| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Asan Medical Center, University of Ulsan, 388-1
Poongnap-dong, Songpa-gu, Seoul 138-736, Korea.
2 Department of Radiology, Soonchunhyang University Hospital, Youngsan-gu, Seoul
140-743, Korea.
3 Department of Radiology, Healthcare System Gangnam Center, Seoul National
University Hospital, Gamman-gu, Seoul 135-948, Korea.
Received December 24, 2004;
accepted after revision March 3, 2005.
This study was supported by a grant (2004-369) from the Asan Institute for
Life Sciences, Seoul, Korea.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty-eight patients underwent gray-scale, color Doppler, and contrast-enhanced sonography examinations and then sonographically guided biopsy. Contrast-enhanced sonography was performed using Levovist at a mechanical index of 1.1-1.4. The performances of the three methods for cancer detection were compared according to biopsy site and patient.
RESULTS. Sensitivity by biopsy site was greater on contrast-enhanced
sonography (68%) than on gray-scale (39%) and color Doppler (41%) sonography
(p 
0.05), whereas the specificity and overall accuracy by biopsy
site (82% and 77% for gray-scale sonography, 84% and 79% for color Doppler
sonography, and 83% and 81% for contrast-enhanced sonography, respectively)
were not different for the three methods (p > 0.05). The
concordance score for sonography and biopsy results by patient was not
different for gray-scale (6.4 ± 1.8), color Doppler (6.3 ± 0.6),
and contrast-enhanced sonography (6.5 ± 0.7) (p = 0.281).
CONCLUSION. Contrast-enhanced sonography could improve only the sensitivity for cancer detection in analysis by biopsy site but did not improve the overall performance of sonography in patients with an indeterminate PSA level and negative digital rectal examination.
Keywords: cancer • prostate • prostate-specific antigen • sonography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Over the past few years, microbubble contrast-enhanced sonography has been introduced as a promising tool that can improve prostate cancer detection. Halpern et al. [1], in an early large study, showed significant improvement of the sensitivity from 38% on gray-scale and color or power Doppler sonography to 65% with contrast-enhanced (CE) sonography. Thereafter, several studies also reported an increased sensitivity of up to 86% [9-12]. However, the actual merit of sonographic contrast material may be in question because the performance of CE sonography varied in those studies. The variable degree of cancer risk in the study populations of previous studies may be largely responsible for the variable results of those studies; patients with a high PSA level and positive digital rectal examination may have a greater chance of having positive sonographic findings than those with an indeterminate risk of prostate cancer [9-14]. Therefore, if CE sonography were shown to have improved performance over gray-scale or color Doppler sonography for the detection of cancer in a strictly designed patient population with indeterminate risk, the impact of CE sonography would become much stronger. On the basis of this hypothesis, we undertook this study to evaluate the usefulness of CE sonography for prostate cancer detection in patients with an indeterminate PSA level (4-10 ng/mL) and a negative digital rectal examination.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Study Population
This study was prospectively designed and performed between August 2003 and
February 2004. Fifty-three consecutive patients with abnormal levels of serum
prostate-specific antigen, ranging from 4 to 10 ng/mL (mean, 7.5 ± 1.8
[SD] ng/mL), were the initial candidates of this study. Of these patients,
five with positive findings on digital rectal examination were excluded from
the study population; therefore, a total of 48 patients (mean age, 62 ±
8 years; range, 52-79 years) were enrolled in this study.
Sonographic Examination
A radiologist with 9 years' experience in endorectal sonography performed
endorectal gray-scale, color Doppler, and CE sonography examinations and then
sonographically guided biopsy. The total time required for the sonography
examination and biopsy was approximately 30-40 min for each patient.
Sonography was performed using a Sequoia 512 unit (Acuson) that provided native tissue harmonic imaging with an EC10-10C5 endorectal probe. Gray-scale images were first obtained with the use of the native tissue harmonic imaging with a transmitted frequency of 10 MHz; then color Doppler sonography images were obtained with a transmitted frequency of 9 MHz and a persistence level of 2. The color window sector width was adjusted to include the entire prostate, and the pulse repetition frequency was set to 1,990 MHz, with a wall filter of 50 MHz. Color Doppler gain was adjusted at the baseline so as not to cause color noise in the background image. After gray-scale and color Doppler sonography were completed, CE sonography was performed using Levovist (SH U 508A, Schering). This contrast agent was prepared by shaking for 5-10 sec with 11 mL of distilled water. After the solution stood for 2 min for equilibration, 12.5 mL of 200 mg/mL (total of 2.5 g) contrast agent suspension was injected manually at a rate of approximately 1 mL/sec via a 20-gauge cannula into an antecubital vein, followed by flushing with an additional 10 mL of physiologic saline at the same injection rate.
Immediately after the injection of the contrast agent, CE sonography was performed with a high mechanical index of 1.1-1.4 using the color Doppler mode. Scanning was continuously performed from the prostate base to the apex and was repeated four times. The time required for each scan varied from 4 to 10 sec according to the prostate volume. To avoid the undesired destruction of microbubbles, we applied an intersweep delay of 20 sec. All images were taken as static cine loops and stored digitally on the hard disk of the sonography unit.
Sonographically guided biopsy was performed after completing the sonography studies. Biopsy was performed at eight sites in each patient; therefore, 384 specimens were collected from the study population. Biopsy sites included the base, the lateral portion of the mid gland, the medial portion of the mid gland, and the apex. At the base and apex levels, the biopsy site was directed toward the lateral portion of the gland to include the outer gland tissue. When all of the gray-scale, color Doppler, and CE sonography concurrently showed abnormality in any site of the prostate, the biopsy site was directed to the abnormal foci. The biopsy specimens were put into separate bottles according to biopsy site, and pathologists observed the presence or absence of cancer cells in each biopsy specimen.
Image Analysis
At each biopsy site, the presence or absence of prostate cancer was
determined on the basis of subjective impression during any of the gray-scale,
color Doppler, or CE sonography examinations. The presence of prostate cancer
was considered when gray-scale sonography showed low echogenicity or when
color Doppler sonography showed increased vascularity. On CE sonography, foci
of strongly increased vascularity were considered to be cancer.
Statistical Analysis
Considering the histologic results to be the reference standard, the
sensitivity, specificity, and overall accuracy for cancer detection were
compared using the z-test for the paired proportion after adjusting for the
effect of clustering [9].
To estimate the degree of concordance between sonography findings and biopsy results, we applied weighted kappa statistics. In addition, the level of concordance between sonography findings and biopsy results in each patient was evaluated using the following scoring system: When the sonography finding and histologic result were concordant at a biopsy site, a score of 1 was given, whereas a score of 0 was given if the results were discordant between sonography and histologic evaluation. Therefore, each patient had a concordance score between 0 and 8. The scores were compared on gray-scale, color Doppler, and CE sonography with the use of the repeated measures of analysis of variance. A p value of less than 0.05 was subject to a post hoc test using the Tukey test.
In every statistical analysis, a p value of less than 0.05 was considered significant.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The results of gray-scale, color Doppler, and CE sonography by biopsy site are shown in Table 2. Biopsy results were concordant with gray-scale sonography results at 296 (77%) of 384 sites, with color Doppler sonography at 305 sites (79%), and with CE sonography at 312 sites (81%). The weighted kappa values for agreement between histologic and sonographic examinations were 0.139 ± 0.048 for gray-scale sonography, 0.183 ± 0.049 for color Doppler sonography, and 0.343 ± 0.047 for CE sonography. In eight patients with one or two positive sites on histologic examination, CE sonography detected cancer that was missed on gray-scale or color Doppler sonography in only one (13%) patient (Figs. 1A, 1B, and 1C). In the other seven patients, no lesions were found on gray-scale, color Doppler, or CE sonography. The sensitivity according to biopsy site was significantly greater for CE sonography (68%) than for gray-scale (39%) or color Doppler (41%) sonography (p = 0.007 for CE sonography vs gray-scale sonography, and p = 0.048 for CE sonography vs color Doppler sonography). The specificity and overall accuracy according to biopsy site were similar for gray-scale, color Doppler, and CE sonography (p > 0.05).
|
|
|
|
Both gray-scale and CE sonography agreed with the biopsy results in 262 sites and disagreed with biopsy results in 40 sites. Gray-scale sonography agreed but CE sonography disagreed with the biopsy results in 33 sites, whereas CE sonography agreed but gray-scale sonography disagreed with the biopsy results in 49 sites. Results of both color Doppler and CE sonography were concordant with the biopsy results in 278 sites and discordant in 47 sites. Only color Doppler sonography was concordant with the biopsy results in 26 sites, and only CE sonography was correct in 33 sites. Both gray-scale and color Doppler sonography were correct in 278 sites and incorrect in 40 sites. Only gray-scale sonography was correct in 26 sites, and only color Doppler sonography was correct in 33 sites. No significant difference was seen between the accuracy of gray-scale and CE sonography and between the accuracy of color Doppler and CE sonography (p > 0.05).
Concordance between sonography and biopsy results by numbers of patients is shown in Table 3. The sum of the concordance score values was 295 for gray-scale sonography, 304 for color Doppler sonography, and 311 for CE sonography. The mean value of the concordance score by patient was 6.4 ± 1.8 for gray-scale sonography, 6.3 ± 0.6 for color Doppler sonography, and 6.5 ± 0.7 for CE sonography, which was not significantly different for the three sonography methods (p =0.281).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
There may be some explanations for the unsatisfactory results of CE sonography in patients with an indeterminate PSA level and negative digital rectal examination. First, our study population might have had a tendency toward low-volume tumors and lower-grade tumors (Gleason score < 6) [15, 16]. Halpern et al. [1] showed a significant correlation between the confidence level for detecting prostate cancer during CE sonography and the Gleason score. It is likely that such tumors might not be detected on CE sonography because their vascular density is not enough for CE sonography to generate signals [1, 17-21]. Second, the efficacy of CE sonography might be underestimated because of the limitations of our study, wherein the histologic results of biopsy were considered to be the reference standard and histologic examination of the entire prostate gland was not performed. Therefore, it is not certain that all cancers were detected by biopsy because approximately 15-35% of all cancers are missed on randomized biopsy [22, 23].
Although CE sonography does not seem to affect the performance of sonography for cancer detection in this study, we suggest that improved sensitivity for cancer detection according to biopsy site may be helpful in guiding sonographically guided biopsy. Previous studies have shown that CE sonographically guided targeted biopsy detects as many cancers as systemic biopsy with fewer than half the number of biopsy cores and that a combination of CE sonography-targeted biopsy and systemic biopsy could increase the cancer detection rate [12, 13].
In this study, no significant difference was seen in the accuracy for cancer detection between gray-scale and color Doppler sonography. This result supports the results of previous studies that showed no demonstrable benefit of color or power Doppler sonography over gray-scale sonography [5, 8].
Contrast agent and scanning setting techniques profoundly affect the performance of CE sonography. A variety of commercially released sonographic contrast agents and scanning techniques exist, including low-intensity ultrasound pulses (low mechanical index) and high-intensity pulses (high mechanical index) [20]. We used Levovist because it is the only contrast agent permitted by the Korean Drug and Food Administration. Levovist is a blood pool agent, and imaging of the microvascular system is generated by provoking the collapse of the microbubbles at a high mechanical index. In this study, a high mechanical index mode (i.e., 1.1-1.4) was used.
When imaging in a high mechanical index mode, it is necessary to wait for reperfusion of the contrast agent before repeated scanning to ensure sufficient signal on the next transmit firing. In this study, the intersweep delay was 20 sec. The optimal interval delay between each ultrasound pulse is not known, but it may depend on the flow velocity of the vasculature of the examined tissue.
Compared with the use of contrast agents suitable for a low-mechanical-index mode, Levovist, which is designed for a high-mechanical-index mode, has both advantages and drawbacks. The use of a high mechanical index can generate more intense signal than a low mechanical index because of the destruction of a greater amount of contrast material [20, 21]. Therefore, Levovist can clearly show hypervascular lesions such as prostate cancer. On the other hand, scanning time is limited with a high-mechanical-index mode because of the fast consumption of contrast material. Therefore, at a high-mechanical-index mode, continuous scanning and real-time evaluation of blood flow are unavailable.
This study is limited by several factors. First, our sample size of 48 patients was small, so we gave more weight to comparison of the three sonography methods by biopsy site than to comparison by patient. For the per-patient comparison, we evaluated only concordance scores between sonography and biopsy results, which showed no significant difference in the three sonography methods. Therefore, although our results showed increased sensitivity for cancer detection by biopsy site, further study with a larger study population should be performed to compare the real accuracy of the three sonography methods on a per-patient basis. Second, our study is limited by review bias that was caused by nonrandomized imaging acquisition by the three sonography methods. Because the observer performed CE sonography while knowing the findings of gray-scale or color Doppler sonography, the sensitivity of CE sonography might be increased. Third, another weakness of this study was that sonographic findings were interpreted by only one observer. This limitation might be inherent to any sonography study because sonographic findings are subject to observer interpretation, and independent observation by multiple reviewers is almost impossible.
A possible controversy over the usefulness of CE sonography may rise in terms of prostate cancer in the inner gland. It is difficult to find inner gland cancer on CE sonography because some adenomas also often show increased vascularity. This study, as well as previous studies, was designed to detect only outer gland cancer; therefore, our results give no information about the usefulness of CE sonography for detecting inner gland cancer.
In conclusion, our study results show that CE sonography could improve only the sensitivity for cancer detection in analysis by biopsy site but that it did not improve the overall performance of sonography in patients with an indeterminate PSA level and negative digital rectal examination. We suggest that CE sonography is still limited as a screening tool for prostate cancer detection. However, improved sensitivity by biopsy site may positively affect the cancer detection rate on sonographically guided biopsy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
