点击显示 收起
1 From the Departments of Radiology (E.J.H., P.A.M., F.F.) and Urology (L.G.G.), Jefferson Prostate Diagnostic Center, Thomas Jefferson University, 132 S 10th St, Philadelphia, PA 19107-5244; and Nycomed Amersham, Oslo, Norway (A.K.A., E.K.H.). Received March 8, 2001; revision requested April 11; revision received and accepted June 5. Supported by a grant from Nycomed Amersham.
| ABSTRACT |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
MATERIALS AND METHODS: Transrectal US was performed in 12 subjects with cancer of the prostate prior to radical prostatectomy. Each gland was evaluated with conventional gray-scale and wide-band harmonic US at baseline and again during intravenous infusion of a microbubble contrast agent. Focal areas of contrast enhancement were identified prospectively in the transverse plane at the base, midgland, and apex of the prostate. US findings were then compared with whole-mount prostatectomy specimens. Baseline and contrast-enhanced findings were compared by using the Wilcoxon signed rank test.
RESULTS: Thirty-one foci of prostate cancer were present at pathologic evaluation, with multiple foci of cancer in 11 of the 12 glands. Three of 10 inner-gland cancers and five of 21 outer-gland cancers were detected at baseline imaging. Diffuse inner-gland enhancement was identified in all subjects during contrast agent infusion. Contrast-enhanced imaging demonstrated an additional five cancer foci in the outer gland (P = .025), for an overall sensitivity of 42% (13 of 31 foci). Seven additional sites of focal contrast enhancement were identified. Five of these sites corresponded to foci of hyperplasia. Two sites were false-positive with no pathologic abnormality. Increased flow was not demonstrated posteriorly in the midline, even when a tumor was present.
CONCLUSION: Contrast-enhanced US of the prostate with Sonazoid can improve sensitivity for the detection of cancers in the outer gland, but it can also demonstrate focal enhancement in areas of benign hyperplasia.
Index terms: Prostate neoplasms, 844.32 • Prostate neoplasms, US, 844.12988 • Ultrasound (US), contrast media
| INTRODUCTION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Results of preliminary studies have demonstrated that contrast material–enhanced US may selectively enhance the microvascularity associated with cancer in the prostate (3). Microvessel density within the prostate is associated with the presence of cancer (4), with metastases (5), with the stage of disease (6–8), and with disease-specific survival (9,10). Quantitative assessment of microvascular density may provide important data that influence therapeutic decision making (11). Preliminary data from several small studies suggest that contrast-enhanced US may improve the detection of prostate cancer (12–15). In a recent study at Thomas Jefferson University, researchers correlated contrast enhancement with sextant biopsy of the prostate and demonstrated that contrast-enhanced intermittent gray-scale US imaging can be optimized to improve the detection of prostate cancer (16).
In previous studies of contrast-enhanced US in the prostate, investigators have compared imaging findings with biopsy results. Unfortunately, approximately 15%–35% of all cancers are missed with standard sextant biopsy (17,18). The purpose of our study was to compare areas of contrast material enhancement in the prostate at ultrasonography (US) with whole-mount radical prostatectomy specimens to determine if the use of contrast material improves the detection rate of prostate cancer.
| MATERIALS AND METHODS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Each subject was evaluated with US at baseline and again during intravenous infusion of Sonazoid (NC100100; Nycomed Amersham, Oslo, Norway). Imaging was performed in the transverse plane with a slow sweep of the transducer from base to apex. All US examinations were performed with the Sonoline Elegra system (Siemens, Issaquah, Wash) with an EC6.5 transrectal probe. Gray-scale imaging was performed at baseline at the fundamental frequency and was then repeated in wide-band harmonic mode by using default settings optimized for harmonic imaging of tissue.
Power Doppler examination of the gland was also performed at baseline imaging. During infusion of contrast material, additional transverse imaging was performed with only wide-band harmonic mode by using default settings optimized for contrast-enhanced harmonic imaging. At least four sequential imaging passes were performed during contrast agent infusion to allow for continuous gray-scale imaging and intermittent imaging with interscan delay times of 1.0, 2.0, and 5.0 seconds. In most subjects, there was sufficient time during contrast agent infusion to perform additional intermittent imaging with interscan delay times of 0.2 and 0.5 second. The mechanical index was initially set to 0.3 for most patients. However, the examiner was allowed to change the mechanical index to optimize visualization of contrast enhancement. Gray-scale gain was adjusted for contrast-enhanced harmonic imaging prior to contrast agent administration, and it was not altered after contrast agent injection. The entire examination was recorded on S-VHS videotape.
Contrast agent infusion was performed via a catheter that was inserted into a vein in the antecubital fossa. Sonazoid is a lipid-stabilized suspension of perfluorocarbon microbubbles with a median diameter of 2.4–2.5 µm. Sonazoid was provided in vials that were reconstituted to yield 2 mL with a concentration of 10 µL of microbubbles per milliliter. A maximum dose of three vials with 60 µL of microbubbles (6 mL) was prepared for each subject. The microbubble infusion was piggybacked onto intravenous tubing with normal saline. This contrast material was infused at an initial rate of 0.015 µL of microbubbles per kilogram per minute. The infusion rate was subsequently titrated to yield the desired level of enhancement. The duration of contrast agent infusion was approximately 10–15 minutes.
US examination was performed by a single experienced radiologist (E.J.H.). A second experienced radiologist (F.F.) was present during nine of the 12 examinations. A map of US findings was created prospectively at the time of imaging. The map was created by consensus opinion with both observers present.
Each lesion visualized with US was marked on a map of the prostate (Fig 1). For purposes of mapping, the gland was divided into base, midgland, and apex. At each of these levels, the gland was further subdivided into inner gland and outer gland with visualization of the surgical capsule. The inner gland was divided into right and left sides. The outer gland was divided into posterior midline, right posterolateral, right anterolateral, left posterolateral, and left anterolateral segments. Thus, each prostate was mapped in 21 segments.
|
Following radical prostatectomy, independent mapping of pathologic results was performed (P.A.M.). Each prostate gland was evaluated with the use of whole-mount transverse sections. Approximately nine to 12 sections were obtained from each gland. Standard hematoxylin-eosin staining of these sections was performed.
The distinction between inner gland and outer gland is often not easily appreciated at pathologic inspection. Nonetheless, the pathologist mapped all visualized cancers with the same map that was used for US mapping. The whole-mount slides from each gland were divided into three groups that were classified as base, midgland, and apical. Many of the visualized lesions overlapped multiple map segments. For each individual lesion within the prostate, major and minor Gleason patterns were defined on a standard scale of 1–5, and a Gleason score was assigned as the sum of these two patterns.
After the independent pathologic and US interpretations were recorded, the two maps were compared. For each abnormality visualized with US that was not present on the pathologic map, the corresponding slides were reviewed to determine whether a benign pathologic process might explain focal US enhancement. A consensus review of the independent pathologic and US interpretations was performed to match US and pathologic results (E.J.H., P.A.M.). For the purpose of subanalysis, the location of each lesion was classified as inner gland or outer gland on the basis of the pathologic map. For lesions that overlapped the boundary between inner and outer gland, if the lesion appeared to have a more inner or outer location, it was classified according to where the majority of the lesion was located. One lesion was evenly distributed between inner gland and outer gland at pathologic interpretation but was visualized only in the outer gland at US examination. This lesion was classified as belonging to the outer gland for the purposes of our analysis, since only the outer-gland portion of the lesion was identified at US.
Statistical analysis was performed to compare the number of cancers detected at baseline imaging and at imaging after contrast agent administration. Because each subject was studied both before and after contrast agent administration, we performed the Wilcoxon signed rank test of equality for matched pairs of observations. The unit of analysis for this comparison was the patient. The null hypothesis was that baseline detection rates and detection rates obtained after administration of contrast material were the same.
To evaluate further the patterns of contrast enhancement, the videotape of each US study was reviewed again after the pathologic results were available. A consensus review was conducted by four of the authors (E.J.H., A.K.A., E.K.H., F.F.). The pattern of contrast enhancement during infusion was noted for the outer gland and inner gland. The intensity of contrast enhancement in the midline of the outer gland was compared with enhancement along the lateral aspect of the gland.
| RESULTS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
US evaluation at baseline demonstrated eight of the 31 cancer foci, including three foci in the inner gland and five foci in the outer gland. At least one focus of cancer was identified at baseline in seven of the 12 subjects. These lesions were all identified on the basis of a focal lesion seen at gray-scale imaging (Fig 2). No additional lesions were found at baseline power Doppler examination.
|
|
|
Although gray-scale enhancement was seen throughout each prostate during contrast agent infusion, the degree of enhancement was not as great in the posterior midline region of the outer gland (Fig 1). Among the outer-gland cancers, seven foci overlapped the posterior midline region at pathologic evaluation. Four of these lesions were detected at contrast-enhanced imaging because of enhancement that was visible in the posterolateral aspect of the prostate (Fig 3). No lesion was detected on the basis of focal enhancement in the posterior midline zone.
Retrospective pathologic interpretation was performed at the seven locations with focal contrast enhancement but with no cancerous findings. These seven lesions were all in the outer gland. Five of these seven lesions were associated with foci of benign hyperplasia. Three of the five hyperplastic lesions had cystic foci, and one of the three was also found to contain a microscopic focus of cancer that was not prospectively appreciated. Two of the seven contrast-enhanced foci contained only normal prostate tissue and must be classified as false-positive US findings (Fig 4).
|
|
| DISCUSSION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Contrast-enhanced US of the prostate with Sonazoid can improve sensitivity for the detection of cancers in the outer gland, but it can also demonstrate focal enhancement in areas of benign hyperplasia. Our sensitivity for outer-gland lesions was doubled from 24% to 48% with contrast agent infusion (P = .025). Among those lesions that were enhanced during contrast agent infusion, the positive predictive value for cancer was 56%. It has been reported that approximately 15%–35% of cancers are missed with conventional sextant biopsy (17,18). The cancer detection rate may be improved by increasing the number of biopsies performed (18,20) or by repeating the biopsy procedure a second or third time (21). However, the results of one recent randomized trial failed to demonstrate any improvement in detection rate by simply increasing the number of biopsy cores obtained (22). Our results suggest that if targeted biopsy is performed with contrast-enhanced US, the detection rate of cancers that are not visible with conventional US may improve.
The Gleason scores of the prostate cancers in this series ranged from 5 to 7, with the majority (24 of 31 cancers) graded as Gleason 6. The absence of higher-grade cancers in this series may be related in part to bias in the referral of patients. At our institution, patients with higher-grade lesions are often enrolled in other treatment protocols prior to referral for radical prostatectomy. Higher-grade tumors tend to have increased microvessel density and should be easier to detect with contrast enhancement (4).
Although contrast-enhanced US can improve the detection rate of cancer, fewer than half of all outer-gland cancers in the present study were detected. Contrast enhancement was limited in the posterior midline segment of the outer gland. This may be related to probe pressure or to a near-field effect (23). Furthermore, contrast-enhanced US was not helpful in the detection of cancers within the inner portion of the gland. The difficulty with detection of inner-gland cancers is probably related to the intense, heterogeneous enhancement pattern associated with benign prostatic hyperplasia (24). Since prostate cancer affects primarily older men, some degree of benign prostatic hyperplasia will be present in almost all subjects with cancer of the prostate.
Fortunately, cancers of the inner gland tend to spread to the outer gland before they metastasize outside the prostate. Thus, a technique that allows detection of only outer-gland cancers will also be useful. Our cancer detection rate, however, remains too low to limit the biopsy procedure to enhancing foci alone. Furthermore, benign hyperplastic lesions are also enhanced with contrast agent infusion. Future research must concentrate on improving the overall enhancement of malignant tissues and developing better methods of distinction for malignant and benign tissues.
Contrast-enhanced imaging of the prostate was performed in both continuous and intermittent imaging modes. Continuous gray-scale US allows real-time imaging of the prostate, and it clearly demonstrates the arrival of contrast material in larger vessels. However, insonation with the rapid frame rate and the high mechanical index commonly used with conventional US destroys much of the contrast agent by rupturing microbubbles before they reach smaller vessels. With intermittent imaging, the ultrasound beam is turned off for longer periods between each image frame, thereby allowing more time for microbubbles to reach the microvascular bed. Intermittent imaging also increases the overall enhancement provided by US contrast agents (25–27). To minimize bubble destruction and to enhance the neovasculature more selectively, we used intermittent US imaging in combination with phase-inversion harmonics and a low mechanical index (3,16).
We believe that a strength of our study was the comparison of US findings with whole-mount slides from radical prostatectomy specimens. Pathologic evaluation of radical prostatectomy specimens allows detection of all cancer foci, not just those cancers found with needle biopsy. On the basis of independent prospective pathologic and US interpretations, we are able to compute a true sensitivity for the detection of cancerous lesions. Furthermore, whole-mount slides can be reviewed retrospectively to identify other pathologic processes that might explain areas of focal enhancement seen with US. We did identify five areas of benign hyperplasia in the outer gland that were associated with contrast enhancement at US.
A limitation of the present study was difficulty with precise three-dimensional correspondence between the US imaging plane and the location of the whole-mount section. US images were obtained with an end-fire probe that must be angled toward the coronal plane as it is moved superiorly to image the base of the prostate. Our US images, especially those obtained toward the base of the prostate, will not represent a true transverse plane. Comparison with pathologic sections is also limited by shrinkage of tissue that takes place during processing of the specimen. Furthermore, there are seldom any precise anatomic landmarks to help identify the precise craniocaudal position of the US imaging plane. An additional limitation of our study was the relatively small study size.
We recognize the above advantages and disadvantages of pathologic comparison with radical prostatectomy specimens. We chose to use an end-fire probe because of its superior imaging characteristics and because, to our knowledge, there is no commercially available side-fire probe that is capable of performing wide-band harmonic imaging and intermittent imaging. Pathologic and US interpretation were limited to only three levels in the craniocaudal dimension to minimize uncertainty with respect to localization in this plane. All imaging sequences were performed in the transverse plane with a standard motion from base to apex, and all procedures were recorded on videotape. These videotapes were available to help resolve any uncertainty in the comparison of US and pathologic results. Since most cancers in the prostate grow along the prostatic capsule, many of the lesions spanned from the base to midgland or from midgland to apex at pathologic interpretation and were therefore easier to line up with our US images.
In conclusion, our small series does demonstrate significantly improved detection of prostate cancer with contrast-enhanced US compared with conventional US techniques. On the basis of these results, we recommend the use of contrast-enhanced US for repeat biopsy in patients with an elevated prostate-specific antigen level and negative sextant biopsy result. However, we cannot presently recommend a limited, targeted biopsy procedure that makes use of contrast-enhanced US, since clinically important cancers may be missed with such a procedure.
| REFERENCES |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|