| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Unité Mixte de Recherche 5639, Centre National de la Recherche Scientifique (CNRS), Magnetic Resonance of Biologic Systems (O.H., C. Delalande, H.T., N.G.), and Hydrology and Environment Laboratory (C.O.), Université Victor Ségalen, 146 rue Léo Saignat, 33 076 Bordeaux, France; the Departments of Anatomic Pathology (C. Deminière) and Urology (S.G.), Hôpital Pellegrin, Bordeaux, France; the Unité Institut National de la Santé et de la Recherche Médicale (INSERM) U489, Paris, France (B.F.); and the Unité INSERM U441, Pessac, France (C.C.). Received September 17, 1999; revision requested November 2; revision received February 22, 2000; accepted February 23. Supported by the Société Française de Radiologie. Address correspondence to N.G. (e-mail: nicolas.grenier@chu-bordeaux.fr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Two experimental rat models of nephropathies were studied: a model of nephrotoxic nephritis induced by means of intravenous injection of sheep anti–rat glomerular basement membrane serum (n = 43) and a model of obstructive nephropathy (n = 6). Imaging sessions were performed with a spectrometer operating at 4.7 T with fast low-angle shot, or FLASH, sequences. Signal intensity was measured in each kidney compartment before and 24 hours after intravenous injection of USPIO (90 µmol of iron per kilogram of body weight). MR findings were compared with histologic data and urine protein levels.
RESULTS: In the nephrotoxic nephritis model 24 hours after injection of USPIO, a significant signal intensity decrease (P < .05) was present only in the cortex where the glomerular lesions were located. In the obstructive nephropathy model, the signal intensity decrease (P < .05) was located in all kidney compartments in response to diffuse interstitial lesions. The decrease in signal intensity was due to iron uptake by either macrophages or mesangial cells gaining endocytic activity and was correlated, in the nephrotoxic nephritis model, to the degree of proteinuria.
CONCLUSION: Twenty-four-hour delayed USPIO-enhanced MR imaging may help identify and differentiate various types of nephropathies.
Index terms: Hydronephrosis, 81.84 • Iron • Kidney, MR, 81.121412, 81.12143 • Magnetic resonance (MR), contrast media, 81.12143 • Nephritis, 81.69
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Currently, detection of this activity, its quantification, and follow-up during treatment are based on findings at renal biopsy. Identification of intrarenal macrophage infiltration with a noninvasive technique could help characterize the kidney disease and its activity and allow a noninvasive method of follow-up.
Because they are superparamagnetic and because they are taken up by phagocytic cells, iron oxide particles are used as a magnetic resonance (MR) contrast agent for exploration of the mononuclear phagocytic system. When injected intravenously, superparamagnetic iron oxide (SPIO) particles have a short blood half-life and are rapidly taken up by the Kupffer cells in the liver and by the spleen (9,10). Ultrasmall SPIO (USPIO) particles have a longer blood half-life, which allows peripheral uptake by macrophagic cells in the lymph nodes (11–13), bone marrow (14,15), and other diseased tissues. Such a pathologic tissue accumulation has been shown in tumor cells adjacent to vessels in experimental gliomas (16) and in myelin of the white substance of the brain in experimental encephalitis (17). In a previous study with use of USPIO in a nephropathic model in rats, MR imaging helped identification of diffuse macrophagic infiltration induced after injection of puromycin aminonucleoside (18).
The aim of this study was to evaluate the role of MR imaging enhanced with USPIO in the identification and differentiation of glomerular versus interstitial macrophagic infiltration in rats on the basis of two experimental models. The first model was the anti–rat glomerular basement membrane (GBM) GN, equivalent to Goodpasture syndrome in humans, that is known to induce specifically glomerular lesions and glomerular macrophagic infiltration (19,20). The second model was experimental obstructive nephropathy, which is known to induce diffuse interstitial lesions with macrophagic infiltration in all kidney compartments (6).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Nephrotoxic anti–rat-GBM GN.—This study involved 43 Sprague-Dawley male rats (150–250 g) divided into six groups: four pathologic groups (in which the nephropathy was induced, with general anesthesia, by intravenously injecting the tail vein with 500 µL of sheep anti–rat-GBM serum) and two control groups (in which isotonic saline solution alone was administered intravenously with the same conditions). The large number of groups in this model was justified by the complex kinetic of the immunologic response with two characteristic phases: heterologous (day 2) and autologous (day 14). The groups were assigned according to the delay between injection of the anti–rat-GBM serum and USPIO-enhanced MR imaging sessions (day of injection was designated day 0): day 2 (n = 10), day 8 (n = 6), day 14 (n = 10), and day 21 (n = 5) as experimental groups; day 2 (n = 6) and day 14 (n = 6) as control groups. The animals had free access to a standard diet and water throughout the entire experimental period.
Animals were sacrificed after the imaging session, followed by kidney ablation for histologic examination and urine sampling for renal function estimation. Urine sampling was performed by means of direct bladder puncture, and the degree of glomerular dysfunction was estimated on the basis of the urinary protein-to-creatinine ratio. This estimation was necessary to evaluate the response of the kidney to the disease, which is known to be variable from one animal to another.
Obstructive nephropathy.—In one group of six Sprague-Dawley male rats (250 g), acute hydronephrosis was induced: with general anesthesia, a small midline incision allowed unilateral ligation of the left ureter at the junction of the lower and middle thirds. The day of ligation was designated as day 0. Only one group was used in this model because the kinetic of immunologic response is monophasic and because each animal served as its own control. USPIO-enhanced MR imaging of the left kidney (pathologic side) and of the right kidney (control side) was performed on day 4. The animals had free access to a standard diet and water throughout the entire experimental period.
Animals were sacrificed immediately after the imaging session, followed by kidney ablation and histologic study. In this model of unilateral hydronephrosis, urine sampling was not performed because proteinuria is not thought to occur. Furthermore, the effect of the ureteral ligation was also supposed to be visible on renal MR images.
Contrast Agent
The USPIO used in this study was in the form of nanoparticles consisting of a nucleus of iron oxide crystal measuring 4–6 nm in diameter coated with dextran (Sinerem; Guerbet, Aulnay-sous-Bois, France). The overall particle diameter was approximately 20–30 nm. Because of the small size of the particles, USPIO uptake by the Kupffer liver cells is minimal (2%), which gives it a long plasma half-life (118 minutes) in the rat (21). The dose of USPIO (90 µmol/kg) was administered intravenously in the tail vein of each animal.
MR Imaging
Two MR imaging sessions were performed with all rats, including a baseline study before intravenous injection of the USPIO and a second study 24 hours after. According to the pharmacokinetics of these particles, the 24-hour delay should allow clearance of all particles from the vascular space by means of macrophagic activity (11).
All sequences were performed with a spectrometer operating at 4.7 T (47-50 Biospec; Brüker Instruments, Karlsruhe, Germany) and equipped with a superconducting solenoid with a 12-cm horizontal bore. Maximal amplitude of the gradients was 193 mT/m, with a ramp time of 180 µsec from zero to maximal amplitude. Signals were transmitted and received by using a curved-form surface coil adapted to rat morphology and manufactured in our research unit. After induction of general anesthesia by means of intraperitoneal injection of pentobarbital (1 mg/100 g), the animals were placed in the lateral position, with the kidney of interest lying in the center of the coil.
Fast low-angle shot, or FLASH, gradient-echo sequences (repetition time msec/echo time msec = 300/12 with Ernst angle) were used to enhance the T2* effects of USPIO. With four signals acquired per image, matrix size of 256 x 256, field of view of 60 mm, and section thickness of 2 mm, voxel size and acquisition time were 0.1 mm3 and 5 minutes, respectively.
In the first model, the induced nephropathy was bilateral; therefore, only one kidney was imaged. In the second model, both kidneys had to be imaged, which necessitated repositioning of the rat during the same session to allow imaging of the opposite kidney with an additional sequence.
Data Analysis
Both qualitative visual analysis and signal intensity measurement were performed. For the latter in all cases, three regions of interest (0.025 cm2 [five pixels]) were positioned on the three renal compartments (inner medulla, outer medulla, and cortex) by consensus of the two authors (C. Delalande, O.H.) who reviewed the various studies. The results were first expressed as the mean plus or minus SD of the signal intensity (SI) normalized to the noise (N) (SInN): SInN = ([SI - N]/N). Previous studies (22) have shown that muscle tissue remains unchanged by contrast agent; therefore, the signal intensity values were then normalized with reference to the muscle (SInM) to cancel the signal intensity fluctuations related to variations in technical parameters between both imaging sequences. The following formula was used: SInM = SInN x (SIM1/SIM2), where SIM1 and SIM2 are signal intensity in the muscle in the first and second imaging sessions, respectively.
Renal Pathologic Conditions
In both models, each kidney of each animal was removed immediately after sacrifice at the end of the second imaging session and divided into several portions. One portion of renal parenchyma was taken for histologic examination with light microscopy (original magnification, x15–100), after which various staining techniques were performed. Initially, Perl stain was used to detect the presence of iron. Hematoxylin-eosin-saffron staining and immunolabeling with ED1 mouse monoclonal antibodies (Serotec, Oxford, England), specific for rat macrophages, were subsequently performed to demonstrate macrophagic infiltration in the kidney and to determine their distribution in the kidney compartments. A second kidney portion was used for electron microscopy (original magnification, x3,000–30,000) to identify iron particles and determine their exact location in the intracellular or extracellular interstitial spaces. A third portion of the kidney sample was taken in day-2 groups (pathologic and control) with anti–rat-GBM GN for total iron measurement in the cortex and the medulla. This measurement was performed with inductively coupled plasma emission spectroscopy (ICP/AES; Variant, Victoria, Australia) after microwave destruction in acid medium (HCl and HNO3). For this procedure, the medulla was not divided into inner and outer compartments because their limits were difficult to distinguish visually.
Statistical Analysis
We used a paired Student t test to compare each group of both models before and 24 hours after injection of USPIO. Differences with a P value less than .05 were considered significant. When the change in signal intensity was significant, the percentage of signal intensity decrease before (SIB) and after (SIA) injection of USPIO was calculated according to the following formula: [(SIB - SIA)/SIB] x 100. In the nephrotoxic model, signal intensity variation (before and after injection of USPIO) was correlated by means of simple correlation analysis to the degree of glomerular dysfunction estimated on the basis of the protein-to-creatinine ratio. When measured, the total iron level was also correlated to signal intensity variation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Anti–rat-GBM GN, which is the equivalent of Goodpasture syndrome in humans, is characterized by two distinct and different phases. The presence of these two phases made it more suitable for our study than the accelerated model, which has less distinct phases. The first phase, called the heterologous phase, begins shortly after injection of the anti–rat-GBM serum and is characterized by severe hematuria and mild proteinuria. A neutrophil influx is observed at 3 hours and rapidly disappears after 24 hours. No macrophage infiltration is seen at this phase. The second phase, called the autologous phase, begins 7 days after injection, with moderate proteinuria. The number of infiltrating macrophages then increases progressively with a peak at day 14 and then decreases slowly by day 28 (20).
The kinetics of signal intensity follow the biphasic evolution of the disease, with two peaks of signal intensity decrease (27.1% and 35.5%) corresponding to the heterologous and autologous phases of the anti–rat-GBM GN model (day 2 and day 14, respectively). The signal intensity decrease at day 2 is explained by the activation, in the glomeruli, of mesangial cells which acquire an endocytic ability (7,8). The accumulation of macrophages is not expected here before day 7. Therefore, the delayed decrease at day 14 corresponds to an actual infiltration of macrophages. The moderate signal intensity decrease at day 8 could be explained by a decrease in activation of mesangial cells and an insufficient infiltration by macrophages. The absence of significant signal intensity decrease noted at day 21 corresponds to the end of the renal disease, as demonstrated by the absence of iron capture by mesangial cells, absence of macrophage infiltration, and disappearance of proteinuria. The degree of proteinuria, a reflection of glomerular damage, was variable in this model, owing to the fact that some rats responded to the serum injection and others did not. The degree of signal intensity decrease related to macrophagic activity, however, was strongly correlated with the degree of proteinuria on days 2 and 14 (r = .89 and .94, respectively). The weaker correlation in the day-8 group (r = 0.7) compared with that in the day-2 and day-14 groups may be due to a less marked signal intensity decrease and to a smaller number of rats (six vs 10). These findings could have important implications in clinical practice for monitoring the progression of disease and response to therapy.
Concerning the total iron content in the day-2 pathologic and control groups, our results showed a significant increase in cortical iron content in the pathologic kidneys, with a strong correlation between iron levels and signal intensity decrease. More surprisingly in this model of selective cortical involvement, the iron content in the medulla of pathologic kidneys was also significantly higher than that of control kidneys. One explanation could be that the limit between the cortex and the medulla was difficult to delineate visually for selective sampling. Therefore, it is likely that there was a certain amount of cortex in the medulla samples, thus explaining the high level of iron in the medulla in the day-2 pathologic group.
In contrast with the first model, hydronephrosis is known to induce an influx of macrophages, which peaks the 2nd day of the obstruction, that is present in all kidney compartments including the interstitium (6). We compared this model with the first one to determine whether different diseases with different modes and location of endocytic activity could be separated by means of this targeting technique. All obstructed kidneys demonstrated diffuse macrophagic infiltration already visible with light microscopy and hematoxylin-eosin-saffron staining. This high and diffuse macrophagic accumulation was identified with MR imaging with a significant and homogeneous signal intensity decrease in the three compartments. In the previous feasibility study (18), in an experimental model of interstitial nephropathy induced with puromycin aminonucleoside, the signal intensity decrease was also diffuse but predominated slightly in the outer medulla. On the contrary, in the hydronephrosis model, which also induces an interstitial type of macrophagic infiltration, the decrease in signal intensity was slightly more pronounced in the cortex. Therefore, it seems that different types of experimental renal diseases show different and reproducible types of intrarenal capture of USPIO that reflect differences in macrophagic infiltration and endocytic activity.
Detection of renal macrophagic infiltration with MR imaging enhanced with USPIO could be of potential interest in humans because macrophages are frequently present in glomeruli in acute proliferative types of GN and because they also play a role in the development of glomerular inflammation (1,23,24). The accumulation of macrophages in the kidney arises from both the influx of circulating blood monocytes and local proliferation (23,25). The presence of glomerular crescents is a known feature of rapidly progressive GN and is associated with a poor prognosis (26). Accumulation of macrophages in the Bowman space is the primary feature in the development of these advanced cellular crescents. Furthermore, as has already been emphasized, macrophage accumulation in the tubulointerstitium also plays a role in renal injury and represents an important pathway of progressive renal injury and functional renal impairment (24). Thus, precise localization of macrophagic infiltration in kidney compartments may have important prognostic implications.
Considering the evolution with time of the inflammatory kidney process, one practical issue would be the appropriate time in the course of the disease to administer the contrast agent. USPIO should be administered when the disease is discovered, to evaluate the degree of phagocytic activity before treatment. In glomerular diseases, for example, the amount of cellular proliferation versus interstitial fibrosis is a key factor in prognosis with the biopsy specimen. Also, during follow-up of the disease, USPIO injection could be used to evaluate the response of cellular infiltration to therapy. Another practical issue would be determination of the delay between contrast agent administration and the imaging session, which is based on the pharmacokinetics of the product. Plasma half-life is a key factor in uptake of particles. Compared with SPIO, the smaller size (20–30 nm) of USPIO particles explains a lower uptake by Kupffer liver cells (2%) and a longer plasma half-life (120 minutes in rats). According to Weissleder et al (11), in rats, all particles are removed from the vascular space in the reticuloendothelial system after a 24-hour delay. In humans, this delay may increase to 48–72 hours owing to the much longer plasma half-life.
The mechanism of USPIO transport to the kidneys is not perfectly known. Two different mechanisms might be involved: first, and most likely, the particles could be taken up directly from the vascular space of the kidney by macrophages or mesangial cells gaining endocytic activity. According to Moore et al (27), this cellular uptake is mediated by fluid-phase endocytosis. Another possibility could be a cellular uptake of the particles by circulating blood monocytes secondarily recruited into the kidney. Also, little is known about the fate of these macrophages, that is, whether they die in situ or emigrate and how long this takes. There is evidence suggesting migration of activated macrophages into the periglomerular interstitium (28). It is likely that iron is released in the body iron pool at the time of macrophage death.
This study has several limitations. First, this is an animal study performed with rats. Although we chose models of nephropathies close to well-known human nephropathies, it is unclear whether the results in this study will be entirely applicable in humans. Second, the study was performed with a 4.7-T magnet dedicated to research that is not available in current practice. We chose a high-field-strength magnet because this allowed increased spatial resolution (by increasing signal-to-noise ratio) and because the magnetic susceptibility of iron particles increases with higher field strength. However, many studies have already been performed with lower field strength (1.5 T) with good results. Third, the USPIO dose in our study was 90 µmol/kg, which is twice the dose recommended in human lymphography. We used this dose to increase the probability of USPIO uptake by phagocytic cells, because the blood half-life of these particles is shorter in rats than in humans. It is likely that the dose might not be such an important issue in clinical practice, because the blood half-life increases dramatically in humans and a lower dose should be as efficient.
In summary, these results show that MR imaging 24 hours after injection of USPIO particles can depict intrarenal signal intensity variations due to iron ingestion by macrophages or by glomerular cells gaining endocytic activity (ie, mesangial cells) and can help localize this endocytic activity precisely in the different kidney compartments. The results of USPIO labeling are variable, depending on the type of renal disease. Moreover, the degree of renal disease appears to be correlated to the degree of endocytic activity, which may have important implications in clinical practice for the follow-up of patients. The value of this technique must be demonstrated in human native and transplanted kidneys.
Practical application: By helping to identify marked intrarenal macrophagic activity, USPIO-enhanced MR imaging could make it possible to identify some nephropathies such as acute proliferative glomerulonephritides at an acute phase. USPIO-enhanced MR imaging could help differentiate between an early active and reversible stage versus a delayed irreversible fibrotic stage. In renal transplantation, it could help distinguish acute rejection versus other causes of renal dysfunction, because acute rejection produces dramatic accumulation of macrophages as part of the immunologic response (2–4). The effect of this type of study on indications for renal biopsy has to be evaluated.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, O.H., N.G.; study concepts, O.H., N.G.; study design, O.H., C.C., N.G.; definition of intellectual content, O.H., N.G.; literature research, O.H., C.C.; experimental studies, O.H., C. Delalande, C. Deminière, C.O., B.F., S.G.; data acquisition, O.H., C. Deminière, H.T., C. Delalande; data analysis, O.H., C. Deminière, C.O., C. Delalande; statistical analysis, O.H.; manuscript preparation, O.H., C.C., N.G.; manuscript editing, O.H., N.G.; manuscript review, O.H., C.C., N.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. Hauger, E. E. Frost, R. van Heeswijk, C. Deminiere, R. Xue, Y. Delmas, C. Combe, C. T. W. Moonen, N. Grenier, and J. W. M. Bulte MR Evaluation of the Glomerular Homing of Magnetically Labeled Mesenchymal Stem Cells in a Rat Model of Nephropathy Radiology, January 1, 2006; 238(1): 200 - 210. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Lutz, D. Weishaupt, E. Persohn, K. Goepfert, J. Froehlich, B. Sasse, J. Gottschalk, B. Marincek, and A. H. Kaim Imaging of Macrophages in Soft-Tissue Infection in Rats: Relationship between Ultrasmall Superparamagnetic Iron Oxide Dose and MR Signal Characteristics Radiology, March 1, 2005; 234(3): 765 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Lutz, C. Seemayer, C. Corot, R. E. Gay, K. Goepfert, B. A. Michel, B. Marincek, S. Gay, and D. Weishaupt Detection of Synovial Macrophages in an Experimental Rabbit Model of Antigen-induced Arthritis: Ultrasmall Superparamagnetic Iron Oxide-enhanced MR Imaging Radiology, October 1, 2004; 233(1): 149 - 157. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. C. Dagher, S. Herget-Rosenthal, S. G. Ruehm, S.-K. Jo, R. A. Star, R. Agarwal, and B. A. Molitoris Newly Developed Techniques to Study and Diagnose Acute Renal Failure J. Am. Soc. Nephrol., August 1, 2003; 14(8): 2188 - 2198. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. C. PATRY, D. C. TREWICK, M. GREGOIRE, M. A. P. AUDRAIN, A. M. N. MOREAU, J.-Y. MULLER, K. MEFLAH, and V. L. M. ESNAULT Rats Injected with Syngenic Rat Apoptotic Neutrophils Develop Antineutrophil Cytoplasmic Antibodies J. Am. Soc. Nephrol., August 1, 2001; 12(8): 1764 - 1768. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
