British Journal of Radiology 74 (2001),411-419 © 2001 The British Institute of Radiology
Renal allograft vasculopathy: ultrasound findings in a non-human primate model of chronic rejection
L Gaschen, DVM, Dr med vet and
H-J Schuurman, PhD
Novartis Pharma AG, Transplantation Research, WSJ 386.526, S.386.526 Kohlenstrasse, 4002 Basel, Switzerland
Correspondence: L Gaschen, DVM, Dr med vet, Rue des Longschamps 44, 2068 Hauterive, Switzerland
 |
Abstract
|
|---|
The purpose was to determine whether decreased cortical flow detected with power Doppler (PD) ultrasound in renal allografts in cynomolgus monkeys marks the presence or onset of chronic renal allograft vasculopathy. The 2D grey scale and PD ultrasound findings of 24 consecutively implanted non-life-supporting renal allografts in cynomolgus monkeys that underwent either 24 h (n=15) or 48 h (n=9) cold ischaemia times were recorded and compared with the results of histology performed every 2 weeks post-operatively. 13 allografts developed vasculopathies, 10 of which had PD scores equal to 1 (severe reduction of cortical flow). A PD score of 1 occurred in only one instance in the group of allografts without vasculopathies and this was due to necrosis. Allografts without vasculopathies otherwise had either PD scores of 3 (normal flow; n=2) or 2 (reduced flow; n=4). Allografts subjected to 48 h cold ischaemia times were smaller than those with 24 h cold ischaemia times (significant at weeks 5–11, p<0.05), but a reduction in graft size associated with vasculopathies occurred infrequently. In conclusion, the finding of reduced renal cortical flow detected by PD ultrasound during serial examination of non-life-supporting renal allografts is highly supportive of a diagnosis of graft vasculopathy due to arteriolar intimal proliferation, and illustrates an excellent method of monitoring changes in cortical perfusion in allografts in animal models. The combination of findings of reduced or absent cortical flow together with severe graft enlargement is highly suggestive of the presence of not only vasculopathies but also tissue damage and degeneration.
|
Introduction
|
|---|
The development of animal transplantation models is currently important to understand the process of allograft rejection as well as its prevention and treatment [1]. Implantation of non-life-supporting allografts (unilateral native kidney nephrectomy) has been applied in the development of a model of chronic renal allograft rejection in the rat [2–4]. Allografts that developed chronic nephropathies showed reduction in renal cortical flow as determined by both MRI and ultrasound serial examinations in the presence of chronic rejection [4]. Decreases in graft volume and hypertrophy of the native kidney were additional findings associated with chronic nephropathies on both MRI and ultrasound. In addition, there was complete agreement in the study by Gaschen et al [4] between the ultrasound and MRI findings compared with the histology results in all 17 cases of chronic allograft nephropathy (CAN) and in all 20 histologically normal grafts. In cases of CAN, decreased flow occurred over time, the first sign being loss of the "blush" of colour shown with power Doppler ultrasound in the renal cortex, followed by progressive loss of flow to more central areas of the kidney. Allografts with severe to end-stage CAN, whether or not accompanied by vasculopathy, received power Doppler scores of 1, which corresponded to a perfusion deficit on MRI of between 75% and 100%. Those grafts with moderate CAN without vasculopathy had power Doppler scores of 2 and corresponded to a perfusion deficit on MRI of 50–75%.
In our own experience, anaesthesia has been a limiting factor when performing MRI studies in monkeys. The contribution of ultrasound, which can be performed on conscious animals, might therefore be more relevant for non-invasively characterizing grafts in this species. Our first experience with power Doppler ultrasound in the cynomolgus monkey allograft model described in this study showed very similar findings in allografts that developed cortical vasculopathies compared with a rat study, i.e. marked reduction in cortical flow detected with serial power Doppler examinations as well as a decrease in renal size.
The superior representation of blood flow in the renal cortex by power Doppler ultrasound compared with conventional colour Doppler ultrasound has been well established [5–7]. Other investigators have recently shown that power Doppler examination of human renal allografts improves clinical diagnostic accuracy [8–11]. Here, we question whether reduced cortical flow detected by serial power Doppler examinations indicates the presence of cortical vasculopathies in chronic renal allograft rejection in the cynomolgus monkey. Vascular changes occuring in chronic renal allograft rejection include thickening of the arteriolar intima with narrowing and obliteration, and few reports describe Doppler changes associated with chronic rejection [12]. We also question whether decreasing allograft size is an additional marker for the presence of chronic rejection, as has been described in the rat model.
|
Materials and methods
|
|---|
In a prospective study, the ultrasound findings of 24 consecutively implanted non-life-supporting renal allografts (unilateral native kidney nephrectomy) in cynomolgus monkeys (Macaca fascicularis) were recorded and compared with the results of histology. All animals were treated once daily with an oral dose of 150 mg kg-1 cyclosporin (Neoral®, Novartis Pharma AG, Basel, Switzerland), starting on day -1 and on day +1 for a total of 14 days. On day 0 (day oftransplantation), an intravenous dose of10–20 mg kg-1 cyclosporin (Sandimmun®; Novartis Pharma AG, Basel, Switzerland) concentrate for infusion (250 mg 5 ml-1) was administered at the end of surgery. From day 15 onwards, the cyclosporin dose was reduced to 100 mg kg-1 orally once daily.
Renal allografts were implanted after either 24 h (n=15) or 48 h (n=9) cold ischaemia times, with preservation in University of Wisconsin solution. Ultrasound examinations and ultrasound guided biopsies were performed every 2 weeks post-operatively starting on day 7, up to a maximum of 100 days. A Siemens Elegra ultrasound unit (Erlangen, Germany) with a 7.5 MHz linear transducer was used for all examinations. Findings were recorded in real-time and were additionally stored on magnetic optical disk. All monkeys were held at our institute, in accordance with Swiss animal welfare regulations, where they were maintained in group housing. The monkeys were 3.5–5 years old and had normal haematology, serum chemistry and urinalysis results as well as negative tuberculosis, viral serology (herpes B, STLV, SIV, SRV type D, hepatitis B), salmonella/shigella and faecal parasite results.
Animals were sacrificed if complete absence of cortical flow was detected with power Doppler ultrasound or if vasculopathies of chronic rejection were present in the biopsy histology.
Ultrasound examination
Four ultrasound parameters were recorded from each examination: (1) per cent increase in graft volume (from the original volume of the donor kidney); (2) cortical thickness in mm; (3)resistive index (RI) of the arcuate artery;and (4) power Doppler (PD) score. For the volume estimations, the left and right kidneys were imaged in the 2D grey scale image in three orthogonal planes. The longitudinal image of the kidney was used when maximum bipolar length was obtained. This view was achieved through a ventrolateral to dorsomedial approach to the kidney, with the transducer positioned just below the last rib. From this view, length and ventral-to-dorsal height measurements at the level of the hilus were obtained. The width measurement was made by rotating the ultrasound head 90° to achieve the transverse view at the level of the renal hilus [13–15]. To estimate cortical thickness, the longitudinal plane in the 2D grey scale image was used to measure the thickness (in mm) of the far cortex midway between the hilus and cranial pole with electronic calipers. Calipers were placed at the outer renal cortical surface and at the corticomedullary junction for this measurement.
Upper limits for each parameter were assigned based on findings in normal renal allografts from a previous study [16]. In that study, normal renal allografts in the cynomolgus monkey with no rejection histologically and a serum creatinine value of less than 200 µmol l-1 had volume increases to a size less than 250% that in the donor, had less than 7 mm cortical thickness, a RI <0.8 and PD scores of 3. An ultrasound score was then assigned to each graft based on the number of abnormalities present (any value above the upper limits); a normal graft (no abnormalities) received a score of 1, and 1 point was added to the base score of 1 for each of the four parameters described above if they were found to be abnormal. A graft with all four abnormal findings would therefore receive a score 5.
Power Doppler
Power gain was optimized by wiping the ultrasound head clean, setting the background to blue and increasing the colour gain until the colour box on the screen was almost uniformly filled with the lowest level of colour [17]. The resulting machine setting was 60 dB. Additionally, a pulse repetition frequency of 880 Hz was used for all examinations. All allografts were located in the right lower abdomen, directly beneath and in contact with the skin. Any attenuation of the power Doppler signal owing to the distance between the graft and the transducer was considered to be constant for all examinations. A PD score (Figure 1
) was assigned based on subjective evaluation as follows: PD=3 (homogeneous or near homogeneous blush of cortical flow extending to the periphery of the cortex); PD=2 (decreased or patchy flow with lack of cortical blushing); and PD=1 (absence of flow in the cortex) (Figure 1).
View larger version (87K):
[in this window]
[in a new window]
|
Figure 1. (a–c) Longitudinal power Doppler images of renal allografts using a dark violet noise floor background colour. Arrows mark the limits of the renal cortex between the corticomedullary junction and the outer border of the kidney in all three images. (a) Power Doppler (PD) score=3. Note the "blush" of flow uniformly filling the renal cortex. (b) PD=2. Rather than a blush of colour, individual vessels and spaces between vessels are easily visible. (c) PD=1. Complete absence of cortical flow is shown in this graft.
|
|
Allograft biopsy and histopathology
Percutaneous ultrasound guided biopsy of the renal allografts was obtained using a Bard® MagnumTM (Covington, GA) biopsy instrument with an 18 G core tissue biopsy needle of 10 cm in length, a 1.9 cm long sample notch and a penetration depth of 2.2 cm. All biopsies were performed using a routine anaesthesia protocol. Anaesthesia was induced with 10 mg kg-1 im ketamine hydrochloride (Narketan®10; Chassot Inc., Bern, Switzerland) after prior (15–60 min) pre-medication with atropine (0.05 mg kg-1 im), which was followed by intubation and administration of O2:N2O at a ratio of 1:1. Isoflurane at 0.2% was administered if muscle relaxation was inadequate.
The biopsy sample was placed in 4% buffered formalin and embedded in paraffin; 3.5 µm sections were made and stained with haematoxylin and eosin as well as periodic-acid Schiff (PAS) for histopathological evaluation. Samples obtained at necropsy were processed in a similar way. In each case, the degree of cellular rejection (absent, marginal, slight, moderate or severe), the presence of tissue damage or degeneration, and the presence of vasculopathy in the form of neointima formation in the renal cortical arterioles was recorded.
Statistical analysis
The ultrasound volume estimations of renal allografts made on the day of euthanasia were compared with the graft weight (g) by calculation of the Pearson correlation coefficient. The t-test was used to test for differences in graft volumes calculated with ultrasound between 24 h and 48 h ischaemia grafts as well as between allografts with and without vasculopathies.
|
Results
|
|---|
Four animals had to be sacrificed on day 3 post-operatively owing to loss of their grafts (two with 24 h ischaemia and two with 48 h ischaemia times) because of severe tissue damage and degeneration as well as severe acute tubular necrosis (ATN). Of the remaining 20 monkeys, 13 developed neointimal formation in the allograft cortex compatible with vasculopathies seen in chronic rejection in clinical transplantation. ATN was detected on day 7 in five animals in addition to those stated above.
Of the 13 allografts that developed vasculopathies, 11 were in the 24 h ischaemia group and 2 were in the 48 h group of animals (Tables 1 and 2). 10 of these allografts with vasculopathies had PD=1 scores (absence of cortical flow), which occcurred in only one instance in the group of allografts without vasculopathies and was due to necrosis (animal 20). Allografts without vasculopathies otherwise had either PD=3 scores (normal flow; n=2) or PD=2 scores (reduced flow; n=4). In contrast, PD=3 (n=2) and PD=2 (n=1) scores were infrequent findings when vasculopathies were diagnosed.
View this table:
[in this window]
[in a new window]
|
Table 1. Power Doppler (PD) and ultrasound (US) scores of 13 allografts that developed vasculopathies. Box marks the examination prior to and at the time vasculopathies were detected histologically 2 weeks later
|
|
In addition, allografts with PD=1 scores had more ultrasound grey scale abnormalities as revealed by a higher ultrasound score. 8 of the 11 grafts with PD=1 had ultrasound scores of 4 or 5. Only three grafts with PD=1 had ultrasound scores of 2–3, which was the predominant finding in the allografts without vasculopathies. We also observed that the development of vasculopathy was signalled by a decrease in the PD score. 11/13 grafts that developed vasculopathies had lower or absent cortical flow (PD scores of 2 or 1), whereas they had been normal in the weeks prior to the development of the vasculopathy. In six of these, the power Doppler score changed within 2 weeks. Grafts that did not develop vasculopathies did not show the same severity of reduction in cortical flow.
In Tables 1 and 2
, the finding of ischaemic/reperfusion damage includes the histological diagnosis of ATN. ATN occurred at an incidence of 44% in grafts having undergone 48 h cold ischaemia times (four out of nine animals including the two that were sacrificed early post-operatively) and in 23% (3 out of 13) of those with 24 h cold ischaemia times. However, only 23% of grafts that developed vasculopathies histologically had ATN occurring post-operatively.
To determine what other factors may have played a role in the differences shown for the PD and ultrasound scores between grafts with and without vasculopathies, we examined whether the degree of cellular rejection and graft size also differed. Figure 2 shows that in vivo volume estimations obtained from ultrasound correlated significantly with actual graft weight at necropsy (Pearson correlation coefficient r=0.91, p<0.001). Figure 3 shows that allografts that underwent 48 h cold ischaemia times were smaller than those subjected to 24 h cold ischaemia times (significantly at weeks 5–11, p<0.05). However, no significant decrease in size in either group could be shown in the same time period. Statistical comparisons of graft sizes at weeks 12 and 14 were not possible owing to the low number of survivors at these time points. However, decreases in graft volume occuring in grafts with vasculopathies did not differ from those without vasculopathies (Figures 4 and 5). Some grafts in both groups decreased in size with time, but others were found to increase. To determine a reason for this finding, the histological diagnosis, specifically the degree of cellular rejection, was compared with the estimated volume from ultrasound for each graft at each time point (Figure 6). Grafts with marginal, slight or no cellular rejection could not be differentiated from each another based on graft volume. However, grafts with moderate and severe increases were significantly larger than those with marginal, slight or no cellular rejection. Additionally, grafts with PD=1 scores showed tissue damage and degeneration as well as cellular rejection in 10/11 cases, and this was seen only infrequently in all other cases.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 2. Comparison of in vivo ultrasound volume estimations of all renal allografts in this study with graft weight, both performed on the day of necropsy. Estimations of graft size with ultrasound correlated significantly (Pearson correlation coefficient r=0.91, p<0.001) with the actual weight of the kidney, showing that such estimations with ultrasound closely represent size changes of the graft and can be used invivo to monitor graft size.
|
|
View larger version (10K):
[in this window]
[in a new window]
|
Figure 6. Volume increases compared with histological degree of cellular rejection. Moderate and severe cellular rejection in allografts was associated with significantly larger volume increases (t-test, *p<0.001) compared with grafts with no rejection as well as those with marginal and slight rejection. No significant differences in volume could be shown between grafts with slight, marginal or no rejection.
|
|
|
Discussion
|
|---|
In this study we show that reduced renal cortical flow detected by power Doppler ultrasound during serial examination of non-life-supporting renal allografts in primates is highly supportive of a diagnosis of graft vasculopathy due to arteriolar intimal proliferation. This illustrates that this non-invasive graft imaging modality is an excellent method of monitoring changes in cortical perfusion in allografts in animal models. Power Doppler signals from the renal cortex appear moderately or severely reduced during the weeks immediately prior to the histological detection of vasculopathies in allografts that are known to already have ongoing degrees of cellular rejection based on serial histology from biopsy samples. Power Doppler findings of reduced cortical flow may have useful applications in monitoring human renal allografts with ultrasound and may indicate reduced vascularization and chronic parenchymal damage.
In addition to the power Doppler findings, 2D ultrasound grey scale imaging showing graft enlargement averaging more than 300% of the size of the donor kidney indicates the presence of moderate to severe degrees of cellular rejection. The combination of findings of reduced or absent cortical flow together with marked graft enlargement is highly suggestive not only of the presence of vasculopathies but also of tissue damage and degeneration. Tissue degeneration in these instances is most likely secondary to cortical ischemia resulting from narrowing of the arteriolar lumen as a consequence of the presence of intimal proliferation [18]. The pathogenesis of chronic renal allograft rejection is multifactorial and factors such as early acute rejection episodes and ischaemia–reperfusion injury have been implicated as possible mechanisms [1, 18]. Early immunological injury and ischaemia–reperfusion injury are thought to lead to tissue destruction of grafts and are the most likely mechanisms that led to the similar findings of tissue damage and degeneration in allografts in this study. Allografts in this primate model of chronic rejection did not decrease in size, as expected from previous such findings in a rat model of chronic rejection. In that model, allografts with vasculopathies also showed severe tissue damage as well as various degrees of cellular rejection in addition to reduction in graft volume and native kidney hypertrophy. The reduction in graft volume in the rat allografts was seen, however, only in the longer experimental endpoints. Lesions such as interstitial oedema concurrent with tissue degeneration may certainly be responsible for enlargement of renal allografts undergoing rejection, but because it is rather difficult to quantify, statistical comparisons were not made.
Serial, non-invasive monitoring with ultrasound is especially valuable in the animal transplantation model that includes only unilateral nephrectomy, since organ function parameters such as serum creatinine and glomerular filtration rate are of no value in providing information on the condition of the allograft. Blood flow assessment by power Doppler ultrasound is very important for the non-invasive characterization of grafts, indicating that perfusion-altering pathology is present. Allografts with reduced flow and or enlargement can be recognized immediately and biopsies for histological examination may be made. Power Doppler ultrasound, by detecting moderate and severe blood flow deficits in the renal allograft cortex, could enable early intervention or follow-up in the pre-clinical setting in animal models potentially in the form of therapeutic trials or simply in determining termination points for organ harvesting prior to loss of the allograft owing to ischaemia and necrosis.
Received for publication September 21, 2000.
Revision received January 24, 2001.
Accepted for publication January 31, 2001.
|
References
|
|---|
-
Cook NS, Zerwes H-G, Rudin M, Beckmann N, Schuurman H-J. Chronic graft loss: dealing with the vascular alterations in solid organ transplantation. Transplant Proc 1998;30:2413–8.[Medline]
-
Schuurman H-J, Beckmann N, Briner U, et al. Chronic rejection of a rat kidney allograft: value of magnetic resonance imaging to assess the graft status and the effect of the somatostatin analog SMS 201-995. Transplant Proc 1996;28:3272–5.[Medline]
-
Beckman N, Joergensen J, Bruttel K, Rudin M, Schuurman H-J. Magnetic resonance imaging for the evaluation of rejection of a kidney allograft in the rat. Transpl Int 1996;9:175–83.[Medline]
-
Gaschen L, Schuuman H, Bruttel K, Tanner M, Beckmann N. MRI and ultrasonographic detection of morphologic and hemodynamic changes in chronic renal allograft rejection in the rat. J Magn Reson Imaging 2001;13:232–41.[Medline]
-
Helenon O, et al. Renal vascular Doppler imaging: clinical benefits of power mode. Radiographics 1998;18:1441–54.[Abstract]
-
Hilborn MD, Bude RO, Murphy KJ, Platt JF, Rubin JM. Renal transplant evaluation with power Doppler sonography. Br J Radiol 1997;70:39–42.[Abstract]
-
Bude RO, Rubin JM, Adler RS. Power versus conventional color Doppler sonography: comparison in the depiction of normal intrarenal vasculature. Radiology 1994;190:853–6.[Abstract/Free Full Text]
-
Trillaud H, Mervill P, Tran Le Linh P, et al. Color Doppler sonography in early renal transplantation follow-up: resistive index measurements versus power Doppler sonography. AJR 1998;171:1611–5.[Abstract/Free Full Text]
-
Sidhu MK, Gambhir S, Jeffrey RB Jr, et al. Power Doppler imaging of acute renal transplant rejection. J Clin Ultrasound 1999;27:171–5.[Medline]
-
Lu MD, Yin XY, Wan GS, Xie XY. Quantitative assessment of power Doppler mapping in the detection of renal allograft complications. J Clin Ultrasound 1999;27:319–23.[Medline]
-
Venz S, Kahl A, Hierholzer J, et al. Contribution of color and power Doppler sonography to the differential diagnosis of acute and chronic rejection, and tacrolimus nephrotoxicity in renal allografts. Transpl Int 1999;12:127–34.[Medline]
-
Deane C. Doppler and color Doppler ultrasonography in renal transplants: chronic rejection. J Clin Ultrasound 1992;20:539–44.[Medline]
-
Barr FJ, Holt PE, Gibbs C. Ultrasonographic measurement of normal renal parameters. J Small Anim Pract 1990;31:180–4.
-
Boag BL, Atilola M, Pennock P. Renal sonographic measurements in the dog preceding and following unilateral nephrectomy. Vet Radiol Ultrasound 1993;34:112–7.
-
Nyland TG, Mattoon JS, Wisner ER. Ultrasonography of the urinary tract and adrenal glands. In: Nyland TG, Mattoon JS, editors. Veterinary diagnostic ultrasound. Philadelphia, PA: WB Saunders, 1995:95–111.
-
Gaschen L, Menninger K, Schuuman H. Ultrasonography of the normal kidney in the cynomolgus monkey (Macaca fascicularis): morphologic and Doppler findings. J Med Primatol 2000;29:76–84.[Medline]
-
Bude RO, Rubin JM. Power Doppler sonography. Radiology 1996;200:21–3.[Free Full Text]
-
Laine J, Holmberg C, Häyry P. Chronic rejection and late renal allograft dysfunction. Pediatr Nephrol 1996;10:221–9.[Medline]
Top
Abstract
Introduction
) vs 48 h ischaemia (
) allografts during the postoperative period (bars indicate standard deviation). A single animal, which had a large graft, represents week 14 in the 24 h group. Otherwise, allografts with 48 h ischaemia times were significantly smaller than those with 24 h ischaemia times at weeks 5–12 (*p<0.05). Too few animals survived at weeks 12 and 14 time points to make statistical comparison.
