Objectives: Although biopsy is the most important method for diagnosing the cause of renal allograft dysfunction, sonoelastography, a new ultrasonography method, can be used to distinguish between the soft or hard nature of lesions. In this study, our aim was to investigate whether sonoelastography could diagnose fibrosis in renal transplant patients.
Materials and Methods: In this prospective study, we included patients over 18 years old who were recommended for clinical biopsy. Sonoelastographic evaluation was made by conducting acoustic radiation force impulse measurements for each patient after they were admitted to the clinic for biopsy. Measurements were performed just before the biopsy procedure. All results were examined by 2 experienced radiologists using the Siemens S3000 Ultrasound Machine (Erlangen, Germany). Comparisons of ultrasonographic values with biopsy results were made with SPSS software (SPSS: An IBM Company, version 20, IBM Corporation, Armonk, NY, USA).
Results: Of the 65 patients included in this study, pathology showed acute T-cell-mediated rejection in 37 patients. There was a significant correlation between the pathologic Banff scores and the sonographic acoustic radiation force impulse values (P = .002), where the degree of Banff increased as the mean acoustic radiation force impulse values elevated. A rise in mean impulse values correlated with increased degree of interstitial fibrosis in renal allografts. Renal parenchymal echogenicity of patients significantly differed by sex (P = .009), with an average renal echogenicity of grade 1 in women and grade 0 in men. Also, a statistically significant difference was found between age of the renal transplant recipient and resistive index values.
Conclusions: Our study showed a significant correlation between Banff degree and the acoustic radiation force impulse values of renal transplant patients. In addition to biopsy, sonoelastography can be beneficial for the diagnosis of fibrosis in renal transplant patients.
Key words : Allograft biopsy, Kidney transplant, Ultrasonography
Chronic kidney disease is a major health problem in developed countries. Renal transplant is the primary method to treat end-stage renal disease.1 Renal transplant procedures have been successfully performed since the 1950s. Despite medical and surgical advances in renal transplant, vascular and nonvascular complications are still encountered. A successful kidney transplant improves quality of life while reducing the risk of mortality. Because a transplant procedure inherently involves a great number of risks, both recipients and donors must be examined in detail before surgery to achieve intended results.1,2
Transplant success is strongly related to the protection of renal allograft function. Allograft biopsy is the “gold standard” for assessing factors that have caused renal dysfunction posttransplant.2 The histopathologic evaluation of a transplanted kidney is useful for accurate diagnosis and treatment.3 However, biopsies also have certain limitations. It is possible to miss a lesion diagnosis when specimens are small and/or isolated. Biopsy evaluations can also be limited by improper amounts of biopsy material, the inability to obtain a renal cortex sample, patchy involvement of the pathology, borderline lesions, treatments applied before biopsy, and chronic parenchymal scars.2 A biopsy procedure is invasive, and there is a possibility of the disintegration of biopsy material, which can make evaluations of renal function for early diagnosis difficult. Therefore, researchers today are working on noninvasive imaging methods instead of biopsy for diagnostic assessment.4,5
The imaging techniques that we use today are helpful in evaluating transplanted kidneys. Ultrasonography and Doppler ultrasonography are useful modalities for diagnosing and follow-up of kidney transplant complications and rejections. With ultrasonography, we can evaluate the size, the collecting system, and stones of the transplanted kidney; the perirenal fluid collections; any occupying solid or cystic lesions in the transplanted kidney; and morphology of the lesions. Furthermore, Doppler ultrasonography can evaluate the vascularity of the kidney. However, these techniques are not always sufficient to assess the transplanted kidney.6
Recently, a new ultrasonography method has been developed. This method, sonoelastography, can distinguish between hard and soft lesions, differently from B-mode ultrasonography, and it can enable the stiffness degree of the tissue to be quantified. Sonoelastography was first used experimentally by Ophir and associates in the 1980s. The first intention of evolving sonoelastography was to detect missed lesions due to similar echogenity using B-mode screening in superficial tissues such as breast, prostate, and thyroid, where palpation is essential.7,8 The sonoelastography method is similar to that of palpation applied during a clinical examination. The stiffness degree of the lesion determined by this method can also be analyzed numerically.
Ultrasonographic elastography imaging can show the stiffness degree of tissues in real time by using different color scales; thus qualitative grading can be practiced visually. In addition, a strain index can be obtained by taking the ratio of the degree of stiffness in normal tissue versus degree of stiffness in a lesion, allowing elasticity maps in which the stiffness degree of the lesion can be evaluated quantitatively.9,10 Elasticity is an essential feature of living tissues. The difference between of elastography versus B-mode ultrasonography in renal allograft studies is that elastography can detect the fibrosis of renal parenchyma by measuring its stiffness. This stiffness degree can help us to find whether there is fibrosis in the transplanted kidney. This imaging modality is a noninvasive ultrasonographic method. It does not involve radiation, and no medication is administered during the procedure.11,12
Sonoelastography techniques differ according to the force used during the procedure (half-static and dynamic) and the method used to gain signals. With the half-static technique, the tissue is stimulated by the ultrasonographic transducer mechanically. With the dynamic technique, the tissue is stimulated by acoustic waves formed by the transducer, without the need of any other force. With both techniques, the response of the tissues to these outer stimuli is measured by different methods. In general, sonoelastography techniques are sorted as strain elastography, acoustic radiation force impulse imaging (ARFI), shear-wave elastography, and transient elastography. Among these, strain elastography is half-static, and the remaining techniques are dynamic elastographic methods.13
With the ARFI technique, short periodic (0.03-0.4 ms) and high energized acoustic pulses generated with the ultrasonographic transducer will cause small amounts of transposition (1-20 μm) in the tissue determined within the region of interest (ROI). As a result of this transposition movement, shear waves are formed, and these waves are detected by the ultrasonographic device using an ultrasonographic correlation method.14
In this study, our aim was to investigate the efficacy of the sonoelastography method for the diagnosis of fibrosis in kidney transplant recipients.
Materials and Methods
After we obtained approval from the local ethics committee, this prospective study was conducted to explore the efficacy of the sonoelastography method for diagnosis of fibrosis in kidney transplant recipients. Patients over the age of 18 years who were recommended for clinical biopsies by general surgery or nephrology departments after a period of 12 months after renal transplant were enrolled in this study.
Of 65 patients included, 15 were female and 50 were male, with a mean age of 38.8 years (range, 18-70 y). Individuals who were less than 18 years old and those who had mental retardation were excluded from this study. All patients were examined by an experienced radiologist using B-mode ultrasonography and ARFI elastography methods. A Siemens Acuson S3000 device (Erlangen, Germany) with a 6C1 curvilinear transducer was used. The findings were saved on the ultrasonographic device to compare with the histopathologic results. The size and collecting system of renal allografts were evaluated by ultrasonography, whereas vascularization was evaluated with Doppler ultrasonography. Sonoelastography measurements were performed just before the biopsy procedure. Therefore, these measurements had no effect on the decision to proceed with biopsy.
During imaging, applying too much pressure was avoided and the pressure was kept as much standard as possible for each patient, considering the rise in elasticity values when the tissue is compressed with the probe. The ROI was placed in the renal cortex, and the shear wave velocity (SWV) was calculated. The areas where the ROIs were placed were strictly in the renal parenchyma, while expelling the major vascular structures and perirenal soft tissues from the ROI. Ten measurements were obtained per transplanted kidney cortex, and mean value of ARFI was calculated (Figure 1 and Figure 2).
On the same day that the ARFI measurements were obtained, ultrasonographic-guided true-cut biopsies were taken from the renal allograft in the interventional radiology department. The sections were examined by an experienced pathologist, blinded to the results of the ARFI evaluation (Figure 1 and Figure 2). Fibrosis was graded on a scale from grade 0 to grade III according to the Banff criteria.15 The Banff classification for renal allografts was developed in Canada in 1993, later improved, and updated in 2008 (Table 1).
The sonoelastographic measurements together with serum creatinine levels, biopsy results, and B-mode and color Doppler ultrasonography results were analyzed with SPSS software (SPSS: An IBM Company, version 20, IBM Corporation, Armonk, NY, USA). Comparisons among patient ARFI values were made by using independent sample t test, analysis of variance (ANOVA), and Spearman and Pearson correlation tests. P < .05 was assumed to be statistically significant.
Of 65 transplant patients included in our study, 15 were women (23.1%) and 50 were men (76.9%). The average age of patients was 38.84 ± 14 years. Of total patients, 51 (78.5%) had living-donor and 14 (21.5%) had deceased-donor kidney transplants.
The mean serum creatinine level of study patients was 2.41 mg/dL (Table 2). The mean resistive index (RI), measured in the interlobar and segmentary arteries of all patients with Doppler ultrasonography, was 0.70. The ARFI values ranged from 1.90 to 4.22 m/s (mean value of 3.15 m/s). The Banff scores of all patients were between grade 0 and III, with mean score of 1.78.
Pathology showed acute T-cell-mediated rejection in 37 patients (61.7%), chronic active humoral rejection in 11 patients (18.3%), inhibitor toxicity of chronic calcineurin in 8 patients (11.7%), acute humoral rejection in 2 patients (3.3%), acute tubulointerstitial nephritis in 1 patient (1.7 %), and thrombotic microangiopathy in 1 patient (1.7%). With use of ANOVA, we observed no significant correlations between the histopathologic findings and mean ARFI values (P = .406).
Our analyses of significance according to male versus female patients had different outcomes. We detected that time elapsed after transplant was longer in men than in women, and Banff score was lower in men than in women. In addition, RI, renal parenchymal echogenity, mean ARFI value, and creatinine level were lower in men. However, among these variables, only the time elapsed since renal transplant and parenchymal echogenity were significantly different according to sex. The time elapsed since renal transplant was 13 years in women and 8 years in men (P = .048). In addition, by t test, we found that the mean renal parenchymal echogenity was grade 1 in women and grade 0 in men (P = .009).
We observed a statistically significant correlation between time after renal transplant and RI values of patients; as time elapsed after transplant increased, the RI values increased (P = .001). However, the time elapsed after transplant had no effect on Banff score, renal parenchymal echogenity, ARFI values, or serum creatinine levels (P > .05).
Pearson test was used to analyze the relationship between patient variables (Table 3). According to our analyses, a statistically significant correlation was shown between serum creatinine levels and Banff scores (P = .003), but the degree of this correlation was weak (r = 0.362). Therefore, we can say that, as serum creatinine levels rose, Banff score and accordingly the degree of interstitial fibrosis increased.
We also found a statistically positive correlation between ARFI values and Banff scores: increases in mean Banff score were correlated with increases in ARFI values (P = .001). The Banff scores and ARFI values of patients included in this study are shown in Table 4. As shown in Table 4, the mean ARFI value of renal allografts with a Banff score of 3 was higher than others. However, mean ARFI value of kidney allografts with a Banff score of 1 was lower than all others (Figure 3). An ANOVA test showed a significant correlation between ARFI values and late-stage (grades II and III) interstitial fibrosis: as fibrosis increased, the mean ARFI value increased. However, we observed no positive correlation between ARFI values and early-stage (grades 0 and I) interstitial fibrosis.
Sonoelastography, a recent ARFI technique, can distinguish between hard and soft tissues, which is different from B-mode ultrasonography; it can also enable quantitative data according to the stiffness degree of the examined tissue. The sonoelastography method mimics palpation applied during a clinical examination. The stiffness degree of the tissue evaluated with this technique can also be analyzed numerically.15,16
In our study, we observed no significant differences between living donors and deceased donors and Banff degree, RI values, renal parenchymal echogenity, ARFI values, serum creatinine levels, or histopathologic diagnosis (P > .05). On the other hand, renal allograft parenchymal echogenities differed significantly according to sex (P < .05). Women demonstrated a mean renal parenchymal echogenity of grade 1, whereas men demonstrated grade 0, meaning that male patients in this study showed normal parenchymal echogenities. This result could be explained by response of patients to transplant medications, which may work better in male patients.
We also observed a significant association between age of the renal allograft and RI values, with RI increasing as time elapsed after transplant increased. Increased RI values (RI > 0.70) are associated with renal failure. A reason that patients included in this study had high mean RI values was because our study patients had renal function disorders, in which patients are clinically suggested to undergo biopsy procedures.
A significant positive correlation was shown between serum creatinine levels and Banff score of patients (P < .05), with increasing serum creatinine levels associated with increasing Banff scores. Renal failure is considered if serum creatinine is above 2.0 mg/dL. Because our study patients were clinically followed for renal function disorders and referred to pathology tests for histopathologic confirmation, their mean serum creatinine levels were already higher than normal. In a prospective study similar to ours from 2010 with 18 patients, the histopathologic findings of renal allografts were compared according to Banff criteria. Fifteen renal transplants subjected to elastographic evaluation by measuring ARFI values showed no correlation between ARFI measurements and RI values and none between ARFI measurements and degree of fibrosis. Eventually, it was concluded that mean ARFI measurements and RI values are independent variables for evaluating the grade of fibrosis in transplanted kidneys.17
In this study, we aimed to evaluate the grade of fibrosis in renal transplants by using ARFI technique as a future way to recognize fibrosis at a modifiable stage without the need for biopsy. The “gold” standard for evaluation of interstitial fibrosis in transplanted kidneys is tissue biopsy. However, biopsy cannot be performed when thrombocytopenia, blood diluting medicine usage, infection, severe hypertension, and blood clotting disorders are present.18,19 We believe that the ARFI method can be used under such circumstances, with additional considerations of it being an inexpensive, easily accessible, and noninvasive method, as well as having no need for contrast administration.
A biopsy gives us information only about a specific region. The specimen may not represent the whole kidney, since fibrosis is not the same in every part of the cortical tissue and even could be absent. On the other hand, with an ARFI technique, we take multiple samples from various parts of the cortex and subsequently calculate the mean values. Therefore, this method is more likely to represent the whole kidney, compared with biopsy.
Renal biopsy has its own complications; therefore, a decision for biopsy should be made according to benefits and risks to the patient. The most common complication encountered in transplant kidney biopsy is hemorrhage. Other complications include arteriovenous fistula, pain, hematoma, and infection.19 Thus, we can say that, for such conditions where biopsy is risky, an ARFI technique has another advantage, which is guidance to biopsy. That is, we would aim to sample the area with the highest ARFI value, increasing the diagnostic accuracy of a biopsy while minimizing biopsy complications.
In our study, ARFI values of renal transplants with Banff score of grade II or grade III were higher than those of grade 0 or grade I. This can be explained with the mean ARFI values being already high in renal transplants, which have higher levels of interstitial fibrosis. We found no positive correlation between ARFI values and early-stage (grade 0 or I) interstitial fibrosis. This may be related to our small sample size or the assumption that, at early stages, fibrosis has not yet diffusely spread, and thus the possibility of cortical biopsy location being normal tissue is considerably high. However, in our study, we observed that, in patients with biopsy results of grade II and grade III interstitial fibrosis (moderate and severe fibrosis), mean ARFI values rose as interstitial fibrosis increased. Therefore, the ARFI technique may be useful for only evaluating moderate and severe grades of fibrosis in renal allografts.
According to our results, the ARFI method can be used to diagnose interstitial fibrosis considering the Banff criteria. This finding is supported by previous work in a larger patient population.20 Renal transplant function and SWV have also been positively correlated in previous work.21 On the other hand, some studies have found no relation between SWV and fibrosis.22,23 That is, previous studies are conflicting concerning evaluation of interstitial fibrosis and its correlation with SWV. In a study of 18 patients, Stock and associates17 described a positive correlation between SWV and fibrosis. Syversveen and associates22,24 and Lee and associates23 reported no significant differences between SWV and fibrosis.
In 2010, a study reported that ARFI values of renal allografts and interstitial fibrosis are not significantly correlated. In that study, the transducer used was not suitable for the patients and their personal anatomy (patients had body mass index > 30 kg/m2). Therefore, the study implied that biopsy was preferred for diagnosing interstitial fibrosis in renal allografts.25
In another similar study from 2010 to 2013 in 91 patients, no significant positive correlation was found between ARFI and interstitial fibrosis. This result was suggested to be because the SWV propagation in ARFI technique was influenced by patient body weight and time elapsed after transplant.23 In 2010, another prospective work similar to our study was performed with 18 renal transplant patients. The histopathologic evaluations of biopsies for fibrosis were made within the limits of Banff criteria, and ARFI measurements of these kidneys were made. The work concluded that the mean ARFI measurements and RI values are significant in revealing the diagnosis of fibrosis.17
In our study, regarding the positive significant correlation found between Banff degree and the ARFI results in sonoelastographic evaluation, we can say that the ARFI technique is a useful method for diagnosis of interstitial fibrosis. A biopsy procedure has some limitations, including sampling mistakes and disparity of evaluation of specimen between different pathologists. In addition, a biopsy cannot be performed in such conditions as thrombocytopenia and when blood diluting medicines are used by the patient. It will be practical to evaluate these patients with the ARFI technique of sonoelastography. In addition, ARFI measurements made before biopsy can be helpful to guide the location of the biopsy procedure.
Our study had a number of limitations, including small patient population and the dependancy of ARFI results to the patient and to the sonographer. In addition, because this technology is new and is not present in every ultrasonography device, there is limited access to this technique and there are few experienced sonographers who are able to produce ARFI results. The quality of the technique can be affected by suboptimal breath holding, movement artifacts, high body mass index of patients, and when optimal supine position cannot be obtained. Our study was also single centered, and we had no control group of ARFI measurements from different centers made by different radiologists. Another limitation was the tissue architecture of a kidney placed in ROI, which is more inhomogeneous compared with the liver. If the ROI is not placed properly in the cortical parenchyma of the kidney, the blood vessels, medulla, and collecting system might be included unintentionally. To minimize this inhomogeneity, new technology developments are needed.
In conclusion, sonoelastography is a useful method for diagnosis of grade II and grade III interstitial fibrosis. In addition, we can increase the accuracy of biopsy by using the ARFI technique as guidance. However, the utility of the ARFI technique in diagnosing fibrosis in renal transplants is controversial considering the previous studies in the literature so far, which have conflicting results. Thus, with improvements of this technology, further studies with larger patient populations are needed to support the effectiveness of the ARFI technique.
DOI : 10.6002/ect.2017.0238
From the 1Department of Radiology, the 2Department of Pathology, and the
3Department of General Surgery, Başkent University Medical Faculty, Ankara,
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Funda Ulu Öztürk, Başkent University Hospital, Department of Radiology, Mareşal Fevzi Çakmak St. No: 45 Bahçelievler 06490, Ankara, Turkey
Phone: +90 555 223 60 70
Table 1. Banff 97 Diagnostic Categories for Renal Allograft Biopsies
Table 2. Laboratory Findings in the 65 Study Patients
Table 3. Comparison of Variables
Table 4. Comparison of Acoustic Radiation Force Impulse Imaging Values and Banff Scores
Figure 1. Acoustic Radiation Force Impulse Imaging Measurements and Biopsy Result of a Patient With Banff Score 0
Figure 2. Acoustic Radiation Force Impulse Imaging Measurements and Biopsy Result of a Patient With Banff Score II
Figure 3. Comparison of Acoustic Radiation Force Impulse Imaging Values Versus Banff Scores