| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (O.R., D.L., P.B., C.P.) and Urology (A.G., X.M.), Pavillons P-V, Edouard Herriot Hospital, 5 place d'Arsonval, 69437 Lyon Cedex 03, France. Received November 9, 1998; revision requested December 21; revision received March 22, 1999; accepted July 1. Address reprint requests to O.R. (e-mail: olivier.rouviere@netcourrier.com).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Forty-one consecutive patients with symptomatic UPJ obstruction in 42 obstructed kidneys underwent renal helical CT and renal intraarterial digital subtraction angiography (DSA; flush aortography and bilateral selective renal injections). The helical CT and DSA images were interpreted in a blinded manner by two readers, and the results were compared.
RESULTS: DSA showed 126 renal arteries in the 41 patients; 56% of patients had supernumerary renal arteries. Helical CT depicted 121 (96%) of these 126 renal arteries prospectively. Retrospectively, 124 (98%) renal arteries were visible on CT images. Twelve (29%) of the 42 kidneys with UPJ obstruction had identifiable arteries crossing the UPJ on DSA images. If DSA is used as the standard of reference, CT angiography was 100% sensitive and 96.6% specific for depicting these crossing arteries.
CONCLUSION: Renal helical CT seems suitable to replace intraarterial DSA in the preoperative assessment of crossing arteries in kidneys with UPJ obstruction.
Index terms: Computed tomography (CT), angiography, 81.12115 • Kidney, CT, 81.12112, 81.12115, 81.12116 • Renal angiography, 81.1243 • Renal arteries, 961.12915, 961.12916 • Renal veins, 966.12915, 966.12916 • Ureter, stenosis or obstruction, 821.84
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
For the past 10 years, endopyelotomy has been the first-line therapy for UPJ obstruction in adult patients. This technique is performed with either endoscopic guidance (percutaneous or retrograde endopyelotomy) or fluoroscopic guidance (cutting balloon technique). For many authors (1,2,7–10), success rates of endopyelotomy are comparable to those of open surgery but with lower morbidity, and a failed endopyelotomy does not jeopardize the success of a subsequently performed pyeloplasty. The treatment of UPJ obstruction with cutting balloon catheters seems to give better results in secondary strictures than the standard endopyelotomy technique but shows higher failure rates in primary strictures (5,11).
It remains unclear whether crossing vessels should be considered a relative contraindication to endopyelotomy. As previous anatomic work has shown that vessels crossing within 1.5 cm of the UPJ are most commonly anterior to the UPJ (12), endourologists typically perform the incision posteriorly or posterolaterally, a location less likely to have vessels. However, the risk of vascular injury is not eliminated with this approach and seems accentuated when the incision is monitored by means of fluoroscopic techniques alone (1,7,8,13,14). Moreover, as the role of crossing arteries in UPJ obstruction is still controversial, the appropriate procedure for treating crossing arteries remains undefined.
Some authors (8,15) have reported good results with endopyelotomy in unselected groups of patients, which suggests that patients with crossing arteries may undergo treatment successfully. Others think that because endopyelotomy cannot correct the relationship of crossing arteries to the UPJ, it may lead to failure of the procedure. According to these authors (1,4,11), preoperative arteriography should be performed routinely, and patients with crossing arteries should undergo open Anderson-Hynes dismembered pyeloplasty.
At our institution, patients with symptomatic primary UPJ obstruction undergo preoperative renal intraarterial digital subtraction angiography (DSA). They undergo Anderson-Hynes pyeloplasty if DSA shows a crossing artery and endopyelotomy if no artery is seen crossing the UPJ.
Contrast material–enhanced helical computed tomographic (CT) angiography, which is less invasive and less costly than DSA, recently has been used to assess the vascular anatomy of potential renal donors (16) or kidneys with UPJ obstruction (17,18).
This study was performed to assess the value of helical CT angiography as a replacement for DSA in the preoperative management of UPJ obstructions.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Their serum creatinine level range was 60–124 µmol/L (mean, 84.5 µmol/L), and their weight range was 31–93 kg (mean, 61.5 kg). Nine patients had ureteral stents at the time of helical CT and DSA. Four patients had undergone previously unsuccessful attempts at UPJ repair (three open pyeloplasties and one endopyelotomy) performed at other institutions. All participating patients signed the appropriate investigational review board consent forms.
The UPJ obstruction involved the right kidney in 21 cases, the left kidney in 19 cases, and both kidneys in one case for a total of 42 obstructed kidneys. One of the patients with left UPJ obstruction had a horseshoe kidney.
Helical CT Angiography
All patients first underwent helical CT angiography. Imaging was performed with use of an SR 7000 CT scanner (Philips Medical Systems, Best, the Netherlands). Initial helical CT scanning was performed with a 10-mm collimation at 10-mm intervals through the kidneys to identify the level of the renal upper poles and the level of the aortic bifurcation. After this initial scanning, 40 mg of furosemide (Lasilix; Hoechst, Frankfurt am Main, Germany) was injected intravenously as a diuretic. Four to 6 minutes later, contrast-enhanced helical CT scanning was performed from the renal upper poles to the aortic bifurcation, if possible. It was performed after injection of 80–120 mL of ioxitalamate (Telebrix 35 [350 mg of iodine per milliliter]; Guerbet, Roissy, France) in an antecubital vein at a rate of 3 mL/sec with a scanning delay of 25–30 seconds, a collimation of 3 mm, a table speed of 3 mm/sec, and a gantry rotation period of 1 second (40–50 rotations). All patients received 120 mL of contrast material except for two: one patient weighing 31 kg received 80 mL and one patient weighing 43 kg received 100 mL.
The patients were asked to hold their breath during the first 30 seconds of the acquisition period and to breathe slowly during the 10–20 remaining seconds. Images were reconstructed at 1-mm increments. These data sets were transferred to a workstation (EasyVision; Philips Medical Systems) where they were viewed in a cine loop. Multiplanar reformations—sagittal, coronal, and oblique—were not obtained systematically but only when the radiologist monitoring the examination online thought that the cine loop display of the transverse images did not depict the anatomic course of the renal vessels with certainty.
The duration of helical CT scanning, from the time the patient was placed in the scanner to the time of discharge from the examination room, was 15–45 minutes (mean, 25 minutes). The duration of transverse image analysis, or interactive cine loop viewing, was approximately 10 minutes. In doubtful cases, 10 additional minutes were needed for calculation and analysis of multiplanar reformations.
Renal DSA
Thirty-five patients underwent renal DSA within 2 hours following CT. In the six remaining patients, DSA was performed 2 days to 6 months later. All the patients underwent imaging with use of a V 3000 angiographic system (Philips Medical Systems). The angiographic procedures were performed through the femoral artery with use of the Seldinger technique. Flush aortography in the anteroposterior projection was performed with a 5-F straight catheter and injection of 30 mL of low-osmolality contrast material (ioxaglate [320 mg of iodine per milliliter], Hexabrix 320; Guerbet) at a rate of 15 mL/sec. Bilateral selective renal injections were performed systematically with a 5-F catheter (Shepherd Hook; Terumo, Tokyo, Japan) in each patient.
The amount of contrast material injected during the angiographic procedure was 50–160 mL (mean, 100 mL). For the six patients who did not undergo DSA the same day as helical CT scanning, an additional 100 mL of ioxitalamate had to be injected intravenously 10–30 minutes before the angiographic examination to opacify the renal collecting system.
The duration of renal DSA, from the time the patient was placed on the angiography table to the time of discharge from the examination room, was 15–80 minutes (mean, 45 minutes).
Image Analysis
Helical CT angiography and renal DSA were monitored online by staff radiologists experienced in each procedure. The examinations were interpreted by one referring radiologist for each procedure (CT angiography, O.R.; DSA, D.L.) in a blinded manner so that reviewers were not aware of the results of the other imaging study.
Renal DSA images and helical CT angiograms were assessed for the following features:
1. Study quality. The overall subjective study quality was assessed by using a four-point scale. This scale was used to evaluate motion artifacts and contrast enhancement features, as follows: 1, standard quality; 2, less than standard quality but did not affect diagnosis and especially did not affect the depiction of the anatomic course of the renal arteries; 3, diagnosis was in doubt and especially the anatomic course of the renal arteries was not clearly depicted (for CT angiography it could be clearly depicted neither on cine loop display of the transverse images nor on multiplanar reformations); 4, inadequate quality for diagnosis.
2. Number of renal arteries supplying each kidney. Arteries with early prehilar bifurcation were regarded as single vessels.
3. Presence or absence of lower pole vessels crossing the UPJ. Only vessels crossing immediately adjacent to the UPJ were regarded as crossing vessels; vessels passing within 1 cm of the UPJ but not immediately adjacent to the UPJ were not considered crossing vessels.
4. Complications of each procedure.
Statistical Evaluation
For the assessment of the number of renal arteries and for the calculation of the sensitivity, specificity, positive predictive value, and negative predictive value of CT angiography in the depiction of crossing arteries, DSA served as the standard of reference. For the depiction of the anatomic course of the renal veins and for the location (anterior or posterior to the UPJ) of the crossing vessels, DSA was not regarded as a standard of reference.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Number of Renal Arteries
DSA with bilateral selective renal injections showed 126 renal arteries in the 41 patients; 56% of patients had supernumerary renal arteries. Nineteen (45%) of the 42 kidneys with UPJ obstruction and 15 (38%) of the 40 nonobstructed kidneys had more than one artery.
CT angiography showed 121 renal arteries. Five arteries were seen at DSA but not at CT angiography. Four supplied a nonobstructed kidney, and one supplied a kidney with a UPJ obstruction but did not cross the UPJ. An upper pole accessory artery was totally outside of the scan volume. Two lower pole accessory arteries originating from the common iliac artery (n = 1) or from the last centimeter of the aorta (n = 1) were partially outside of the scan volume, but their insertion into the kidney was visible in retrospect. In the fourth kidney, which had four arteries, a small lower pole accessory artery was missed because of motion artifacts. It was not clearly visible in retrospect. The fifth kidney had a lower pole accessory artery that was depicted as an early branching vessel at CT angiography. DSA with selective injections showed there were two independent arteries. In retrospect, their separate origins were visible on transverse CT images.
Thus, if DSA with selective injections is used as the standard of reference, CT angiography could prospectively depict 121 (96%) of the 126 renal arteries. Retrospectively, 124 (98%) renal arteries were visible on CT images.
UPJ Crossing Vessels
On DSA images, 12 (29%) of the 42 kidneys with UPJ obstruction had identifiable arteries crossing the UPJ (Figs 1, 2).
|
|
|
|
|
|
|
|
Thus, if DSA is used as the standard of reference, CT angiography was 100% sensitive and 96.6% specific for depicting crossing arteries.
Complications
No major complication occurred at CT angiography or DSA. After CT angiography, four patients had renal colic that rapidly resolved after antispasmodic treatment. Five patients had transient vagal reaction during DSA.
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Helical CT theoretically has many advantages over DSA. It is less expensive and less invasive; it can be performed easily on an outpatient basis, thus allowing decisions about the surgical technique in advance; and it can depict nonvascular abnormalities.
Helical CT seems to provide an accurate depiction of arterial supply to the kidneys. Results of previous studies (16,20–23) have shown that there was agreement between helical CT and conventional angiographic anteroposterior projections in the number of renal arteries depicted in 89%–100% of kidneys. However, some of the discrepancies between the two modalities could be due to angiographic errors, as selective renal injections were not performed systematically (16,20–23). When compared with selective renal angiography or with surgical findings, helical CT proved to depict 96%, as in our study, to 100% of renal arteries (20,24). One of the main causes of CT errors is that some polar arteries are not included in the scan volume because of insufficient scan coverage, because patients suspend respiration differently for the helical CT scan than for the localizing scan obtained before the administration of contrast material, or both (16). However, this problem of scan coverage is not an important drawback for the preoperative assessment of UPJ obstructions, as only vessels close to the UPJ are important to image.
Study results (17,18) have shown that helical CT angiography could not only localize the UPJ and the proximal ureter in almost all patients with UPJ obstruction, even in the absence of a ureteral stent, but also depict the vascular anatomy at the UPJ. However, since most of the patients undergo endopyelotomy, the literature provides only few correlations between helical CT and surgical findings.
Intraarterial angiography is the current standard of reference in the preoperative evaluation of the renal arterial anatomy. Our work shows that there is a good correlation between helical CT angiography and DSA for diagnosing crossing arteries in patients with UPJ obstruction. When helical CT shows no vessel around the UPJ, the diagnosis is easy, and a crossing artery can be ruled out with certainty. Most of the time, crossing arteries are also easily recognized by viewing the transverse sections in a cine loop, which allows good depiction of the anatomic course of the vessels. However, in some cases, vessels are found close to the UPJ, but it is difficult to know whether they cross adjacent to the UPJ. In such cases, multiplanar reformations are useful for delineating the relationship of those vessels to the UPJ and allow a more certain diagnosis.
The potential need for high-quality multiplanar reformations, at least in some cases, emphasizes the importance of a good imaging protocol. We think that thin (3 mm or less), largely overlapping transverse sections are mandatory. We agree with Quillin et al (17) that an early, or arterial, phase of imaging should be preferred to a delayed, or equilibrium, phase because vessels are more clearly depicted and because arteries and veins can be differentiated more reliably. The UPJ is usually easily localized without opacification of the renal collecting system and, most of the time, an excretory phase of imaging is useless. However, in some instances, when the anatomy is difficult to localize in an earlier phase of imaging, an excretory phase of imaging may be helpful in identifying the UPJ.
CT exploration is more difficult when the renal pelvis is poorly distended because normal renal vessels are close to the UPJ and may be misdiagnosed as crossing vessels. We think that pelvic distention by means of diuretic injection provides better conditions of exploration, and, in our series, we used it systematically. However, it is probably not necessary in all cases and could also be used selectively, depending on the degree of pelvic distention on the initial nonenhanced scans. Diuretic injection in patients with UPJ obstruction occasionally may lead to renal colic pain, but this complication was rare in our experience and, when present, resolved rapidly after antispasmodic treatment.
Low-osmolality contrast materials have been shown to induce less patient motion than high-osmolality contrast materials during helical scanning (25). As helical CT angiography requires rapid iodine injection and high-quality reformations, we think that low-osmolality contrast materials should be preferred. We recently changed our imaging protocol and now systematically use low-osmolality contrast materials for CT angiography, even in low-risk patients. Breathing is one of the main sources of motion at CT scanning, and motion artifacts also depend on the scanning time. Our imaging protocol was chosen to obtain high-quality reformations and broad coverage in the z direction, but it necessitates a long data acquisition time. Decreasing the z direction coverage, as only the UPJ is important to image, slightly increasing the pitch ratio, or both probably would have reduced motion artifacts in the patients in our study without dramatically impairing the diagnostic quality of the studies. Thus, the best imaging protocol is still to be determined.
In this study, we focused on arteries that cross immediately adjacent to the UPJ because our referring urologists think that open Anderson-Hynes dismembered pyeloplasty is more suitable than endopyelotomy for patients with such crossing arteries.
Other urologists, especially those who use fluoroscopy-guided cutting balloon catheters, may be interested in vessels that may complicate endopyelotomy. In these instances, all the vessels (arteries and veins) within 1–2 cm of the UPJ are of interest, and their anatomic location (anterior or posterior to the UPJ) should be depicted accurately. DSA, even with oblique projections, does not always clearly depict the anatomic position (anterior or posterior to the UPJ) of the crossing arteries and should not be regarded as a standard of reference for this particular topic. Therefore, our study does not directly address the accuracy of CT in such an indication. However, we think that transverse CT images are of particular interest in distinguishing arteries that cross the collecting system anteriorly from those that cross it posteriorly. Moreover, CT depicts the anatomic course of renal veins more accurately than does DSA (16,20,22).
Therefore, our study shows that CT angiography is as accurate as DSA in depicting crossing arteries when arteries that may be the cause of obstruction are considered. We also think that CT angiography is probably more appropriate than DSA for depicting the anterior or posterior position of renal vessels (arteries and veins) within 1–2 cm of the UPJ when vessels that may cause bleeding at endopyelotomy are considered.
Other imaging modalities are now available to define the vasculature adjacent to the UPJ. Endoluminal ultrasonography allows for an area of approximately 2 cm to be imaged around the UPJ (26). In a small series of 13 patients (18), it showed a good correlation with helical CT data. It is less expensive than helical CT but necessitates general anesthesia and fluoroscopy-guided ureteral retrograde catheterization. Therefore, it is performed at the time of the surgical repair of the UPJ and does not allow the surgical technique to be discussed and scheduled in advance. Magnetic resonance (MR) angiography also will probably provide noninvasive imaging of the kidneys with UPJ obstruction. It has already been successfully used to assess the vascular anatomy of potential renal donors (27,28). The depiction of crossing arteries at MR angiography has proved feasible (29), but further studies are mandatory to assess the accuracy of this technique in such an indication.
We think that none of the current UPJ imaging techniques can discriminate crossing arteries that are the direct cause of obstruction from those that are not. Therefore, in the absence of a randomized trial comparing the outcome of endopyelotomy and open pyeloplasty, it remains controversial whether depicting a crossing artery is important for surgical management. However, the newly available imaging techniques, such as helical CT and maybe MR angiography, will probably help in settling this question, as it now seems possible to perform noninvasive imaging of the vasculature adjacent to the UPJ in each patient undergoing endopyelotomy and to correlate these data with the outcome of the procedure.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantor of integrity of entire study, O.R.; study concepts, D.L.; study design, O.R., D.L.; definition of intellectual content, O.R., D.L.; literature research, O.R.; clinical studies, A.G., X.M.; data acquisition, O.R., D.L., P.B., C.P.; data analysis, O.R., P.B.; statistical analysis, O.R.; manuscript preparation, O.R.; manuscript editing, O.R., D.L.; manuscript review, O.R., D.L., P.B., A.G., X.M.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. P. Lawler, T. W. Jarret, F. M. Corl, and E. K. Fishman Adult Ureteropelvic Junction Obstruction: Insights with Three-dimensional Multi-Detector Row CT RadioGraphics, January 1, 2005; 25(1): 121 - 134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Frauscher, O. Rouvière, D. Lyonnet, P. Berger, C. Pangaud, A. Gelet, and X. Martin Value of Contrast-enhanced Color Doppler Imaging for Detection of Crossing Vessels in Patients with Ureteropelvic Junction Obstruction Dr Rouviere and colleagues respond: Radiology, December 1, 2000; 217(3): 916 - 917. [Full Text]
|
|
|
|
|
|
A. Kawashima, C. M. Sandler, R. D. Ernst, E. P. Tamm, S. M. Goldman, and E. K. Fishman CT Evaluation of Renovascular Disease RadioGraphics, September 1, 2000; 20(5): 1321 - 1340. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
