| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University of Michigan Health System, B1-132 Taubman
Center 0302, 1500 E Medical Center Dr., Ann Arbor, MI 48109-0302.
2 Department of Pathology, University of Michigan Health System, Ann Arbor, MI
48109-0302.
3 Department of Urology, University of Michigan Health System, Ann Arbor, MI
48109-0302.
Received July 13, 2004;
accepted after revision September 27, 2004.
Address correspondence to E. M. Caoili.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Of 370 MDCT urography examinations performed over an 18-month period, 18 patients were diagnosed with 27 renal collecting system or ureteral urothelial neoplasms at endoscopic biopsy (n = 8) or surgery (n = 19). Initial MDCT reports were reviewed to determine the sensitivity of original reviewers in detecting these neoplasms. Two radiologists also retrospectively reviewed these scans and characterized the CT appearance of the neoplasms on both axial CT and 3D reformatted images. Findings at retrospective review were correlated with pathology results to determine whether any CT features could be used to predict tumor grade.
RESULTS. Eighteen of 27 neoplasms were prospectively identified on MDCT urography, and an additional six neoplasms were detected on retrospective review. Three ureteral neoplasms could not be visualized. The 24 retrospectively detected neoplasms had three distinct MDCT appearances: circumferential urothelial wall thickening (n = 14), small masses (> 5 mm in maximal diameter) (n = 5), and large masses (> 5 mm in maximal diameter) (n = 5). All detected lesions could be seen on axial excretory phase images provided wide window settings were reviewed; however, only six were detected on 3D reconstructions. MDCT urography appearance did not correlate with tumor grade.
CONCLUSION. MDCT urography is a promising technique for detecting upper urinary tract neoplasms. The static 3D reconstructions used in this study are insufficient for visualization. Axial image review remains essential for tumor identification.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although CT has been shown to be superior to excretory urography in identifying urolithiasis [4] and renal masses [5], excretory urography is still used routinely by most urologists to evaluate the renal collecting system and ureter. Use of excretory urography for detecting urothelial abnormalities is not ideal, however, because it has limited sensitivity, failing to detect up to 40% of upper tract urothelial malignancies [6-10].
Recently, several investigators have suggested that MDCT urography, with its ability to acquire thinly collimated data sets that can be used to create excellent quality 3D images of the urinary tract, can replace or even improve on excretory urography in identifying urothelial abnormalities and disorders [11-13]. The purpose of our study was to review the MDCT urography appearance of pathologically proven transitional cell carcinomas of the renal collecting system and ureter and to correlate the MDCT urography findings with pathology findings. We performed a retrospective review of our experience with MDCT urography to determine the frequency with which upper tract urothelial neoplasms were detected. We also evaluated the MDCT urography appearance of these neoplasms and correlated their appearance with pathologic tumor grade and stage.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our study group of 18 patients included 16 men and two women with a mean age of 70 years (range, 58-88 years). Twelve patients (67%) had a history of bladder cancer, half of whom underwent radical cystectomy, and the other half of whom were managed with conservative therapy consisting of transurethral resection of the bladder tumor and intravesical immunotherapy. Six patients (33%) had no history of urinary tract malignancy but presented with recurrent hematuria. Twenty-seven foci of upper tract urothelial neoplasms were surgically confirmed in these 18 patients at endoscopic biopsy (eight tumors) or surgery (19 tumors). Eleven patients had a single focus, and seven had multiple foci (five with two foci, and two with three foci). The mean time from MDCT urography to pathology diagnosis was 34.4 days (range, 1-90 days). Five patients underwent excretory urography within 1 year of MDCT urography. The mean duration of time between the two studies was 73 days (range, 6-293 days).
|
5 or > 5 mm). Findings at MDCT urography were then
compared with and correlated with excretory urography results if available and
with surgery and pathology reports. The appearances of the tumors were
correlated with the pathologic type and histologic grade on the basis of the
2003 World Health Organization and International Society of Urological
Pathology consensus classification
[14]. The CT appearances of
the tumors were also correlated with tumor stage.
MDCT Urography Technique
Examinations were performed with MDCT scanners (LightSpeed QX/i version
1.3, GE Healthcare). MDCT urography consisted of four imaging phases:
unenhanced, nephrographic, early excretory, and delayed excretory. The
unenhanced phase was obtained through the abdomen and pelvis, at 4 x
3.75 mm collimation, and was reconstructed at a thickness of 5 mm.
Nephrographic phase images were obtained 100 sec after the initiation of IV
contrast material injection (150 mL of Omnipaque 300 [iohexol], Amersham
Health) administered at a rate of 3 mL/sec. Nephrographic phase images were
obtained from the diaphragm through the kidneys using a 4 x 2.5 mm
configuration and reconstructed at 5-mm thickness. Between April 2000 and
January 2002, early and late excretory phase images were obtained at a 200-
and a 300-sec delay. Beginning in February 2002, the early excretory phase was
obtained at a 300-sec delay and the late excretory phase at a 450-sec delay.
The following technique was used for all excretory phase series: 4 x
1.25 mm configuration, reconstructed at an image thickness of 2.5 mm and at
50% overlapping (1.25 mm) intervals using 120 kVp and 120-280 mA.
Three-dimensional reconstructions of the two excretory phase scans were
created at an independent workstation (Advantage Windows 3.1) by CT
technologists. Coronal and bilateral 25° coronal oblique projections were
created using maximum-intensity-projection, average-intensity-projection, and
volume-rendering algorithms. The maximum-intensity-projection and
average-intensity-projection images were produced by electronically selecting
a large slice thickness (usually > 50 mm) that included the kidneys and the
ureters. Volume-rendered images were created from all axial image data except
back-cut or front-cut editing that was frequently used to remove overlying
ribs. Opacity curves for the volume-rendered images were chosen to
preferentially show excreted contrast material in the renal collecting system,
ureter, and bladder.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Three lesions were not identified on prospective or retrospective interpretation, one located in a mid ureter and two in the distal ureter. In one of these cases, the ureteral segment in which the tumor was located was neither opacified with excreted contrast material nor distended, possibly explaining the failed detection (Figs. 4A, and 4B). In the second, the involved ureteral segment was opacified but not distended, whereas in the third, the ureteral segment was both opacified and distended.
|
|
The retrospective reviewers detected 24 neoplasms on the axial images, 20 of which could be seen on the axial images when standard soft-tissue windowing was used. However, four small masses were visible only when the axial images were viewed with nonstandard wide window settings (Figs. 5A, and 5B). In comparison, only six of the 24 cancers (three producing urothelial thickening, two large masses, and one small mass) could be detected by each reviewer on the 3D images, with both reviewers identifying the same six neoplasms (Figs. 6A, and 6B).
|
|
|
|
Correlation of MDCT Urography Appearance with Pathologic Type and Grade
According to the 2003 World Health Organization and International Society
of Urological Pathology consensus classification
[14], urothelial neoplasms are
divided into two groups—papillary or flat—according to their
growth pattern. They are also classified according to histologic grade, which
affects prognosis. Both papillary and flat urothelial neoplasms can be further
divided into noninvasive and invasive lesions. Noninvasive papillary
urothelial lesions are further divided into papillary hyperplasia, papilloma,
papillary urothelial neoplasm of low malignant potential, and low-grade and
high-grade neoplasm, based on the degree of epithelial differentiation.
Invasive papillary lesions can be classified as low grade or high grade. Flat
noninvasive urothelial lesions can be further divided into hyperplasia,
urothelial atypia of unknown significance, low-grade dysplasia, and carcinoma
in situ (high-grade dysplasia), where carcinoma in situ is a documented
precursor of invasive malignancy
[14].
The MDCT urography appearance of upper tract neoplasms could not be used to reliably predict tumor growth pattern or grade. Of the various types of tumors occurring in this series, all types produced urothelial wall thickening. Of 14 neoplasms producing wall thickening, nine were found to have papillary and five to have flat growth patterns. Five of these 14 neoplasms were high-grade (all five of which were papillary), two low-grade (both papillary), two of low malignant potential (both papillary), and five were carcinoma in situ (all flat).
Of the five lesions that were small masses, three were papillary and two flat. Only one of the five was a high-grade tumor, however, and three were low-grade. The grade of the fifth lesion was not determined because it was located adjacent to a high-grade tumor and its grade was therefore not relevant to patient treatment and prognosis.
Most large masses were papillary and high grade. Of the five large masses, four were papillary and one flat. Three of these neoplasms were high grade, one had low malignant potential, and the last was a carcinoma in situ.
Of the three lesions that could not be detected on MDCT urography, two were flat and one was papillary. Two of these tumors represented carcinoma in situ, but the grade of the third was not determined (again because of its close proximity to a high-grade lesion).
Correlation with Tumor Stage
The depth of tumor invasion could be assessed only for the 19 neoplasms
treated with nephroureterectomy and not in the eight foci confirmed at
endoscopic biopsy. Only 16 of these tumors were identified in the
retrospective review. Nine tumors produced urothelial wall thickening. Of
these nine tumors, four were flat lesions and represented noninvasive
carcinoma in situ. The remaining five tumors were papillary lesions. Two cases
of papillary urothelial thickening represented noninvasive tumors of low
malignant potential. Three lesions were the high-grade papillary type. Two
invaded the lamina propria (stage T1)
(Fig. 7) and one invaded the
muscularis propria (stage T2). Four tumors produced small masses. Two were
flat low-grade lesions that were noninvasive, one was a papillary high-grade
lesion with no invasion (stage Ta), and one was not graded. Three tumors
produced large masses. One was a noninvasive, papillary neoplasm of low
malignant potential, and the remaining two large masses were papillary
high-grade tumors with invasion of the adjacent renal parenchyma (stage T3)
(Figs. 8A, and
8B). Parenchymal invasion was
suspected for both of these tumors on MDCT urography.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several early studies have indicated that conventional CT can detect at least some upper tract uroepithelial neoplasms [17-21], although the early reported sensitivity of CT suggested that it had substantial limitations [22-24]. McCoy et al. [22] reported that CT detected only 50% (17/34) of tumors. After excluding suboptimal examinations, sensitivity increased to 68% (15/22). The missed tumors were located in the calices (n = 5), pelvis (n = 1), and ureter (n = 1). Lower stage tumors, defined as ranging from neoplasms with no invasion to those invading the muscularis propria, were even less likely to be identified on CT, with a detection rate of 45% (vs higher-grade tumors, which had a detection rate of 88%). Those investigators also reported that all cases of carcinoma in situ (n = 8) were not seen on CT (as was also the case for excretory urography and retrograde pyelography).
The results of a report by Scolieri et al. [23] were also disappointing. Those authors found that CT was not useful in diagnosing upper tract urothelial tumors. CT did not identify 27% (10/37) of surgically proven urinary tract tumors. In that study, neoplasms in the renal pelvis (n = 1) and the ureter (n = 9) were not detected. All missed neoplasms were low grade. Those investigators concluded that CT did not affect their management of patients with suspected upper tract disease.
With the advent of helical CT, a much greater number of images can be obtained using more thinly collimated images. Thus, helical CT can be expected to improve the detection of small tumors. MDCT has offered further refinements in spatial and temporal resolution over standard helical CT. On the newest MDCT scanners, some configurations offer data sets composed of images that have a thickness of 1 mm or less. Experience with the ability of MDCT scanners to detect urinary tract abnormalities has been promising [11-13]. The results of our series suggest that MDCT urography can detect most upper tract transitional cell carcinomas. In our investigation, 89% of malignant upper tract foci in our study population of 18 patients were detectable with MDCT urography.
Urothelial neoplasms have been described previously in conventional CT
studies as having two distinctly different CT patterns: intraluminal masses
and ureteral wall thickening
[19,
22]. In our series, the
detected urothelial malignancies also had these two appearances; however, we
further subdivided our masses into large (> 5 mm) and small ( 5 mm), to
determine how well MDCT urography could specifically identify even small
neoplasms.
In our series, small urothelial tumors could be detected, although most were seen only in retrospect, and most of these could be seen only on the axial images and only when these images were viewed with nonstandard wide window settings. Small urothelial neoplasms are often not detected on images with standard soft-tissue window settings because they are obscured by the dense contrast material in the renal collecting system and ureter. The use of wide window settings is important to maximize the detection of small intraluminal lesions.
Three-dimensional reconstructions offer potential advantages in that they can summarize the data from a large number of axial images into fewer images that can be further manipulated at a workstation. If the review of MDCT urography images could be reduced merely to evaluation of the 3D images (without a loss in study accuracy), the time required to review each study might be decreased dramatically. Unfortunately, our study has shown that certain commonly used 3D image reconstruction techniques (volume rendering of the entire data set, and thick-slab average- and maximum-intensity-projection images) are insufficient. In our series, only a small percentage (25%) of the urothelial tumors detected on MDCT urography could be seen on these generated 3D images, even in retrospect. Our experience suggests that review of the axial data set is still necessary to optimize detection of upper tract urothelial neoplasms. This is particularly the case when the neoplasms produce small masses and ureteral or renal pelvis wall thickening, in which cases the urinary tract lumen is not distorted. Similar to excretory urography, volume-rendered, average-intensity-projection, and maximum-intensity-projection images of thick slabs cannot depict these abnormalities reliably. In fact, in the small subset of patients who underwent both excretory urography and MDCT urography, three foci appearing as wall thickening were not seen on excretory urography. All these techniques best show only the contrast material excreted into a normal-appearing urinary tract lumen.
Our study showed that tumors, both papillary and flat, and both high grade and low-grade, may present as small masses, large masses, or urothelial wall thickening, although most small masses were low-grade tumors and most large masses were high grade. Still, there is nothing about the MDCT urography appearance of upper tract tumors that can be used consistently to predict tumor type, growth pattern, or grade. Many types of urothelial tumors can produce all three types of abnormalities on MDCT urography.
Despite previous publications to the contrary, we found that the urinary tract was frequently abnormal in areas in which carcinoma in situ was present, permitting detection of abnormalities in even these patients. The abnormality that was present in most of these cases of carcinoma was urothelial wall thickening (without any luminal narrowing or irregularity).
Our investigation included only a small number of patients and therefore could not focus on the ability of MDCT urography to stage upper tract urothelial tumors. Although our series included 18 patients, accurate staging could be performed only in the 13 patients who subsequently underwent nephroureterectomy. In the remaining five patients, diagnosis was obtained at ureteroscopy, which can provide a definitive diagnosis of malignancy but is limited in its ability to accurately stage malignant foci. In the subset of lesions that were staged, we found two renal pelvis lesions with parenchymal invasion correctly staged at MDCT urography. In the remainder of the high-grade neoplasms, none of the lesions showed invasion past the muscularis on imaging studies or when the pathology report was reviewed. Similar to prior reports on the ability of CT to stage urothelial tumors [17, 22-24], our study found that MDCT urography could not distinguish stages Ta, T1, and T2. However, unlike prior investigations, MDCT urography could detect many low-stage (Ta-T2) lesions.
Although MDCT urography is a promising technique, imaging problems are encountered. Not all urothelial tumors will be identified. Three such neoplasms were missed even retrospectively in this series. In two of these cases, at the time that the excretory phase images were acquired the ureteral segment in which the neoplasm was located was not opacified or distended in one patient and was poorly opacified but not distended in one patient. Poor visualization of the involved urinary tract segment certainly may have contributed to our failure to detect these lesions. Indeed, nono-pacification of ureteral segments is a continuing problem in MDCT urography. Regardless, occasional transitional cell carcinomas will not be detected even when the ureteral segments in which they are located are well opacified and distended, as was the case in this series in the third patient whose transitional cell carcinoma was not detected.
Our study has several important limitations. It was a retrospective review that included a small number of patients. It is possible that we have overreported our sensitivity in detecting urothelial tumors because some patients with neoplasms that we did not detect may have not yet been diagnosed. In addition, an inherent selection bias exists in our MDCT urography population that is composed of patients for whom there was a high clinical suspicion of urinary tract disease after consultation with a urologist. Finally, our series evaluated the sensitivity of MDCT urography in detecting neoplasms only when our specific technique was used. It is quite possible that other techniques, including other 3D reconstruction approaches, may be more sensitive. For example, the 3D images in our study were not isotropic because all examinations were obtained with either a 4- or an 8-MDCT scanner. It is our hope that as CT hardware and software continue to improve, so will the ability of CT to detect smaller transitional cell carcinomas. Further investigation into the potential role of high-resolution isotropic 3D imaging (created from axial images on 16- to 64-MDCT scanners during MDCT urography) would also be of great interest.
In conclusion, MDCT urography performed on 4- and 8-MDCT scanners can detect upper tract urothelial malignancies, whether they present as large masses, small masses, or urothelial wall thickening. Small lesions, which were once thought to be detectable only with excretory urography, can be visualized on axial CT images if these images are viewed with nonstandard wide window settings. MDCT urography can even identify some renal collecting system or ureteral segments in which carcinoma in situ is present (by identifying wall thickening around an otherwise normal-appearing portion of the urinary tract). MDCT urography can identify papillary and flat, high-grade and low-grade malignancies (although it cannot determine tumor type, growth pattern, or grade on the basis of tumor size or morphology). MDCT urography may also have a role in staging some tumors (showing renal parenchymal invasion in some neoplasms of the intrarenal collecting system). We believe that given the ability of MDCT urography to image the urinary tract and the remainder of the abdomen and pelvis, it can play an increasing role in managing patients with suspected urinary tract neoplasms.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Kawamoto, K. M. Horton, and E. K. Fishman Transitional Cell Neoplasm of the Upper Urinary Tract: Evaluation with MDCT Am. J. Roentgenol., August 1, 2008; 191(2): 416 - 422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Takahashi, A. Kawashima, J. F. Glockner, R. P. Hartman, B. C. Leibovich, A. C. S. Brau, P. J. Beatty, and B. F. King Small (<2-cm) Upper-Tract Urothelial Carcinoma: Evaluation with Gadolinium-enhanced Three-dimensional Spoiled Gradient-Recalled Echo MR Urography Radiology, May 1, 2008; 247(2): 451 - 457. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Jinzaki, A. Tanimoto, H. Shinmoto, Y. Horiguchi, K. Sato, S. Kuribayashi, and S. G. Silverman Detection of Bladder Tumors with Dynamic Contrast-Enhanced MDCT Am. J. Roentgenol., April 1, 2007; 188(4): 913 - 918. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
