Spiral CT of the Kidneys: Current Imaging Strategies 2001
Elliot K. Fishman, M.D.
Spiral computed tomography (CT) provides many advantages in evaluation of the kidney. Spiral scanning essentially eliminates data misregistration, and allows for visualization of the entire kidney during peak contrast enhancement or in any specific phase of enhancement desired (1,2). The generated dataset is free from motion artifact, which improves multiplanar reconstructions as well as allowing three dimensional reconstruction of the data sets. These factors are of vital importance for the evaluation and characterization of small renal tumors (3).
Spiral CT increases sensitivity in the detection of subtle asymmetries of the cortical nephrogram, which is useful in diagnosing disorders such as renal artery stenosis, renal vein thrombosis, acute pyelonephritis, and renal obstruction. This syllabus provides a detailed analysis of spiral CT evaluation of renal pathology, illustrating the advantages of spiral scanning in the evaluation of both renal inflammatory diseases and renal masses. Such innovative applications such as renal donor evaluation, renal and ureteral stone detection and 3D imaging are also addressed.
Spiral CT of the Kidney: Imaging Protocols
Dynamic CT scanning of the kidneys has long been accepted as the preferred method for evaluation of renal pathology (4-6). Bolus intravenous enhancement aids in the detection and characterization of renal masses and parenchymal disease, allows for assessment of tumor vascularity, and helps to differentiate juxtarenal from intrarenal processes. Dynamic scanning is most sensitive during the early cortical nephrogram phase of renal enhancement, when corticomedullary differentiation is maximal (5). However, conventional scanners are quite limited in obtaining images during this early phase of enhancement. A major advantage of spiral CT is the ability to image the entire length of the kidneys during this ideal period of enhancement, in conjunction with routine dynamic evaluation of the liver and pancreas (7). Newer software for extended spiral capabilities provides optimal contrast enhanced images of the kidney during routine evaluation of the entire abdomen.
The protocol for spiral scanning of the kidneys continues to be one of great controversy with protocols varying between institutions (8). Our current protocol for a dedicated kidney study requires a localizing topogram. and an initial axial section is then selected in the upper third of the liver. Subsequently, 100-120 cc of non-ionic contrast (Omnipaque-350, Nycomed Amersham, Princeton, NJ) is peripherally administered through a power injector, ideally at a rate of 3 cc/sec. Spiral scanning begins within 30 seconds for arterial phase imaging and within 60 seconds for the cortical medullary phase following the initiation of contrast injection. The table incrementation rate is 5 mm/sec, with 5mm collimation at 280 mAs and 120 kVp (Siemens Somatom Plus-4). Overlapping reconstruction is usually obtained using 3-4 mm interscan spacing. Although representative multiplanar and three-dimensional (3D) analysis can be performed with these scan parameters in cases where evaluation of vascular anatomy is the focus such as renal transplant donor evaluation or suspected renal artery stenosis then a different scanning protocol is selected. This protocol with often require imaging during both the arterial and corticomedullary phases of enhancement in order to obtain key functional data or to image both arterial and venous anatomy. For example in cases of renal donor evaluation we will obtain 3mm thick sections with a table speed of 3-5 mm and reconstruct the data at 1-3 mm increments. In this case dual phase spiral CT is routinely utilized. In the case of renal artery stenosis we will obtain only arterial phase images but use 2mm collimation and reconstruct the data at 1 mm intervals.
With MDCT we routinely will use 3 mm sections of the kidney reconstructed at 3 mm intervals. When CT angiography and 3D imaging is used then we routinely acquire 1.25 mm thick sections reconstructed at 1mm intervals.
Certain clinical situations require other alterations of the routine renal spiral protocol. Initial pre-contrast scanning is performed if renal or ureteral calculi are clinically suspected. In most cases recent articles suggest that no additional scans are needed. In cases of suspected transitional cell carcinoma, renal infection, or obstruction, delayed scans are obtained to evaluate the opacified collecting system. In cases of infection the presence of focal inflammatory changes may only be seen on these delayed images (9,10). Although a question can be raised about the use of multi-phase studies in each patient we tend to be a bit more pragmatic based on a busy daily clinical workload. Although it might be easy to suggest that a complete examination of the kidney requires a non-contrast scan, an arterial phase acquisition, a corticomedullary phase acquisition, a nephrographic phase acquisition and an excretory phase acquisition we do not routinely do this in each patient. We select specific protocols based on the clinical problem to solve and the questions to answer. Let us now look at these decisions and some of the controversies in more detail.
THE NORMAL KIDNEY
Following bolus intravenous injection, the renal parenchyma demonstrates four distinct phases of enhancement: the cortical angiogram, the glomerulogram, the cortical nephrogram, and the tubular nephrogram (11). The first two phases occur within a few seconds and are not routinely imaged. Routine spiral CT scanning occurs during the peak of the third phase of renal enhancement, the cortical nephrogram phase. At this time (approximately 40 to 70 seconds after injection), contrast fills the cortical capillaries, peritubular spaces, and proximal convoluted tubules, resulting in exquisite corticomedullary differentiation (5, 11, 12). The density of the cortical nephrogram and the degree of corticomedullary differentiation obtained on the spiral scan depends on the patient's renal function and cardiac output, as well as on the rate and volume of contrast administration (11). In our experience, bolus injection at 2 to 3 cc/sec usually results in cortical enhancement ranging from 100 to 200 HU.
Unlike conventional dynamic scanning, spiral CT is routinely able to image the entire length of the kidneys during this ideal period of enhancement. This can be helpful to discriminate normal cortical variants such as a prominent column of Bertin or dromedary hump from an abnormal renal mass. Spiral CT provides increased sensitivity in the detection of subtle asymmetry of the cortical nephrogram, which reflects abnormal hemodynamics in the setting of either abnormal perfusion or delayed tubular transit (12).
The renal arteries are confidently displayed on spiral CT, and this has application in the evaluation of renal artery stenosis or thrombosis (13,14). Most spiral scans also provide sufficient opacification of the renal veins, although opacification may be delayed in patients with poor renal function. The inferior vena cava is normally non-opacified on early phase spiral scans. In patients with renal cell carcinoma and suspected vascular invasion, delayed scans may to be obtained to better visualize the renal veins and inferior vena cava.
Following spiral data acquisition, conventional delayed images or an additional spiral study may be obtained during the pyelogram (excretory) phase of renal enhancement, which routinely begins three to four minutes following injection. At this time, contrast normally fills the renal collecting system and ureter. We routinely obtain delayed scans in evaluation for suspected tumor or calculi in the collecting system, or in cases of obstruction.
Although our initial spiral CT protocols focused on the corticomedullary phase as the optimal time for data acquisition there have been numerous articles looking specifically at lesion detection rates and phase of contrast enhancement. Cohan et al. analyzed 33 patients where corticomedullary phase scans (40 seconds after initiation of injection), and nephrographic phase scans (median of 125 seconds after injection) were compared for lesion detection. The nephrographic phase images detected nearly four times as many lesions when they were located in the renal medulla. Corticomedullary phase images also could result in false positive diagnosis. Birnbaum et al. analyzed 30 patients with indeterminate renal masses with corticomedullary and nephrographic phase studies and found that nephrographic phase studies were more accurate in lesion discrimination and favored this phase for lesion analysis. The specific value of the nephrographic phase were those cases with relatively hypovascular masses where enhancement was not noted in the corticomedullary phase of the study. Szolar et al. found similar results especially for the smaller (<3cm) renal cell carcinomas. They found that the number of masses smaller than 3.0 cm detected on corticomedullary-phase scans were statistically significantly fewer than those on nephrographic phase scans. False positive results only occurred on the corticomedullary-phase scans. The authors concluded that if only a single phase study was to be obtained then it should be the nephrographic phase.
RENAL NEOPLASTIC DISEASE
RENAL CELL CARCINOMA
CT is widely accepted as the preferred modality for detection and staging of renal cell carcinoma (15-22). The typical carcinoma presents as a bulky, heterogeneous mass that demonstrates inhomogeneous enhancement, yet remains relatively hypodense compared to the normally enhancing renal parenchyma although this will vary depending on the phase of data acquisition discussed above. Cystic renal cell cancers are differentiated from complicated benign cysts by their thick irregular wall and tumor nodularity. Local spread of renal cell carcinoma typically occurs into the renal vein, paraaortic lymph nodes, or perinephric spaces.
Spiral CT offers many advantages in the evaluation of renal cell carcinoma. The uniform peak vascular opacification that is routinely achieved with spiral scanning improves both detection and characterization of the typically hypervascular renal cell carcinoma. Prognostic significance may also follow, in that tumor size and vascularity may correlate with a renal cell carcinoma's ability to metastasize (23). Optimal vascular opacification also enables the detection of extension of a renal cell carcinoma into the renal vein and inferior vena cava. Venous extension of renal cell carcinoma appears to be more common with right sided tumors, with approximately 10% of renal cell cancers invading the inferior vena cava (24, 25). Tumor thrombus presents as a luminal filling defect, and vascular collaterals can be occasionally be demonstrated. Delayed scans may improve visualization of the inferior vena cava, which can normally be nonopacified on an early spiral scan.
A particular problem with conventional CT in the precise delineation of the degree of tumor extension into the inferior vena cava. This is particularly true when thrombus approaches the right atrium. Currently, MR imaging is the most useful modality for tumor thrombus evaluation (26). However, multiplanar reconstruction and 3D imaging obtained from the motion-free spiral dataset may prove as effective in the depiction of tumor thrombus extension. Three dimensional models of the kidneys and their vascular supply can also be helpful in surgical planning for patients eligible for partial nephrectomy. In our institution spiral CT with 3D mapping in routinely obtained in all difficult cases of potential partial nephrectomy.
A major advantage of spiral CT over conventional scanning is the increased frequency of detection of small renal cell carcinomas. This advantage results in part from data continuity provided by a single-breath-hold acquisition, which eliminates breathing misregistration and partial volume averaging. This allows for increased visualization of smaller lesions that may otherwise have been missed. Peak contrast enhancement also helps to differentiate small tumors from brightly enhancing normal parenchyma.
Diagnosis of the small renal cell carcinoma also depends, in part, on the detection of "significant" contrast enhancement. Currently, only relatively large increases in attenuation value (>20 HU) indicate true enhancement on conventional CT, in part because smaller increases of attenuation are artificially produced by partial volume averaging (27). By eliminating this potential source of error, spiral scanning may allow for a decrease in the currently accepted "threshold" level of significant enhancement, thereby increasing our specificity in the characterization of the small renal mass.
Many of these small (< 3 cm.) renal cell carcinomas are discovered incidentally. Because of disagreement about their clinical course, detection of these lesions poses a problem in clinical management (18). In the past, these were largely felt to represent adenomas with a low potential for metastasis. At present, most pathologists believe these lesions should all be considered carcinomas (16, 20). Because there is no effective cure for advanced renal cell cancer, a potential increase in early detection of small renal cell carcinomas with spiral scanning could provide a valuable role in increasing overall patient survival.
Finally, spiral CT provides an excellent means for following the patient who is status post-partial nephrectomy. Thinly collimated slices, with optimal contrast enhancement and absence of artifacts due to partial voluming and motion, permit clear differentiation of postoperative change from small foci of tumor recurrence. Detection of sites of distant metastases including the lung and liver are more accurately detected with spiral CT as compared with conventional CT scanning.
Renal lymphoma invariably results from hematogenous dissemination. Lymphoma can demonstrate multiple appearances on CT (6). Most commonly, the kidneys show diffuse enlargement, with other variations including bilateral focal masses, a single dominant mass, or direct invasion by adjacent adenopathy. Lymphomatous masses typically are homogeneous, and demonstrate minimal contrast enhancement (10-20 HU). The appearance of focal lymphoma is indistinguishable from renal cell carcinoma. Small or subtle lesions are detected on spiral CT with increased frequency, accuracy, and confidence.
Although there currently is no formal published results, we feel that corticomedullary phase imaging is ideal for renal imaging in patients with lymphoma. We do not routinely do multiphase imaging in these patients unless the initial phase of study is indeterminate or if primary renal pathology is suspected.
Angiomyolipomas are focal hamartomas classically seen in middle-aged females or patients with tuberous sclerosis (28). Spiral scanning can help in the detection of subtle foci of fat within a renal mass. This is considered almost pathognomonic for angiomyolipoma, although recent reports have described subtle fat deposits in renal cell carcinoma (29, 30). Improved lesion characterization provided by the virtual elimination of partial volume averaging can help to differentiate angiomyolipomas from small cysts. Angiomyolipomas can grow quite large and appear almost entirely eccentric to the renal parenchyma. Spiral scanning can aid in demonstrating that the mass is intrinsic to the kidney, thus ruling out retroperitoneal liposarcoma.. Finally, accurate evaluation of larger lesions (greater than three or four centimeters) can be helpful for potential presurgical planning.
TRANSITIONAL CELL CARCINOMA
Transitional cell carcinoma accounts for approximately 5% of all renal tumors. Patients frequently present with hematuria. In conjunction with intravenous urography, CT has an important role in the evaluation and staging of transitional cell carcinoma. Several distinct patterns of involvement are described on conventional and spiral CT (31). Urban et al. recently reviewed the CT appearance of transitional cell carcinomas and divided them into early stage (stage I and II) and late stage (stage III and IV) disease. Early stage disease consists of a central solid mass that expands centrifugally with compression of the renal fat. On noncontrast studies the lesions are in the 5-30H range and will only minimally enhance after contrast injection. This enhancement is best seen with spiral CT. Spiral CT can not reliably distinguish stage I from stage II disease but can distinguish from the more advanced tumors. Delayed scans are very helpful in defining the true extent of tumor as contrast can pool around the tumor mass on these views.
The key feature in separating early from late stage disease is renal parenchymal invasion. In stage III disease there is tumor infiltration into the renal parenchyma or the peripelvic fat. Stage IV disease is defined by local extrarenal extension, lymph node involvement, or distant metastases. Both stage III and stage IV disease can be confused with an infiltrating renal cell carcinoma.
When performing a spiral CT scan of the kidneys, careful attention must be directed to the renal pelvis. Routine spiral scanning does not opacify the renal collecting system. Therefore, for patients with any subtle suggestion of intraluminal filling defect, or in studies performed for evaluation of hematuria or known transitional cell carcinoma, delayed scans are necessary for complete evaluation of the renal pelvis and ureter. While delayed scanning markedly improves detection and conspicuity of transitional cell carcinoma, we have found that most tumors can be suggested even before obtaining the delayed scan.
A variety of other neoplasms can commonly involve the kidney. Wilms tumor is the most common pediatric renal tumor, usually presenting as a large mass which can invade the renal vein and inferior vena cava. Collateral vessels may be demonstrated. Tumor extension is accurately demonstrated using multiplanar reconstruction generated by the spiral dataset. Oncocytomas usually demonstrate features identical to renal cell carcinoma, although demonstration of a central scar may suggest the diagnosis. These tumors are hypervascular although this alone is not of significant value in the differentiation from renal cell carcinoma. Most, if not all, patients still require nephrectomy to exclude renal cell carcinoma.
The use of spiral CT for the routine staging of the oncologic patient has led to an increase in the number of incidentally discovered renal metastases. Metastases may present as single or multiple masses which typically are hypovascular. In many cases the lesions do not distort the renal outline and so can easily be overlooked on non-contrast scans. Some have advocated partial nephrectomy as treatment for small metastases, which are detected with increased frequency on spiral scanning (20).
RENAL CYSTIC DISEASE
THE SIMPLE CYST
Renal cysts are round and usually cortically-based. On CT, cysts demonstrate sharp smooth margins with the normal renal parenchyma. Attenuation values are near that of water (-10 to 20 HU), with no increase after the administration of contrast. Because of peak contrast enhancement, spiral CT is ideally suited to help define the walls and clearly define the smooth transition with normal, uniformly enhancing renal cortex.
Conventional CT can result in errors in the diagnosis of the simple cyst (32). Errors can result from partial volume averaging of small (< 1 cm) cysts or those that are small and completely intrarenal, both of which may falsely elevate the attenuation number. Another common error is the impression of a thick cyst wall that occurs when the parenchymal beak is imaged in cross-section (33). By virtually eliminating partial volume averaging, spiral scanning can increase our specificity in the diagnosis of the simple renal cyst, thereby avoiding the erroneous diagnosis of an indeterminate renal mass. Similarly, indeterminate masses found on ultrasound or conventional CT can be confidently characterized on spiral CT (3). Bosniak and Rofsky in a review of the problems associated with the detection and characterization of small renal masses felt "that a normal contrast enhanced CT scan of the kidneys with use of 5mm sections, obtained with a helical scanner, for all practical purposes rules out a clinically significant renal parenchymal neoplasm."
Renal cysts are associated with syndromes such as tuberous sclerosis or von Hippel-Lindau disease. In tuberous sclerosis, the cysts are often small, rarely exceeding three centimeters in diameter (34). Multiple angiomyolipomas are characteristic of tuberous sclerosis. Patients with von Hippel-Lindau disease can also demonstrate multiple renal cysts; however, these cysts may be a precursor to malignancy. Tumors in this setting often present as masses arising from the cyst wall, and are frequently bilateral. Careful evaluation is necessary in these patients, as tumors are often small (< 2 cm) and very difficult to distinguish from other cysts (35).
THE COMPLICATED CYST
Complicated cysts can be seen in the setting of superimposed infection or hemorrhage. The benign, complicated cyst can be difficult to differentiate from a cystic and septated renal cell carcinoma. In general, if the septae are thin (1 mm or less) and smooth, a benign diagnosis is made with confidence. The septations should not demonstrate enhancement. Minimally complicated cysts may demonstrate calcification in the septations or cyst wall (36). Infected cysts can demonstrate markedly thickened walls that occasionally calcify (32).
If septations measure greater than 1 mm in thickness or are associated with solid or enhancing elements at their attachment to the cyst wall, malignancy must be considered (36). Because of data continuity provided from a single breath hold acquisition, spiral CT can visualize the margins of cystic lesions with great precision. This can be helpful in the detection of subtle nodularity or enhancement in the wall of a cystic neoplasm (3).
Hyperdense renal cysts can be confused with solid masses on CT. Most hyperdense cysts result from hemorrhage or proteinaceous debris, with hemorrhage complicating approximately 6% of simple cysts (32). Two-thirds of patients with polycystic kidney disease have hyperdense cysts 18, ( 37, 38).
Parapelvic cysts are likely lymphatic in origin. They can be single or multiple, and commonly present in the perihilar region of the kidney. The CT characteristics are those of the simple cyst. Parapelvic cysts can mimic hydronephrosis or the renal pelvis on dynamic spiral CT evaluation. Delayed scans, which demonstrate characteristic compression of the renal pelvis, are necessary to establish the proper diagnosis.
POLYCYSTIC KIDNEY DISEASE
Polycystic kidney disease usually manifests clinically during the third or fourth decade. Patients can present with flank pain, hematuria, hypertension, infection, or a palpable abdominal mass. Some patients are asymptomatic. Spiral CT demonstrates enlarged kidneys with multiple, bilateral cysts of varying size. Cysts can often also be demonstrated in the liver, spleen and pancreas. The hyperdense cyst (60-90 HU) is a common complication, seen in up to two-thirds of patients (37, 38). It usually results from prior post-traumatic hemorrhage, followed by clot retraction and protein concentration (39, 40). Cyst wall calcification then results in approximately 25% of patients, usually those with more advanced disease. This finding alone is not worrisome for neoplasm unless an associated soft tissue mass also is present (41).
Flank pain with hematuria is a common clinical dilemma in a patient with polycystic kidney disease. This may result from cyst hemorrhage, calculi (found in 20-36% of patients), or infection (41). Spiral CT is ideally suited for differentiating these complications. We have found the early dynamic images obtained during spiral CT to be more sensitive than conventional delayed scanning in the detection of secondary pyelonephritis especially in these patients. Very rarely, neoplasm can complicate polycystic kidney disease (42, 43).
ACQUIRED CYSTIC DISEASE OF UREMIA
While the exact etiology is unclear, cysts can complicate renal failure, typically three or more years after the onset of hemodialysis. Uremic cysts can regress after renal transplantation (44). The kidneys in acquired cystic disease of uremia are typically small, and the cysts are commonly intrarenal in location. There is a significant increased incidence of renal carcinomas in these patients. Spiral CT can help differentiate simple uremic cysts from a solid tumor.
RENAL INFLAMMATORY DISEASE
Acute pyelonephritis is usually a clinical diagnosis requiring no imaging. Patients may be referred for CT evaluation if there is a poor clinical response and an abscess is suspected or if the diagnosis is in doubt. Ascending retrograde infection typically spreads through the collecting ducts into the renal parenchyma (45). Hematogenous infection may involve the kidney in patients with tuberculosis, fungemia, or staphylococcal sepsis. In pyelonephritis, increased interstitial pressure results in obstruction of the tubules and intense focal vasoconstriction of the blood vessels. Involvement may be focal or diffuse, and the infected kidney is generally enlarged. Enhanced CT scans demonstrate focal areas of striated or wedge-shaped perfusion abnormality, resulting in a characteristic "patchy" nephrogram (46).
Poorly enhanced areas of focal pyelonephritis can mimic a renal mass on conventional CT. Spiral scanning is probably more specific in differentiating infection from tumor. Spiral CT is also helpful in detecting subtle cases of acute pyelonephritis. Abnormalities only visualized during early dynamic contrast enhancement such as loss of the normal, sharp corticomedullary differentiation and delayed appearance of the cortical nephrogram, are clues to the diagnosis of underlying infection. Delayed views of the infected kidney may demonstrate a dense nephrogram or a striated nephrogram.
In the absence of direct visualization of thrombus or vascular collaterals, acute renal vein thrombosis can sometimes mimic conventional CT findings of acute pyelonephritis (47). Both entities can demonstrate renal enlargement and poor renal function. In acute pyelonephritis, low attenuation wedge-shaped defects extend to the renal cortex, with ill-defined or poorly enhancing parenchymal lesions occasionally seen (48, 49). In renal vein thrombosis, focal low attenuation areas are better defined, representing the renal medulla surrounded by high attenuation cortex (47). Spiral scanning during peak contrast enhancement can help to differentiate these patterns.
In the patient with suspected acute pyelonephritis or other renal infection the sequencing of the spiral CT examination is crucial. Select non-contrast scans are obtained in most cases to exclude the possibility of renal calculi. A spiral CT acquisition in the corticomedullary phase is obtained and may clearly define the presence and extent of infection. Delayed scans are also obtained in the excretory phase, which may in fact define the lesion a bit more specifically especially if a striated nephrogram is detected. Kawashima et al. has looked into the problem of renal infection with CT and found that there is variability between cases and no single set of rules as to the optimal image sequence has yet been defined.
In patients with chronic pyelonephritis, spiral CT demonstrates focal parenchymal scars with underlying calyceal distortion. This is typically most pronounced in the polar regions of the kidney. The kidneys are often small, and renal calculi are frequently seen.
In addition, chronic scarring can produce an irregular, lobulated cortical surface that may mimic a renal mass on sonography or conventional CT. Spiral CT nicely demonstrates the normally enhancing cortex, and thus improves detection of focal areas of cortical scarring or thinning. Thus, spiral CT helps to distinguish focal scarring from a renal mass. Delayed scans are often helpful to demonstrate the underlying blunted calyces, and to confirm the location of renal calculi.
Xanthogranulomatous pyelonephritis is an uncommon form of chronic inflammatory disease, felt to result from obstruction in the presence of infection. CT findings here include a staghorn calculus, an absent or markedly diminished nephrogram, and dilated calyxs (48). Extension to the perirenal space and psoas compartment is readily demonstrated by spiral CT.
Abscess formation can complicate pyelonephritis or an infected cyst. Lesions are low in attenuation, and classically demonstrate an enhancing, thick and irregular wall, especially when chronic (49). Spiral CT improves lesion characterization due to bolus contrast enhancement. Furthermore, the virtual elimination of partial volume averaging provided by spiral scanning can aid in the detection of smaller lesions and in the demonstration of gas bubbles which are virtually pathognomonic for renal abscess formation. A perinephric abscess can result from extension through the renal capsule.
RENAL ARTERY STENOSIS
Renal artery stenosis is implicated in approximately 5% of patients with hypertension. With currently available technology, neither color duplex ultrasound nor magnetic resonance angiography has proven adequate in screening patients for significant renal artery stenosis (50-53). Both multiplanar and 2D and 3D imaging using the spiral CT dataset may provide a reliable, non-invasive screening method for direct visualization of renal artery stenosis. Since imaging occurs during peak arterial opacification, spiral CT clearly defines the location and extent of luminal narrowing.
Brink et al. investigated the optimization of protocols for renal artery evaluation and found that critical renal artery stenosis was best depicted with 2mm collimation, 2-4mm table speed (pitch of 1-2) and data reconstruction at 1mm intervals. Several articles from leading academic institutions have found spiral CT angiography to be equal in accuracy to classic renal angiography. Because of its non-invasiveness (i.e. complication free) as well as lower cost and speed spiral CT angiography is currently the preferred choice of examination for renal artery evaluation.
On spiral CT the renal artery and any narrowing can be clearly defined. Other indirect, corroborative evidence of renal artery stenosis is deduced by noting a small and smooth, non-obstructed kidney. Spiral CT may demonstrate an asymmetric delay in the appearance of the cortical nephrogram. The cortical nephrogram is then prolonged on delayed images, resulting from decreased glomerular filtration (12). Frequently, other clues to the diagnosis of renal vascular disease are present on the CT scan, such as severe abdominal aortic atherosclerosis or aneurysm formation.
Renal infarcts are commonly the result of cardiac emboli. The kidney in acute global infarction is rarely enlarged. A characteristic rim of enhancement, representing collateral capsular perfusion, is seen surrounding a hypodense, non-functioning infarcted kidney (57). The rim usually measures no more than 4 mm in depth (12). Occasionally, vermiform and globular medullary enhancement is demonstrated; while this finding is of unknown origin, it probably reflects collateral blood flow (58).
Renal infarcts can also be focal, presenting as wedge-shaped defects that extend to the capsular surface. Over time, atrophy occurs in the vascular distribution of the infarction. Segmental infarctions of the anterior or posterior renal arteries demonstrate a characteristic appearance on CT, with the larger ventral branch supplying the anterior lateral portion of the kidney and the smaller dorsal branch supplying the posterior medial aspect. The optimal enhancement of normal renal cortex afforded by spiral technique highlights areas of renal infarction that could be missed on conventional CT.
RENAL VEIN THROMBOSIS
The CT diagnosis of renal vein thrombosis is based on direct visualization of an intraluminal filling defect within an enlarged vein. In patients with poor renal function or decreased renal perfusion, care must be taken not to misdiagnose normal delayed filling of the renal veins for thrombus. Delayed scans are often helpful here in questionable cases. Indirect signs of renal vein thrombosis include demonstration of collateral vessels, renal enlargement, thickening of Gerota's fascia, and retroperitoneal hemorrhage (12). The use of spiral CT with narrow collimation and close interscan reconstruction intervals is ideal for mapping the patency of the renal vein. Acquisition of data at around 40 seconds after injection appears to be an ideal protocol in our experience.
Spiral CT is sensitive in detection of the asymmetric, abnormally delayed cortical nephrogram seen in patients with renal vein thrombosis. This delay results from diminished glomerular filtration and slowed tubular transit due to compression of the tubules by interstitial edema. Delayed scans may be obtained when the renal veins are opacified to confirm the presence if intraluminal thrombus. Perhaps, most importantly, spiral scanning has the potential to be more sensitive than conventional CT in the demonstration of underlying etiologies of renal vein thrombosis, such as renal cell carcinoma.
SMALL VESSEL DISEASE
Many diseases can involve the intrarenal vasculature (49). These include the collagen vascular diseases, radiation nephritis, and hypertension. The spiral CT appearance is non-specific in these entities with parenchymal scarring frequently seen. The involved kidneys may be small and poorly functioning, with marked cortical thinning. Areas of hemorrhage may result from rupture of microaneurysms in patients with polyarteritis nodosa, Wegener's granulomatosis, lupus, or intravenous drug abuse. Arterial phase imaging is ideal for visualization of these small aneurysms. Radiation nephritis may be acute or chronic, with renal failure ensuing from tubular and glomerular degeneration, as well as vascular fibrinoid necrosis.
One or more accessory renal arteries are present in up to 40% of normal patients. These usually arise distal to the major artery. Accessory renal branches are best appreciated with spiral CT angiography. Three-dimensional CT angiography with volumetric reconstruction is ideal in these patients and can detect accessory renal arteries with an accuracy equal to classic angiography.
Common venous anomalies include the circumaortic and retroaortic renal vein (59, 60). The retroaortic left renal vein is seen in approximately 2% of patients (60). Demonstration of these vascular anomalies is of vital importance for pre-surgical planning in patients who are potential renal donors as well as in patients scheduled for resection of renal tumors. Dual phase spiral CT can define the venous map as well as the arterial map in a single examination
OTHER VASCULAR DISEASE
Renal artery aneurysms can be seen in patients with arteriosclerosis or as a result of septic emboli (61). Congenital arteriovenous malformations are rare. Acquired AVM's most commonly result from prior renal biopsy. Dynamic scanning can reveal enlarged feeding vessels and collateral circulation.
Conventional CT has been shown to be sensitive in the detection of renal calculi (62, 63). Theoretically, smaller calculi should be detected with increased frequency and sensitivity on spiral CT. Lack of breathing misregistration ensures continuous coverage of the entire kidney which improves detection of small stones. Density measurements of small renal calculi are more accurate, as partial volume averaging is virtually eliminated. Olcott et al. found that "3D spiral CT depicted calculi more sensitively than traditional techniques and provided new information and improved accuracy in the evaluation of nephrolithiasis."
In performing a spiral CT for suspected stone disease, precontrast scans should be obtained. In most cases, no additional scan sequences will be needed. However, even on enhanced studies alone, one is usually able to distinguish renal calculi from potentially obscuring contrast especially if images are acquired before contrast excretion in the corticomedullary phase. Delayed scans may then be obtained to help localize calculi within a calyx or the renal pelvis.
Early scanning during the cortical nephrogram phase of enhancement can be useful in the detection of subtle obstruction (12). Asymmetry in the appearance of the cortical nephrogram on spiral CT, manifested by early delay and late persistence, signifies the presence of either abnormal perfusion or tubular transit (12, 64). In this setting the finding of a dilated collecting system on delayed scans strongly suggests active obstruction.
However, acute cases of obstruction may demonstrate little to no calyceal dilatation, making spiral CT findings similar to those of renal artery stenosis, renal vein thrombosis, or pyelonephritis. On later images, the nephrogram can appear mottled or striated, and hyperdense (64, 65). Delayed images of the collecting system and ureter may help define the cause of obstruction at any point from renal pelvis to the bladder. Suspicious areas can then be re-scanned with narrow collimation spiral CT to help characterize the obstructing lesion.
Recently one of the most investigated areas of spiral CT applications has been for the detection of ureteral calculi and ureteral obstruction. Although specific imaging techniques have varied between institutions most investigators now agree that with narrow collimation (3-5mm) and close interscan reconstruction (1-3mm) that spiral CT is the gold standard for calculi detection. Some authors including Sommer et al. have stressed the value of reformatted multiplanar images in these cases. Others such as Katz et al. have focused on the secondary signs of ureteral obstruction including hydronephrosis, hydroureter, perinephric stranding, and periureteral edema.
Early fusion of the lower poles of the kidneys can result in the horseshoe kidney, This is found in 1 in 600 people and represents a very common renal fusion anomaly (69). The lower poles fuse anterior to the aorta and inferior vena cava via an isthmus of functioning renal tissue or fibrous tissue. The majority of patients with horseshoe kidney demonstrate multiple, bilateral renal arteries. Other fusion anomalies include crossed fused ectopia and the pancake kidney. Spiral scanning can help define the relationship of the abnormal kidney to the major vessels. Other common congenital anomalies include renal agenesis, renal duplication, and malrotation.
Spiral CT has a potential role in the evaluation of the renal transplant patient. The most important application here may lie in the evaluation of vascular complications, which are a significant cause of graft dysfunction and occur in up to 10% of patients. Common types of vascular complications include arterial and venous stenosis or occlusion, arteriovenous fistulas, and pseudoaneurysms (70). Scanning during peak vascular enhancement increases detection of these potential complications. Other common complications include peri-transplant fluid collections and obstruction.
The role of spiral CT for the non-invasive evaluation of potential living related renal transplant donors is another developing application. At our institution renal transplants are being harvested by a laparoscopic approach rather than with classic open surgery. Patients have less post procedure complications as well as return to work and full activity at an earlier date. Dual phase 3D spiral CT has proven to be a viable alternative to angiography and has become the study of choice at our institution. The use of volumetric 3D rendering has been critical to our success in these patients.
Spiral CT represents an exciting advancement in evaluation of renal disease, offering several technical advantages in evaluation of the kidney. Scans can be obtained during specific phases of renal contrast enhancement, densitometry measurements are accurate, and minimal respiratory artifact results in less potential errors in diagnosis. As a result of the newer techniques, smaller cancers are detected with increased frequency, which will likely impact upon overall patient survival. Subtle asymmetries of the cortical nephrogram can be readily detected, increasing sensitivity, and oftentimes specificity in the diagnosis of renal obstruction, renal artery stenosis, or pyelonephritis. findings from partial volume averaging of benign lesions with normal parenchyma are minimized. Finally, improved lesion characterization should decrease the incidence of the diagnostically frustrating indeterminate renal mass. Newer vascular applications as well as 3D imaging suggest that the role of spiral CT will continue to increase in the near future.
1. Kalender WA, Polacin A. Physical performance characteristics of spiral CT scanning. Med Phys 1991; 18: 910-915.
2. Kalender WA, Seissler W, Klotz E, Vock P. Spiral volumetric CT with single-hold-technique, continuous transport, and continuous scanner rotation. Radiology 1990; 176:181-183.
3. Silverman SG, Seltzer SE, Adams DF, Tumeh SS, Allegra DP, MellinsHZ. Spiral CT of the small indeterminate renal mass: results in 48 patients (abstr). Radiology 1991;181(P):125.
4. Burgener FA, Hamlin DJ. Contrast enhancement in abdominal CT: bolus vs. infusion. AJR 1981;137:351-358.
5. Ishakawa I, Onouchi Z, Saito Y, Kitada H, Shinoda A, Ushitani K, Tabuchi M, Suzuki M. Renal cortex visualization and analysis of dynamic CT curves of the kidney. J Comput Assist Tomogr 1981;5:695-701.
6. Lee JKT, Sagel SS, Stanley RJ. Computed body tomography with MRI correlation. 2nd Ed. New York: Raven Press, 1989; 627-639.
7. Zeman RK, Fox SH, Silverman PM, et al. Helical (Spiral) CT of the abdomen. AJR 1993; 160:719-725.
8. Bluemke DA, Fishman EK. Spiral CT of the liver. AJR 1993;160:787-792.
9. Ney DR, Fishman EK, Niederhuber JE. Three-dimensional display of hepatic venous anatomy generated from spiral computed tomography data: preliminary results. J Digit Imaging 1992; 5:242-245.
10. Fishman EK, Wyatt SH, Ney DR, Kuhlman JE, Siegelman SS. Spiral CT of the pancreas with multiplanar display. AJR 1992; 159:1209-1215.
11. Boijsen E. Anatomic and physiologic considerations. In Abrams HL (ed): Abrams Angiography, 3rd edition. Boston:Little Brown and Co, 1983; 1118-1119.
12. Birnbaum BA, Boskniak MA, Megibow AJ. Asymmetry of the renal nephrogram of CT: significance of the unilateral prolonged cortical nephrogram. Urol Radiol 1991;12:173-177.
13. Napel S, Marks MP, Rubin GD, et al. CT angiography with spiral CT and maximum intensity projection. Radiology 1992; 185:607-610.
14. Rubin GD, Dake MD, Napel SA, McDonnell CH, Jeffrey RB. Three-dimensional spiral CT angiography of the abdomen: initial clinical experience. Radiology 1993; 186:147-152.
15. Zimmer WD, Williamson B Jr, Hartman GW, Hattery RR, O'Brien PC. Changing patterns in the evaluation of renal masses: economic implications. AJR 1984;143:285-289.
16. Curry NS, Schabel SI, Betsill WL. Small renal neoplasms: diagnostic imaging, pathologic features, and clinical course. Radiology 1986;158:113-117.
17. Balfe DM, McClennan BL, Stanley RJ, Weyman PJ, Sagel SS. Evaluation of renal masses considered indeterminate on computed tomography. Radiology 1982:142:421-428.
18. Levine E, Huntrakoon M, Wetzel LH. Small renal neoplasms: clinical, pathologic, and imaging features. AJR 1989:153:69-73.
19. Smith SJ, Bosniak MA, Megibow AJ, Hulnick DH, Horii SC, Raghavendra BN. Renal cell carcinoma: earlier discovery and increased detection. Radiology 1989;170:699-703.
20. Amendola MA, Bree RL, Pollack HM, Francis IR, Glazer GM, Jafri SZH, Tomaszewski JE. Small renal cell carcinoma: resolving a diagnostic dilemma. Radiology 1988;166:637-641.
21. Yamashita Y, Takahashi M, Watanabe O, Yoshimatsu S, Ueno S, Ishimaru S, Kan M, Takano S, Ninomiya N. Small renal cell carcinoma: pathologic and radiologic correlation. Radiology 1992;184:493-498.
22. Dunnick NR. Renal lesions: great strides in imaging. Radiology 1992;182:305-306.
23. McClennan BL. Computed tomography in the diagnosis and staging of renal cell carcinoma. Semin Urol 1985;3:111-131
24. Goncharendo V, Gerlock Jr. AJ, Kadir S, Turner B. Incidence and distribution of venous extension in 70 hypernephromas. AJR 1979;133:263-265.
25. Madayag MA, Ambos MA, Lefleur RS, Bosniak MA. Involvement of the inferior vena cava in patients with renal cell carcinoma. Radiology 1979;133:321-326.
27. Foster WL, Roberts L, Halvorsen RA, Dunnick NR. Sonography of small renal masses with indeterminant density characteristics on computed tomography. Urol Radiol 1988;10:59-67.
28. Bret PM, Bretagnolle M, Gaillard D, Plauchu H, Labidie M, Lapray JF, Roullard Y, Cooperberg P. Small, asymptomatic angiomyolipomas of hte kidney. Radiology 1985;154:7-10.
29. Strotzer M, Lehner KB, Becker K. Detection of fat in a renal cell carcinoma mimicking angiomyolipoma. Radiology 1993;188:427-428.
30. Helenon O, Chretien Y, Paraf F, Melki P, Denys A, Moreau JF. Renal cell carcinoma containing fat: demonstration with CT. Radiology 1993;188:429-430.
31. Baron RL, McClennan BL, Lee JKT, Lawson T. Transitional cell carcinoma of the pelvis and ureter- CT evaluation. Radiology 1982;144:125-130.
32. Hartman DS. Cysts and cystic neoplasms. Urol Radiol 1990;12:7-10.
33. Segal AJ, Spitzer RM. Pseudothick-walled renal cyst by CT. AJR 1979;132:827-828.
34. Mitnick JS, Bosniak MA, Hilton S, Raghavendra BH, Subramanyam BR, Genieser NB. Cystic renal disease in tuberous sclerosis. Radiology 1983;147:85-87.
35. Levine E, Collins DL, Horton WA, Schimke RN. CT screeening of the abdomen in von Hippel-Lindau disease. AJR 1982;139:505-510.
36. Bosniak MA. The current radiological approach to renal cysts. Radiology 1986;158:1-10.
37. Levine E, Grantham JJ. High-density renal cysts in autosomal dominant polycystic kidney disease demonstrated by CT. Radiology 1985;154:477-482.
38. Meziane MA, Fishman EK, Goldman SM, Friedman AC, Siegelman SS. Computed tomography of high density renal cysts in adult polycysitc kidney disease. J Comput Assist Tomogr 1986;10:767-770.
39. Sussman S, Cochran ST, Pagani JJ, McArdle C, Wong W, Austin R, Curry N, Kelly KM. Hyperdense renal masses: a CT manifestation of hemorrhagic renal cysts. Radiology 1984;150:207-211.
40. Fishman MC, Pollack HM, Argre PH, Banner MP. Case report. High protein content: another cause of CT hyperdense benign renal cyst. J Comput Assist Tomogr 1983;7:1103-1106.
41. Levine E, Grantham JJ. Calcified renal stones and cyst calcifications in autosomal dominant polycystic kidney disease: clinical and CT study in 84 patients. AJR 1992;159:77-81.
42. Gabow PA. Autosomal dominent kidney disease--more than a renal disease. Am J Kidney Dis 1985;5:104-111.
43. Gregoire JR, Torres VE, Holley KE, Farrow GM. Renal epithelial hyperplasia and neoplastic proliferation in autosomal dominant polycystic kidney disease. Am J Kidney Dis 1987;9:27-38.
44. Levine E, Grantham JJ, Slucher SL, Greathouse JL, Krohn BP. CT of acquired cystic kidney disease and renal tumors in long-term dialysis patients. AJR 1984;142:125-131.
45. Roberts JA. Pyelonephritis, cortical abscess, and perinephric abscess. Urol Clin North Am 1986;13:637-645.
46. Soulen MC, Fishman EK, Goldman SM, Gatewood OMB. Bacterial renal infection: role of CT. Radiology 1989; 171:703-707.
47. Glazer GM, Francis IR, Gross BH, Amendola MA. Computed tomography of renal vein thrombosis. J Comput Assist Tomogr 1984;8:288-293.
48. Goldman SM, Hartman DS, Fishman EK, Finizio JP, Gatewood OMB, Siegelman SS. CT of xanthogranulomatous pyelonephritis: radiologic-pathologic correlation. AJR 1984;142:963-969.
49. Dunnick NR, McCallum RW, Sandler CM. Textbook of Uroradiology. Baltimore: Williams and Wilkins, 1991.
50. Desberg AL, Paushter DM, Lammert GK, Hale JC, Troy RB, Novick AC, Nally Jr. JV, Weltevreden AM. Renal artery stenosis: evaluation with color doppler flow imaging. Radiology 1990;177:749-753.
51. Middleton WD. Doppler evaluation of renal artery stenosis: past, present, and future. Radiology 1992;184:307-308.
52. Berland LL, Koslin DB, Routh WD, Kellar FS. Renal artery stenosis: prospective evaluation of diagnosis with color duplex US compared with angiography. Radiology 1990;174:421-423.
53. Kim D, Edelman RR, Kent KC, Porter DH, Skillman JJ. Abdominal aorta and renal artery stenosis: evaluation with MR angiography. Radiology 1990;174:727-731.
54. Schwartz RB, Jones KM, Chernoff DM, et al. Common carotid artery bifurcation: evaluation with spiral CT. Work in progress. Radiology 1992; 185:513-519.
55. Castillo M. Diagnosis of disease of the common carotid artery bifurcation: CT angiography vs. catheter angiography. AJR 1993;161:395-398.
56. Dillon EH, van Leeuwen MS, Fernandez MA, Mali WPTM. Spiral CT angiography. AJR 1993;160:1273-1278.
57. Glazer GM, London SS. CT appearance of global renal infarction. J Comput Assist Tomogr 1981;5:847-850.
58. Malmed AS, Love L, Jeffrey RB. Medullary CT enhancement in acute renal artery occlusion. J Comput Assist Tomogr 1992;16:107-109.
59. Beckmann CF, Abrams HL. Circumaortic venous ring: incidence and significance. AJR 1979;132:561-565.
60. Reed MD, Friedman AC, Nealey P. Anomalies of the left renal vein: analysis of 433 CT scans. J Comput Assist Tomogr 1982;6:1124-1126.
61. DuBrow RA, Patel SK. Mycotic aneurysm of the renal artery. Radiology 1981;138:577-582.
62. Segal AJ, Spataro RF, Linke CA, Frank IN, Rabinowitz R. Diagnosis of nonopaque calculi by computed tomography. Radiology 1978;129:447-450.
63. Federle MP, McAninch JW, Kaiser JA, Goodman PC, Roberts J, Mall JC. Computed tomography of urinary calculi. AJR 136:255-258.
64. Samin A, Becker JA. CT nephrogram in acute obstructive uropathy. Urol Radiol 1991;12:178-180.
65. Bigongiari LR, Davis RM, Novak WG, Wicks JD, Kass E, Thornbury JR. Visualization of the medullary rays on excretory urography in experimental ureteric obstruction. AJR 1977;129:89-93.
66. Federle MP, Kaiser JA, McAninch JW, Jeffrey RB, Mall JC. The role of computed tomography in renal trauma. Radiology 1981;141:455-460.
67. Sandler CM, Toombs BD. Computed tomographic evaluation of blunt renal injuries. Radiology 1981;141:461-466.
68. Bretan PN, McAninch JW, Federle MP, Jeffrey RB. Computerized tomographic staging of renal trauma: 85 consecutive cases. J Urol 1986;136:561-565.
69. Moore KL: The developing human: clinically oriented embryology, ed. 4. Philadelphia, WB Saunders, 1988.
70. Dodd III GD, Tublin ME, Shah A, Zajko AB. Imaging of vascular complications associated with renal transplants. AJR 1991;157:449-459.