| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
by the American Institute of Ultrasound in Medicine J Ultrasound Med 27:45-53 • 0278-4297 Detection of Bladder Tumors With 3-Dimensional Sonography and Virtual Sonographic CystoscopyDepartments of Radiology (E.K., A.K., A.K.P., H.A.) and Urology (I.O., F.F.), Faculty of Medicine, Firat University, Elazig, Turkey. Address correspondence to Ercan Kocakoc, MD, Department of Radiology, Faculty of Medicine, Firat University, 23119 Elazig, Turkey. E-mail: ercankocakoc{at}yahoo.com
Objective. Bladder tumors are among the most common types of malignant neoplasms of the urinary tract. The purpose of this study was to evaluate the potential value of 3-dimensional (3D) sonography and sonographic cystoscopy in detection of bladder tumors. Methods. Thirty-one patients with suspected or known bladder tumors were included this study. All patients underwent 3D sonography and conventional cystoscopy within 15 days. The number, size, location, and morphologic features of the lesions were evaluated on gray scale, 3D virtual, and multiplanar reconstruction images obtained from the patients. The results of 3D sonographic cystoscopy were compared with the findings from conventional cystoscopy, which was considered the reference standard. Results. Twenty-eight (90.3%) of 31 3D virtual sonographic cystoscopic studies had good or excellent image quality. Conventional cystoscopy revealed 47 lesions in 22 of 28 patients; 3D sonographic virtual cystoscopy showed 41 (87.2%) of 47 lesions. Three-dimensional virtual sonography alone had sensitivity of 96.2%, specificity of 70.6%, a positive predictive value of 93.9%, and a negative predictive value of 80% for tumor detection. The combination of gray scale sonography, multiplanar reconstruction, and 3D virtual sonography had sensitivity of 96.4%, specificity of 88.8%, a positive predictive value of 97.6%, and a negative predictive value of 84.2% for tumor detection. Conclusions. Three-dimensional sonography is a promising alternative noninvasive technique for use in detection of bladder tumors, their localization, and perivesical spreading. The location, size, and morphologic features of the tumors shown on 3D sonography agreed well with the findings of conventional cystoscopy.
Key Words: bladder • neoplasm • sonography • 3-dimensional virtual cystoscopy Abbreviations: CT, computed tomography • MPR, multiplanar reconstruction • MR, magnetic resonance • NPV, negative predictive value • PPV, positive predictive value • 3D, 3-dimensional • 2D, 2-dimensional
Bladder tumors are among the most common neoplasms of the urinary tract, accounting for 6% of all malignancies in men and 2% of those in women.1 In the United States, it is the 7th leading cause of cancer death in men and the 10th in women.2 A patient with a bladder tumor usually has hematuria, and the initial evaluations are a cytologic analysis of urine and cystoscopy with or without a biopsy.3 The role of cross-sectional imaging in the initial evaluation of bladder cancer is limited, and computed tomography (CT) and magnetic resonance (MR) imaging are usually performed to evaluate extravesical extension or to stage the tumor.3,4 Excretory urography has been reported to have sensitivity of only 23% for bladder tumors less than 5 mm in diameter, and it is usually performed to detect synchronous tumors of the upper urinary tract, which occur in approximately 2% of patients with bladder cancer.5,6 Recent advances in computer technology and display techniques (eg, spiral and multidetector CT and MR imaging with rapid image acquisition and 3-dimensional [3D] rendering) have led to the development of virtual endoluminal views of hollow organs similar to those obtained with conventional endoscopy. The virtual endoscopy technique has been used to evaluate many hollow organs. Vining et al7 were the first to apply this technique in the detection of bladder cancers, after several studies including CT or MR virtual endoscopy of the bladder had been published.3,8–12 Some reports already suggested the use of virtual cystoscopy based on 3D sonographic data for various applications, but no comprehensive assessment of the 3D sonographic technique in bladder tumors has yet been made. The purpose of this study was to evaluate the potential value of 3D sonography and sonographic cystoscopy (volume-rendered imaging) in detection of bladder tumors.
Thirty-one consecutive patients (29 men and 2 women; mean age ± SD, 64.6 ± 10.5 years; range, 43–85 years) with suspected or known bladder tumors were included in the study. Written informed consent was obtained from each patient, and the study was approved by our local Ethics Committee. All patients were examined by 3D sonography followed by conventional cystoscopy within 15 days. Sonographic examinations were performed with an Aplio SSA-770A system (Toshiba Medical Systems Co, Ltd, Tokyo, Japan) and a 2-to 5-MHz multifrequency broadband convex transducer. Initially, routine gray scale sonography of the pelvis was performed. For any lesions seen on the gray scale examination, the number, location, and size were recorded. Approximately 1 hour before the sonographic examination, 500 to 750 mL of water was given to the patients orally, and the examination was performed with the bladder full. The 3D sonographic examinations were performed by 1 of 2 experienced radiologists (at least 4 years of ultrasound experience) independently using a freehand technique with a single axial sweep of the ultrasound beam across the bladder. The examiner moved the transducer with a smooth, steady motion and only the angle of the transducer was changed. After scanning, the software automatically created 3D volume data sets. To examine the bladder surface, volume-rendered, surface-rendered, maximum-intensity projection, and minimum-intensity projection algorithms were used. The whole surface of the bladder was examined from anterior to posterior at different angles while displayed on the monitor of the machine. Pathologic findings were recorded as single images. After the examination, multiplanar reconstruction (MPR) with different planes was performed manually and reviewed. For each patient, the 3D sonographic examination and image reconstruction procedures were completed within 10 to 15 minutes. All patients underwent conventional cystoscopy under general anesthesia in our Department of Urology. One urologist who was unaware of the prior sonographic examination results performed all cystoscopic examinations using a rigid cystoscope (22F; field of view, 30°–70°; Karl Storz, Tuttlingen, Germany). On cystoscopy, the urethra was visualized on the way to its entry into the bladder. The entire bladder mucosa and the orifices of the ureters were examined in detail. In the bladder, any papillary or solid lesions and areas of edema and hyperemia were noted if present. Later, during resection with a 26F resectoscope (Karl Storz 27050 SL), the tumoral lesions were resected and cauterized. Still images were interpreted separately by 2 radiologists, who were blinded to the cystoscopic findings, and discrepancies were discussed until a consensus was achieved. Observers first assessed the 3D virtual sonographic (volume-rendered) images alone. Three-dimensional virtual images with MPR and gray scale sonographic images were assessed 7 to 10 days later. A 3-point scale was used for lesion detection (1, tumor definitely present; 2, tumor definitely absent; and 3, results uncertain). The number, size, location, and morphologic features of the lesions were evaluated on gray scale, 3D virtual, and MPR images. The lesions were recorded as polypoid, sessile, or wall thickening. A lesion that was taller than its width was considered polypoid, and a lesion that was wider at the base was defined as sessile. A lesion was characterized as wall thickening when there was elevation of the bladder wall without a discrete mass. The clinical contribution and added value of 3D virtual images were also assessed subjectively by each observer with a 3-point scale (1, no change in diagnosis due to 3D virtual images, not useful; 2, 3D virtual images provided some additional information or allowed easier lesion localization but no change in diagnosis; and 3, 3D virtual images necessary for correct diagnosis), as previously described.8 In terms of quality, 3D images were assessed subjectively and were categorized as follows: nondiagnostic if the presence of numerous artifacts made less than 30% of the bladder visible for analysis; good if the image quality was good despite some artifacts; and excellent if the image quality was excellent with no artifacts. Two radiologists reexamined the 3D still sonographic images together in a different session; a consensus was achieved for all lesions, and these agreed-on lesions were used for a reliability analysis (sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]).
The results of 3D sonographic cystoscopy were compared with the findings of conventional cystoscopy, which was considered the reference standard. Two-dimensional (2D) sonographic results were compared with the findings of 3D sonography and conventional cystoscopy. The McNemar test was used for comparing these results. Statistical analysis was performed with SPSS version 10.1 software (SPSS Inc, Chicago, IL). The weighted
The image quality of 3D virtual sonographic cystoscopy was good or excellent in 28 (90.3%) of 31 patients. Images in 3 patients were suboptimal because of inadequate bladder distension. Two of these patients had wall thickening, and 1 had a sessile tumor on gray scale sonographic and MPR images. Results from these 3 patients were not included in the study.
Conventional cystoscopy revealed 47 lesions in 22 of the remaining 28 patients. Three-dimensional sonographic virtual cystoscopy showed 41 (87.2%) of these 47 lesions. Morphologically, the most frequent type of tumor was polypoid (78.7% of lesions), followed by sessile (14.9%) and diffuse wall thickening (6.4%) (Figures 1
Two patients had benign prostate hyperplasia mimicking polypoid tumors on 3D virtual images. However, 3D MPR images clearly showed the nature of these lesions (Figure 4  ). One patient appeared to have a polypoid lesion on 3D virtual sonography, which on cystoscopy was diagnosed as a large bladder hematoma. On gray scale sonography, 12 lesions were less than 1 cm in diameter; 11 lesions were 1 to 2 cm; and 17 lesions were greater than 2 cm. The mean tumor diameter in the short axis on gray scale sonography was 17.3 ±15.1 mm (range, 4–68 mm). One patient had a small polypoid lesion (4 mm) that was not visualized on gray scale sonography; 3D virtual images showed this lesion clearly (Figure 5). In 2 patients with normal 3D virtual sonographic findings, cystoscopy revealed 4- and 2-cm polypoid tumors.
Our findings showed that conventional gray scale 2D sonography had sensitivity of 93%, specificity of 66.7%, a PPV of 93%, and an NPV of 66.7% for tumor detection. Three-dimensional virtual sonography had sensitivity of 96.2%, specificity of 70.6%, a PPV of 93.9%, and an NPV of 80% for tumor detection. The combination of gray scale sonography, MPR, and 3D virtual sonography increased the diagnostic performance for tumor detection to sensitivity of 96.4%, specificity of 88.8%, a PPV of 97.6%, and an NPV of 84.2%. Using the McNemar test, we could not find a statistically significant difference between these observations (for gray scale and 3D sonography, P =.4795; for 3D sonography and the combination of gray scale sonography, MPR, and 3D sonography, P = .6171). Adding the gray scale sonography and MPR to 3D virtual sonography did not significantly alter sensitivity and PPV values but increased specificity and NPV values. Interobserver agreement was found to range from moderate to perfect (  = 0.44– 0.93) for overall interpretation (Table 1).
Many radiologic imaging techniques have been used to detect and evaluate bladder tumors, but none have been found to be reliable in detection of bladder cancer. Cystoscopy therefore remains necessary.3,7,9 Cystoscopy, however, has some limitations; it is invasive, time-consuming, and expensive, requires sedation or anesthesia, and sometimes leads to iatrogenic injury. Also, evaluation of lesions located in the base or neck of the bladder or in the diverticulum is difficult because of the limited field of view of the cystoscope.9,14,15 Because bladder tumors have a high tendency toward multifocality and recurrence, it is important to use diagnostic techniques that are sensitive and that enable visualization of the entire organ.12 The recent introduction of virtual endoscopy enables evaluation of bladder tumors. The 3D images generated from volumetric data obtained with helical CT or MR imaging3,7–12,14,15 and CT virtual cystoscopy were used first for this purpose. Vining et al7 insufflated carbon dioxide via a Foley catheter into the bladders of 3 patients (2 of whom had bladder tumors) and obtained views of tumors similar to those provided by conventional cystoscopy. Since that study, others have focused on the success and sensitivity of this technique.3,9,14,15 Insufflation of air or carbon dioxide into the bladder requires catheterization and has some limitations, such as patient discomfort and failure of the procedure in severe urethral stenosis. Contrast media have been used in some studies, permitting virtual cystoscopy to be performed as part of routine contrast-enhanced abdominal CT.9,16 Fenlon et al14 detected all lesions smaller than 10 mm in their study of 13 patients. Narumi et al15 identified 77% of lesions smaller than 10 mm. Song et al3 reported 60% sensitivity for lesions 5 mm and smaller. There are some reports in the literature regarding MR imaging-based virtual endoscopy of the bladder.8,11,12 Lammle et al12 reported a 90.9% detection rate for bladder tumors overall and a 100% detection rate for tumors 1 cm or larger in diameter. Using MR cystoscopy, Bernhardt et al11 reported sensitivity of 88.9% and specificity of 96.6% for lesions smaller than 10 mm and 100% sensitivity and specificity for lesions larger than 10 mm. Virtual CT cystoscopy provides many advantages. It permits accurate lesion localization because of its wide field of view (360°) and multi-planar capability, measures tumor size reliably, and can depict anatomic landmarks outside the bladder. It can be performed when conventional cystoscopic examination is not suitable, such as in patients with severe urethral strictures or in the presence of active bleeding.3,9,15 Evaluation of transverse and virtual views allows detection of extravesical invasion and involvement of other pelvic organs.3,9 The base of the bladder, the anterior wall of the bladder neck, and the diverticulum can be visualized.9 Virtual cystoscopy has some limitations, the most important being its inability to show flat or intramural lesions (carcinoma in situ), which appear as subtle mucosal color changes on conventional cystoscopy. The morphologic appearance alone is not sufficient to differentiate tumors from inflammation or fibrosis.3,9,14 Another limitation is that virtual cystoscopy does not allow biopsy for histologic evaluation. Also, obtaining images in both supine and prone positions with very thin slices requires a higher radiation dose for virtual CT cystoscopy. The main advantage of 3D over 2D sonography is that the 3D images can be reconstructed from the data obtained in a single sweep across the examined organ. The exact relationship between anatomic structures is accurately recorded in the 3D images.17 Three-dimensional sonography has unlimited viewing perspectives and planes, and different viewing algorithms allow the data to be displayed with a variety of techniques such as surface rendering, volume rendering, and MPR. Three-dimensional sonography has been shown to be more reliable and repeatable than gray scale sonography in evaluating anatomic structures and disease entities.17–19
In this study, we showed the feasibility of 3D sonography and virtual sonographic cystoscopy in detection of bladder lesions larger than 5 mm. To our knowledge, a study applying these methods in detection of bladder tumors has not been reported previously. Our results are comparable with those obtained with CT or MR cystoscopy. Our relatively lower specificity and NPV may be related to a small number of true-negative cases. Our interobserver agreement was moderate to perfect for overall interpretation, and our relatively lower Three-dimensional sonography with virtual sonographic cystoscopy has some advantages over other virtual techniques. Catheterization is not necessary, so there is no risk of catheter-related trauma or infection. Three-dimensional sonography is inexpensive, can be performed in real time, and can be performed at the bedside. By post processing the volumetric sonographic data, high-quality MPR images can be obtained without prolonging the scan time. These are useful in identifying the origin of the tumor and its relationship with the ureters and in revealing any extravesical invasion. Three-dimensional sonography allows large data sets to be obtained, and the amount of normal tissue surrounding a lesion usually aids in the diagnosis. If an area is undersampled, subtle anomalies might be missed.17 However, larger data sets require more powerful computers and are more prone to motion and reconstruction artifacts. In 3D sonography, artifacts may occur during image acquisition or reconstruction. These are usually caused by unplanned patient motion or transducer movement during acquisition. In this study, these artifacts were especially prominent at the anterior wall of the bladder, and we missed a 2-cm polypoid lesion because of a pronounced artifact. There are many artifacts seen on sonography (eg, shadowing, reverberations, and loss of the signal from adjacent bowel gas); these may not be obvious on rendered images but may be clear on MPR images. In general, the faster one can acquire the data, the less likely it is that artifacts will occur. The liberal use of warm gel is recommended to help minimize artifacts.17 We finished the sonographic examinations without appreciable artifacts in about 90% of cases. We completed each examination in 10 to 15 minutes, including the review of volume-rendered 3D sonographic and MPR data sets. Although the individual acquisition time is shorter than in 2D sonography, reconstruction and viewing may be time consuming, especially when volume calculations are preferred.22 One limitation to our study was that in our patients, the presence of bladder neoplasia was known or suspected; therefore, gray scale and 3D sonographic examinations were performed with a high level of suspicion for tumors. A second limitation was that we did not have enough true-negative cases to calculate the exact specificity. Third, we obtained the volumes swept out using a handheld technique rather than an automated technique. With the handheld technique, the entire bladder might not fit into the volume, but we did not encounter that problem. In conclusion, we found that 3D sonography clearly showed the normal intraluminal structures of the bladder, and the tumor location, size, and morphologic features shown on 3D sonography agreed well with the findings of conventional cystoscopy. Therefore, 3D sonography appears to be comparable with CT and MR imaging in providing virtual cystoscopy for investigation of bladder cancer. Virtual sonographic cystoscopy may be a useful alternative for screening and follow-up, particularly if conventional cystoscopy cannot be performed because of severe urethral strictures or prostate enlargement.
Received June 25, 2007, from the Departments of Radiology (E.K., A.K., A.K.P., H.A.) and Urology (I.O., F.F.), Faculty of Medicine, Firat University, Elazig, Turkey. Revision requested July 18, 2007. Revised manuscript accepted for publication September 12, 2007. We thank Greg Hammond for assistance with manuscript preparation.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Top
Abstract
Introduction
