| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Genitourinary Imaging |
1 From the Departments of Radiology (E.J.H., F. Frauscher, F. Forsberg, L.N.N., P.O.) and Urology (S.E.S., L.G.G.), Jefferson Prostate Diagnostic Center, Thomas Jefferson University, 132 S 10th St, Philadelphia, PA 19107-5244. Received May 23, 2001; revision requested June 25; revision received July 27; accepted August 24. Address correspondence to E.J.H. (e-mail: ethan.halpern@mail.tju.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-two patients were evaluated in the left lateral decubitus position with gray-scale, color Doppler, and power Doppler transrectal US. Up to four directed biopsy specimens were obtained on the basis of gray-scale and Doppler US findings, and modified sextant biopsy followed. Analysis of variance and the Wilcoxon signed rank test were used to evaluate the distribution of Doppler signals within the prostate. Three healthy volunteers with no known prostate disease were also examined in supine and both decubitus positions.
RESULTS: In the patient group, both color and power Doppler US demonstrated increased flow on the left side of the prostate, with greater flow toward the base of the gland (P < .002). Consequently, 62 of 90 directed-biopsy cores were obtained in the left base and midgland. The positive biopsy rate for directed biopsy was not significantly different from that of sextant biopsy (P = .4). Seven patients had cancer that was identified with sextant biopsy, but only four cancers were identified with directed biopsy. Each of the three healthy volunteers demonstrated increased Doppler flow on the dependent side when the subject was in the lateral decubitus position.
CONCLUSION: The positive yield of directed biopsy was similar to the yield of sextant biopsy. On the basis of observations made in healthy volunteers, the authors conclude that flow asymmetry in patients who underwent biopsy may have been related to patient position.
© RSNA, 2002
Index terms: Prostate, biopsy, 844.1261, 844.12985 • Prostate neoplasms, US, 844.12984, 844.12985, 844.32
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Standard sextant biopsy of the prostate involves six specimen cores that are obtained systematically from various regions of the prostate (2). Approximately 15%–35% of cancers are missed with conventional sextant biopsy (3,4). Color Doppler ultrasonography (US) with directed biopsy may increase the number of cancers detected (5–10). Power Doppler US has been suggested as a promising technique to increase the detection of prostate cancer (11–13). These techniques, however, have not proven sufficient sensitivity for directed biopsy to replace systematic biopsy of the prostate (14–16).
Sensitivity of Doppler US for slow flow in small arteries is increased at higher frequencies of insonation (17,18). In the past 5 years, authors of most articles on Doppler US of the prostate have used frequencies in the range of 5–7 MHz. Commercially available transrectal transducers now provide greatly increased flow sensitivity, with Doppler US frequencies as high as 9 MHz. We hypothesized that directed biopsy with high-frequency Doppler US would be sufficiently sensitive to tumor vascularity to replace standard sextant biopsy. The present study, therefore, was performed to evaluate cancer detection with directed biopsy of the prostate on the basis of high-frequency Doppler US findings, and to determine the effect of patient position on the observed flow pattern.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Gray-scale, color Doppler, and power Doppler transrectal US were performed with each patient in the left lateral decubitus position. All examinations were performed with an EC10C5 end-fire probe and the Sequoia 512 system (Acuson, Mountain View, Calif). For gray-scale US, the probe center frequency was 10.0 MHz. For color and power Doppler US, the center probe frequency was 9.0 MHz. Color and power gain were adjusted as follows: Gain was increased until clutter was observed and then reduced just enough to remove clutter from the image of the prostate. Transrectal gray-scale US examination followed a standard sequence of transverse imaging from base to apex, followed by sagittal imaging from right to left. Doppler US examination followed a standard sequence of transverse imaging from base to apex. Each examination began with gray-scale US, followed by color Doppler US and, finally, power Doppler US.
Gray-scale and Doppler US findings were recorded prospectively for each sextant biopsy site in the prostate. Laterally directed sextant sites were defined per the modified sextant biopsy protocol suggested by Stamey (19). A five-point rating scale was used to classify each site for gray-scale, color Doppler, and power Doppler US findings. The gray-scale score was based on the presence of an echotexture abnormality or a contour deformity. Abnormalities in echotexture included a definite hypoechoic lesion or an area of heterogeneous echotexture. Contour deformity was defined as a focal bulge of the prostate contour. Color and power Doppler US images were evaluated for the presence of increased flow within the parenchyma of the prostate. The amount of flow within each sextant was judged by visual inspection of the color pixel density. Gray-scale and Doppler US abnormalities were judged primarily on transverse US images, and the contralateral half of the gland was used for comparison. Each five-point-scale assessment was a prospective, subjective assessment made by one of four physicians (E.J.H., S.E.S., L.N.N., P.O.) who performed the examinations. The scale was graded as follows: 5, definitely abnormal (focal hypoechoic mass at gray-scale US or obvious increase in flow at Doppler US); 4, probably abnormal (probable hypoechoic mass at gray-scale US or mild increase in flow at Doppler US); 3, indeterminate (abnormal gray-scale echotexture without definite mass or subtle increase in flow at Doppler US); 2, probably normal (heterogeneity at gray-scale US or minimal increase in Doppler flow that might simply represent artifact); and 1, definitely normal (homogeneous gray-scale appearance with symmetric Doppler flow pattern).
Prostate biopsy was performed immediately after gray-scale and Doppler US evaluation. The same physician who performed the diagnostic examination also performed the biopsy procedure. A maximum of four directed-biopsy specimens was obtained from each patient on the basis of gray-scale and Doppler US findings. Directed-biopsy sites were chosen to include up to four of the most abnormal areas in the prostate on the basis of the previously described subjective rating scale. Patients with no abnormal sites at gray-scale and Doppler US did not undergo directed biopsy. Directed biopsy was followed by a modified sextant biopsy protocol with six spatially distributed biopsy cores (19). Sextant biopsy specimens were obtained as peripherally as possible at the base, midgland, and apex, without regard to gray-scale and Doppler US findings. An 18-gauge core biopsy needle was used to obtain all specimens. Each biopsy core was marked as "directed" or "sextant" and labeled according to gland location (left or right, base, midgland, or apex).
Early in the course of this study, it became apparent that focally increased flow was present in many patients (predominantly on the left side of the prostate) and was not associated with the presence of malignancy. To evaluate further the possibility that this finding might be related to patient position, three authors of this article, aged 39–40 years and with no prior history of prostate disease, volunteered to have their prostates evaluated. Serum PSA values were not available for these subjects. Each subject provided written informed consent to undergo a US evaluation procedure approved by the institutional review board. Each subject was evaluated in the left lateral decubitus, right lateral decubitus, and supine positions. Both color and power Doppler US were performed with the subject in each position by using a transverse sweep from base to apex. A 1-minute equilibration period was allowed in each position before the start of imaging. Doppler US was performed at 9 MHz, just as in the initial clinical protocol.
The purpose of evaluating these subjects was to demonstrate whether a change in patient position might alter the observed Doppler flow pattern within the prostate. The use of only three subjects was not intended to establish a statistically significant trend of flow pattern as a function of position. Rather, the small number of subjects was used to test the feasibility of our hypothesis—that patient position might influence the observed flow pattern. Since these subjects did not undergo sextant biopsy, the flow pattern was not rated on a 1–5 scale for each sextant. Rather, for each position—left decubitus, right decubitus, and supine—the flow pattern was rated by subjective visual inspection of color pixel density as symmetric or focally increased on one side. The Doppler flow pattern of each subject was assessed by consensus of the remaining two volunteer subjects.
Statistical analysis was performed to evaluate the distribution of Doppler signals in our biopsy population. To determine whether there was significant effect on the basis of location within the prostate, analysis of variance was performed on the Doppler signal ratings with location and patient as explanatory variables. On obtaining a significant F-test value, additional paired comparisons were performed with the Wilcoxon signed rank test to determine the location of this effect (left base vs right base, left midgland vs right midgland, left apex vs right apex, left base vs left apex, and right base vs right apex). To compensate for multiple comparisons, an adjusted P value of 0.05/5 = .01 was established as the cutoff for a statistically significant result (Bonferroni adjustment).
Pathologic reports for all sextant and directed core biopsies were reviewed. Each core was classified as benign or malignant, and a Gleason score was recorded for each malignant core. The positive biopsy yield was computed individually for sextant and directed biopsy cores. Because these biopsy data were clustered by patient, conditional logistic regression was performed to compare the positive biopsy yield for directed and sextant cores and to compensate for the lack of statistical independence among multiple cores within each subject. The regression model used pathologic findings as the dependent variable and biopsy type (directed vs sextant) as the independent variable.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
On the basis of prior studies of color and power Doppler US, one would expect a higher positive yield of cancer from Doppler US–directed biopsy compared with that of systematic sextant biopsy. Kelly et al (6) described an overall positive predictive value of 77% (65 of 84 cases) with color Doppler US and a positive predictive value of 94% (30 of 32 cases) for the most vascular lesions in the prostate. Newman et al (7) demonstrated a positive predictive value of 62% (21 of 34 biopsy sites) for increased color Doppler flow in the peripheral zone. Cornud et al (14) demonstrated a positive predictive value of 84% for directed biopsy of hypervascular foci. Okihara et al (13) suggested that power Doppler US–directed biopsy could reveal 98% of cancers and reduce the biopsy rate by 60%. The initial objective of our study was to demonstrate that directed biopsy with high-frequency Doppler US would be sufficiently sensitive to tumor vascularity to replace standard sextant biopsy. However, the positive biopsy yield of our directed-biopsy approach (9%) was similar to the yield of sextant biopsy (8%) in this study. Among seven cancers found with sextant biopsy, only four were detected with directed biopsy. Doppler US-directed biopsy did not reveal cancer in any additional patients in our study. Clearly, the major determinant of prostatic blood flow in our study was patient position. If there is increased Doppler flow around malignant foci, this pattern is overwhelmed by the positional asymmetry of blood flow with the patient in the left lateral decubitus position. We conclude that Doppler US evaluation of the prostate should not be performed with the patient in a decubitus position.
Patients are often placed in the left lateral decubitus position in reported studies of Doppler US evaluation of the prostate. To examine the normal vascular anatomy of the prostate with color Doppler US, Neumaier et al (20) evaluated 35 infertility subjects with a 6-MHz Doppler US frequency. More recently, Leventis et al (21) evaluated 40 healthy subjects with power Doppler US by using a 6-MHz Doppler US frequency. Symmetric flow patterns were noted in both of these studies, in which subjects were placed in the left lateral decubitus position. Newman et al (7) and Lavoipierre et al (8) evaluated biopsy patients in the left lateral decubitus position with a 7-MHz Doppler US frequency and did not note a positional effect. Shigeno et al (9) evaluated biopsy patients in the left lateral decubitus position with a 5-MHz Doppler US frequency and did not note a positional effect. Patient position is not described in the remaining references cited in this article. The Doppler US frequency of 9 MHz employed in the present study was higher than that used in any of the cited references (range, 5–7.5 MHz). It is likely that in prior studies, imaging performed at a lower frequency identified fewer vessels and vessels of larger caliber. The effect of position on Doppler flow patterns may not be as obvious in larger vessels.
The anatomic explanation for the effect of patient position is not obvious. The primary arterial supply to the prostate derives from prostaticovesical arteries that enter at the base of the prostate. The prostaticovesical artery on each side is a branch of the internal iliac artery that supplies both urethral and capsular arteries (20). The urethral arteries extend along the prostatic urethra and are responsible for the vascular appearance of the periurethral zone. The capsular arteries are responsible for the normal radial pattern of blood flow from the prostatic capsule toward the urethra. Increased Doppler flow at the base relative to the apex likely results from the presence of larger and more numerous vessels near the prostatic base. With the patient in a decubitus position, it is possible that the effect of gravity simply shifts more flow to the dependent side. Such an explanation has been offered for pulmonary venous return during transesophageal Doppler echocardiography (22). Alternatively, movement of the prostate toward the dependent side may stretch the vessels that supply the contralateral side and limit inflow. A true anatomic understanding of the positional dependence of prostatic blood flow will require further investigation with measurement of flow pattern in the vessels that supply the prostate.
The premature termination of our clinical study after recruiting only 32 patients represents a limitation of our study. Although we would have liked to recruit 100 patients, a review of findings in the first 32 patients demonstrated an obvious predilection for flow on the left side of the prostate. Premature termination of a protocol introduces a theoretical potential for bias due to patient selection. However, we found no obvious cause for the left-sided predilection of flow other than patient position. Furthermore, positional changes in the flow patterns of our healthy subjects convinced us that our original observation was related to patient position rather than some form of selection bias. On the basis of these considerations, we believed that we would not be ethically justified in subjecting more patients to directed biopsy in the decubitus position.
Although we do not claim to understand the anatomic basis for positional changes in the prostatic flow pattern, this finding has clear clinical implications. If Doppler US is used to identify hypervascular neoplasms within the prostate, it should not be performed with the patient in a decubitus position. Given the more symmetric flow pattern observed in our healthy subjects in the supine position, it seems more reasonable to perform Doppler US evaluation of the prostate with the patient in a supine position.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, E.J.H.; study concepts and design, E.J.H., F. Frauscher; literature research, E.J.H., F. Frauscher; clinical studies, all authors; experimental studies, E.J.H., F. Frauscher, F. Forsberg; data acquisition, all authors; data analysis/interpretation, E.J.H., F. Frauscher; statistical analysis, E.J.H.; manuscript preparation, E.J.H.; manuscript definition of intellectual content, E.J.H., F. Frauscher; manuscript editing, revision/review, and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. P. Ives, L. G. Gomella, and E. J. Halpern Effect of Dutasteride Therapy on Doppler US Evaluation of Prostate: Preliminary Results Radiology, October 1, 2005; 237(1): 197 - 201. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. J. Halpern, F. Frauscher, S. E. Strup, L. N. Nazarian, P. O'Kane, and L. G. Gomella Prostate: High-Frequency Doppler US Imaging for Cancer Detection Radiology, October 1, 2002; 225(1): 71 - 77. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
