| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Stanford University School of Medicine, Stanford,
CA.
2 Present address: Department of Radiology, Oregon Health and Science
University, MC L340, 3181 SW Sam Jackson Park Rd., Portland, OR 97201.
3 Department of Radiology, VA Palo Alto Health Care System, Palo Alto, CA.
Received June 3, 2005;
accepted after revision March 22, 2007.
Address correspondence to L. C. Chow
(chowl{at}ohsu.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Five hundred consecutive patients referred for evaluation of possible urinary tract abnormalities (327 for painless hematuria) underwent CT urography with unenhanced scanning of the abdomen and pelvis and scanning during concurrent nephrographic and excretory phase enhancement produced by administration of a split contrast bolus. The enhanced abdomen scan was obtained with abdominal compression; the enhanced pelvis scan was obtained after release of compression. Findings from axial sections and coronal maximum intensity projections were correlated with clinical follow-up and, as available, with laboratory and other imaging studies including cystoscopy, ureteroscopy, urine cytology, surgery, and pathology. Follow-up management for each patient was determined by the clinical judgment of the referring physician.
RESULTS. CT urography identified 100% of pathologically confirmed renal cell carcinomas (n = 10) and uroepithelial malignancies involving the renal collecting system or ureter (n = 8). An additional nine renal masses were identified for which no pathologic proof has yet been obtained, including eight subcentimeter solid renal masses and one multiloculated lesion. Fourteen of 19 confirmed cases of uroepithelial neoplasm involving the bladder were identified. CT urography yielded one false-positive for bladder tumor, two false-positives for ureteral tumor, and one patient with a bladder mass who refused further evaluation. CT urography yielded sensitivity and specificity of 100% and 99% and 74% and 99% and positive predictive value and negative predictive value of 80% and 100% and 93% and 99% for the renal collecting system and ureter and bladder, respectively. CT urography was ineffective in identifying 11 cases of noninfectious cystitis. CT urography also depicted numerous other congenital and acquired abnormalities of the urinary tract.
CONCLUSION. Split-bolus MDCT urography detected all proven cases of tumors of the upper urinary tract, yielding high sensitivity and specificity. The split-bolus technique has the potential to reduce both radiation dose and the number of images generated by MDCT urography.
Keywords: bladder • CT • kidney • ureter • urography
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The application of CT, however, in the evaluation of the urothelium has been hampered by poor depiction of the renal collecting system and ureters by various factors, including limited longitudinal resolution, poor urinary tract distention, and obscuration of the urothelium by dense excreted contrast material. With current advances in MDCT and careful attention to the specifics of CT protocol design, each of these obstacles is surmountable. Early work in this area has shown that it is possible to combine the benefits of EU with those of cross-sectional imaging into a single CT study termed "CT urography," which depicts both the renal parenchyma and the collecting system and ureters [1-5].
Published work in this area has used the capabilities of MDCT scanners to image the abdomen with thin sections before and after contrast administration during corticomedullary, nephrographic, and excretory phases, often resulting in three or four imaging passes during the course of a single examination [2, 3, 5]. Clearly, such methods raise concern for the total radiation dose being imparted to patients and also result in large numbers of axial source images that must be interpreted. The purpose of this study was to evaluate the utility of a split-bolus CT urography technique that depicts the kidneys during synchronous nephrographic and excretory phases, reducing the number of imaging passes required per examination, for the detection of urinary tract malignancy and other potential causes of hematuria.
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MDCT Urography Acquisition Technique
All CT urography was performed on LightSpeed MDCT scanners (4-MDCT
LightSpeed QX/i or 8-MDCT LightSpeed Ultra, GE Healthcare) using similar
techniques (Table 1).
|
All patients were asked to drink 900 mL of water while in the waiting area approximately 20 minutes before scanning. IV contrast material (Omnipaque 300 [iohexol], GE Healthcare) was administered with a split-bolus technique as follows: 40 mL was administered at a rate of 2 mL/s after the unenhanced phase. After a 4-minute delay, an additional 80 mL was administered at 2 mL/s and the abdominal compression device was inflated. The contrast-enhanced, breath-hold abdominal phase images were acquired 120 seconds after the second contrast bolus, yielding images in synchronous nephrographic and excretory phases of enhancement. Subsequently, the compression device was removed, and contrast-enhanced pelvic images were obtained to show the pelvic ureters. A small overlap in coverage of the abdominal and pelvic phase images was prescribed to ensure that there were no gaps in coverage resulting from slight changes in the degree of inspiration between the two breath-holds. Digital scout images were obtained before the unenhanced phase and immediately after the contrast-enhanced abdomen and pelvis phases.
Image Reconstruction
In addition to axial images, coronally reformatted
maximum-intensity-projection (MIP) and average-intensity-projection images
were generated in all cases as follows. Individual thick-slab MIP and
average-intensity-projection images of the enhanced right and left kidneys and
proximal ureters in a double-oblique plane truly coronal to the kidneys were
generated. Enhanced thick-slab MIP and average-intensity-projection images
including both kidneys and ureters were generated (Fig.
1A,
1B). Stacks of sliding
thin-slab MIP images (5-mm thick, 2-mm overlap) in a double-oblique plane
truly coronal to the kidneys were generated from both unenhanced and
contrast-enhanced data sets (Fig.
2A,
2B,
2C,
2D). Thick slab coronal and
sagittal MIP and average-intensity-projection images of the pelvic ureters
were generated.
|
|
|
|
Additional reformatting with volume rendering and curved planar reformation was performed on occasion on an as-needed basis but was not performed routinely. All image reconstructions were generated on an Advantage Workstation (GE Healthcare).
Image Interpretation and Clinical Follow-Up
All images, including axial source images and all reconstructions, were
primarily interpreted in soft-copy format on a PACS workstation (Centricity,
GE Healthcare) by one of six attending radiologists in the abdominal imaging
section with a minimum of 6 years of experience in interpreting abdominal CT.
Images were viewed with different window and level settings appropriate for
evaluation of the renal parenchyma or for the collecting structures and
ureters. The results of MDCT urography were retrospectively compared with the
results of other imaging examinations such as cystoscopy; retrograde studies
and ureteroscopy; laboratory studies including urine cytology; and, when
available, surgery and pathology. In patients who had a negative workup,
clinical follow-up was obtained through review of patient medical records to
determine whether urinary tract abnormalities were ultimately diagnosed after
the initial workup. The retrospective review of the medical chart was
performed in part by all four authors. Mean followup duration was 468 days
with a median follow-up duration of 433 days. Sensitivity, specificity, and
positive and negative predictive values of MDCT urography were calculated for
the detection of pathologically proven urothelial malignancies.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
MDCT urography resulted in two false-positive findings for a ureteral tumor. In one case, circumferential wall thickening of the distal left ureter yielded only inflammatory changes on repeated ureteroscopic biopsy (Fig. 5). In the second case, a ureteral filling defect was thought to be a papillary tumor on ureteroscopy, but on biopsy, the tissue was consistent with a scar.
|
|
Renal Cell Carcinoma
Ten of 10 pathologically confirmed renal cell carcinomas (Fig.
7A,
7B) were identified by MDCT
urography, yielding a sensitivity of 100%. The mean size of the neoplasms was
4.6 cm with a range of 1.6-14.0 cm. Seven of these patients were ultimately
staged at T1 N0 M0, two at T2 N0 M0, and one at T3a N0 M0. One of these
patients had undergone sonography only 16 days earlier, which showed normal
kidneys. Two patients underwent intraoperative laparoscopic sonography that
confirmed the presence of a renal tumor. In one patient, pathologically
confirmed synchronous renal cell and transitional cell carcinomas were
identified. Eight additional small solid renal masses (all < 1 cm) with
radiologic appearances consistent with renal cell carcinoma were identified on
MDCT urography in six patients (two had bilateral lesions). No pathologic
proof has been obtained for these cases because the patients were either poor
surgical candidates or opted for conservative management. The presence of one
small renal lesion worrisome for malignancy on MDCT urography could not be
confirmed with additional imaging workup. In this patient with underlying
polycystic kidney disease, a 12-mm exophytic renal lesion appeared to enhance
by 25 H. Neither transabdominal nor intraoperative sonography showed a solid
renal mass.
|
|
|
Congenital Anomalies
MDCT urography showed congenital anomalies in 15 patients. Duplication of
the right collecting system was seen in five patients (Fig.
9A,
9B), the left collecting
system in seven patients, and bilateral duplication in two patients. Two
patients had a horseshoe kidney, one of which was associated with bilateral
hydronephrosis.
|
|
Miscellaneous
One patient was found to have papillary necrosis resulting from sickle cell
disease (Fig. 10A,
10B). Renal tubular ectasia
(Fig. 11) was identified in
four patients.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The proliferation of MDCT examinations has raised concern about radiation exposure. This is of particular relevance with CT urography in which repeated acquisitions may be performed. The average effective radiation dose from CT urography has been estimated to be in the range of 15-35 mSv, with actual values depending on the specific protocol used [11, 12]. Other previously reported CT urography techniques require multiple contrast-enhanced acquisitions, resulting in a minimum of three imaging passes through the abdomen [2-5, 12, 13]. Our technique reduces the total radiation dose by eliminating one or more imaging passes. By administering IV contrast material in two boluses separated by a suitable time delay, nephrographic and excretory phases are acquired in a single imaging pass. With all other scanning parameters held constant, this technique clearly reduces the effective radiation dose when compared with MDCT urography protocols using a single contrast bolus.
The majority of our patients underwent CT urography as part of the workup for hematuria. In many medical centers, a combination of EU and unenhanced and contrast-enhanced abdominopelvic CT is used in the assessment of such patients. The split-bolus protocol requires only two full imaging passes through the abdomen and pelvis and thus exposes patients to a similar amount of radiation as a standard unenhanced and contrast-enhanced abdominopelvic CT. By eliminating the additional radiation exposure from EU, the split-bolus CT urography protocol could theoretically result in a radiation dose reduction compared with a workup performed with a combination of EU and CT.
Another often-cited issue with MDCT is the large number of images produced. An added advantage of the split-bolus protocol is that fewer images for interpretation are generated because one acquisition sequence is eliminated. On the other hand, reducing redundancy in the contrast-enhanced nephrographic and excretory phases may reduce the likelihood of visualization of the entire collecting system. Segments of the ureter that may be unopacified or in peristalsis during the contrast-enhanced acquisition will not have another chance to be captured during a second imaging pass. Our CT urography protocol incorporates the use of digital scout images obtained immediately after the contrast-enhanced abdomen and pelvis phases, to augment the likelihood of visualizing the entirety of the ureters [14], with only a minimal incremental increase in radiation dose, which is far less than an additional CT pass would incur.
One of the greatest challenges of MDCT urography remains obtaining images with good distention and opacification of the renal collecting systems and ureters. Proposed techniques for optimizing urinary tract distention include the use of prone imaging, abdominal compression, and either IV or oral hydration [1-5, 13, 15]. Prone imaging has shown little or no advantage over the other techniques [2-4]. We have opted to use abdominal compression in all patients for whom compression is not contraindicated because of its acknowledged benefits when used in conjunction with conventional EU. Some controversy has arisen as to the necessity for abdominal compression with CT urography, with a recent study suggesting that compression does not improve urinary tract visualization [13].
In addition to compression, all patients in this study were hydrated with oral water before scanning to aid urinary tract distention and lower the density of intraluminal contrast material. A prospective comparison of IV versus oral hydration techniques in 40 patients undergoing compression MDCT urography revealed no statistically significant difference in intraluminal density, streak artifact from luminal contrast material, or urinary tract distention or opacification [15].
Finally, the scanning delay time for excretory phase imaging must be considered for optimal urinary tract visualization. Caoili et al. [13] showed an advantage to excretory phase imaging at 450 seconds versus 300 seconds. This would support excretory phase imaging performed with an effective delay time of 420 seconds for the kidneys and upper ureters and 450 seconds for the pelvic ureters with the split-bolus protocol used in this study. Since this study was conducted, we have adjusted our technique to 80 mL for the first bolus and 60 mL for the second bolus, with an interbolus delay of 6 minutes, with the objective of improving opacification, but further investigation is required to determine the optimal allocation of contrast dose between the first and second boluses and the optimal delay between boluses.
Limited data exist on the sensitivity of MDCT urography for detecting urothelial lesions. In a retrospective review of patients with surgically proven transitional cell carcinoma of the upper urinary tract, preoperative MDCT detected 88% (15/17) of renal collecting system tumors and 94% (14/15) of ureteral tumors [16]. In a series of 106 patients with hematuria, unexplained hydronephrosis, or both, Cowan and McCarthy [17] showed MDCT urography to be superior to retrograde ureteropyelography, conventional EU, and sonography in the detection of urothelial abnormality.
In a retrospective study of 65 patients by Caoili et al. [5], MDCT urography found 15 of 16 foci of transitional cell carcinoma, with the one missed lesion occurring at the base of the bladder. A much lower incidence of urothelial carcinoma was found in our patient population, with 24 pathologically proven cases in a series of 500 patients. In the detection of upper urinary tract urothelial neoplasms, our study showed 100% sensitivity, confirming the value of this technique for the detection of upper urinary tract neoplasms. The marked difference in incidence of urothelial neoplasms between these two studies most likely results from the difference in patient populations. Whereas the study by Caoili et al. included only patients at high risk for urinary tract disease (42 of 65 patients [65%] with a prior history of urothelial neoplasm), our study evaluated a screening population, with only 2.8% (14/500) with a prior history of urothelial neoplasm.
There were two false-positive studies in this series for upper urinary tract urothelial tumors. In one case, repeated ureteroscopic biopsy of circumferential urothelial thickening revealed only inflammatory changes. Interestingly, in the other case, direct visualization with ureteroscopy also yielded a false-positive result with visualization of a papillary ureteral filling defect that was initially detected by MDCT urography and thought to be neoplastic by both studies. Although these two cases were considered examples of false-positive findings for tumor, true morphologic abnormalities did exist in both.
Our CT urography protocol was 74% sensitive in the detection of bladder tumors. Although CT has been shown to detect bladder lesions with sensitivities up to 90-95% using specialized virtual cystography and reconstruction techniques [18, 19], our protocol was not similarly optimized for visualizing the bladder because cystoscopy is routinely performed on patients with hematuria at our institution. Small tumors at the ureteral orifices were missed, possibly due to the normal protrusion that is often seen in that region. Mixing artifacts within the bladder can also result in false-positive and false-negative interpretations. Despite the improvements in CT spatial resolution, the use of an anatomic imaging approach will not provide the ability to identify the presence of flat bladder tumors such as carcinoma in situ. Conventional cystoscopy still remains the gold standard for evaluation of the bladder urothelium, and patients with hematuria should ideally undergo both CT urography and conventional cystoscopy.
The split-bolus protocol also accurately detected all instances of renal cell carcinoma, suggesting that synchronous acquisition of nephrographic and excretory phases does not compromise the ability to visualize the renal parenchyma. One caution regarding this imaging protocol is that early phase contrast-enhanced images, such as during the arterial or corticomedullary phase, are not obtained. Thus, this protocol may be suboptimal for the accurate staging of tumors once they are detected, and such patients may require an additional study for staging. Given the low incidence of malignancy in our series, however, we believe that the benefit of reduced radiation dose to most patients who had no malignancy outweighs the inconvenience of possible reimaging to the few patients with malignancy.
We have found that it remains imperative that the axial source images are viewed and interpreted and that appropriate tailored window and level settings be used when evaluating the collecting system and ureters so that dense intraluminal contrast material does not obscure fine urothelial detail and, potentially, small urothelial lesions. In addition to axial images, certain simple reconstructions can be of great benefit. In particular, simple sliding thin-slab (3 mm) and thick-slab (35-50 mm) MIP or average-projection images of each kidney in a double-oblique plane to the patient but truly coronal to the kidney in question provide the most intuitive display of the data.
Although the MIP algorithm emphasizes the densely enhanced collecting system, the nature of the MIP algorithm may result in obscuration of lesions lower in attenuation than surrounding high-attenuation structures. This is particularly true for small lesions such as small urothelial tumors and is exacerbated by the increasing thickness of the MIP slab. Thus, one must be cautioned to not interpret the MIP images, particularly the thick-slab MIPs, in isolation. Despite this limitation, the utility of these images lies in the ability to quickly convey overall anatomic features in an intuitive fashion. Thus, the images are useful for communicating information to busy clinicians who may not be inclined to scroll through all of the axial source images.
Our referring clinicians and urologists have found these images to be most helpful in clarifying anatomic relationships in the setting of renal masses and visualizing the extent of diffuse renal processes such as medullary nephrocalcinosis, acute or chronic pyelonephritis, papillary necrosis, or renal tubular ectasia. Such images may be particularly helpful to the urologist who is planning nephron-sparing surgery. These images can also be helpful in the detection of small or subtle lesions. In our series, one tiny (3 mm) renal pelvic transitional cell carcinoma was initially missed on the axial images but detected on the coronal oblique reconstructed thin-section images.
Although this study includes the largest patient series for CT urography reported in the literature at the time of writing, the low incidence of urothelial tumors limits our ability to assess the effectiveness of this technique in detecting urothelial neoplasms. A larger multicenter study will likely be necessary to fully evaluate CT urography for this capacity. Furthermore, this study did not directly compare the split-bolus protocol with other CT urography protocols or with EU. Another weakness of this study was that primary interpretations of the CT urography studies were reviewed and a blinded reinterpretation of the studies was not performed. Finally, not all patients received confirmatory studies and, indeed, often the CT urography results were used to guide subsequent evaluation. The relatively long average clinical follow-up duration (468 days) for patients with negative CT urography results should help to mitigate the lack of a pathologic gold standard in many cases; however, sensitivity and specificity calculations should be used with caution in light of this bias.
In summary, we have shown that a split-bolus protocol for CT urography can be used successfully to evaluate for urinary tract calculi, renal abnormalities, and urothelial lesions in one simple, noninvasive examination. Although this study did not include a direct comparison between this technique and other MCDT urography protocols, our data suggest that the sensitivities and specificities in detecting a variety of abnormalities are comparable to other protocols. Perhaps more important, the strong negative predictive values provided by this CT urography protocol suggest that a urinary tract malignancy is highly unlikely in the setting of a negative CT urogram.
The split-bolus protocol reduces radiation dose to patients and results in a smaller number of images for interpretation compared with other MDCT urography protocols. When compared with hybrid CT and EU strategies for the evaluation of hematuria, split-bolus CT urography provides equivalent if not superior visualization of the upper urinary tracts, reduced radiation exposure, and added logistic convenience. Given the significant advantages of this protocol, we believe that it has potential as an alternative to existing MDCT urography protocols and EU for the evaluation of the urinary tract.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J Richenberg Haematuria Imaging, March 1, 2008; 20(1): 57 - 72. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
