(Radiographics. 2002;22:833-846.)
© RSNA, 2002
Atypical Low-Signal-Intensity Renal Parenchyma: Causes and Patterns1
Jun Yong Jeong, MD,
Seung Hyup Kim, MD,
Hak Jong Lee, MD and
Jung Suk Sim, MD
1 From the Department of Radiology and Institute of Radiation Medicine, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea; and the Clinical Research Institute, Seoul National University Hospital. Presented as an education exhibit at the 2000 RSNA scientific assembly. Received June 8, 2001; revision requested July 17; final revision received February 12, 2002; accepted February 28. Supported in part by the 2000 BK21 Project for Medicine, Dentistry, and Pharmacy. Address correspondence to S.H.K. (e-mail: kimsh@radcom.snu.ac.kr).
 |
Abstract
|
|---|
Certain renal diseases manifest as low signal intensity of the renal parenchyma on magnetic resonance images. Sometimes, the appearance is sufficiently characteristic to allow a specific radiologic diagnosis to be made. The causes of this finding can be classified into three main categories on the basis of the pathophysiology: hemolysis, infection, and vascular disease. The first category includes paroxysmal nocturnal hemoglobinuria (PNH), hemosiderin deposition in the renal cortex from mechanical hemolysis, and sickle cell disease. The second category includes hemorrhagic fever with renal syndrome (HFRS). The third category includes acute renal vein thrombosis, renal cortical necrosis, renal arterial infarction, rejection of a transplanted kidney, and acute nonmyoglobinuric renal failure with severe loin pain and patchy renal vasoconstriction. These disease processes have different patterns of low signal intensity. PNH, hemosiderin deposition from mechanical hemolysis, and sickle cell disease involve the entire cortex including the columns of Bertin. HFRS involves the medulla, especially the outer medulla, whereas cortical necrosis involves the inner cortex including the columns of Bertin. In renal vein thrombosis, low-signal-intensity lesions involve the outer medulla, an appearance resembling that of HFRS. Wedge-shaped low-signal-intensity regions involving both the cortex and the medulla are seen in arterial infarction.
© RSNA, 2002
Index Terms: Hemorrhagic fever, 811.659 • Kidney, diseases, 811.65 • Kidney, infarction, 811.77 • Kidney, necrosis, 811.654 • Renal veins, thrombosis, 811.751 • Sickle cell disease (SS, SC), 811.651
|
LEARNING OBJECTIVES FOR TEST 4
|
|---|
After reading this article and taking the test, the reader will be able to:
- List the spectrum of renal diseases that cause low signal intensity on T2-weighted MR images.
- Describe the pathophysiologic mechanism of low signal intensity on T2-weighted MR images.
- Recognize the MR imaging findings of each renal disease that causes low signal intensity.
|
Introduction
|
|---|
Although intravenous urography, ultrasonography (US) with Doppler imaging, and computed tomography (CT) remain the primary imaging modalities for evaluation of the kidney, magnetic resonance (MR) imaging may provide additional information that can lead to a better understanding of the changes in the renal parenchyma in some disease processes.
In normal, well-hydrated kidneys, the renal cortex has a shorter T1 than the medulla and therefore appears more intense than the medulla on T1-weighted MR images (1). The loss of this normal corticomedullary differentiation is a nonspecific finding that may be seen in a variety of renal disorders in which the cortex and the medulla appear isointense (2). Loss of signal intensity of the renal parenchyma on T2-weighted images is unusual but can be a relatively specific feature at MR imaging. Usually, low signal intensity is caused by calcification or ossification, hypercellularity with scant cytoplasm, fibrocollagenous stroma, high-protein material, paramagnetic substances such as melanin, free radicals, and hemorrhage (3). The low signal intensity in most of these conditions is based on the fact that there is little free hydrogen ion in the lesion or on the paramagnetic properties of the deposited substances.
The patterns of signal intensity changes in hemorrhage are complex. Generally, the paramagnetic properties of iron alter the usual signal intensity characteristics of tissue. Recognition of iron within an organ is necessary for accurate differential diagnosis with MR imaging (4,5). When blood extravasates into tissue, oxyhemoglobin molecules within the red blood cells lose oxygen, becoming deoxyhemoglobin. Within a few days, the ferrous (Fe2+) iron in hemoglobin converts into ferric (Fe3+) iron, forming methemoglobin, which cannot carry oxygen. Subsequently, methemoglobin is digested by macrophages and the iron is ultimately stored as hemosiderin. Each of these iron-containing substances has different magnetic properties that affect its appearance on MR images. Deoxyhemoglobin and methemoglobin are paramagnetic, hemosiderin is superparamagnetic, whereas oxyhemoglobin is diamagnetic, having no paramagnetic effect (5–7). They result in signal intensity changes at T1- and T2-weighted MR imaging that can facilitate or complicate diagnosis of renal lesions.
Hemosiderin has paramagnetic effects and shortens predominantly T1 and T2 relaxation times at low and high concentrations, respectively (8,9). At low concentrations, T1 shortening is maximal; it may appear as slightly increased signal intensity on T1-weighted images and decreased signal intensity on T2-weighted images. A higher concentration of hemosiderin dramatically decreases T2. Because T1-weighted images are influenced by changes in T2 as well as in T1, signal intensity can be decreased on T1-weighted images as well. Therefore, the appearance of hemosiderin on T1-weighted images varies; it may appear as minimally increased signal intensity, as loss of corticomedullary differentiation, or as marked low signal intensity, depending on the amount of hemosiderin deposition. T2-weighted images are more sensitive than T1-weighted images for detection of hemosiderin (10–12). Moreover, Tanaka et al (11) concluded that T2*-weighted imaging is superior to T2-weighted imaging in detection of renal cortical hemosiderosis because of the higher magnetic susceptibility effect and shorter examination times.
The diseases that cause low signal intensity of the renal parenchyma on MR images can be divided into three main categories on the basis of the pathophysiology: hemolysis, infection, and vascular disease. The first category includes paroxysmal nocturnal hemoglobinuria (PNH), hemosiderin deposition in the renal cortex by mechanical hemolysis, and sickle cell disease. The second category includes hemorrhagic fever with renal syndrome (HFRS). The third category includes acute renal vein thrombosis, renal cortical necrosis, renal arterial infarction, rejection of a transplanted kidney, and acute nonmyoglobinuric renal failure with severe loin pain and patchy renal vasoconstriction.
Familiarity with the spectrum of findings in the renal parenchyma resulting from hemorrhage and iron deposition is necessary for diagnosis and better understanding of the pathogenesis of these renal diseases (13).
We have obtained spin-echo or turbo spin-echo MR images (SPECTRO-20000 2.0 T [Goldstar Medical Systems, Seoul, Korea] or Signa 1.5 T [GE Medical Systems, Milwaukee, Wis]). The repetition time (msec)/echo time (msec) value was 500–700/20–30 for T1-weighted images and 2,000–2,500/60–85 for T2-weighted images. Gadopentetate dimeglumine (Magnevist [Schering, Berlin, Germany]; 0.1 mmol/kg) was used as a contrast medium for enhanced T1-weighted images.
|
Low Signal Intensity Caused by Hemolysis
|
|---|
Paroxysmal Nocturnal Hemoglobinuria
PNH is an acquired myelodysplastic, hematopoietic stem cell disorder caused by an increased sensitivity to complement, resulting in acute or chronic intravascular hemolysis. Factors that have been reported to precipitate hemolysis include infections, reactions to drugs and immunizations, transfusion, surgery, and exercise; frequently, the cause remains unknown (10,14). Major clinical features include hemoglobinuria, iron deficiency anemia, bleeding, and venous thrombosis (15). PNH can cause chronic renal failure as a result of microinfarctions due to repeated episodes of microvascular thrombosis rather than to the tubular damage from hemosiderin deposition (14) or a direct nephrotoxic effect of iron (16). The disease occurs predominantly in adults and is rare in children and adolescents, in whom the prognosis is worse. The typical histologic finding in the kidney is deposition of hemosiderin in epithelial cells of the proximal convoluted tubules in the renal cortex (Fig 1d) (14).
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. PNH. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla, causing reversed corticomedullary differentiation. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the signal intensity of the renal medulla is normal, although that of the cortex remains low. Note the abnormal corticomedullary differentiation. (c) Nonenhanced CT scan shows slight increased attenuation of the renal cortex compared with that of the medulla, making corticomedullary differentiation possible. (d) High-power photomicrograph (original magnification, x400; hematoxylin-eosin stain) of a biopsy specimen from the kidney shows numerous brownish hemosiderin granules scattered in the tubular epithelial cells.
|
|
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. PNH. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla, causing reversed corticomedullary differentiation. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the signal intensity of the renal medulla is normal, although that of the cortex remains low. Note the abnormal corticomedullary differentiation. (c) Nonenhanced CT scan shows slight increased attenuation of the renal cortex compared with that of the medulla, making corticomedullary differentiation possible. (d) High-power photomicrograph (original magnification, x400; hematoxylin-eosin stain) of a biopsy specimen from the kidney shows numerous brownish hemosiderin granules scattered in the tubular epithelial cells.
|
|
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1c. PNH. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla, causing reversed corticomedullary differentiation. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the signal intensity of the renal medulla is normal, although that of the cortex remains low. Note the abnormal corticomedullary differentiation. (c) Nonenhanced CT scan shows slight increased attenuation of the renal cortex compared with that of the medulla, making corticomedullary differentiation possible. (d) High-power photomicrograph (original magnification, x400; hematoxylin-eosin stain) of a biopsy specimen from the kidney shows numerous brownish hemosiderin granules scattered in the tubular epithelial cells.
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1d. PNH. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla, causing reversed corticomedullary differentiation. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the signal intensity of the renal medulla is normal, although that of the cortex remains low. Note the abnormal corticomedullary differentiation. (c) Nonenhanced CT scan shows slight increased attenuation of the renal cortex compared with that of the medulla, making corticomedullary differentiation possible. (d) High-power photomicrograph (original magnification, x400; hematoxylin-eosin stain) of a biopsy specimen from the kidney shows numerous brownish hemosiderin granules scattered in the tubular epithelial cells.
|
|
MR imaging is the best imaging modality for demonstrating hemosiderin deposition in the renal cortex (12). Visual comparison of the signal intensity of lesions with that of the paraspinous musculature has been reported to be useful in assessment of parenchymal iron deposition in other diseases (17). The MR imaging findings of PNH include low signal intensity in the renal cortex not only with T1- and T2-weighted sequences (Fig 1a, Fig 1b) but also with gradient-echo and inversion-recovery sequences (10). This is due to deposition of hemosiderin in the epithelial cells of the proximal convoluted tubules in the renal cortex (Fig 1d) (12). This low cortical signal intensity is more conspicuous on T2-weighted images due to the adjacent high-signal-intensity medulla, which is not affected in PNH and therefore has a normal high-signal-intensity appearance. Theoretically, gradient-echo imaging has superior sensitivity to spin-echo imaging in detection of hemosiderin because it uses magnetic gradients to refocus the transverse magnetization. Gradient-echo imaging is also faster (11). At nonenhanced CT, higher attenuation of the renal cortex can be seen, but it is less conspicuous than at MR imaging (Fig 1c).
Unlike hemolytic disorders such as sickle cell disease, in which iron from extravascular hemolysis is mainly deposited in the liver and spleen, PNH and mechanical damage from heart valves result in intravascular hemolysis sufficient to saturate the plasma haptoglobin with iron. The hemoglobin passes through the glomerulus and is reabsorbed in the proximal tubules and stored as hemosiderin. Intravascular hemolysis does not result in hemosiderin accumulation elsewhere in the body (18).
In contrast to other hemolytic anemias such as sickle cell disease, autoimmune hemolytic anemia, and hereditary spherocytosis, which are characterized by heavy deposition of iron in the liver and spleen, PNH usually demonstrates normal iron concentration in the liver and spleen (although patients with PNH who have received multiple transfusions may have slightly increased hepatosplenic levels of iron). As a consequence, decreased signal intensity within the liver and spleen occurs less often in patients with PNH than in conditions such as hemochromatosis and transfusion hemosiderosis or as a consequence of hepatic or portal vein thrombosis or long-standing biliary cirrhosis (13). In the absence of cortical calcification, diffuse low signal intensity in the renal cortex indicates intravascular hemolysis rather than systemic iron overload. When transfusional siderosis is absent, low signal intensity confined to the kidney suggests PNH, whereas low signal intensity limited to the spleen and liver is characteristic of transfusional hemosiderosis and hemochromatosis (19,20).
The MR imaging findings of PNH in the kidney are almost opposite to those seen in HFRS, in which medullary hemorrhage results in low signal intensity of the medulla (21).
Hemosiderin Deposition by Mechanical Hemolysis
Although hemosiderin deposition in the renal cortex by mechanical hemolysis due to a malfunctioning prosthetic cardiac valve has a different pathogenesis from PNH, the pattern of renal iron deposition and the MR imaging findings are identical (Fig 2).
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Intravascular hemolysis due to a prosthetic cardiac valve. (a) Axial T1-weighted MR image (550/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the renal cortex has lower signal intensity than the medulla. (Reprinted, with permission, from reference 18.)
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Intravascular hemolysis due to a prosthetic cardiac valve. (a) Axial T1-weighted MR image (550/30) obtained at 2.0 T shows that the signal intensity of the renal cortex is lower than that of the medulla. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows that the renal cortex has lower signal intensity than the medulla. (Reprinted, with permission, from reference 18.)
|
|
Lee et al (18) reported that the signal intensity of the renal cortex was much lower than that of the medulla on T2-weighted images due to deposition of hemosiderin in proximal convoluted tubules by intravascular hemolysis. Although the incidence is low, mechanical hemolysis due to a malfunctioning prosthetic cardiac valve should be included in the differential diagnosis as a possible cause of low signal intensity of the renal cortex on T1- and T2-weighted MR images. At MR imaging, PNH is indistinguishable from mechanical hemolysis (Figs 1, 2) (18). Similarly, any other cause of intravascular hemolysis such as hemolytic anemia can cause low signal intensity within the renal cortex due to hemosiderin deposition (22).
Sickle Cell Disease
Sickle cell disease is a common hereditary disorder in African-Americans. Although sickle cell disease is primarily a hematologic disorder, several renal syndromes have been referred to as sickle cell nephropathy (23).
Sickle cell nephropathy is known to involve all parts of the kidney parenchyma (24). Cortical changes include dilatation and engorgement of glomerular capillary tufts, glomerular sclerosis, increased mesangial matrix, and iron deposition in the glomerular epithelium secondary to sickled erythrocytes (24). Although engorgement and dilatation of capillaries and edema and focal scarring of the papillary stroma associated with dilatation and congestion of peritubular and pelvic mucosal capillaries are the pathologic changes in the renal medulla, they are not likely to be observed on MR images (24).
Patients with mild sickle cell disease who have not undergone transfusion do not develop iron overload in the renal cortex because extravascular hemolysis in the reticuloendothelial system (eg, the spleen) by defective red blood cells is the main pathophysiology (13,19). However, acute hemolytic crises or intravascular hemolysis contributes up to one-third of the hemolysis in sickle cell disease and causes hemosiderin and ferritin storage in the renal proximal convoluted tubule (13,19). This explains the low signal intensity of the renal cortex on T2-weighted MR images (24).
The MR imaging findings of sickle cell disease are the same as those of PNH and mechanical damage from heart valves, although in mild cases they will not be seen (13,19,24). Differentiating among the various causes of intravascular hemolysis is impossible with MR imaging (Figs 1, 2) (13). On T2-weighted MR images, low signal intensity of the spleen due to extravascular hemolysis is noted in all sickle cell disease patients irrespective of their transfusion history (19). However, low signal intensity of the liver or pancreas can be seen only in patients who have received multiple blood transfusions (19).
Renal failure in sickle cell disease is not dependent on iron deposition. When renal failure develops in sickle cell disease, involvement of the renal glomeruli, distal tubules, or renal medulla and renal infection are more likely causes (19).
|
Low Signal Intensity Caused by Infection
|
|---|
HFRS, formerly referred to as Korean hemorrhagic fever, is an acute infectious, rodent-transmitted disease caused by the genus Hantavirus and clinically characterized by fever, visceral hemorrhage, and a variable degree of renal failure. The Seoul virus and the Hantaan virus are the most important etiologic agents of HFRS in Korea (25). HFRS occurs primarily in Asia and Europe and is a major public health problem in Far Eastern Asia, especially in Korea, where the mortality rate is 3%–5% (26).
There are five relatively well-defined, successive clinical stages: febrile, hypotensive, oliguric, diuretic, and convalescent. The renal changes begin in the hypotensive or oliguric phase and persist during the diuretic phase (27).
The central derangement of HFRS seems to be vascular dysfunction (28), but the exact pathophysiologic mechanism is not understood. The most prominent pathologic features of HFRS are congestion and hemorrhage with hemorrhage, hemosiderin formation, and cellular reaction in the renal medulla, right atrium of the heart, and anterior lobe of the pituitary gland (Fig 3c, Fig 3d) (29,30).
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. Images of an isolated kidney specimen from a 21-year-old man who died of HFRS. (a) T1-weighted MR image (500/30) shows that the low-signal-intensity medulla is clearly differentiated from the cortex. (b) T2-weighted MR image (2,000/80) shows low signal intensity of the whole medulla, resulting in marked abnormal corticomedullary contrast. (c) Photograph of the pathologic specimen shows severe hemorrhage and congestion confined to the medulla with an intact cortex. (d) Photomicrograph (original magnification, x100; hematoxylin-eosin stain) shows hemorrhage and intertubular congestion in the medulla, which are the typical microscopic findings of HFRS. Note that there is no hemorrhage around the glomeruli and proximal tubules. (Reprinted, with permission, from reference 32.)
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. Images of an isolated kidney specimen from a 21-year-old man who died of HFRS. (a) T1-weighted MR image (500/30) shows that the low-signal-intensity medulla is clearly differentiated from the cortex. (b) T2-weighted MR image (2,000/80) shows low signal intensity of the whole medulla, resulting in marked abnormal corticomedullary contrast. (c) Photograph of the pathologic specimen shows severe hemorrhage and congestion confined to the medulla with an intact cortex. (d) Photomicrograph (original magnification, x100; hematoxylin-eosin stain) shows hemorrhage and intertubular congestion in the medulla, which are the typical microscopic findings of HFRS. Note that there is no hemorrhage around the glomeruli and proximal tubules. (Reprinted, with permission, from reference 32.)
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. Images of an isolated kidney specimen from a 21-year-old man who died of HFRS. (a) T1-weighted MR image (500/30) shows that the low-signal-intensity medulla is clearly differentiated from the cortex. (b) T2-weighted MR image (2,000/80) shows low signal intensity of the whole medulla, resulting in marked abnormal corticomedullary contrast. (c) Photograph of the pathologic specimen shows severe hemorrhage and congestion confined to the medulla with an intact cortex. (d) Photomicrograph (original magnification, x100; hematoxylin-eosin stain) shows hemorrhage and intertubular congestion in the medulla, which are the typical microscopic findings of HFRS. Note that there is no hemorrhage around the glomeruli and proximal tubules. (Reprinted, with permission, from reference 32.)
|
|
View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3d. Images of an isolated kidney specimen from a 21-year-old man who died of HFRS. (a) T1-weighted MR image (500/30) shows that the low-signal-intensity medulla is clearly differentiated from the cortex. (b) T2-weighted MR image (2,000/80) shows low signal intensity of the whole medulla, resulting in marked abnormal corticomedullary contrast. (c) Photograph of the pathologic specimen shows severe hemorrhage and congestion confined to the medulla with an intact cortex. (d) Photomicrograph (original magnification, x100; hematoxylin-eosin stain) shows hemorrhage and intertubular congestion in the medulla, which are the typical microscopic findings of HFRS. Note that there is no hemorrhage around the glomeruli and proximal tubules. (Reprinted, with permission, from reference 32.)
|
|
The congestion and hemorrhage are most prominent in the outer or subcortical zone of the medulla and extend deep into the medullary tissues (Figs 3, 4) (30). Microscopic features of the kidney in HFRS are renal swelling, intense congestion and hemorrhage in the renal medulla, and acute tubular necrosis (Fig 3d) (27,31).
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Convalescent phase HFRS in a 46-year-old man. (a) Axial T1-weighted MR image (550/30) obtained at 2.0 T shows low signal intensity in the renal medulla and higher signal intensity in the cortex. The low-signal-intensity line (arrows) along the lateral margin of the right kidney is due to chemical shift artifact. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows nearly the same low signal intensity in the outer renal medulla and high signal intensity in the cortex and inner medulla.
|
|
View larger version (75K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Convalescent phase HFRS in a 46-year-old man. (a) Axial T1-weighted MR image (550/30) obtained at 2.0 T shows low signal intensity in the renal medulla and higher signal intensity in the cortex. The low-signal-intensity line (arrows) along the lateral margin of the right kidney is due to chemical shift artifact. (b) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T shows nearly the same low signal intensity in the outer renal medulla and high signal intensity in the cortex and inner medulla.
|
|
The low signal intensity is noted in the renal medulla, either the outer portion or the entire medulla, irrespective of the clinical phase, on both T1- and T2-weighted images but is more prominent on T2-weighted images (Figs 3, 4) (21). Reversed corticomedullary differentiation on T2-weighted images (bright cortex, dark medulla) with preserved corticomedullary distinction on T1-weighted images is another MR imaging finding (Figs 3a, 4a). These findings are opposite to those seen in PNH (Fig 1), in which cortical hemosiderin deposition results in a diminution or reversal of corticomedullary demarcation on T1-weighted images and an enhancement of abnormal corticomedullary demarcation on T2-weighted images (8,21).
The most prominent finding at MR imaging, a well-defined zone of low signal intensity in the outer medulla, correlates well with the histopathologic findings of HFRS (Fig 3c). The cause of the low signal intensity of the medulla is congestion and hemorrhage of the outer medulla (21). MR images obtained in the oliguric and diuretic phases show various degrees of globular renal swelling; in the convalescent phase, the kidneys revert to their normal shape (21).
At first, deoxyhemoglobin, intracellular methemoglobin, and hemosiderin are the causes of low signal intensity on T2-weighted images (6,21,32). Deoxyhemoglobin is the cause of low signal intensity on T1-weighted images (6,21,32). In addition, after the convalescent phase, persistent low signal intensity in the medulla in HFRS may be caused by intertubular fibrosis (6,21,32). The Seoul variant of the viral infection has less bleeding, vascular changes, and renal derangement but more severe liver dysfunction than HFRS caused by the Hantaan variant (25). Accordingly, the MR imaging findings of HFRS caused by the Seoul virus may be less conspicuous than those of HFRS caused by the Hantaan virus (25,32).
Although low signal intensity in the outer medulla at MR imaging may be seen in renal parenchymal changes due to renal vein thrombosis (Fig 5) (33), it is a fairly characteristic MR imaging feature of HFRS in the appropriate clinical circumstances. MR imaging can be helpful in diagnosis of HFRS, especially in the early phase of HFRS, when differentiation from other diseases such as leptospirosis, murine typhus, or scrub typhus is difficult (21). However, definitive diagnosis depends on serologic tests. MR imaging can also be used to detect complications of HFRS, including retroperitoneal hemorrhage and rupture of the kidney (21).
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Renal vein thrombosis. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows a swollen right kidney with diminished corticomedullary contrast. (b) Axial T2-weighted MR image (2,500/100) obtained at 2.0 T shows low signal intensity in the outer part of the medulla. (Reprinted, with permission, from reference 33.)
|
|
View larger version (86K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Renal vein thrombosis. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows a swollen right kidney with diminished corticomedullary contrast. (b) Axial T2-weighted MR image (2,500/100) obtained at 2.0 T shows low signal intensity in the outer part of the medulla. (Reprinted, with permission, from reference 33.)
|
|
|
Low Signal Intensity Caused by Vascular Disease
|
|---|
Acute Renal Vein Thrombosis
Renal vein thrombosis is most common in patients with membranous glomerulonephritis. However, there are other predisposing conditions, including membranoproliferative glomerulonephritis, diabetic nephropathy, lupus nephritis, amyloidosis, sarcoidosis, sickle cell anemia, transplantation, malignancy, dehydration, and sepsis; infants of diabetic mothers are also predisposed to renal vein thrombosis. It is also seen in retroperitoneal malignancies, abdominal tumors, trauma, hypercoagulable states, and retrograde extension of a thrombus from the inferior vena cava (34).
Because the duration of the thrombosis is a factor in the effectiveness of therapy, it is important to make the diagnosis and to determine the site of the thrombus as early as possible in the least invasive manner. However, renal vein thrombosis can be difficult to diagnose with less invasive techniques, including excretory urography, US, CT, and radionuclide imaging (35–39).
MR imaging has been suggested as a possible noninvasive diagnostic modality for deep vein thrombosis. MR imaging findings in the kidney with acute renal vein thrombosis include renal swelling, indistinct corticomedullary differentiation at T1-weighted imaging, and prolongation of the T1 and T2 relaxation times of the renal cortex and medulla with resulting low signal intensity (Figs 5, 6) (36,40). There are additional imaging findings, such as obliteration of the renal sinus fat and compression of the renal collecting systems, marked attenuation of the renal veins, multiple venous collateral vessels in the perinephric region, and dilatation of the gonadal vein (35).
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Renal vein thrombosis in a transplanted kidney. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows a swollen transplanted kidney with poorly defined low signal intensity in the medulla. (b) Coronal T1-weighted MR image (500/30) obtained at 2.0 T shows the same features as in a. (c) Axial T2-weighted MR image (2,500/100) obtained at 2.0 T shows prominent areas of low signal intensity in the renal medulla. (Reprinted, with permission, from reference 33.)
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Renal vein thrombosis in a transplanted kidney. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows a swollen transplanted kidney with poorly defined low signal intensity in the medulla. (b) Coronal T1-weighted MR image (500/30) obtained at 2.0 T shows the same features as in a. (c) Axial T2-weighted MR image (2,500/100) obtained at 2.0 T shows prominent areas of low signal intensity in the renal medulla. (Reprinted, with permission, from reference 33.)
|
|
View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Renal vein thrombosis in a transplanted kidney. (a) Axial T1-weighted MR image (500/30) obtained at 2.0 T shows a swollen transplanted kidney with poorly defined low signal intensity in the medulla. (b) Coronal T1-weighted MR image (500/30) obtained at 2.0 T shows the same features as in a. (c) Axial T2-weighted MR image (2,500/100) obtained at 2.0 T shows prominent areas of low signal intensity in the renal medulla. (Reprinted, with permission, from reference 33.)
|
|
In obstruction of the renal vein, the inner part of the medulla is less involved than the cortex and the outer part of the medulla because the outer part of the medulla is especially vulnerable to hypoxic injury (41). A low-signal-intensity band in the outer part of the medulla was noted in patients with acute renal vein thrombosis, similar to the findings in patients with HFRS (Figs 3, 4) (21). The pathologic basis of the low signal intensity in the medulla is nearly the same as that in HFRS (21).
According to the results of an experimental study of MR imaging with ligation of the renal vein to cause acute renal vein thrombosis (33,42,43), on T1-weighted images, the involved kidney enlarged maximally on the third day and returned to its initial size within 2 weeks. Decreased signal intensity of the renal parenchyma was noted on the first day. On T2-weighted images, four concentric layers of alternating low and high signal intensity were visualized. In the pathologic specimen, intense congestion and hemorrhage were found in the medulla—mainly in the outer part—and the subcapsular cortex (33,42,43). Fibrosis, hemorrhage, and calcification are contributing factors to the signal intensity loss on T2-weighted images (34). The loss of corticomedullary differentiation on T1-weighted images is the first radiologic feature of renal vein thrombosis, appearing on the first day after the event, and this worsens over time.
MR imaging can be useful in delineation of the parenchymal changes associated with renal vein thrombosis. MR imaging can be considered when venography is contraindicated or the findings at other imaging studies such as US or CT are equivocal (34,44).
Renal Cortical Necrosis
Renal cortical necrosis is a rare cause of acute renal failure that is usually associated with third-trimester obstetric hemorrhage, particularly abruptio placentae. Other causes of renal cortical necrosis include severe traumatic shock, septic shock, transfusion reaction, severe dehydration, venom toxin, and hemolytic uremic syndrome; it can also occur as a complication of renal transplantation. Although vasospasm and disseminated intravascular coagulation have been proposed as the primary event causing renal cortical necrosis, the pathogenesis still remains unclear (45).
Low signal intensity of the inner renal cortex and the columns of Bertin with every MR imaging sequence is the major characteristic finding of renal cortical necrosis (Fig 7a) (45). Swelling of both kidneys, corticomedullary differentiation on T2-weighted images instead of T1-weighted images (Fig 7a), and increased signal intensity of the cortex on T2-weighted images are other features of renal cortical necrosis at MR imaging (45). The area of nonenhanced low signal intensity of the renal cortex corresponds to the hypoattenuating zone observed on contrast material–enhanced CT scans (Fig 8) and to the histologic zone of cortical necrosis (46–49).
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Acute renal cortical necrosis. (7a) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T on the 30th day of renal cortical necrosis shows that both kidneys are shrunken and the low-signal-intensity cortex and columns of Bertin are prominent. The low-signal-intensity line (arrows) along the lateral margin of the right kidney is due to chemical shift artifact. (7b) Plain radiograph of the kidneys shows diffuse calcification along the cortical margin, which is the cause of the low signal intensity on MR images. (7c) Nonenhanced CT scan shows diffuse calcification involving the renal cortex and the columns of Bertin, which causes the low signal intensity on MR images. (Reprinted, with permission, from reference 45.)
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Acute renal cortical necrosis. (7a) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T on the 30th day of renal cortical necrosis shows that both kidneys are shrunken and the low-signal-intensity cortex and columns of Bertin are prominent. The low-signal-intensity line (arrows) along the lateral margin of the right kidney is due to chemical shift artifact. (7b) Plain radiograph of the kidneys shows diffuse calcification along the cortical margin, which is the cause of the low signal intensity on MR images. (7c) Nonenhanced CT scan shows diffuse calcification involving the renal cortex and the columns of Bertin, which causes the low signal intensity on MR images. (Reprinted, with permission, from reference 45.)
|
|
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Acute renal cortical necrosis. (7a) Axial T2-weighted MR image (2,000/80) obtained at 2.0 T on the 30th day of renal cortical necrosis shows that both kidneys are shrunken and the low-signal-intensity cortex and columns of Bertin are prominent. The low-signal-intensity line (arrows) along the lateral margin of the right kidney is due to chemical shift artifact. (7b) Plain radiograph of the kidneys shows diffuse calcification along the cortical margin, which is the cause of the low signal intensity on MR images. (7c) Nonenhanced CT scan shows diffuse calcification involving the renal cortex and the columns of Bertin, which causes the low signal intensity on MR images. (Reprinted, with permission, from reference 45.)
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Acute renal cortical necrosis. In another patient, MR imaging was performed at 1.5 T on the 10th day after massive upper gastrointestinal bleeding. Contrast-enhanced CT scan of the kidneys shows a nonenhancing low-attenuation rim with an enhanced outer cortex and medulla.
|
|
The thin rim of enhanced subcapsular tissue on MR images persists because of its separate capsular blood supply, including collateral circulation from extrarenal arteries. With time, the renal swelling decreases but the low-signal-intensity rim involving the inner cortex and the columns of Bertin does not change or slightly thickens (45). Contrast-enhanced CT demonstrates characteristic findings of renal cortical necrosis, including a nonenhancing cortical rim that correlates with the histopathologic findings of the disease (Figs 7c, 8) (48,49). However, MR imaging is better at demonstrating the extent of the disease because it does not entail use of iodinated contrast media.
Arterial Ischemia and Infarction
Infarction of the kidney can result from various causes, including thromboembolism, renal artery thrombosis, vasculitis, shock, and trauma.
On both T1- and T2-weighted MR images, the signal intensity of the infarcted area is usually lower than that of the noninfarcted area (Fig 9), although signal intensity may be higher in hemorrhagic infarcts (Fig 10) (50). Corticomedullary differentiation can be obliterated or less conspicuous in infarcted areas on T1-weighted images (Fig 9a) (50–53), and postcontrast T1-weighted images clearly demonstrate wedge-shaped infarcted areas (Fig 9c, Fig 9d). The extent and distribution of the infarcted areas at MR imaging correlate well with CT and angiographic findings (Fig 9c–Fig 9e) (50). Other findings such as a subcapsular hematoma or fluid collection, mass effect or a focal area of renal enlargement, and a thickened renal fascia can also be observed on MR images (50–53).
View larger version (114K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Renal arterial infarction. MR imaging was performed at 2.0 T 12 days after onset of right flank pain. (a) Axial T1-weighted MR image (500/30) shows areas of faint low signal intensity in the right kidney. Diminished corticomedullary differentiation is also noted. (b) Axial T2-weighted MR image (2,000/80) shows irregular low-signal-intensity lesions along the periphery of the right kidney. (c) Postcontrast axial T1-weighted MR image (500/30) clearly shows nonenhancing infarcted regions and enhancing noninfarcted areas in the right kidney. Note that the signal intensity of the noninfarcted areas of the right kidney is higher than that of the uninvolved left kidney. This finding is also seen in b. (d) Postcontrast coronal T1-weighted MR image (500/30) shows the same features as in c and the entire extent of the lesion. (e) Right renal arteriogram shows occlusion of arterial branches and nephrographic defects in the peripheral renal cortex. Note the embolus within the renal artery (white arrow) and the prominent capsular artery (black arrows). (Reprinted, with permission, from reference 50.)
|
|
View larger version (107K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Renal arterial infarction. MR imaging was performed at 2.0 T 12 days after onset of right flank pain. (a) Axial T1-weighted MR image (500/30) shows areas of faint low signal intensity in the right kidney. Diminished corticomedullary differentiation is also noted. (b) Axial T2-weighted MR image (2,000/80) shows irregular low-signal-intensity lesions along the periphery of the right kidney. (c) Postcontrast axial T1-weighted MR image (500/30) clearly shows nonenhancing infarcted regions and enhancing noninfarcted areas in the right kidney. Note that the signal intensity of the noninfarcted areas of the right kidney is higher than that of the uninvolved left kidney. This finding is also seen in b. (d) Postcontrast coronal T1-weighted MR image (500/30) shows the same features as in c and the entire extent of the lesion. (e) Right renal arteriogram shows occlusion of arterial branches and nephrographic defects in the peripheral renal cortex. Note the embolus within the renal artery (white arrow) and the prominent capsular artery (black arrows). (Reprinted, with permission, from reference 50.)
|
|
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Renal arterial infarction. MR imaging was performed at 2.0 T 12 days after onset of right flank pain. (a) Axial T1-weighted MR image (500/30) shows areas of faint low signal intensity in the right kidney. Diminished corticomedullary differentiation is also noted. (b) Axial T2-weighted MR image (2,000/80) shows irregular low-signal-intensity lesions along the periphery of the right kidney. (c) Postcontrast axial T1-weighted MR image (500/30) clearly shows nonenhancing infarcted regions and enhancing noninfarcted areas in the right kidney. Note that the signal intensity of the noninfarcted areas of the right kidney is higher than that of the uninvolved left kidney. This finding is also seen in b. (d) Postcontrast coronal T1-weighted MR image (500/30) shows the same features as in c and the entire extent of the lesion. (e) Right renal arteriogram shows occlusion of arterial branches and nephrographic defects in the peripheral renal cortex. Note the embolus within the renal artery (white arrow) and the prominent capsular artery (black arrows). (Reprinted, with permission, from reference 50.)
|
|
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9d. Renal arterial infarction. MR imaging was performed at 2.0 T 12 days after onset of right flank pain. (a) Axial T1-weighted MR image (500/30) shows areas of faint low signal intensity in the right kidney. Diminished corticomedullary differentiation is also noted. (b) Axial T2-weighted MR image (2,000/80) shows irregular low-signal-intensity lesions along the periphery of the right kidney. (c) Postcontrast axial T1-weighted MR image (500/30) clearly shows nonenhancing infarcted regions and enhancing noninfarcted areas in the right kidney. Note that the signal intensity of the noninfarcted areas of the right kidney is higher than that of the uninvolved left kidney. This finding is also seen in b. (d) Postcontrast coronal T1-weighted MR image (500/30) shows the same features as in c and the entire extent of the lesion. (e) Right renal arteriogram shows occlusion of arterial branches and nephrographic defects in the peripheral renal cortex. Note the embolus within the renal artery (white arrow) and the prominent capsular artery (black arrows). (Reprinted, with permission, from reference 50.)
|
|
View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9e. Renal arterial infarction. MR imaging was performed at 2.0 T 12 days after onset of right flank pain. (a) Axial T1-weighted MR image (500/30) shows areas of faint low signal intensity in the right kidney. Diminished corticomedullary differentiation is also noted. (b) Axial T2-weighted MR image (2,000/80) shows irregular low-signal-intensity lesions along the periphery of the right kidney. (c) Postcontrast axial T1-weighted MR image (500/30) clearly shows nonenhancing infarcted regions and enhancing noninfarcted areas in the right kidney. Note that the signal intensity of the noninfarcted areas of the right kidney is higher than that of the uninvolved left kidney. This finding is also seen in b. (d) Postcontrast coronal T1-weighted MR image (500/30) shows the same features as in c and the entire extent of the lesion. (e) Right renal arteriogram shows occlusion of arterial branches and nephrographic defects in the peripheral renal cortex. Note the embolus within the renal artery (white arrow) and the prominent capsular artery (black arrows). (Reprinted, with permission, from reference 50.)
|
|
View larger version (119K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Renal infarction. MR imaging was performed at 2.0 T 2 days after infarction of the left kidney. (a) Axial T1-weighted MR image (450/30) shows poor corticomedullary contrast in the left kidney and slightly lower signal intensity in the anterior portion of the left kidney. (b) Postcontrast axial T1-weighted MR image (450/30) shows a nonenhancing area of infarction (arrows). Note that the signal intensity of the noninfarcted posterior portion of the left kidney (arrowheads) is much higher than that of the uninvolved right kidney. (Reprinted, with permission, from reference 50.)
|
|
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Renal infarction. MR imaging was performed at 2.0 T 2 days after infarction of the left kidney. (a) Axial T1-weighted MR image (450/30) shows poor corticomedullary contrast in the left kidney and slightly lower signal intensity in the anterior portion of the left kidney. (b) Postcontrast axial T1-weighted MR image (450/30) shows a nonenhancing area of infarction (arrows). Note that the signal intensity of the noninfarcted posterior portion of the left kidney (arrowheads) is much higher than that of the uninvolved right kidney. (Reprinted, with permission, from reference 50.)
|
|
According to the results of an animal experiment with pathologic correlation (54), the changes in signal intensity on MR images correlated with the histologic findings and, to an extent, with the age of the infarction. Six hours after arterial ligation, the kidney showed low signal intensity on both T1- and T2-weighted images with the pathologic features of mild interstitial edema and hemorrhage. From 1 day to 2 weeks after renal artery ligation, the signal intensity increased progressively with both sequences due to the initiation of coagulation necrosis and the sequential changes of interstitial hemorrhage in the infarcted area. Two weeks later, the signal intensity of the infarcted area decreased again with the pathologic findings of cortical atrophy and organizing fibrosis. Postcontrast T1-weighted images clearly demonstrated the extent of the lesion (54).
The noninfarcted portion of the involved kidney may have higher signal intensity than the opposite uninvolved kidney on T2-weighted images and postcontrast T1-weighted images (Fig 10) (50). This appearance may be attributable to acute tubular necrosis and interstitial edema at the margin of the infarct or reperfusion injury following a brief period of ischemia and subsequent migration of a central occluding embolus from the main renal artery into peripheral branches (55). The low signal intensity on both T1- and T2-weighted images has been explained by lack of blood perfusion in the early phase and organization of the infarcts in the late phase (56,57). According to the results of an experimental study in dogs, the main pathologic features of the infarcted kidney were ischemic tubular damage with prominent interstitial edema in the early stage (up to 7 days) and organization and maturation of the infarct beginning on the 7th day and being well advanced by 17 days after occlusion of the renal arteries (57). MR imaging, especially postcontrast imaging, can demonstrate the extent of the infarction with an accuracy comparable to that of CT or angiography without the danger of iodinated contrast media (51).
Rejection of a Transplanted Kidney
Rejection of a transplanted kidney has been reported to show low signal intensity in some cases (58–60). Although loss of corticomedullary differentiation at MR imaging is the most sensitive indicator in the early evaluation of rejection, when this entity is combined with diffuse hemorrhagic necrosis or cortical necrosis, low signal intensity may appear in the involved area (58–60).
Acute Nonmyoglobinuric Renal Failure
Acute nonmyoglobinuric renal failure with severe loin pain and patchy renal vasoconstriction occurs in young, healthy persons following strenuous exercise and is often associated with a history of taking analgesics prior to exercise. It has been reported to show high signal intensity on T1-weighted images and variable signal intensity on T2-weighted images (61). The pathophysiologic mechanism is renal vasoconstriction at the level of an arcuate artery or interlobar artery resulting in focal ischemia and infarction (62–66). In view of this mechanism and the results of animal experiments related to renal infarction (54), MR imaging performed after the 7th day following the event will likely show low signal intensity on both T1- and T2-weighted images.
|
Conclusions
|
|---|
The kidney is a major organ involved by many different disease entities with systemic manifestations. Many of them can be diagnosed with intravenous urography, US, and CT. However, because many of these patients have renal dysfunction, iodinated contrast media can aggravate already impaired renal function. Under these circumstances, MR imaging can be helpful. Some of these diseases may produce low-signal-intensity renal parenchyma at MR imaging. This appearance is related to congestion, hemorrhage, fibrosis, and calcification in the renal parenchyma. When we encounter these findings, awareness and understanding of the disease entities, their underlying pathogenesis, and pathologic findings may help us reach a specific diagnosis.
|
Footnotes
|
|---|
Abbreviations: HFRS = hemorrhagic fever with renal syndrome,
PNH = paroxysmal nocturnal hemoglobinuria
|
References
|
|---|
- Hricak H, Crooks L, Sheldon P, Kaufman L. Nuclear magnetic resonance imaging of the kidney. Radiology 1983; 146:425-432.[Abstract/Free Full Text]
- Leung AW, Bydder GM, Steiner RE, Bryant DJ, Young IR. Magnetic resonance imaging of the kidneys. AJR Am J Roentgenol 1984; 143:1215-1227.[Abstract/Free Full Text]
- Atlas SW. Intraaxial brain tumors. In: Atlas SW, eds. Magnetic resonance imaging of the brain and spine. New York, NY: Raven, 1991; 223-326.
- Janick PA, Hackney DB, Grossman RI, Asakura T. MR imaging of various oxidation states of intracellular and extracelluar hemoglobin. AJNR Am J Neuroradiol 1991; 12:891-897.[Abstract]
- Bradley WG. Hemorrhage and brain iron. In: Stark DD, Bradley WG, eds. Magnetic resonance imaging. 2nd ed. St Louis, Mo: Mosby–Year Book, 1992; 721-738.
- Gomori JM, Grossman RI. Mechanisms responsible for the MR appearance and evolution of intracranial hemorrhage. RadioGraphics 1988; 8:427-440.[Abstract]
- Saini S, Frankel RB, Stark DD, Ferrucci JT. Magnetism: a primer and review. AJR Am J Roentgenol 1988; 150:735-743.[Free Full Text]
- Mulopulos GP, Turner DA, Schwartz M, Murakami ME, Clark JW. MRI of the kidneys in paroxysmal nocturnal hemoglobinuria. AJR Am J Roentgenol 1986; 146:51-52.[Free Full Text]
- Stark DD, Moseley ME, Bacon BR, et al. Magnetic resonance imaging and spectroscopy of hepatic iron overload. Radiology 1985; 154:137-142.[Abstract/Free Full Text]
- Kim SH, Han MC, Lee JS, Kim S. Paroxysmal nocturnal hemoglobinuria: case report of MR imaging and CT findings. Acta Radiol 1991; 32:315-316.[Medline]
- Tanaka YO, Anno I, Itai Y, Abe T. Paroxysmal nocturnal hemoglobinuria: MRI findings. J Comput Assist Tomogr 1993; 17:749-753.[Medline]
- Roubidoux MA. MR of kidney, liver, and spleen in paroxysmal nocturnal hemoglobinuria. Abdom Imaging 1994; 19:168-173.[CrossRef][Medline]
- Roubidoux MA. MR imaging of hemorrhage and iron deposition in the kidney. RadioGraphics 1994; 14:1033-1044.[Abstract]<
Top
Abstract
LEARNING OBJECTIVES FOR TEST...
