| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Jefferson Prostate Diagnostic Center, Thomas
Jefferson University, 132 S. 10th St., Philadelphia, PA 19107-5244.
2
Bristol-Myers Squibb Medical Imaging, 331 Treble Cove Rd., N., Billerica, MA
01862.
3
Department of Urology, Jefferson Prostate Diagnostic Center, Thomas Jefferson
University, Philadelphia, PA 19107-5244.
Received September 6, 2001;
accepted after revision October 15, 2001.
Supported by a grant from the DuPont Pharmaceuticals Corporation.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Forty patients were evaluated with harmonic gray-scale sonography. The evaluation was performed before administration of contrast agent, during continuous IV infusion of perflutren lipid microspheres, and again during bolus administration of the microspheres. Sextant biopsy sites were scored prospectively on a six-point scale for suggestion of malignancy at baseline during contrast infusion and after bolus administration. An additional directed core was obtained at 20 of the sextant biopsy sites based on contrast-enhanced imaging.
RESULTS. Cancer was identified in 30 biopsy sites in 16 of the patients (40%). A suspicious site identified during contrast-enhanced transrectal sonography was 3.5 times more likely to have positive biopsy findings at than an adjacent site that was not suggestive of malignancy (p < 0.025). When a suspicious site was evaluated with an additional biopsy core, the site was five times more likely to have a biopsy with positive findings than a standard sextant site (p < 0.01). We found no difference in diagnostic accuracy between continuous infusion of contrast material and bolus administration.
CONCLUSION. Contrast-enhanced transrectal sonography improves the sonographic detection of malignant foci in the prostate. The performance of multiple biopsies of suspicious enhancing foci significantly improves the detection of cancer. There is no advantage to additional examination of the gland after bolus administration of contrast material.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
A recent report revealed improved detection of prostate cancer with targeted biopsy of the prostate using contrast-enhanced color Doppler sonography [7]. A prospective study of harmonic gray-scale imaging in 60 patients reported significantly improved sonographic identification of prostate cancer during infusion of a microbubble sonographic contrast agent [8]. Our study was performed to determine whether the positive yield of prostate biopsy might be further improved with supplemental targeted biopsies using contrast-enhanced harmonic gray-scale imaging of the prostate. Our secondary aim was to compare bolus administration of contrast material with an infusion technique for the identification of prostate cancer.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Over a period of 5 months (from July to November 2000), 40 patients participated in this protocol. Patients ranged in age from 40 to 83 years (mean age, 61 years). Thirty-six patients were included on the basis of an elevated PSA (n = 35) or PSA velocity (n = 1), with a range of 3.6 to 36.5 ng/mL, and a mean of 8.9 ng/mL. Four patients were included on the basis of abnormal findings on a digital rectal examination. During a single visit, each patient in the study underwent a standard sonographic examination of the prostate, a repeated examination during contrast infusion, a repeated examination during bolus administration of contrast, and sextant biopsy of the prostate. The senior author performed all examinations. Examination time was approximately 30 min per patient.
Sonographic examination was performed with the Sonoline Elegra system (Siemens Medical Systems, Issaquah, WA), using software version 5.0, which provides wideband harmonic pulse inversion imaging on the EC6.5 transrectal probe. Gray-scale imaging was performed in both transverse and sagittal planes at the fundamental frequency and repeated in wideband harmonic mode with default settings optimized for contrast harmonic imaging. Four additional transverse harmonic imaging passes were performed in intermittent imaging mode with interscan delay times of 0.2, 0.5, 1.0, and 2.0 sec. The intermittent imaging mode on our system controls the frame rate by setting a fixed time interval between the transmit pulses for each frame. A longer interscan delay will allow more contrast material to traverse into the microvasculature of the imaging plane. The mechanical index was initially set to 0.3, but it was adjusted during the examination to optimize visualization of contrast enhancement. Gray-scale gain was adjusted for baseline imaging and was not altered after contrast injection. The entire continuous and intermittent imaging sequence was performed at baseline and repeated during contrast infusion. The sonographic technique used during bolus administration of contrast material was chosen on the basis of enhancement that was visualized during infusion. In general, continuous imaging or imaging with a 0.2-sec interscan delay was used for bolus imaging. The entire examination was recorded on super-VHS videotape.
The sonographic contrast agent used in this study, perflutren lipid microsphere (Definity; DuPont Pharmaceuticals, Billerica, MA), is a sterile, nonpyrogenic suspension of liposome encapsulated perfluoropropane microbubbles. Definity is composed of a blend of three phospholipids contained in a matrix of sodium chloride, propylene glycol, and glycerin in water. Definity is supplied in a vial that contains the phospholipids as well as perfluoropropane gas. The microbubble agent is supplied in a standard-size vial that is prepared by shaking the vial with the aid of a shaking device (Vialmix; Espe, Seefeld/Oberbay, Germany). For the purposes of infusion, two vials (2.6 mL total) of the agent were prepared immediately before its infusion and diluted in a 50-mL bag of normal saline, yielding a concentration of 49.4 µL/mL. The contents of this bag were infused at an initial rate of 4 mL/min. For the purpose of bolus administration, two equal doses were prepared using up to 10 µL/kg of contrast agent, with the total dose limited to 50 µL/kg. Each bolus was hand-injected over approximately 10 sec and followed by a 10-mL saline flush.
Sextant biopsy was performed after completion of the imaging protocol. When no visible abnormality was found, a standard sextant biopsy was performed by taking three samples from each side of the gland: one sample each at the base, mid gland, and apex. Biopsy specimens were directed preferentially to the lateral portion of the gland to sample outer gland material. When an abnormality was present either at baseline sonography or during contrast-enhanced transrectal sonography, the biopsy specimen from the corresponding sextant was directed toward the visualized abnormality. A second biopsy was performed whenever the original sextant core was thought to be inadequate (n = 4), or at a suspicious site when the examining physician was not certain that the first biopsy passed through the focus of abnormality (n = 16). A total of 20 biopsy sites were sampled twice.
The site of each biopsy specimen was evaluated and rated during the procedure at baseline, again during contrast infusion, and again after bolus administration. Ratings were obtained by consensus of the first two authors. A six-point rating scale was used to force a decision between benign and malignant at each site and to record a degree of certainty for this decision (6 = malignant definite, 5 = malignant possible, 4 = malignant indeterminate, 3 = benign indeterminate, 2 = benign possible, 1 = benign definite). The baseline score was a subjective impression based on gray-scale findings, such as the presence of an echotexture abnormality or a contour deformity. The score of each biopsy site after contrast agent administration was based on baseline findings as well as the level of visualized enhancement during contrast infusion. When baseline gray-scale findings and postinfusion enhancement were discordant, the degree of contrast enhancement was weighted more strongly in determining the postcontrast score. Baseline scores were assigned before contrast infusion. Postinfusion scores were assigned before bolus administration. Postbolus scores were assigned before biopsy. All complications of the procedure and physical complaints by the patient were recorded.
Pathologic evaluation of biopsy cores was the reference standard for calculation of sensitivity and specificity. Each biopsy core was evaluated by a pathologist for the presence of cancer. A Gleason score was recorded for each positive biopsy core. Sensitivity and specificity of baseline sonography and contrast-enhanced sonography were computed both by biopsy site and by patient. The six-point sonographic grading scheme was divided in halves for this calculation. Scores of 1-3 were classified as benign, and scores of 4-6 were classified as malignant. The comparison of baseline and contrast-enhanced results required a paired analysis, which was performed with a McNemar chi-square test for symmetry using Stata 6.0 software (Stata Corporation, College Station, TX).
To determine whether contrast-enhanced transrectal sonography was useful in selecting biopsy sites in patients with biopsy findings positive for cancer, logistic regression was used to calculate an odds ratio for the presence of cancer on the basis of gray-scale interpretation, contrast infusion, and bolus imaging. To compensate for the lack of independence among the six observations in each prostate, we used the conditional form of logistic regression, with multiple sites matched by patient. The pathology result (presence or absence of cancer) served as the outcome variable for the regression model. Univariate analysis was performed, in which the subjective score for suggestion of cancer (at baseline, during infusion, or during bolus administration) served as the independent variable for the regression model.
To determine whether any benefit was derived from additional biopsy cores at sites with suspicious enhancing lesions, we performed a comparison between those 219 sites that were evaluated with a single core and the 20 sites that were evaluated with two cores each. To compensate for the lack of independence among multiple biopsies from individual patients, we performed statistical comparison with conditional logistic regression. As before, the pathology result served as the outcome variable for the regression model. The independent variable for this analysis classified each biopsy core as a single core site or a multiple core site.
To determine whether prostatitis was a significant cause of false-positive results with contrast-enhanced transrectal sonography, we tabulated the prevalence of prostatitis for benign biopsy cores obtained from areas of increased contrast enhancement. The prevalence of prostatitis in this group of cores was compared with the prevalence of prostatitis among all biopsy cores.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One patient complained of back pain after infusion of the contrast agent, and therefore he was not evaluated with the bolus technique. His back pain subsided several minutes after the infusion was stopped, and the patient proceeded directly to biopsy. In one patient, only five of the six sextant cores were obtained because he was unable to tolerate the biopsy procedure. All of the remaining patients completed the entire imaging protocol. There were no other complications.
Gray-scale enhancement was identified in every patient during contrast infusion and again during bolus administration. Although the degree of enhancement was much greater during bolus administration, the delineation of suspicious foci was not significantly better after bolus administration. The change in confidence ratings postbolus is summarized in Table 1. In 94% of sites, the use of bolus imaging made no difference or only minimal difference in the confidence score. A typical patient with postinfusion and postbolus enhancement is illustrated in Figure 1A,1B,1C.
|
|
|
|
Sensitivity and specificity for detection of prostate cancer are calculated
with pathologic interpretation of the biopsy core specimens as the reference
standard. For baseline imaging, sensitivity was 33% (10/30) with a specificity
of 77% (160/209). For infusion imaging, sensitivity was 40% (12/30) with a
specificity of 81% (169/209). For bolus imaging, sensitivity was 43% (12/28)
with a specificity of 79% (161/205). The differences in sensitivity and
specificity at baseline and during contrast infusion or bolus administration
were not significant (p 
0.15, McNemar chi-square test for
symmetry).
Conditional logistic regression yielded an odds ratio of 2.3 for the localization of a cancer on the basis of gray-scale criteria (p = 0.15). The odds ratio increased to 3.5 for localization of a cancer during contrast infusion (p = 0.024), and to 3.7 for localization of a cancer during bolus administration (p = 0.022). Conditional logistic analysis confirms that contrast-enhanced transrectal sonography does provide a significant improvement in the localization of cancer.
Among 20 biopsy sites with two specimen cores, nine cancers were identified (45%). Among 52 sites identified as suspicious during infusion, positive biopsies were obtained in eight of 16 sites with two specimens (positive predictive value 50%). Among 219 biopsy sites evaluated with a single core specimen, only 21 cores with positive findings were obtained (10%). Thus, a suspicious site evaluated with two cores was five times more likely to have positive biopsy findings than a standard sextant site. Conditional logistic regression analysis demonstrated a significantly higher positive biopsy rate from sites sampled with two cores (odds ratio 17, p < 0.01).
Among 239 biopsy specimens in this study, we identified 35 biopsy specimens with pathologically diagnosed prostatitis (14.6%). Among 40 false-positive sites identified during contrast infusion, seven cores were diagnosed as demonstrating prostatitis (17.5%). Among 44 false-positive sites identified during contrast bolus administration, nine cores were diagnosed as demostrating prostatitis (20.5%).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although the infusion of contrast agent provides an extended time for diagnostic evaluation, we reasoned that a bolus technique with greater contrast enhancement would result in better diagnostic accuracy for the detection of prostate cancer. Improved sonographic visualization of hypervascular liver lesions has been shown with bolus administration of a microbubble contrast agent [9]. We evaluated the gland with a bolus injection after the infusion, so that bolus imaging could be directed to evaluation of any suspicious or questionable areas. Our results suggest no additional diagnostic advantage with bolus administration over infusion of contrast material. Although there is greater enhancement with bolus administration, we found no improvement in the distinction between benign and malignant foci.
The presence of prostatitis has been described as a cause for false-positive results on color Doppler evaluation of the prostate. Newman et al. [10] reported that 23% (14/62 biopsy sites) of false-positive findings on color Doppler sonography were identified as prostatitis on pathologic—histologic examination. Prostatitis was pathologically diagnosed in 14.6% of biopsy specimens in our study. Among our false-positive cases with contrast-enhanced endorectal sonography, prostatitis was present in no more than 20.5%. There was no other pathologic explanation for the apparent false-positive results in the remainder of the cases. Thus, although there is considerable overlap in contrast-enhanced endorectal sonography patterns between cancer and prostatitis, it appears that prostatitis is not the major cause of false-positive suspicious foci during contrast-enhanced endorectal sonography.
We based our study on a comparison of endorectal sonographic findings with needle biopsy results. Unfortunately, sampling error during biopsy may result in a false-negative biopsy that is interpreted as a false-positive contrast-enhanced transrectal sonographic finding. Furthermore, the use of contrast-enhanced endorectal sonography to target the sextant biopsy sites introduces workup bias into the study [11, 12]. The issue of workup bias is compounded by the fact that we obtained multiple biopsies from 20 sites on the basis of sonographic findings. Nonetheless, targeted biopsy of selected sites was needed to determine whether there was an advantage to additional biopsy cores based on contrast-enhanced transrectal sonography. The requirement for sextant biopsy in all cases ensured that many biopsy sites were included that did not have sonographic abnormalities. Furthermore, because biopsies were directed to any focus of abnormality seen at baseline, with infusion, or after bolus administration, a similar bias would be expected for un-enhanced, infusion, and bolus imaging.
An important limitation of this study is the subjective quantification of contrast enhancement. Although our two reviewers were unaware of pathology results (that were not available at the time of their assessment), no quantified measure of enhancement was obtained. Computation of a fractional moving blood volume is possible with power Doppler sonography and may allow objective quantification of vascular volume [13, 14]. No similar technique is available for harmonic gray-scale imaging. Nonetheless, we wanted to study the detection of prostate cancer with gray-scale harmonic imaging because of its greater spatial and temporal resolution compared with Doppler imaging. In the future, use of image-based, computer-assisted quantitative assessment of the enhancement kinetics may improve the value of contrast-enhanced harmonic imaging. Huber at al. [15] showed, using computer-assisted quantification, that breast carcinomas and benign breast lesions behave differently after microbubble agent injection, in both degree and dynamics of enhancement. For the purposes of our study, however, we relied on a subjective evaluation of enhancement patterns. Although a quantitative technique might provide a more objective assessment, subjective evaluation of enhancement is the norm for most clinical imaging studies and is more likely to succeed as a clinically practical means for targeted biopsy of the prostate.
Our small study suggests that contrast-enhanced endorectal sonography improves the localization of prostate cancer and that the addition of multiple directed biopsy cores based on contrast-enhanced transrectal sonography results in a significantly higher positive biopsy rate. Contrast enhancement of vessels in the prostate may be easier to see with color Doppler imaging, but superior spatial and temporal resolution are obtained with gray-scale harmonic imaging. Furthermore, intermittent gray-scale imaging should allow more selective imaging of contrast material in smaller vessels [16]. For this technique to gain widespread acceptance among clinicians, however, the distinction between malignant and benign tissue must be improved. Newer bubble agents or imaging technologies may improve the distinction between malignant and benign tissues. Future research in contrast-enhanced imaging of the prostate must focus on selectively improving the enhancement of malignant foci. Further clinical trials are needed to better define the role of contrast-enhanced transrectal sonography during biopsy of the prostate and to compare contrast-enhanced color Doppler—targeted biopsy with harmonic gray-scale imaging.
Acknowledgments
We thank Sharon Molotsky and Donna George for technical and nursing
support.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. S. Adler, S. Fealy, J. R. Rudzki, W. Kadrmas, N. N. Verma, A. Pearle, S. Lyman, and R. F. Warren Rotator Cuff in Asymptomatic Volunteers: Contrast-enhanced US Depiction of Intratendinous and Peritendinous Vascularity Radiology, September 1, 2008; 248(3): 954 - 961. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Yi, J. K. Kim, S. H. Park, K. W. Kim, H. S. Kim, J. H. Kim, H. W. Eun, and K.-S. Cho Contrast-Enhanced Sonography for Prostate Cancer Detection in Patients with Indeterminate Clinical Findings. Am. J. Roentgenol., May 1, 2006; 186(5): 1431 - 1435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Beyersdorff, A. Winkel, B. Hamm, S. Lenk, S. A. Loening, and M. Taupitz MR Imaging-guided Prostate Biopsy with a Closed MR Unit at 1.5 T: Initial Results Radiology, February 1, 2005; 234(2): 576 - 581. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. H. Arger, S. B. Malkowicz, K. N. VanArsdalen, C. M. Sehgal, A. Holzer, and S. M. Schultz Color and Power Doppler Sonography in the Diagnosis of Prostate Cancer: Comparison Between Vascular Density and Total Vascularity J. Ultrasound Med., May 1, 2004; 23(5): 623 - 630. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
