Comparative study of novel endoluminal ultrasonography and conventional transurethral ultrasonography in staging of bladder cancer

Authors


Yuji Saga md, Department of Urology, Asahikawa Medical College, 2-1-1-1 Midorigaoka Higashi, Asahikawa 078-8510, Japan.
Email: ysaga@asahikawa-med.ac.jp.

Abstract

Abstract  Background:  Recent advances in ultrasonic techniques have improved the image quality and diagnostic accuracy for staging of bladder cancer. The aim of this study was to assess the feasibility and usefulness of endoluminal ultrasonography (ELUS) in staging of bladder cancer, and to compare them with those of conventional transurethral ultrasonography (TUUS).

Methods:  From 2000 to 2002, 19 patients with bladder cancer were evaluated by ELUS and TUUS before transurethral resection or biopsy. Clinical staging using ELUS, TUUS, computed tomography (CT) and magnetic resonance imaging (MRI) was compared with the results of pathological staging.

Results:  In 16 of 19 patients, both ELUS and TUUS were able to diagnose tumor stage. In the remaining three patients, both methods were unable to evaluate stage of tumor. In two of these patients, this inability to evaluate tumor state was caused by a difficulty in depicting the tumor base in rectangular scanning. In the remaining patient, the inability to evaluate tumor stage was caused by a difficulty in recognizing the normal muscularis because of edema around the tumor base. Both diagnostic accuracies of ELUS and TUUS were 84%, which were superior to those of CT (44%) and MRI (82%).

Conclusions:  Endoluminal ultrasonography and TUUS were equally useful for staging diagnosis of bladder cancer. Because the ELUS probe is very small in diameter and can be manipulated under direct vision, it is superior to the TUUS in safety and in fine visualization. However, the main limitations of ELUS include an inability to evaluate the depth of invasion of large tumors and an inability to visualize the tumor base in the position of the bladder neck.

Introduction

It has long been recognized that the stage of a bladder cancer is the most important factor for planning treatment and predicting prognosis. In addition, it is well established that pathological staging is the gold standard among various staging procedures. Recent advances in computed tomography (CT, e.g. spiral scanning) and in magnetic resonance imaging (MRI, e.g. dynamic imaging procedures) have considerably improved diagnostic accuracy in staging bladder cancers.1,2

Several ultrasonic techniques have been used to assess bladder cancers, including transabdominal, transrectal  and  transurethral  ultrasonography  (TUUS).3,4 The former two techniques can detect extravesical involvement, but are not helpful in determining the degree of bladder wall invasion. The latter is more valuable in evaluating the stage of tumors confined to the bladder wall.4 More recently, endoluminal ultrasonography (ELUS) using a high frequency, miniature transducer has become available for clinical use. However, few studies have evaluated the usefulness of ELUS in determining the depth of bladder tumor invasion.5,6 We focused on the feasibility and limitations of ELUS for clinical staging of bladder cancers, compared with TUUS.

Methods

From 2000 to 2002, we treated 19 patients with bladder cancer for whom both clinical staging with ultrasound and pathological staging were performed. Informed consent was obtained from all patients. The patient group comprised 12 men and seven women and the mean age was 72 years (range 58–83 years). Patients were scheduled for transurethral resection (15), total cystectomy (3) or partial cystectomy (1) without preoperative chemotherapy or radiation therapy. The preoperative diagnostic work-up of the patients involved abdominal CT (18) with a helical scanner and dynamic MRI (11).

Before transurethral resection of a bladder tumor (TUR-Bt) or transurethral biopsy, both ELUS and TUUS were performed in all patients. ELUS was performed using a probe with an extra fine transducer (4.5 Fr or 6.0 Fr in diameter, 20 MHz, Aloka, Tokyo, Japan), which was inserted endoscopically. The ELUS scanning images were depicted by the SSD-560 Intracavitary Diagnostic Ultrasound System (Aloka, Tokyo, Japan). The probe was passed through the working channel of the cystoscope and was guided into an appropriate position under direct vision. TUUS was performed with a 24 Fr ASU-65 transurethral probe (Aloka, Tokyo, Japan) having a frequency of 7.5 MHz. The TUUS scanning images were depicted by an SSD-2000 Diagnostic Ultrasound System (Aloka, Tokyo, Japan). Ultrasonographical staging was performed in accordance with the staging system of the Japanese Urological Association.7 With either ELUS or TUUS, it was difficult to distinguish tumors confined to the mucosa from those that had infiltrated the lamina propria. Therefore, stage Ta was included in stage T1. Similarly, it was difficult for the two systems to distinguish tumors infiltrating the superficial muscularis from those infiltrating more deeply. Therefore, stages T2b to T4 were included in stage T2.

The results of the imaging classification were compared to the histopathological diagnosis of the surgically excised lesion. Sensitivity, specificity, correct diagnostic accuracy, positive predictive value (PPV) and negative predictive value (NPV) were calculated. Pathological staging was performed in accordance with the latest staging system of the Japanese Urological Association.7

Results

Endoluminal ultrasonography was in some cases able to distinguish three layers in the normal bladder wall: the hyperechogenic mucosa, the hypoechogenic submucosa and the hyperechogenic muscle. However, in most cases, the hypoechogenic submucosal layer was not recognized by ELUS. Alternatively, TUUS never distinguished these three layers. Using both ELUS and TUUS, the tumors were depicted as heteroechogenic masses that were less echogenic than the hyperechogenic mucosa and muscle. Typical examples of superficial or muscle-invasive tumors depicted by ELUS and TUUS are shown in Figure 1. Based on the results of both ELUS and TUUS, the hyperechogenic muscle layer was not interrupted by hypoechogenic tumor base in a superficial cancer (Fig. 1a,b). In contrast, muscle layers were partially (Fig. 1c,d) or completely (Fig. 1e,f) interrupted by tumor bases in muscle-invasive cancers.

Figure 1.

(a) Endoluminal ultrasonography (ELUS) and (b) transurethral ultrasonography (TUUS) scans of a papillary tumor of stage pTa. Both scans demonstrate a heterogeneous mass 2 cm in size with a short stalk. (c) ELUS and (d) TUUS scans of a papillary tumor of stage pT2a. Both scans demonstrate a heterogeneous mass 4 cm in size with a broad base. Tumor is infiltrating the inner half of the muscle layer. Note the normal echo reflections from the outer half of the muscle layer. Continuity of the outer layer of the bladder wall is not interrupted. (e) ELUS and (f) TUUS scans of a non-papillary tumor of stage pT3. Both scans demonstrate a heterogeneous mass 4 × 2 cm in size with a broad base. Continuity of the bladder wall is interrupted.

Special attention was required in both ELUS and TUUS scanning because the intramural ureter was recognized as a hypoechogenic zone mimicking a defect in the muscle layer (Fig. 2a). Under circumstances where the tumor base was located adjacent to the ureteral orifice, it was essential to identify the underlying intramural ureter and to distinguish this from the muscular invasion. A typical example of a superficial tumor mimicking muscular invasion is shown in Figure 2b,c. ELUS was superior to TUUS in visualizing the intramural ureter and in distinguishing the muscular invasion, because of better mobility and traceability of the ELUS transducer.

Figure 2.

(a) Transurethral ultrasonography (TUUS) scan demonstrating hypoechogenic zones (arrow) derived from the two intramural ureters. (b) Endoluminal ultrasonography and (c) TUUS scans of a papillary tumor of stage pTa. Both scans demonstrate a heterogeneous mass 4 cm in size with a narrow stalk. Note the interrupted hyperechogenic muscle layer beneath the tumor stalk (arrow), caused by an intramural ureter.

Histopathological examination showed urothelial (transitional cell) carcinoma in 16 patients (84%) and adenocarcinoma in three (16%).

Both ELUS and TUUS were performed in 19 patients. In 16 of these patients, both ELUS and TUUS were able to diagnose tumor stage. Moreover, in these 16 patients, results of ELUS staging were perfectly consistent with the results of TUUS staging. In these patients, all tumors were correctly diagnosed by both ELUS and TUUS regardless of pathological staging (Table 1). In three of the 19 patients, neither ELUS nor TUUS was able to evaluate tumor stage. In two of these patients, the inability to evaluate tumor state was caused by a difficulty in depicting the tumor base in rectangular scanning. In the remaining patient, the inability to evaluate tumor stage was caused by a difficulty in recognizing the normal muscularis because of edema around the tumor base.

Table 1.  Correlation between ultrasonographic and pathological findings in 19 patients
Diagnostic imagingPathological stageDiagnostic accuracy (%)
Superficial (Ta–T1)Invasive (T2–T3)
  1. ELUS, endoluminal ultrasonography; TUUS, transurethral ultrasonography.

ELUS  84.2
 Superficial (T1)90
 Invasive (T2)07
 Not determined30
TUUS  84.2
 Superficial (T1)90
 Invasive (T2)07
 Not determined30

The results of the imaging classification, including CT and MRI, were compared to the histopathological diagnoses. Sensitivity, specificity, correct diagnostic accuracy, PPV and NPV of each imaging method are summarized in Table 2. Both ELUS and TUUS were superior to CT and MRI in specificity, accuracy and PPV. Sensitivity and NPV of all imaging methods were 100%.

Table 2.  Diagnostic accuracy of ELUS, TUUS, CT and MRI in the detection of muscular invasion in bladder cancer
Diagnostic methodNumber of patientsSensitivity (%)Specificity (%)Accuracy (%)PPV (%)NPV (%)
  1. CT, computed tomgraphy; ELUS, endoluminal ultrasonography; MRI, magnetic resonance imaging; PPV, positive predictive value; NPV, negative predictive value; TUUS, transurethral ultrasonography.

ELUS191007584100100
TUUS191007584100100
CT18100 944 64100
MRI111006782 71100

The most common sites affected by tumors were adjacent to the ureteral orifice (in nine patients, 47%) and the posterior walls (in eight patients, 42%). The bladder neck and the inside of the vesical diverticulum were involved in one patient each (5%). The rate of correct staging in ELUS at each tumor site is summarized in Table 3. Tumors in the bladder neck or intradiverticulum were difficult to scan perpendicularly to the base of tumor, resulting in failure to depict the tumor base, as described above. A major drawback of ELUS was its difficulty in depicting the tumor base existing in the bladder neck or intradiverticulum, regardless of good flexibility in its transducer.

Table 3.  Correlation between tumor characteristics and the rate of correct staging in endoluminal ultrasonography
CharacteristicsPathologically superficial (≤pT1)
No. correct stage/No. patients
Pathologically invasive (≥pT2)
No. correct stage/No. patients
Rate of correct staging (%)
  1. Tumor size was represented by either the minor axis of the tumor or the diameter of the stalk.

Tumor site
 Adjacent to the uteral orifice7/72/2100
 Posterior wall2/35/5 87.5
 Inside the diverticulum0/10/0 0
 Bladder neck0/10/0 0
Tumor size (cm)†
 <1.05/50/0100
 1.0–1.92/52/2 57.1
 2.0–2.91/12/2100
 3.0–3.90/02/2100
 ≥4.01/11/1100
Tumor shape
 Papillary, pedunculated3/40/0 75
 Papillary, broad basic6/82/2 80
 Non-papillary, broad basic0/05/5100

Correlations between tumor size, represented by either the minor axis of the tumor or the diameter of the stalk, and the rate of correct staging in ELUS were not observed (Table 3); difficulties experienced in depicting and classifying tumors by ELUS were not significantly affected by tumor size.

Moreover, correlations between tumor shape and the rate of correct staging in ELUS were not observed (Table 3). In papillary shaped and pathologically superficial tumors, the rate of correct staging in ELUS was not so high (Table 3). When investigating the rate of correct staging subdivided by both tumor shape and size, ELUS tended to offer correct staging in any sizes of papillary and pedunculated tumors, as well as non-papillary and broad basic tumors (Table 4). In contrast, in papillary and broad basic tumors, the rate of correct staging in ELUS was lower in small tumors than in large ones. The cause of the low proper diagnosis rate of the tumors included in this category could not be accounted for by their size or shape; tumor site was the most important factor influencing correct staging in ELUS.

Table 4.  Correct staging rate of endoluminal ultrasonography subdivided by tumor shape and tumor size
Rate of correct stagingTumor size (cm)
<1.01.0–1.92.0–2.93.0–3.9≥4.0
Tumor shape
Papillary, pedunculated1/10/11/10/01/1
Papillary, broad basic4/42/41/10/01/1
Non-papillary, broad basic0/02/21/12/20/0

Discussion

Accurate clinical staging is needed for optimum treatment of patients with bladder cancer. Thus far, CT, MRI and transabdominal ultrasonography have been helpful in detecting extravesical involvement, but these techniques are not accurate in determining the depth of invasion of the bladder wall. Recent advances in CT and MRI have improved the image quality and diagnostic accuracy for staging of bladder cancer. The overall accuracies of spiral CT and dynamic MRI have been reported to be as good as 87–95% and 60–78%, respectively.1,8 Similarly, advances in ultrasonic techniques such as TUUS and ELUS have improved diagnostic accuracy in the staging of bladder tumors to 82–93% and 88%, respectively.4,6,9

In the present study, total accuracies of TUUS and ELUS were 84% each. All 19 patients safely underwent both ELUS and TUUS procedures. In 16 of the 19 patients, the two examinations were in total agreement regarding ultrasonographic staging, and results were also consistent with pathological staging. Therefore, the diagnostic accuracy of the two examinations is equally high for clinical staging of bladder cancer.

Tumor site was correlated with the diagnostic accuracy of ELUS and TUUS in the present series. Because it was difficult for an intravesical probe to perform perpendicular scanning of the base of tumors in the bladder neck or intradiverticulum, it was difficult to obtain good images of the tumor base for accurate diagnosis of tumor stage. This problem is inevitable, because both ELUS and TUUS are able to scan only cross-sectional images against the longitudinal axis of the body trunk, and the scanning images of the tumor never become perpendicular.

Tumor size had no effect on the accuracy of staging with either ELUS or TUUS in the present series. For TUUS, this result is consistent with that of a previous report;4 however, for ELUS this results is in contrast with the results of Horiuchi et al., who identified tumor size as a factor affecting ELUS assisted staging of bladder cancer.6 Horiuchi et al. reported that whenever tumor diameter was more than 2 cm, the possibility of staging error increased. Generally, penetration of the ultrasonic beam in a 20-MHz probe is limited to approximately 2 cm.10 Because the ELUS probe is flexible, it is easy to locate the probe in appropriate positions under direct vision, and contact scanning with the tumor surface can clearly depict the base of the tumor. Therefore, correct staging by ELUS can be obtained in bladder tumors larger than 2 cm in size as well as tumors less than 2 cm in diameter. Using both ELUS and TUUS, we obtained correct staging of a tumor with no stalk and a diameter of 4 cm in both axes (Fig. 1e,f).

In the present study, tumor shape had no effect on the accuracy of staging with either ELUS or TUUS. Generally, papillary and pedunculated tumors might be superficial with high probability. In contrast, non-papillary and broad basic tumors might be invasive with high probability. With the exception of the patients that were included in these categories, 8 of 10 cases that categorized as papillary and broad basic were correctly diagnosed by both ELUS and TUUS in our series. ELUS and TUUS were particularly useful for staging the tumors of papillary and broad basic shape.

Although additional studies with sufficient numbers of cases are needed, results from the present study suggest that the usefulness of ELUS and TUUS is almost equal for diagnosis of bladder cancer. These broad-spectrum procedures do not require any special preparations or a separate session of their own, and can be accomplished at cystoscopy or TUR-Bt. The clinical benefits of ELUS compared with TUUS include ease of control of position of the probe under direct vision by cystoscopy, avoiding complications such as bladder and prostate injury, and visualizing capability of fine structures. In contrast, the main limitations of ELUS include difficulty in evaluating the depth of invasion by large tumors and in providing good images of the tumor base in the bladder neck and intradiverticulum.

Ancillary