7	14	---
Deceased donor, n (%)	5 (45*)	3 (100)	2 (29)	---
Median donor age (y)	34	31	36
Cause of renal failure:   − Genetic kidney disease	3 (27*)	2 (67)	1 (14)	---
- CAKUT	6 (55*)	1 (33)	5 (71)	---
− Others	1 (9*)	0 (0)	1 (14)	---

^†IQR= Inter-quartile range; BMI= Body mass index; CAKUT=Congenital anomalies of the kidney and urinary tract; USE= Ultrasound elastography; MRE = Magnetic resonance elastography; eGFR=estimated glomerular filtration rate.

^*Percent from total transplant recipients only (not including controls).

Kidney transplant recipients vs. controls

For USE, we found that the mean stiffness of IFTA allografts (23.4 kPa) was significantly higher than stable allografts (13.7 kPa, p=0.05) and controls (9.1 kPa, p=0.02). The same pattern was observed for MRE where the mean stiffness of IFTA allografts (5.6 kPa) was higher than stable allografts (4.4 kPa, p=0.04) and controls (3.6 kPa, p=0.01). Similarly, the mean stiffness of stable allografts was significantly higher than controls (p=0.03 for USE and p=0.02 for MRE). Box and whisker plots comparing the median and IQR between groups are illustrated in Figure 2.

Figure 2:

Box and Whisker plot showing mean of (A) USE and (B) MRE stiffness values (in kPa) for healthy controls, stable allografts, and G1 IFTA allografts.

†USE=Ultrasound elastography; MRE= Magnetic resonance elastography; G1=Grade 1; IFTA = Interstitial fibrosis and tubular atrophy.

Elastography and pathology correlation

Ten participants (6 transplant recipients and 4 controls) underwent both USE and MRE scans with a total of 14 kidneys (6 transplant and 8 native) analyzed. Understanding the imaging techniques are not directly comparable we performed correlation analysis on stiffness values and found there was a very strong correlation between USE and MRE stiffness measurements (r= 0.97; 95% CI 0.96 – 0.99; p<0.01) (Figure 3).

Figure 3:

Pearson correlation between USE and MRE stiffness measurements (in kPa).

†USE=Ultrasound elastography; MRE= Magnetic resonance elastography.

Seven transplant recipients underwent USE scans (3 with “stable” allografts and 4 with IFTA+ on biopsy reports) and 9 transplant recipients underwent MRE scans (3 with “stable” allografts and 6 with IFTA on biopsy reports). The ROC analysis (Figure 4) for elastography modalities yielded high AUC curves for both USE (AUC = 0.91; 95% CI 0.69 – 1) and MRE (AUC = 0.89; 95% CI 0.65 – 1); however, the findings were not significant (p = 0.077 for USE and p=0.071 for MRE). The optimal stiffness cut-off value for USE and MRE to predict kidney transplant fibrosis was 13.8 kPa and 4.6 kPa, respectively. Both values yielded a sensitivity of 100% but the specificity of USE (72%) was slightly higher than MRE (67%). Representative images of USE and MRE stiffness maps with color coded spectrums are illustrated in Figure 5.

Figure 4:

ROC curve showing sensitivity and specificity of (A) USE and (B) MRE stiffness values in differentiating stable from fibrosed allografts among transplant recipients. Cut-off values: USE 4.6kPa, MRE 13.8 kPa

†ROC= Receiver operating characteristic; USE= Ultrasound elastography; MRE= Magnetic resonance elastography.

Figure 5: USE and MRE renal stiffness maps.

Representative images for USE (top row) with ROI delineated and MRE (bottom row) renal stiffness maps shown for (A) a 9-year-old female healthy control participant, (B) a 12-year-old female patient with stable allograft, and (C) a 12-year-old male patient with G1 IFTA allograft.

†MRE= Magnetic resonance elastography; USE=Ultrasound elastography; ROI=Region of interest; G1=Grade 1; IFTA = Interstitial fibrosis and tubular atrophy.

Even though this pilot study was not powered to find differences in subgroups to analyze the presence of rejection among IFTA allografts, for both US and MR elastography, we compared the mean stiffness of allografts with rejection and allografts without rejection. There was no difference between these means (19.2 versus 18.8 kPa in US elastography and 5.5 versus 5.3 kPa in MRE, respectively; p>0.05).

Discussion:

In our study, we found a positive correlation between kidney elasticity and the presence of histological fibrosis, where stiffness values tended to be higher in fibrosed allografts than stable ones or native kidneys. Our findings contrast with the only available pediatric kidney transplant elastography study which showed that there was no correlation between stiffness parameters and the degree of fibrosis [16]. However, it is important to mention that, despite having a larger sample size, their study was limited by poor interobserver variability, biopsy sampling error in a large proportion of their cases, and a low success rate in linear probe elastography measurement [16].

Previously published studies evaluating elastography assessment of renal stiffness in adults have also shown conflicting results, both for native kidneys and allografts (Supplement Table A1). These discrepant findings have led the European Federation for Ultrasound in Medicine and Biology (EFSUMB) to point out that no definitive recommendations for the application of elastography in renal assessment can be made to date [22]. Most of the studies showed a positive correlation between stiffness parameters and fibrosis grades based on the assumption that as renal fibrosis progresses, the renal architecture becomes stiffer leading to faster propagation of the shear waves which corresponds to higher elastography stiffness values [14, 23–27]. Our findings stand in line with this hypothesis. Conversely, some studies that had a negative correlation between elastography and biopsy findings argued that, because the kidney is a highly vascular organ, renal perfusion, rather than IFTA, is the predominant factor influencing renal stiffness. While some of these studies have provided histologic evidence of arteriolar wall thickening, it is noteworthy to mention that none of these studies have conducted quantitative analysis of the renal perfusion (e.g., through dynamic renal imaging) in their samples to validate their findings [28, 29]. On the contrary, Chhajer et al. who showed a positive correlation between elastography stiffness values and Banff scores found no correlation between Doppler resistive indices and stiffness or biopsy parameters [24].

Lee et al. was among the very few studies that found no correlation between shear wave velocity and renal fibrosis or dysfunction. However, their allograft protocol biopsy to a cohort of living donor recipients was performed within 6 months in most of their patients thereby excluding all kidneys with a transplant age of more than one year, thus limiting potential cases with graft fibrosis [30].

There are many possible reasons for the inconsistent USE and MRE findings reported in kidney transplant elastography literature. Some of these factors are attributed to the anatomical and mechanical properties of the kidney transplant itself which limit elasticity measurements such as anisotropy, heterogeneity of the tissues (cortex, medulla, and sinuses), degree of hydronephrosis, and variability in renal depth and mobility [31]. These variations could potentially explain the significant differences in the stiffness values we found between the stable allografts and the native kidneys, which were also reported in the previous pediatric kidney transplant elastography study [11]. The superficial location of the transplanted kidneys likely rendered them more sensitive to pressure variations from the probe/paddle where a slight pressure from the probe/paddle dramatically increased the stiffness of the kidney cortex [32]. This negative correlation between the elastography stiffness and the depth of the graft position has been documented in previous studies and it was attributed to the damping of the amplitude of the shear waves as they hit intervening structures [33].

Other factors are related to renal parenchymal segmentation in elastography which often involves areas from both the cortex and the medulla. This could be a source for potential error since biopsy specimens are only obtained from the renal cortex and not the medulla. Previous attempts to obtain segments from the cortex only yielded inadequate wave information for stiffness calculation because the mean thickness of the renal cortex is approximately six millimeters, which is only about two voxels at the current image resolution [34]. Differentiating the mechanical properties of the cortex and medulla using wave propagation is still under technical development but may require raising the imaging resolution (to 256² or 512²), increasing the vibration frequency (> 100 Hz) to accurately depict structures with smaller wavelengths, or utilizing 3D imaging techniques to acquire thinner slices (< 2 mm) while maintaining approximate isotropic image resolutions [14].

It is also possible that the nature of the biopsy specimen itself might be flawed. Kirpalani et al. acknowledged this possibility after noticing that elasticity values correlate positively but not strongly with fibrosis severity (Spearman rho =0.67; P=0.01) [35]. Their rationale was that tissue fibrosis is heterogeneously distributed throughout the renal parenchyma, so biopsy, even though considered the reference standard, may not truly reflect the fibrotic burden of the whole kidney because of its small sample size and potential for sampling error. Since localizing the exact biopsy site on the elastography scan is not feasible, comparing whole kidney stiffness with Banff fibrosis scores of a small biopsy specimen may not always reveal a uniform and robust correlation [35].

The main strength of our study is that we were able to show high accuracy for both USE and MRE in renal stiffness assessment with an excellent correlation between their measurements. This not only demonstrates the feasibility that MR and US imaging can be performed in a pediatric transplant population on their day of the biopsy with no additional time added to their visit (studies performed within 2–2.5hrs, time required to be in the facility for screening procedure purposes), but mainly that both modalities could have the potential to evaluate IFTA in kidney allografts. While the scanning protocol for both modalities could be modified to generate 3D imaging sequences thus having a better marking of the renal cortex, each modality seems to be superior in a certain aspect with regard to graft assessment. Ultrasound is relatively quicker, less claustrophobic, and can be performed when MRI is contraindicated. In addition, contrast-enhanced ultrasound, already proven to be superior to color Doppler, can provide quantitative analysis of renal microvascular perfusion for a better understanding of its influence on renal stiffness without causing nephrotoxicity, even in the incidence of graft rejection [36–38]. MRE can propagate shear waves with higher frequency which can provide better estimate for the stiffness of smaller and deeper structures [39]. Other non-contrast MRI sequences such as diffusion-weighted imaging (DWI) have been reported to reveal the imaging features of the renal parenchymal microstructure with excellent cortico-medullary differentiation by showing regional differences in apparent diffusion coefficient (ADC) measurements [40, 41]. These sequences can be combined with elastography to optimize the stiffness maps by subtracting medullary regions thus obtaining more accurate cortical stiffness values. If clinically approved, elastography not only would be useful for early detection of IFTA in renal transplants, which would individualize the need for immunosuppression therapy, but also could be used as a screening tool to evaluate fibrosis in potential donors to identify best candidates prior to the transplantation procedure.

There are some limitations to our study. The main limitation is the small sample size of this pilot study. This limitation has been described in many previous publications in the literature addressing the same topic because of the overall low prevalence of renal transplant recipients, more so in the pediatric population [14, 16, 25, 27]. Validation of the reliability of the measurements can only be achieved when larger cohort samples are obtained. Other limitations include the absence of correlations between elastography parameters and other confounding factors that could potentially affect the renal tissue stiffness such as age and gender [42]. Age correlation with elasticity values in pediatric native kidneys was controversial; Lee et. al. found that elasticity strongly correlates with age in children younger than five years [43], but a much larger cohort conducted by Grass et. al. [44] found no correlation between age and elasticity; correlation with donor age was not available for these studies. No variations in elasticity values across genders were found in the pediatric literature [44, 45]. Additionally, testing the stiffness values in more advanced stages of fibrosis (e.g., Grade 2 and 3 IFTA) was not feasible because these cases were not found in our study. However, our main goal was to detect fibrosis at its earliest stages to understand and create the expectation of allograft half-life by enabling elastography as a monitoring tool of allograft fibrosis.

Conclusion:

While each modality has specific advantages, our study has shown that both MRE and USE have the potential to be utilized as non-invasive tools for quantitative measurement of the degree of fibrosis in renal allografts with a strong correlation between their measurements. Ongoing investigation using larger sample sizes and modified imaging protocols may still be needed to ultimately ascertain if these new imaging biomarkers can be clinically validated as screening tools for detection of low-grade severity of renal pathology post-transplantation.

Table 2:

Stiffness Parameters between healthy controls and transplant recipients with “stable” and “IFTA+” allografts

	MRE	USE
Mean stiffness ± SD (kPa)  Controls	3.6 ± 0.5	9.1 ± 2.3
Stable allografts	4.4 ± 0.5	13.7 ± 3.0
IFTA + allografts	5.6 ± 0.8	23.4 ± 6.2
t-test comparison (p-value)* Controls vs stable allografts	0.036	0.023
Controls vs G1 IFTA allografts	0.026	0.001
Stable vs G1 IFTA allografts	0.042	0.045
Three subject group comparison (p-value)**	<0.001	<0.001

^†USE=Ultrasound elastography; MRE= Magnetic resonance elastography; G1=Grade 1; IFTA= interstitial fibrosis and tubular atrophy.

^*Comparison of the mean stiffness values between each pair of subject groups was done using independent t-test.

^**Comparison of the mean stiffness values between the three subject groups was done using one-way ANOVA.

Data statement:

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.