| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (E.J.H., D.G.M., R.J.W., E.K.O.) and Surgery (M.J.M., G.A.W.), Thomas Jefferson University Hospital, 132 S 10th St, Philadelphia, PA 19107-5244. From the 1999 RSNA scientific assembly. Received August 24, 1999; revision requested October 5; revision received October 25; accepted November 2. Address correspondence to E.J.H. (e-mail: Ethan.Halpern@mail.tju.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-five living renal donors underwent preoperative contrast material–enhanced CT angiography and gadolinium-enhanced MR angiography. Each study was interpreted by two independent radiologists blinded to all other studies and to interpretations provided by other reviewers. Eighteen kidneys had surgical correlation.
RESULTS: CT demonstrated 33 supernumerary arteries in 19 patients, bilateral solitary arteries in 16 patients, and 18 proximal arterial branches in 16 patients. MR demonstrated 26 supernumerary arteries in 15 patients, bilateral solitary renal arteries in 20 patients, and 21 proximal arterial branches in 16 patients. Interobserver agreements for MR (
= 0.74) and CT ( = 0.73) were similar to the agreement between MR and CT ( = 0.74). Among the kidneys chosen for nephrectomy, one small accessory artery and one proximal arterial branch were missed with CT and MR. Two of the accessory arteries suggested at CT were not found at nephrectomy. By averaging data for both modalities, supernumerary arteries were present in 49% of kidney donors and were bilateral in approximately 17%. Proximal arterial branches were present in 46% of kidney donors.
CONCLUSION: Preoperative CT and MR angiography of the renal arteries in renal donors demonstrate substantial agreement. Interobserver disagreement in the interpretation of CT and MR angiograms is related to 1–2-mm-diameter vessels.
Index terms: Computed tomography (CT), angiography, 81.12116 • Kidney, blood supply, 961.13 • Kidney, transplantation, 81.4551 • Magnetic resonance (MR), vascular studies, 81.12142 • Renal arteries, CT, 961.12916 • Renal arteries, MR, 961.12942
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Anatomic assessment of the donor kidney is performed prior to transplantation to aid in selection of which kidney to use and to plan the surgical approach. The increasing use of laparoscopic donor nephrectomy makes this preoperative diagnosis even more important since the details of arterial and venous anatomy may be more difficult to appreciate during laparoscopic surgery. In the past, urography or ultrasonography (US) was used to evaluate for renal masses and stones and ureteral anatomy, and an arteriogram was obtained to identify the number, position, and patency of the renal arteries, as well as to determine the presence of proximal branches of the main renal artery. These imaging techniques have been supplanted largely by computed tomographic (CT) or magnetic resonance (MR) angiography. Replacing both arteriography and urography or US with a single CT or MR examination provides a less costly and less invasive evaluation of donor kidneys.
Findings of several studies have demonstrated successful application of CT angiography (3–7) for preoperative evaluation of renal donors. Initial reports (8–11) with regard to non–gadolinium-enhanced MR angiography suggested that accessory arteries are inadequately depicted with time-of-flight and phase-contrast techniques. Investigators in one study (12) compared helical CT with phase-contrast MR angiography and concluded that CT is more accurate. More recent clinical data (13) suggest that gadolinium-enhanced MR angiography is superior to the older nonenhanced MR techniques. However, to our knowledge, there has been no published article in which contrast-enhanced CT has been compared with gadolinium-enhanced MR for the evaluation of renal donors.
In the current study, we compared CT angiography performed after the administration of iodinated contrast material with MR angiography performed after the administration of gadolinium-containing contrast material for the preoperative evaluation of kidney donors. The goal of the study was to demonstrate in a paired comparison the relative advantages and disadvantages of CT versus MR for the preoperative evaluation of living renal donors.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CT Technique
CT evaluation was performed with a helical HiSpeed Advantage scanner (GE Medical Systems, Milwaukee, Wis). All examinations were performed with intravenous injection of iohexol (Omnipaque 300; Nycomed Amersham, Oslo, Norway), a nonionic contrast medium.
After an initial scout image was obtained, preliminary imaging of the abdomen was performed with administration of a small contrast material bolus. Transverse images were obtained, starting 8–10 seconds after injection of 20 mL of contrast material. This localizer series served to test the intravenous catheter, to establish the timing delay between injection and aortic enhancement, to localize the kidneys and the aortic bifurcation, to practice the timing of the breath hold with the patient, and to screen for renal stones. Arterial phase scans—the angiographic series—were next obtained from the upper poles of the kidneys through the aortic bifurcation following the injection of 130 mL of contrast medium at 4 mL/sec. All patients were evaluated with 3-mm collimation at a pitch of 1.3–1.5, with 260–280 mAs. Targeted reconstruction images were obtained from the helical data with a 22-cm field of view at 1.5-mm intervals.
To evaluate the renal parenchyma in the nephrographic phase of contrast medium enhancement, an additional helical series was obtained through the kidneys at approximately 1 minute after injection. This series was obtained with 5-mm collimation at a pitch of 2. Finally, a delayed scout image was obtained to evaluate the ureters.
MR Technique
MR images were obtained with a HiSpeed Signa 1.5-T magnet (GE Medical Systems; software version 5.6). Precontrast images of the renal parenchyma were obtained with a T2-weighted single-shot fast spin-echo technique by using an infinite repetition time and an effective echo time of approximately 100–180 msec. To establish the proper timing for MR angiography, a test injection of 2 mL of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) was used followed by a 20-mL isotonic saline flush. A single, fast multiplanar spoiled gradient-echo 2-cm-thick sagittal image was obtained of the abdominal aorta; this was repeated once per second for 45 seconds after injection. The arrival time of the contrast medium was noted for this sequence.
Contrast-enhanced three-dimensional breath-hold fast spoiled gradient-echo images were obtained. Parameters included a repetition time (approximately 6 msec) and echo time (approximately 2 msec) set to the minimum possible, with a 45° flip angle and 0.5 signal acquired. Fat saturation was employed with a chemically selective reduced flip angle inversion pulse by using an inversion time of 32 msec (6/2/32).
Coronal images were obtained with a 3-mm section thickness but were zero-fill interpolated to yield contiguous images at 1.5-mm increments. A precontrast series was obtained with these settings; the images depicted ureters by means of enhancement from the test bolus. Following the intravenous administration of 28 mL of gadolinium-containing contrast medium and a 20-mL saline flush, the angiographic sequence was repeated three times in rapid succession. The start of the first postcontrast sequence was timed with the arrival time of contrast medium after the test injection. Each of the three sequences required approximately 25 seconds, with an additional 10–15 seconds for the patient to breathe between sequences. Venous enhancement was often visible with all three sequences but was most optimal with the second sequence.
Image Analysis
Two independent radiologists reviewed the images from each CT examination (E.J.H., R.J.W.) and each MR examination (D.G.M., E.K.O.) at a picture archiving and communication system, or PACS, workstation with cine capability (Cannon USA; Los Angeles, Calif). Each of the four reviewers is an experienced cross-sectional imager and had participated in a prior clinical comparison of cross-sectional angiography and conventional arteriography (14). Images from each CT and MR examination were interpreted by a primary reviewer at the time of the examination and were subsequently interpreted by a second independent radiologist. Reviewers were blinded to all other images and to the interpretations provided by the other reviewers. Each primary reviewer created maximum intensity projection (MIP) renderings for every study subject at the time of the initial clinical review. MIP images were rendered for both CT and MR with a Windows Advantage workstation (GE Medical Systems) and were archived for subsequent review by the second independent radiologist.
A study worksheet was completed independently by each reviewer for each examination; the worksheet detailed the numbers and sizes of renal arteries found on each side. For each artery, the presence of any stenosis or of a proximal branch was noted. The distance of each proximal branch from the aorta was recorded. For statistical analysis, any branch within 2.0 cm from the aorta was classified as a proximal branch. Renal venous anatomy was evaluated for the presence of accessory veins, retroaortic veins, and circumaortic veins. The number of veins and ureters associated with each kidney was recorded.
A transplant surgeon (M.J.M. or G.A.W.) completed a separate worksheet for each subject who underwent nephrectomy. The surgical worksheet identified the number of arteries and size of each artery and the presence of any proximal branch. Any unexpected surgical findings were described. The surgeon commented on whether CT and/or MR images were used to determine which kidney was chosen for nephrectomy and whether there was a diagnostic advantage to either CT or MR.
Statistical Evaluation
The frequency of multiple renal arteries and proximal arterial branches was tabulated on the basis of the readings of the primary reviewer for each CT and MR angiogram. Each frequency value was computed "by kidney" and "by subject." Since each subject represents an independent observation, 95% CIs were computed only for the "by subject" frequencies, but not for the "by kidney" frequency because the imaging findings for the two kidneys within each subject are not independent.
To demonstrate the degree of interobserver agreement for the presence of multiple renal arteries and the presence of proximal (within-2-cm) branches, a statistic was computed. The statistic was computed to demonstrate two types of agreement: intermodality agreement between the two primary readers and interobserver agreement within each modality. Data were analyzed by calculating 95% CIs and statistics.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Supernumerary Arteries
According to the reading by the primary reviewer, CT demonstrated 33 supernumerary arteries involving 26 kidneys in 19 patients and bilateral solitary arteries in 16 patients. The primary and secondary reviewers agreed on the number of arteries to 57 of 70 kidneys. The secondary reviewer suggested one additional artery to nine kidneys and one fewer artery to four kidneys (Table 1). Interobserver agreement for the presence of multiple arteries to a kidney was indicated by a statistic of 0.73. All discordant readings were among arteries measured as less than or equal to 2 mm in diameter.
|
statistic of 0.74. All of the discordant readings were among arteries classified as less than or equal to 2 mm in diameter.
|
statistic of 0.74.
|
Figure 1 demonstrates MIP renderings for a subject with multiple renal arteries as depicted by both CT and MR. Although larger supernumerary arteries were visible on both (a) the original transverse or coronal images and (b) the MIP renderings, smaller supernumerary vessels were often difficult to visualize on MIP renderings. Very small vessels were best defined with careful evaluation of the original transverse CT or coronal MR image by using the cine feature of the workstation.
|
|
= 0.65) and in 58 of 70 kidneys on MR angiograms ( = 0.53). On the basis of our results with both CT and MR, proximal arterial branches are present in 46% of kidney donors. The intermodality agreement for the presence of proximal arterial branches was indicated by a statistic of 0.54.
Venous and Ureteral Anatomy
Circumaortic left renal veins were identified in four subjects on CT and in eight subjects on MR angiograms. The presence of a large retroaortic venous component in one of these patients contributed to the decision to opt for a right nephrectomy. One subject with a retroaortic left renal vein and no vein anterior to the aorta was identified with both modalities. Multiple right renal veins were identified in eight subjects with CT and seven subjects with MR. Single ureters were identified in all of the study subjects.
Surgical Correlation
Surgical correlation was available for 18 kidneys in which nephrectomy was performed. Twelve left nephrectomies were performed, including two laparoscopic nephrectomies. Six right nephrectomies were performed. Twenty-three renal arteries were transplanted, including three right-sided supernumerary arteries and two left-sided supernumerary arteries. Twenty-two of the 23 vessels were identified with both CT and MR. One 2-mm upper pole accessory artery to a left kidney was interpreted as a proximal upper pole branch of the main renal artery on CT and was not detected on MR angiograms (Fig 2). There was no adverse effect to the patient.
|
|
|
|
Two of the very small (1-mm) accessory arteries suggested at CT were not found at nephrectomy (Fig 3). Transplant surgeons observed that both CT and MR were helpful in selecting which kidney to transplant and in deciding whether to use an open or a laparoscopic approach. All of the transplantations were successful with respect to surgical technique. No diagnostic advantage was noted for either CT or MR.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although there was substantial agreement between CT and MR angiographic findings, there were intermodality and intramodality differences in interpretation with regard to small (
2-mm) accessory arteries and proximal branches. These disagreements may result, in part, from anisotropic volume acquisition with lower z-axis resolution of CT data and lower anterior-to-posterior resolution of MR data (15). However, since intermodality agreement between CT and MR findings was similar to interobserver agreement within CT and within MR, our findings suggest that the choice of reviewer is at least as important as the choice of diagnostic modality.
A major limitation in the design of this study is the lack of a standard of reference. Although surgical correlation was available, surgeons usually chose to avoid the kidney with more complex anatomy. When there was a disagreement between CT and MR findings on the presence of an accessory artery or a proximal branch, the kidney in question was most often not chosen for nephrectomy. This choice introduces an unavoidable element of bias into our series with surgical correlation.
Nephrectomy was performed in two subjects in whom CT and MR findings disagreed on a left upper pole accessory artery. No accessory renal artery was found in these two subjects. For one of these subjects, the surgical report noted an adrenal artery with a branch to the upper pole of the kidney (Fig 3). When such vessels are found during nephrectomy, they are transiently occluded, and, if no ischemic changes are noted, these vessels are ligated. Thus, the two false-positive arteries in our series likely represent adrenal arteries that contacted the upper pole of the kidney. In this instance, modern cross-sectional imaging may be depicting vessels that are not clinically important.
To our knowledge, the frequency of accessory renal arteries in this study is higher than that reported for either an autopsy series or arteriography. In 1958, Merklin and Michels (16) collected accounts of the experiences of 45 authors who examined 10,987 kidneys at autopsy. As summarized in a current textbook (17, table 17.5), their series demonstrated a single renal artery without proximal branching in 71.1% of kidneys. Proximal arterial branches were noted in less than 20% of kidneys. Fewer than 30% of kidneys demonstrated more than one renal artery arising from the aorta (17). In a series of 400 patients (18), the prevalence of multiple renal arteries at aortography was reported at 25%.
A mean of our CT and MR results suggests that multiple renal arteries supply approximately one-third of all kidneys and are found in about 49% of renal donors (95% CI: 32%, 65%). Proximal arterial branches were present in the arteries supplying 28% of all kidneys in our series and were found in 46% of renal donors (95% CI: 29%, 62%). As demonstrated in Figure 3, some of these smaller arteries may not be of clinical importance. Furthermore, it is likely that the visualization of these extremely small branches results in greater interobserver disagreement for CT and MR than for conventional arteriography. Although some of the small vessels seen at CT and MR may represent false-positive findings, the relatively good interobserver and intermodality agreements in our study suggest that some of these vessels are real arteries that are simply not appreciated at conventional arteriography or at autopsy (19).
The reviewers for this study evaluated both the original transverse or coronal images and the MIP renderings. We chose to include MIP images for each subject even though all of the anatomic information obtained with CT and MR is present on the original transverse or coronal images. The MIP renderings, which are created from the transverse, sagittal, and coronal sections, provide a more global view of the anatomy, although they often fail to demonstrate the smaller vessels accurately (Figs 1, 2). Our reviewers evaluated all images but tended to defer finer anatomical questions to review of the original transverse images for CT and coronal images for MR. The rendered MIP images were slightly more helpful for MR than for CT. This may be due to the fact that renal arteries are usually easily followed on transverse CT images.
Initially, our surgical colleagues requested the MIP images for all cases. However, through the course of the study it became clear that there were often important anatomic details that were not clearly demonstrated on MIP renderings. It is possible that some of the newer volume-rendering techniques may improve the utility of three-dimensional renderings and decrease the number of images that must be reviewed (20). Our current experience suggests that all of the original sections should be scrutinized for small accessory arteries and branches.
In the past, renal venous anatomy has not been a major factor in surgical planning for renal donation. However, venous anatomy may be increasingly important because of the preference for laparoscopic left nephrectomy over right nephrectomy. As demonstrated in our series, circumaortic and retroaortic left renal veins are found in a substantial minority of donors. Our reviewers were all in agreement with respect to the larger retroaortic venous structures. Disagreement between CT and MR findings with respect to the presence of circumaortic left renal veins was related to differing interpretations of very small retroaortic venous structures. The presence of retroaortic or lumbar branches, as well as their sizes and locations, is valuable information to the surgeon. At our institution, large retroaortic renal veins and complex arterial anatomy are both considered relative contraindications to the laparoscopic approach. Both CT and MR angiography may be used to depict renal venous anatomy.
In the present study, we compared contrast-enhanced CT and MR angiography for the preoperative evaluation of renal donors. CT and MR provide similar information for the preoperative planning of renal donation. Although there was substantial agreement between these modalities, disagreements were found in the interpretation of small arteries and veins. With respect to differences between CT angiography and MR angiography, interobserver intramodality variation was as important as variation related to the choice of modality. If CT and MR angiography are to be successful diagnostic tools for the evaluation of vascular anatomy, radiologists should be trained in cross-sectional vascular imaging to minimize interobserver variation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, E.J.H.; study concepts, all authors; study design, E.J.H.; definition of intellectual content, all authors; literature research, E.J.H.; clinical studies, all authors; data acquisition, all authors; data analysis, E.J.H.; statistical analysis, E.J.H.; manuscript preparation, E.J.H.; manuscript editing and review, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Holden, A. Smith, P. Dukes, H. Pilmore, and M. Yasutomi Assessment of 100 Live Potential Renal Donors for Laparoscopic Nephrectomy with Multi-Detector Row Helical CT Radiology, December 1, 2005; 237(3): 973 - 980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. J. M. Kock, J. N. M. Ijzermans, K. Visser, S. M. Hussain, W. Weimar, P. M. T. Pattynama, G. P. Krestin, and M. G. M. Hunink Contrast-Enhanced MR Angiography and Digital Subtraction Angiography in Living Renal Donors: Diagnostic Agreement, Impact on Decision Making, and Costs Am. J. Roentgenol., August 1, 2005; 185(2): 448 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Kim, J. H. Kim, S.-J. Bae, and K.-S. Cho CT Angiography for Evaluation of Living Renal Donors: Comparison of Four Reconstruction Methods Am. J. Roentgenol., August 1, 2004; 183(2): 471 - 477. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Matsuki, H. Kani, F. Tatsugami, S. Yoshikawa, I. Narabayashi, S.-W. Lee, H. Shinohara, E. Nomura, and N. Tanigawa Preoperative Assessment of Vascular Anatomy Around the Stomach by 3D Imaging Using MDCT Before Laparoscopy-Assisted Gastrectomy Am. J. Roentgenol., July 1, 2004; 183(1): 145 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-K. Kim, S.-Y. Park, H.-j. Kim, C.-S. Kim, H.-J. Ahn, T.-Y. Ahn, and K.-S. Cho Living Donor Kidneys: Usefulness of Multi-Detector Row CT for Comprehensive Evaluation Radiology, December 1, 2003; 229(3): 869 - 876. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kawamoto, R. A. Montgomery, L. P. Lawler, K. M. Horton, and E. K. Fishman Multidetector CT Angiography for Preoperative Evaluation of Living Laparoscopic Kidney Donors Am. J. Roentgenol., June 1, 2003; 180(6): 1633 - 1638. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Hussain, M. C. J. M. Kock, J. N. M. IJzermans, P. M. T. Pattynama, M. G. Myriam Hunink, and G. P. Krestin MR Imaging: A "One-Stop Shop" Modality for Preoperative Evaluation of Potential Living Kidney Donors RadioGraphics, March 1, 2003; 23(2): 505 - 520. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. K. Willmann, S. Wildermuth, T. Pfammatter, J. E. Roos, B. Seifert, P. R. Hilfiker, B. Marincek, and D. Weishaupt Aortoiliac and Renal Arteries: Prospective Intraindividual Comparison of Contrast-enhanced Three-dimensional MR Angiography and Multi-Detector Row CT Angiography Radiology, March 1, 2003; 226(3): 798 - 811. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Liem, M. C. J. M. Kock, J. N. M. Ijzermans, W. Weimar, K. Visser, and M.G. M. Hunink Living Renal Donors: Optimizing the Imaging Strategy--Decision- and Cost-effectiveness Analysis Radiology, January 1, 2003; 226(1): 53 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Israel, V. S. Lee, M. Edye, G. A. Krinsky, M. T. Lavelle, T. Diflo, and J. C. Weinreb Comprehensive MR Imaging in the Preoperative Evaluation of Living Donor Candidates for Laparoscopic Nephrectomy: Initial Experience Radiology, November 1, 2002; 225(2): 427 - 432. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Jha, S. J. Korangy, S. M. Ascher, J. Takahama, P. C. Kuo, and L. B. Johnson MR Angiography and Preoperative Evaluation for Laparoscopic Donor Nephrectomy Am. J. Roentgenol., June 1, 2002; 178(6): 1489 - 1495. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U J Schoepf, C R Becker, R D Bruening, B M Ohnesorge, A Huber, L-G Haw, H Hildebrandt, and M F Reiser Multislice CT angiography Imaging, December 15, 2001; 13(5): 357 - 365. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Glockner Three-dimensional Gadolinium-enhanced MR Angiography: Applications for Abdominal Imaging RadioGraphics, March 1, 2001; 21(2): 357 - 370. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
