Genitourinary Imaging Original Research

Transcription

1 Raman et al. MDCT of Normal and Variant Renal natomy Genitourinary Imaging Original Research 105.fm 11/30/06 C M E D E N T U R I C L I M G I N G JR 2007; 188: X/07/ merican Roentgen Ray Society Y Steven S. Raman 1 Suwalee Pojchamarnwiputh 1,2 Kobkun Muangsomboon 1,3 Peter G. Schulam 4 H. lbin Gritsch 4 David S. K. Lu 1 O F Raman SS, Pojchamarnwiputh S, Muangsomboon K, Schulam PG, Gritsch H, Lu DSK Keywords: anatomy, genitourinary tract imaging, kidney disease, kidney transplantation, living liver donor, MDCT DOI: /JR Received June 10, 2005; accepted after revision December 7, Department of Radiology, David Geffen School of Medicine at the University of California at Los ngeles, BL-428 CHS/Box , Los ngeles, C ddress correspondence to S. Raman (sraman@mednet.ucla.edu). 2 Present address: Department of Radiology, Chang Mai University, Chang Mai, Thailand. 3 Present address: Department of Radiology, Siriraj Hospital, Mahidol University, Bangkok, Thailand. 4 Department of Urology, University of California at Los ngeles, Los ngeles, C. Surgically Relevant Normal and Variant Renal Parenchymal and Vascular natomy in Preoperative 16-MDCT Evaluation of Potential Laparoscopic Renal Donors OBJECTIVE. Using 16-MDCT, we describe and quantify the frequency and types of renal anatomic variants and findings relevant for preoperative evaluation and surgical planning for potential laparoscopic renal donors. MTERILS ND METHODS. On 16-MDCT, 126 consecutive potential donors underwent scanning before contrast administration and after IV power injection of nonionic contrast material during the arterial, nephrographic, and excretory phases. On a 3D workstation, CT images were evaluated retrospectively in consensus by three abdominal imagers. The number and branching pattern of bilateral renal arteries and veins, including anomalies of the inferior vena cava and lumbar gonadal axis, were categorized along with the frequency of incidental findings of the renal parenchyma and collecting system. RESULTS. Major arterial variants including supernumerary and early branching arteries were present in 16% and 21%, respectively, of left kidneys and 22% and 15%, respectively, of right kidneys. Major and minor venous variants were detected in 11% and 58% of left kidneys and 24% and 3% of right kidneys. Late confluence of the venous trunk was identified in 17% of left kidneys and 10% of right kidneys. Incidental parenchymal and urothelial abnormalities, most commonly cysts and calyceal calcifications, were identified in 30% of the kidneys. Other relevant incidental findings included focal infarcts, cortical scars, atrophic scarred kidney, and bilateral papillary necrosis. Urothelial variants included bilateral simple ureteroceles and rightsided complete duplicated collecting system. CONCLUSION. 16-MDCT angiography and urography allow confident detection and classification of a variety of anatomic and incidental anomalies relevant to the preoperative selection of potential laparoscopic renal donors and to surgical planning. ince 1995, laparoscopic nephrectomy has supplanted traditional S open donor nephrectomy as the preferred method to harvest a kidney for living donor transplantation [1]. dvantages of the laparoscopic technique include decreased morbidity, decreased recovery time, decreased pain, and improved cosmesis [2]. Laparoscopic donor nephrectomy entails a steep learning curve and can be technically challenging because of the limited surgical field of view. Certain areas in the laparoscopic operative field, such as the posterior aspect of the renal veins, are difficult to visualize when compared with the operative field in open donor nephrectomy [3 5]. The left kidney is preferred for laparoscopic living donor nephrectomy because of its relative technical ease of removal and flexibility afforded by the longer left venous pedicle [2]. relative drawback of left-sided donor kidneys, however, is the high proportion of venous variants, including collateral venous pathways related to the left renal vein [6]. Knowledge of these variants is relevant because these potentially large and confounding vessels related to the lumbar and gonadal veins usually anastomose to the posterior aspect of the renal veins where laparoscopic visualization is limited. Preoperative knowledge of venous variants may help the surgeon anticipate these anomalies and avoid inadvertent ligation of transection of these vessels, which may cause unanticipated hemorrhage in the laparoscopic field. Thus, laparoscopic surgeons demand more information from preoperative crosssectional imaging with regard to arterial and venous anomalies both for donor and kidney selection and for surgical planning. For example, a less desirable right donor nephrectomy may be performed if complex vascular anat- JR:188, January

2 Raman et al. TBLE 1: CT Parameters for Preoperative Evaluation of 126 Potential Kidney Donors Scanning Parameters Unenhanced Scan rterial Phase omy (e.g., multiple arteries or veins) is present in the left kidney. Preoperative imaging also helps identify the less normal kidney (i.e., with incidental findings such as a small stone or hemorrhagic cyst), which is usually chosen in living donor transplantation. To date, however, a comprehensive radiologic assessment of the types and frequency of surgically relevant bilateral arterial, venous, and parenchymal variants using contemporary MDCT equipment and protocols is lacking in the donor population. Knowledge of the types and frequency of variant vascular anatomy, especially venous anatomy, has been derived largely from autopsy and surgical series [7]. Prior studies assessing the role of CT have used older scanners including single- [8 16], dual-, or quad-detector scanners [7, 17 24]. Most studies in the radiology literature have been primarily designed to assess the performance of imaging in comparison with surgery rather than to describe the types and frequency of renal arterial and venous variants. By design, most studies are limited mainly to the left side and are inherently biased toward donors with the simplest anatomy because those with complex arterial anatomy are excluded. Description of venous anatomy has been limited to major variants, such as circumaortic left renal vein or duplicate inferior vena cava (IVC). The frequency of variant venous anatomy related to the lumbar gonadal axis, directly relevant for laparoscopic surgical planning, has not been well described. The advent of 16-MDCT has enabled improved temporal and spatial resolution, decreased motion and partial volume artifacts, and near isotropic data acquisition. With improved postprocessing software, the data Nephrographic Phase Time None Bolus tracking 85 s after arterial phase Excretory Phase 5 min after nephrographic phase Region scanned T11 to iliac crest T11 S2 T11 S2 T11 to pubic symphysis kvp Effective ms Table speed (mm) Gantry rotation (s) Collimation (mm) Reconstruction Thickness (mm) Space (mm) may be displayed to maximum advantage using a variety of techniques. lthough 16- MDCT has been shown to be excellent for detection of normal and variant anatomy in a subset of patients undergoing left-sided donation [25], this sample is biased toward a donor cohort largely free of complex vascular anomalies. However, this sample is biased toward a healthy population and does not account for the spectrum of right- and left-sided variants. To our knowledge, no study has described and quantified the frequency of incidental renal anomalies and the bilateral prevalence of arterial and venous variants, especially the incidence of minor venous variants, in a laparoscopic donor population on 16-MDCT. The purpose of this study was to describe and quantify the frequency of bilateral laparoscopically relevant renal vascular, parenchymal, and urothelial variants using preoperative 16-MDCT angiography and urography in a cohort of potential renal donors. Materials and Methods Patients n institutional review board exemption was obtained for this study. From June 2003 to pril 2004, 126 consecutive living renal donors (57 men and 69 women; age range, years; mean age, 39.7 years) underwent preoperative IV contrast enhanced 16-MDCT examination before the planned laparoscopic nephrectomy. ll patients were enrolled in a renal transplantation program in our institution. MDCT Protocol Images of all 126 patients were obtained using a 16-MDCT scanner (Sensation 16, Siemens Medical Solutions) with the same protocol consisting of images acquired in the unenhanced, arterial, nephrographic, and excretory phases. fter fasting for at least 3 hours, each patient ingested 250 ml of water during a 15- to 20-min period before scanning began to enable improved distention of the collecting system. n explanation of the CT procedure and the breathing instructions were then given to each patient by the CT technologist. ll phases were performed during expiration. Unenhanced CT scans were obtained from the 11th thoracic vertebral body to iliac crest with the parameters shown in Table 1. These scans were used to localize the kidneys and to provide diagnostic information (e.g., show renal calculi and provide baseline attenuation measurements of renal masses or other unexpected findings). Using a power injector, ml of 350 mg/ml nonionic contrast material (iohexol [Omnipaque 350, GE Healthcare]) was IV injected a rate of 4.0 ml/s via an 18-gauge peripheral line in an antecubital vein. The estimated dose was determined on the basis of patient weight as follows: weight of less than 100 lb (45 kg), 100 ml; lb (45 90 kg), 125 ml; and greater than 200 lb (90 kg), 150 ml. The start time of arterial phase scanning was determined using automatic bolus tracking (Smart Prepare, Siemens Medical Solutions). Scanning was initiated 5 seconds after a threshold of 150 H was reached in the region of interest within the abdominal aorta just above the kidneys. For the arterial phase, scanning was initiated approximately 20 seconds after bolus injection, and the area of coverage included T11 to the S2 level to allow coverage of the common iliac artery bifurcation. For the nephrographic phase, scanning began 85 seconds after the end of the arterial phase with the same scanning parameters (Table 1). For the excretory phase, scanning began 5 minutes after the nephrographic phase using the parameters in Table 1, and the area scanned ranged from above the kidneys to the bladder base. We decreased the milliampere-seconds setting (ms) by 50% to decrease the radiation dose to the donor. We also administered 250 ml of saline IV after the nephrographic scan and before the excretory phase scan to help decrease excreted iodine concentration of contrast medium and increase urine volume. During the 5-minute delay before scanning donors were asked to rotate three times on the CT table to decrease layering and improve the homogeneity of bladder opacification. Image Processing and nalysis The volumetric imaging data were reviewed on a workstation with 2D and 3D capability (Vitrea 2, Vital Images) by three abdominal imaging radiologists individually. ny discrepancies were resolved in consensus to serve as a gold standard. For each 106 JR:188, January 2007

3 MDCT of Normal and Variant Renal natomy CT examination, the reviewers studied the source axial images, which were supplemented by multiplanar reformations (MPRs), volume rendering, and maximum intensity projections (MIPs) as necessary. The reviewers edited CT volume data sets to create optimal 3D MDCT angiography and MDCT urography images in real time. Coronal, sagittal, and curved MPR images were used to evaluate vascular anatomy. For 3D MDCT angiography, volume-rendering techniques were usually used, but MIP rendering was also used as an adjunct display. The reviewers described and categorized the number and branching patterns of arteries and veins; the presence of arterial abnormalities, including mural stenosis and plaque; and the presence of incidental congenital and acquired renal abnormalities. Renal artery evaluation In general, knowledge of vascular anomalies is needed for donor and kidney selection and for preoperative planning. For assessment of the arteries, we retrospectively evaluated the number, branching pattern, and morphology of the renal arteries bilaterally. Supernumerary renal arteries were those that had a separate origin from the aorta or iliac arteries that was independent of the main renal arteries. In laparoscopic living renal donor transplantation, identifying the distance between the aorta and the takeoff of the first right or left renal arterial branch from the main renal artery is important because that length determines the number of arterial anastomoses to be performed in the recipient. short neck, also known as an early branching renal artery, exists when the first renal arterial branch takes off between 1 and 2 cm from the origin of the renal artery and is determined by surgical preference [8, 9, 20, 23]. In this instance, either an arterioarterial anastomosis may be performed after harvest of the donor kidney, before transplantation, or two separate arterial anastomoses may be performed in the recipient. t our institution, we define an early branch of the left side to be present when the first branch originates within 2.0 cm from the left lateral wall of the aorta. On the right side, an early arterial bifurcation exists when the first branch arises proximal to the right lateral margin of the IVC (on the undersurface of the IVC) (Fig. 1). We also evaluate the main and branch renal artery morphology with attention to mural calcification, stenosis, asymmetric mural thickness, and beading pattern related to presumed medical fibromuscular dysplasia. Renal vein evaluation For venous evaluation, we retrospectively evaluated the number and anastomotic pattern of renal veins with attention to lumbar and gonadal veins. s with the arteries, the number and branching pattern of veins are important for laparoscopic donor nephrectomy and anastomoses in the recipient. If branches of the renal vein coalesce near the lateral wall of the aorta (termed a late confluence ), then a short neck exists and, analogous to an early arterial branch, two separate veins may be resected. This may result in a back table venovenous anastomosis or possibly two venous anastomoses in the recipient. t our institution, a late venous confluence is diagnosed on the left side when venous branches converge within 1.5 cm from the left lateral wall of abdominal aorta and on the right side when venous branches converge within 1.5 cm of the confluence with the IVC (Fig. 2). Renal venous anomalies were subclassified as major or minor. Major renal venous anomalies on the left side included variants that affected recipient venous anastomosis, such as duplicated IVC or circumaortic left renal vein, or that altered donor laparoscopic dissection, as in cases of retroaortic left renal vein. Major right-sided venous anomalies included primarily supernumerary veins. Minor venous anomalies bilaterally were those that influenced donor laparoscopic dissection planning but did not alter recipient venous anastomoses and included anomalies associated with the lumbar, gonadal, adrenal, or retroperitoneal veins, including large gonadal and lumbar veins (> 5 mm) and their associated confluence with the main or branch renal veins. Preoperative knowledge of minor variants (a venous road map ) enables our laparoscopic surgeons to anticipate these often large and confusing vessels during dissection posterior to the renal vein. Renal parenchymal and urothelial evaluation Knowledge of these incidental anomalies is used for donor and kidney selection. In general, the presence of congenital fusion anomalies or complex cystic (Bosniak classification of IIf IV) or solid (angiomyolipoma or renal cell carcinoma) renal lesions precludes an individual from donation. Incidental findings such as tiny nonenhancing lesions, cysts, nonobstructing 3-mm-or-less stones, most urothelial congenital anomalies, and cortical scars may be considered in kidney selection because the less normal-appearing kidney is selected for harvest, thus the donor retains the more normal kidney. Unenhanced images provided a baseline attenuation measurement for the evaluation of incidental renal lesions and for the detection of urolithiasis. Nephrographic phase images were primarily used to detect and characterize renal lesions and to assess the presence of cortical abnormalities such as scars. Excretory phase images were used to evaluate the anatomy and associated abnormalities of the calyces, infundibula, renal pelvis, ureters, and bladder. Results On consensus review, the MDCT images were evaluated as technically satisfactory in all 126 donors. rterial phase images were delayed in two patients due to technical problems (scanner malfunction). Despite the delay, the renal artery and vein could be identified in those two donors and both were included in the study. rterial natomy and Variants The number and percentage of left and right kidneys with single and multiple renal arteries are described in Table 2. Early branching of the left renal arteries (< 2 cm from the aorta) was present in 26 (21%) of 126 kidneys. Early branching of the right renal arteries was identified in 19 (15%) of 126 kidneys. Overall, the percentage of variants varied only slightly with respect to side. Incidental rterial bnormalities In three donors, we found incidental arterial wall abnormalities. The first of these donors was a 63-year-old man who had calcifications within the wall of the abdominal aorta and at the ostia of both renal arteries without associated significant luminal stenosis. This donor had no history of hypertension or diabetes, and the findings were attributed to age-related and age-appropriate incidental atherosclerosis. The second donor was a 49- year-old asymptomatic woman who had a subtle beadlike appearance of the middle third of the bilateral renal arteries, presumptively diagnosed as medial fibromuscular dysplasia (Fig. 3). The diagnosis was difficult on the axial images and required coronal oblique MPR. The third asymptomatic donor had a slightly dilated posterior arterial branch that was of unclear significance. ll three donors had neither a history of urologic abnormalities nor known medical conditions. Venous natomy and Variants In general, the frequency of surgically relevant venous variants was greater than the frequency of arterial variants. The number and percentage of left and right kidneys with single and multiple renal veins are described in Table 3. Major left renal venous anomalies (i.e., those directly associated with the IVC) including supernumerary veins, such as circumaortic left renal vein (Fig. 4), retroaortic left renal vein, and duplicated IVC (Fig. 5), were identified in 14 (11%) of 126 left kidneys (Tables 3 and 4). Three of 10 cases with a circumaortic left renal vein had a 5-mm-or-larger lumbar vein or gonadal vein draining into either the preaortic or the retroaortic branch of the circumaortic left renal vein. In one donor, a 5-mm lumbar vein and a 5-mm gonadal vein connected with the retroaortic component. JR:188, January

4 Raman et al. Fig. 1 Oblique axial volume-rendered images show early branching of renal arteries in kidney donor candidates. = anterior, P = posterior., 56-year-old woman. Image shows early branch of right main renal artery (arrow). B, 46-year-old man. Image shows early branch of left renal artery (arrow). Fig. 2 Oblique coronal volume-rendered images of renal vein and late venous confluence in kidney donor candidates., 46-year-old woman. Image shows late venous confluence of right renal vein (arrow). B, 50-year-old woman. Image shows late venous confluence of left renal vein (arrow). Minor left renal venous anomalies were detected in 73 (58%) of 126 cases, including 5- mm-or-larger gonadal or lumbar veins draining into the main renal vein (Fig. 6) in 50 cases (40%) or into the branch renal veins in 18 cases (14%). Other rare variants were found including two cases of multiple draining gonadal veins, two cases of large draining veins from a renal vein to the hemiazygos vein, and one case of a communicating renal vein to splenic vein in a donor without known cirrhosis (Table 4). late confluence of left renal venous trunks (< 1.5 cm from the left lateral wall of the aorta) was present in 21 cases (17%). In 16 of the 21 cases with late venous confluence, a 5-mm-or-larger gonadal vein or lumbar vein was also present. In 14 of those 16 cases, these veins drained into a branch of the main renal vein rather than into the main renal vein itself. ll major right-sided venous anomalies were supernumerary veins, which occurred in 30 (24%) of 126 kidneys (Table 3). Supernu- B B 108 JR:188, January 2007

5 MDCT of Normal and Variant Renal natomy TBLE 2: Number of Renal rteries in the Left and Right Kidneys of 126 Potential Kidney Donors No. of Renal Left Kidney Right Kidney rteries No. % a No. % a Total a Percentage of kidneys with one or more arteries. merary veins more commonly present on the right side (158) than on the left side (136). On the right side, both late confluence of venous trunks and minor variants, which included 5- mm-or-larger gonadal vein draining into right renal veins, were uncommon, presenting in 13 (10%) of 126 kidneys and four (3%) of 126 kidneys, respectively. Renal Parenchyma and Urothelial Variants Both renal parenchymal and urothelial enhancement were deemed to be adequate and diagnostic in all studies obtained during the nephrographic and excretory phases, respectively. Parenchymal and urothelial abnormalities were identified in 34 left kidneys (27%) and 41 right kidneys (33%). These were mostly small incidental lesions, such as Bosniak I or II cysts, or nonenhancing low-density lesions less than 5 mm in diameter. Focal areas of cortical volume loss, variable cortical enhancement, or both were detected in eight kidneys. These were presumed to be focal infarcts in three kidneys, cortical scars in four kidneys, and both scarring and atrophy in one kidney (Fig. 7). Incidental nonobstructing calyceal calcifications were found in eight kidneys in asymptomatic donors without underlying metabolic abnormalities on further serum and urinary metabolic evaluation (i.e., no subclinical hypercalcemia, hypercalciuria, or uric acid abnormality). One of these donors was subsequently treated by extracorporeal shock wave lithotripsy (ESWL). One potential donor had incidentally detected bilateral irregular medullary contrast collections on the excretory phase images, findings that are consistent with bilateral papillary necrosis (Fig. 7B). She had no history of diabetes or other known underlying risk factors on further medical investigation. Her ureters were also mildly dilated, presumably because of chronic compression by an enlarged leiomyomatous uterus. One donor had simple bilateral ureteroceles, and another had a complete uncomplicated duplication of the right renal collecting system. Cohort Characteristics In our cohort of 126 potential kidney donors (252 kidneys), 70 successfully donated. Of these, 67 donors underwent left-sided laparoscopic nephrectomy, one underwent rightsided laparoscopic nephrectomy, and two underwent right open nephrectomy. Of the remaining 56 cases, 26 cases were excluded because of complex vascular anatomy, arterial problems (atherosclerosis, medial fibromuscular dysplasia), papillary necrosis, other renal scarring, stone, or medical and psychosocial factors. The reminding 30 donors are awaiting potential transplantation. Radiation Dose The weighted CT dose index ranged from to 16.8 mgy for the unenhanced scans, mgy for the arterial phase scans, mgy for the nephrographic phase scans, and mgy for the excretory phase scans. B Fig year-old woman with presumed medial fibromuscular dysplasia. and B, Coronal volume-rendered image () shows subtle irregularity of left renal artery and more prominent mural irregularity at midportion of right renal artery, which is also seen in vessel view image (B). JR:188, January

6 Raman et al. TBLE 3: Number of Renal Veins in the Left and Right Kidneys of 126 Potential Kidney Donors No. of Left Kidney Right Kidney Veins No. % No. % a b Total a Three of 116 cases were single retroaortic left renal vein. b ll cases were circumaortic left renal vein. Discussion Donor laparoscopic nephrectomy, a minimally invasive alternative to open nephrectomy, although preferable, is more technically challenging than open nephrectomy in part because of the limited surgical field of view. Technical success is influenced in part by careful donor selection, which is based on many factors including high-quality imaging. Preoperative knowledge of renal vascular, parenchymal, and urothelial anatomy based on imaging has always been important for donor and kidney selection. However, the advent of near-isotropic CT data, enabled by 16-MDCT and the multiplanar display enabled by sophisticated postprocessing software, helps laparoscopic surgeons to anticipate variant anatomy intraoperatively and avoid potential donor complications, such as hemorrhage, and potential recipient problems, such as graft ischemia and urine leak [12]. In the assessment of living renal donors, conditions precluding donation include congenital fusion anomalies, bilateral multiple (> 3) arteries or veins, bilateral arterial or venous aberrant supply or drainage (e.g., iliac vessels), asymptomatic diffuse renal disease, or incidental discovery of a solid mass in the kidneys or other organs. In general, the less normal kidney is selected for harvesting. If a donor is acceptable, the left kidney is usually harvested because of its longer pedicle. However, the right kidney may be used if left-sided vascular anatomy is complex or if the right kidney contains more relative abnormalities (scars, infarcts, Bosniak II cystic lesions). Finally, a description of the anastomosing pattern of the lumbar gonadal venous pattern is important for laparoscopic surgeons to identify them easily at surgery. In this study, for the first time, we have provided a comprehensive assessment of the types and frequency of bilateral renal vascular, parenchymal, and urothelial anomalies using stateof-the-art imaging and contemporary 2D and 3D postprocessing techniques. Unlike the findings of older studies, our findings have direct clinical relevance for practicing radiologists and urologists because a cohort of mostly young, asymptomatic potential renal donors was studied. The results of this study provide an overall perspective on the frequency of various clinically relevant anomalies in a donor population. Knowledge of these anomalies is necessary for appropriate donor and kidney selection and operative planning, as outlined earlier. To date, arterial anomalies and branching patterns have been relatively well described on catheter angiography [18], single-detector CT or 4-MDCT angiography [8 24], and 16-MDCT angiography [25]. However, characterization of venous anomalies relevant to laparoscopic donor nephrectomy has been restricted to only major variants in a few series [19, 21, 23]. Characterization of minor venous variants, which is important for surgical planning, has been described previously [25], with a few reports relying largely on autopsy and surgical series [7, 26, 27]. Fig year-old female kidney donor candidate. Coronal volume-rendered image obtained with 16-MDCT scanner shows circumaortic left renal vein. Fig year-old man with duplicated inferior vena cava (IVC) joining left renal vein (arrowhead). Coronal volume-rendered image shows right IVC (black arrow) and left IVC (white arrow). S = superior. 110 JR:188, January 2007

7 MDCT of Normal and Variant Renal natomy TBLE 4: Major and Minor Venous nomalies in the Left Kidneys of 126 Potential Kidney Donors nomalies No. (%) Major venous anomalies 14 (11) Circumaortic left renal vein 10 a (8) Prominent GV or LV 3 Separate retro- and preaortic veins 2 Retroaortic left renal vein 3 (2) Prominent LV 1 Duplicated inferior vena cava 1 Minor and other collateral venous anomalies 73 (58) Prominent GV or LV 68 (54) Prominent GV or LV connecting to main renal vein 50 (40) Prominent GV or LV connecting to venous branch 18 (14) Large hemiazygos vein connecting to renal vein 2 b Splenorenal shunt 1 Multiple draining gonadal veins 2 Note GV = gonadal vein, LV = lumbar vein. a Treated as supernumary vein in Table 3. b One was associated with late venous confluence and the other was associated with retroaortic renal vein. One important observation in this donor cohort was that although the overall frequency of venous variants was higher on the left side than on the right side kidneys, the frequency of major venous anomalies, such as supernumerary veins, was higher for rightsided kidneys (24% vs 8%). This is important for kidney selection because the longer leftsided vascular pedicle is preferred. dditional graft-related procedures, such as venovenous anastomoses, that may increase graft ischemia time (and resulting graft dysfunction) might be less commonly needed in leftsided kidneys. lso, the common major leftsided venous anomalies such as circumaortic left renal vein, retroaortic left renal vein, and duplicated IVC draining into the left renal vein were relatively easy to characterize, especially if imaging was extended to the common iliac bifurcation and standard and curved MPR images were reviewed [25]. In contrast, the frequency of minor renal venous anomalies related to the size and anastomotic pattern of the gonadal and lumbar veins was more common on the left side than on the right side (58% vs 3%, respectively). These variants and their patterns of anastomoses, although difficult to describe on axial images, are relatively easily identified on coronal and curved MPR and volume-rendered images. Our laparoscopic surgeons now believe that the intraoperative road map available to them in the form of axial, MPR, and processed volume-rendered images is important for identifying the gonadal vein intraoperatively, anticipating often complex variants, and minimizing potential hemorrhage from inadvertent transection. The lumbar gonadal venous branches are important landmarks for the laparoscopic surgeons because those branches trace the gonadal insertion into the main renal vein, branch renal vein, or lumbar vein. They generally transect the main left renal vein just lateral to its insertion with the lumbar gonadal complex. However, lumbar and gonadal veins are often difficult to see laparoscopically especially if they join the undersurface of the main renal vein, where laparoscopic visualization is limited [18]. In prior surgical series [26, 27], lumbar venous branches joined the undersurface of the left main renal vein in up to 43% of the subjects. In our series, 5-mmor-larger gonadal or lumbar veins drained into the main renal vein in 50 cases (40%) or into branch renal veins in 18 cases (14%) for left kidneys. In many cases, the gonadal vein joined the lumbar branch before eventual insertion into a left renal vein. Preoperative knowledge of the late left venous confluence helps laparoscopic surgeons anticipate two venous transections if they cannot gain control around the short main renal vein segment distal to the anastomosis. In our series, this variant was present in 17% (21 of 126 of left kidneys) but was complicated by the presence of a 5-mm-orlarger lumbar or gonadal vein in 81% (17 of 21) of cases. Unusual variants including unnamed retroperitoneal branches, branches to the hemiazygous system, and splenic vein branches were all present. These may be safely transected when recognized. In prior series, multiple renal arteries were present in up to 26 32% of left kidneys and 23 29% of right kidneys [3, 25 28]. In our cohort, supernumerary arteries were less common overall and were only slightly more prevalent on the right side (22% of right and 16% of left renal arteries). lthough we have previously seen examples of arteries arising from the common iliac vessels, in this cohort all supernumerary renal arteries arose from the abdominal aorta without aberrant iliac branches. Similar to prior reports, early branching was more common in left renal arteries (21%) than in right renal arteries (15%). t our institution, defining the branching pattern of renal arteries is important for laparoscopic nephrectomy because our surgeons prefer at least a 2-cm neck free of arterial branches. If unrecognized, injury to these vessels may result in ischemic or hemorrhagic complications. Laparoscopic criteria extend the definition of an early branch to 2 cm from the previously defined criterion of 1.5 cm from the aorta defined for open nephrectomy [11]. Only three asymptomatic potential donors had arterial wall abnormalities detected on MDCT that precluded them from donation. One had bilateral ostial atherosclerosis with mild renal luminal stenosis. Further workup with renal arterial pressure measurement was not performed. nother asymptomatic donor presented with subtle mural beading of the mid renal artery that was presumed to be medial fibromuscular dysplasia. lthough this morphologic finding has been described on catheter angiography, it was seldom described in series evaluating older CT technology. Detection of these anomalies is likely enabled by the nearisotropic 16-detector data sets. The third donor had a slightly dilated posterior arterial branch. These three donors were excluded because of concern over the potential long-term complications (e.g., earlier onset hypertension) in the donor [29] and potential for unknown complications in the recipient due to medial fibromuscular dysplasia in the graft. In general, donors are screened before undergoing imaging to exclude those with known renal abnormalities. To detect and characterize incidental renal and urothelial abnormalities, we also perform high-resolu- JR:188, January

8 Raman et al. C E Fig. 6 Renal vein and draining gonadal vein or lumbar vein in kidney donor candidates., 46-year-old woman. Coronal volume-rendered image shows left gonadal vein (white arrow) connected with left main renal vein (black arrow). B, 35-year-old woman. Coronal volume-rendered image shows left gonadal vein (arrow) connected with inferior renal venous branch (arrowhead). C, 36-year-old woman. Oblique coronal maximum-intensity-projection image shows two right renal veins (black arrows) and right gonadal vein (white arrow) draining into inferior renal vein. S = superior. D, 50-year-old woman. Oblique volume-rendered axial image shows large lumbar vein (arrow) connected with left main renal vein. = anterior. E, 31-year-old woman. Oblique volume-rendered axial image shows large lumbar vein (white arrow) anastomosing with posterior renal venous branch (black arrow). Common trunk joined main renal vein within 1 cm of aorta (late confluence). P = posterior. B D 112 JR:188, January 2007

9 MDCT of Normal and Variant Renal natomy Fig. 7 Parenchymal and urothelial abnormalities in kidney donor candidates., 54-year-old man with cortical scars. Coronal maximum-intensity-projection image obtained during excretory phase of left kidney shows cortical thinning (arrows), which indicates presence of scars and hydrocalyces adjacent to scar. B, 51-year-old woman with bilateral papillary necrosis. Coronal volume-rendered image on excretory phase CT shows small contrast collections in renal papillary pyramids (arrows). tion MDCT urography, which includes unenhanced, nephrographic, and excretory phase imaging. Unenhanced CT urography images are used to locate the renal arteries, to detect renal calculi, and to serve as the baseline to evaluate incidental renal lesions. Focal nonobstructing calyceal calcifications were detected in eight donor kidneys on unenhanced scans. No metabolic disorders involving uric acid, calcium, or oxalate were detected in these donors on further evaluation. Parenchymal abnormalities were present in 34 (27%) left kidneys and 41 (33%) right kidneys of 126 donors. ll were either cysts or low-density lesions smaller than 5 mm without clear enhancement. None of these donors was excluded. In all donors who underwent nephrectomy, lesions were determined to be cysts at surgery. One potential donor who had atrophic scarred kidneys was excluded. Incidental urothelial variants and abnormalities were uncommon. In our cohort, one donor with nonobstructed duplications and one with bilateral simple ureteroceles successfully underwent nephrectomy. One potential donor with incidentally detected renal papillary necrosis was subsequently excluded from donation because the finding was unexplained even on further questioning and evaluation. The CT urogram enables a high-resolution evaluation of urothelial anatomy. Limitations of this study must be acknowledged. surgical gold standard could not be used because of the study design and goals; thus, three-reviewer individual and consensus reviews were used to establish the true diagnosis. lthough most donors underwent primarily left nephrectomy in this series, a surgical standard would be impractical in a descriptive study because only relatively simple variants would be included and only a small cohort of right renal variants would be included. lso, many anomalies, such as medial fibromuscular dysplasia, could not be proven because donors were asymptomatic. However, in prior studies this diagnosis was also essentially morphologic. Finally, we acknowledge that the addition of excretory phase helical CT is not universally performed because of radiation concerns. We used a low-dose protocol to decrease the dose. However, evaluation of the entire urothelium is important because urothelial variants and pathologic conditions may affect the decision about whether to use the potential donor kidney. Introduction of high-resolution (1,024 1,024) CT topograms may change our practice in the future. This basic donor protocol may be widely applicable and would also be useful in the preoperative assessment of patients undergoing partial nephrectomy. In summary, using a state-of-the-art CT protocol, we described and characterized the bilateral frequency of a wide variety of renal vascular, parenchymal, and urothelial anomalies relevant for renal donor and kidney selection and for laparoscopic operative planning. ngiography and urography using 16-MDCT allow excellent assessment of vascular parenchymal and urothelial variants in potential renal donors to enable donor and kidney triage and to aid in surgical planning for laparoscopic nephrectomy. B JR:188, January

10 Raman et al. References 1. Ratner LE, Ciseck LJ, Moore RG, Cigarroa F, Kaufman H, Kavoussi LR. Laparoscopic live donor nephrectomy. Transplantation 1995; 60: Jacobs SC, Cho E, Dunkin BJ, et al. Laparoscopic live donor nephrectomy: the University of Maryland 3-year experience. J Urol 2000; 164: Ratner LE, Kavoussi LR, Schulam PG, Bender JS, Magnuson TH, Montgomery R. Comparison of laparoscopic live donor nephrectomy versus the standard open approach. Transplant Proc 1997; 29: Ratner LE, Kavoussi LR, Sroka M, et al. Laparoscopic assisted live donor nephrectomy: a comparison with the open approach. Transplantation 1997; 63: Fabrizio MD, Ratner LE, Kavoussi LR. Laparoscopic live donor nephrectomy: pro. Urology 1999; 53: Satyapal KS. The renal veins: a review. Eur J nat 2003; 7: Kawamoto S, Montgomery R, Lawler LP, Horton KM, Fishman EK. Multi-detector row CT evaluation of living renal donors prior to laparoscopic nephrectomy. RadioGraphics 2004; 24: Rubin GD, lfrey EJ, Dake MD, et al. ssessment of living renal donors with spiral CT. Radiology 1995; 195: Platt JF, Ellis JH, Korobkin M, Reige K. Helical CT evaluation of potential kidney donors: findings in 154 subjects. JR 1997; 169: Cochran ST, Krasny RM, Danovitch GM, et al. Helical CT angiography for examination of living renal donors. JR 1997; 168: Pozniak M, Balison DJ, Lee FT Jr, Tambeaux RH, Uehling DT, Moon TD. CT angiography of potential renal transplant donors. RadioGraphics 1998; 18: Smith P, Ratner LE, Lynch FC, Corl FM, Fishman EK. Role of CT angiography in the preoperative evaluation for laparoscopic nephrectomy. RadioGraphics 1998; 18: Smith P, Fishman EK. Three-dimensional CT angiography: renal applications. Semin Ultrasound CT MR 1998; 19: Del Pizzo JJ, Sklar GN, You-Cheong JW, Levin B, Krebs T, Jacobs SC. Helical computerized tomography arteriography for evaluation of live renal donors undergoing laparoscopic nephrectomy. J Urol 1999; 162: Platt JF, Ellis JH, Korobkin M, Reige K, Konnak JW, Leichtman B. Potential renal donors: comparison of conventional imaging with helical CT. Radiology 1996; 198: Patil UD, Ragavan, Nadaraj, et al. Helical CT angiography in evaluation of live kidney donors. Nephrol Dial Transplant 2001; 16: Rubin GD, Shiau MC, Schmidt J, et al. Computed tomographic angiography: historical perspective and new state-of-the-art using multi detector-row helical computed tomography. J Comput ssist Tomogr 1999; 23[suppl]:S83 S Rydberg J, Kopecky KK, Tann M, et al. Evaluation of prospective living renal donors for laparoscopic nephrectomy with multisection CT: the marriage of minimally invasive imaging with minimally invasive surgery. RadioGraphics 2001; 21[spec. no.]:s223 S Sahani DV, Rastogi N, Greenfield C, et al. Multidetector row CT in evaluation of 94 living renal donors by readers with varied experience. Radiology 2005; 235: Johnson JE, Loveday EJ, rcher LJ, et al. Preoperative evaluation of live renal donors using multislice CT angiography. Clin Radiol 2005; 60: Rajamahanty S, Simon R, Edye M, Butt K, Eshghi M. ccuracy of three-dimensional CT angiography for preoperative vascular evaluation of laparoscopic living renal donors. J Endourol 2005; 19: Kim JK, Park SY, Kim HJ, et al. Living donor kidneys: usefulness of multi-detector row CT for comprehensive evaluation. Radiology 2003; 229: Kawamoto S, Montgomery R, Lawler LP, Horton KM, Fishman EK. Multidetector CT angiography for preoperative evaluation of living laparoscopic kidney donors. JR 2003; 180: Kawamoto S, Lawler LP, Fishman EK. Evaluation of the renal venous system on late arterial and venous phase images with MDCT angiography in potential living laparoscopic renal donors. JR 2005; 184: Raman SS, Pojchamarnwiputh S, Muangsomboon K, Schulam PG, Gritsch H, Lu DSK. Utility of 16- MDCT angiography for comprehensive preoperative vascular evaluation of laparoscopic renal donors. JR 2006; 186: Baniel J, Foster RS, Donohue JP. Surgical anatomy of the lumbar vessels: implications for retroperitoneal surgery. J Urol 1995; 153: Pollak R, Prusak BF, Mozes MF. natomic abnormalities of cadaver kidneys procured for purposes of transplantation. m Surg 1986; 52: Satyapal KS, Haffejee, Singh B, Ramsaroop L, Robbs JV, Kalideen JM. dditional renal arteries: incidence and morphometry. Surg Radiol nat 2001; 23: Cragg H, Smith TP, Thompson BH, et al. Incidental fibromuscular dysplasia in potential renal donors: long-term clinical follow-up. Radiology 1989; 172: JR:188, January 2007