Value of Power Doppler and 3D Vascular Sonography as a Method for Diagnosis and Staging of Prostate Cancer

Transcription

1 European Urology European Urology 44 (2003) Value of Power Doppler and 3D Vascular Sonography as a Method for Diagnosis and Staging of Prostate Cancer J.L. Sauvain a,*, P. Palascak b, D. Bourscheid c, C. Chabi b, A. Atassi b, J.M. Bremon a, R. Palascak b a Medical Imaging Center, 6 passage Jules Didier, Vesoul 70000, France b Urology Division, Paul Morel Hospital, Vesoul 70000, France c Internal Medicine and DIM Division, Paul Morel Hospital, Vesoul 70000, France Accepted 11 March 2003 Abstract Objectives: To compare the value of Power Doppler Sonography (PDS) and B mode sonography in the diagnosis of prostate cancer and to assess the value of PDS to specify capsular effraction of the cancer. Patients and Methods: 323 patients were investigated: 41 control subjects allowed the establishment of normal vascular semiology and 282 patients with suspected cancer (PSA >4 ng/ml). Power Doppler Sonography with 3D reconstruction was used to describe Power Doppler Sonography features of normal or abnormal vessels. Three types of blood supply(a: regular avascular posterior peripheral margin, b: irregular avascular posterior peripheral margin, c: vessels crossing the posterior peripheral margin) were described as a function of the presumed stage of cancer (a: intraprostatic, b: undetermined, c: extraprostatic). Comparison with histology was performed on random biopsies without Doppler (282 cases) (median PSA level ¼ 15.8 ng/ml), on second biopsies indicated with PDS (72 cases), and radical prostatectomy specimens (63 cases). Results: A cancer was diagnosed in 157 of the 282 patients (55.7%) with suspected cancer. The overall sensitivity of PDS in the initial diagnosis of prostatic cancer was 92.4% and its specificity was 72% (versus 87.9% and 57.6% for sonography alone respectively). The negative predictive value of PDS was elevated to 80.6% ( p < 0:0001). Targeting area presenting abnormal blood flow in any part of the prostate was useful to detect isoechoic or lesions in patients with first negative biopsy results (in 41 of 72 targeted patients with first negative biopsies with PDS a cancer was diagnosed: 58% of these cancers had less than 3 positive biopsies and 34% only one positive biopsy). The 3 vascular types a, b, c were evaluated prospectively in the detection of capsular effraction. The presence or absence of vessels crossing the capsule to determine an extracapsular extension was a significant sign ( p < 0:0001). Capsular effraction was detected in 3 of the 27 cases (11%) of type a cancer and in 16 of the 18 cases (87%) of type c cancer. Conclusion: PDS improves the accuracy of echographic imaging in the diagnosis of cancer. Combining first sextant biopsies and targeted areas presenting abnormal blood flow using PDS can increase cancer detection with an optimized number of biopsy cores. The risk of extracapsular involvement can be evaluated by the presence of vessels perforating the capsule. # 2003 Elsevier Science B.V. All rights reserved. Keywords: Prostate Doppler; Prostate biopsy; Prostate cancer 1. Introduction Abbreviations: US, ultrasonography; PDS, Power Doppler Sonography; 3D PDS, three dimensional vascular reconstruction with Power Doppler Sonography; PSA, prostate specific antigen; PPV, positive predictive value; NPV, negative predictive value * Corresponding author. Tel. þ address: jlfsauvain@aol.com (J.L. Sauvain). Prostate cancer is a common disease in older men. Prostatic tumors range from small slow growing lesions to aggressive tumors that metastasize rapidly and there is controversy about which screen detected lesion will become clinically significant [1]. Current methods of /03/$ see front matter # 2003 Elsevier Science B.V. All rights reserved. doi: /s (03)

2 22 J.L. Sauvain et al. / European Urology 44 (2003) screening involve measurement of PSA [2 4] followed by transrectal ultrasound scanning and biopsy [5]. Gray scale US is limited as a screening test [6 8] and lacks adequate specificity and sensitivity. With gray scale ultrasonography prostate cancer more often presents visible findings like a hypoechoic area or can sometimes appear echogenic, but cancer is also isoechoic and invisible (50% of non palpable tumor in some radical prostatectomy series) [9]. A hypoechoic area can be found in prostatitis lesions. Color Doppler improves the echographic diagnostic predictive value and specificity in cancer by virtue of demonstration of abnormal blood flow [10]. Color Doppler sonography has also been shown to be an adjunct to gray scale US improving the accuracy of cancer detection especially for isoechoic cancer and cancer with elevated Gleason score and allowed more precise orientation and guidance of biopsies [11 13]. So other authors found endorectal ultrasonography and Color Doppler imaging useful to differentiate low risk invisible and hypovascular tumors from high risk visible and hypervascular tumors [14]. However gray scale and Doppler Imaging did not reveal prostatic cancer with sufficient accuracy to avoid sextant biopsy [10] and only few urologists used echographic findings to perform biopsies [15]. Power Doppler Sonography (PDS) [16 18] is more sensitive to slow flow and less angle dependent than Color Doppler [19]: the result is that blood flow measurement allows a better analysis of intraprostatic vessels without variation in color or intensity according to flow orientations. Some authors suggested using Power Doppler for targeted biopsies when the number of biopsies must be limited [15]. Power Doppler also allows three dimensional reconstruction of blood flow: its interest was shown when used with contrast agent to perform biopsies [20] but is unknown in prostate cancer stage. The purpose of this study was to compare the value of Power Doppler Sonography (3D PDS) with three dimensional reconstruction and B mode sonography in the diagnosis of prostate cancer and to assess the value of PDS to specify capsular involvement. 2. Materials and methods 2.1. Patients 323 patients were investigated between January 1997 and October One set of 41 subjects allowed the establishment of referring intraprostatic blood flow semiology with PDS. These subjects at least 35 years old without clinical symptoms related to prostate disease and with an absence of abnormal findings at both digital rectal examination and gray scale transrectal US with a prostate volume less than 15 g were examined on the basis of microscopic hematuria or sterility. The second set included 282 patients with suspected cancer, ranged from 49 to 87 years old with a median age of 70.5 years and presented an abnormal elevated serum PSA level greater than 4 ng/ ml (Table 1) [3,21] (median PSAwas 15.8 ng/ml graded from 4.7 to 500). An initial and referring exploration with Power Doppler Sonography was performed in all patients before first random sextant biopsies without PDS. 72 patients with negative initial biopsies were maintaining six month later an abnormal PSA level: repeat directed biopsies with PDS were indicated. The results (Table 1) of these two set of biopsies were compared to the initial and referring Power Doppler exploration. 157 cancers were diagnosed, 116 after the first set of random biopsies without PDS and 41 after the second set of repeat directed biopsies with PDS. 101 cancers were palpable and 56 non palpable. 63 patients with clinically localized cancer underwent a radical prostatectomy (Table 2) Technique used for Power Doppler mode and 3D vascular imaging The transrectal gray scale US and PDS was performed by two experienced radiologists with echo Doppler equipment: ATL HDI 3000 and ATL HDI 5000 coupled to a wide band transrectal probe C 9-5 mz with the Doppler frequency set at 6 MHz (Advanced Technology Laboratories, Bothell, WA, USA). In addition to the transrectal ultrasonography, exploration using Power Doppler mode was made on all patients by exerting slight pressure on the probe. Prior control settings were selected in advance and used in the same manner at every examination. The Doppler gain was optimized using the technique described by Rubin [19]: the appropriate value for which there was no background noise was 80% of the maximum value. The PRF (Pulse Rate Frequency) was adjusted to 700 Hz standard value. Filtering and persistence were preset to their maximum values for an average line density, thus allowing use of a maximum color palette covering the entire gland. Three-dimensional reconstruction with Power Doppler of intraprostatic blood flow imaging (3D PDS) was performed for a better understanding of the prostate perfusion. 3D PDS was obtained with the capture of 11 Power Doppler US images after scanning a Table 1 Comparison between results of 282 random biopsies without Doppler and 72 random biopsies with adjunction directed biopsies in lesions with increase blood flow detected by Doppler according to PSA value Patients with suspected cancers Cancers confirmed with biopsies Cancers confirmed with first sextant biopsies Cancer confirmed with PDS biopsies 4 < PSA < (48.6%) 49 (35.8%) 29 (59%) 20 (41%) PSA > (51.4%) 108 (74.5%) 87 (80.5%) 21 (19.5%) Total (55.6%) 116 (74%) 41 (26%)

3 J.L. Sauvain et al. / European Urology 44 (2003) Table 2 Results of PDS in 63 localized cancers subjected to radical prostatectomy and prospectively evaluated Type a Type b Type c Number Mean Gleason score PSA Range Mean Median Sextant <50% 15/27 (55.5%) 6/18 (33.3%) 3/18 (16.6%) Grade 4 or 5 13/27 (48%) 8/18 (44%) 14/18 (77%) Prostate weight Mean 35 g 38 g 42 g Range Clinical stage TNM T1c T2a þ b T2c Anatomical stage pt pt2a 4 0 pt2b or pt2c pt3a or pt3b 13 þ 2Mþ 6 10 pt3c suspicious volume. Then all these images were added in a three dimensional reconstruction allowing us to make an angular movement of 15 degree. On the final image only Power Doppler acquisitions were present without gray scale data Gray scale US and Power Doppler imaging studies Gray scale ultrasonography allowed to search for any abnormal area or nodule (peripheral or central prostate) and revealed its echogenecity (hypo-, hetero- or hyperechoic), outline appearance (regular or irregular) and the presence of an associated peripheral bending. Power Doppler mode located flow in periphreal prostatic capsular vessels, parenchymal vessels and vessels anastomosing with extraprostatic vessels (Fig. 1). Wherever a nodule or abnormal area was found, the presence or absence of posterior prostatic margin blood flow was noted. Also a search around the lesion for detectable vessels allowed a mass effect to be assessed, when a gap in the perfusion was present (Fig. 2). 3D PDS looks for area or intraprostatic volume with an increase in the number of detectable vessels or asymmetry in the distribution of blood flow in each investigated case. A vascular density in the suspected area or volume was noted. Three vascular types of intracapsular and peripheral blood flow were prospectively evaluated in the 63 patients who underwent a radical prostatectomy (Fig. 3): Type a: perfusion of prostatic posterior margin was absent, effaced by the tumor mass effect. A regular, hypoechogenic margin with a thickness of between 1 and 2 mm was found behind the most peripheral detectable vessel. Extracapsular involvement was presumed to be absent. Cancer without measurable blood flow was considered as type a. Type b: the perfusion in posterior margin is also absent but the margin was irregular and exhibits localized thinning. There was no detectable anastomosis with the extraprostate vascular system with respect to the lesion. In this type, investigation of extraprostatic spread is difficult and could not be determined. Type c: the posterior peripheral margin was absent in places and the vessels having tumoral origin left the prostate gland to create direct anastomoses with the periprostatic vessels. In this lesion type, extracapsular involvement was presumed to be present. Results with PDS and gray scale were cosigned by the two radiologists A file was set up for each patient which contained the Fig. 1. Normal capsule vascularization on Power Doppler Sonography and in 3D vascular imaging. (a) Axial section: Visualization of 3 vessels in posterior peripheral margin on axial section and placed symmetrically on 2D imaging. The capsule is raised in 3D by its vessels. (b) Axial section: posterior peripheral vessels of left and right pedicular origin on Power Doppler and in 3D PDS. (c) Axial section: Anastomosis between periurethral and posterior peripheral vessels.

4 24 J.L. Sauvain et al. / European Urology 44 (2003) obtained with a 18 gauge core biopsy system. The specimens were labeled according to their location. The number of sextants involved by the tumor was recorded Pathological analysis Pathological analysis was performed by experienced pathologist. Radical prostatectomy was realized with McNeal method (Standford University) for the prospective study. The pt classification was used to assess intra- or extraprostatic tumor spread. If a few tumor cells were present exterior to the prostate capsular penetration was classified as focal or established if more extensive extraprostatic spread was present [22]. A Gleason score was determined for all tumors. 6 patients with a single positive margin observed in one place without periprostatic tissue and no signs of capsular penetration were classified pt2 with one positive margin [23]. Fig. 2. Mass effect of a small differentiated avascular cancer of the transitional zone that was the subject of an echo-guided punch biopsy: (a) Axial section using Power Doppler: small hypoechogenic avascular nodule: the vessels are effaced by the mass effect at the back of the lesion. (b) Using 3D Power Doppler: a vascular gap is noted. set of echographic analysis criteria of the lesion. The results of DRE, PSA level were available at the conclusion of the imaging but all explorations were done prior availability of the pathology reports. The radiologist was blinded to the pathologic findings Prostatic biopsies series The ultrasound machine with Power Doppler was not available in the urologic center and initial sextant random biopsies were performed by urologic surgeons without Doppler, blinded to the PDS results (3 to 4 biopsies in each lobe according to volume). Urologic surgeons and trained physicians performed the second set of repeat biopsies using PDS in the Radiology department without anesthesia. Seven to eight cores were taken, one per sextant and one or two in suspected area or volume of the prostate when anomalous vessels with 3D PDS were present (Fig. 4). When no anomalous perfusion was present two biopsies in each peripheral zone and two in the transition zone were performed. Biopsy specimens were 2.6. Statistical analysis Gray scale and PDS findings were correlated with the results of pathological examinations. A true positive gray scale US exploration was defined as one gray scale abnormality in the same area as a positive biopsy. A true negative gray scale US exploration was defined as the absence of malignancy in all biopsies. Identical rules were applied for PDS. Calculation of sensitivity, specificity, positive and negative predictive values and accuracy of gray scale and Power Doppler US in diagnosis of cancer was made and the statistical significance was calculated with the w 2 test with p < 0:05 considered to indicate the level of significance. Discordance between the two methods was calculated using the McNemar test. In the prospective study of extracapsular involvement with Power Doppler we assessed the risk for each situation (positive, indefinite, negative test). We calculated sensitivity, specificity, predictive values and accuracy in case of presence or absence of detectable vessels going across the capsule to determine extraprostatic spread. We analyzed the presence or absence of extracapsular extension by using the w 2 test. The significance threshold was set at Results 3.1. Normal subjects (n ¼ 41) From the anatomical point of view, the prostate gland is vascularized principally by a branch of the internal iliac artery, the vesico-prostatic artery which is divided Fig. 3. The three vascular types encountered (Power Doppler using different power settings) in three cases of localized cancer of identical size and appearance with PSA lower than 10, having undergone radical prostatectomy. (a) Slightly vascularized hypoechogenic nodule presenting no posterior peripheral vessels, classified according to our criteria as type a: no extraprostatic spread. (b) Vascularized hypoechogenic nodule with an intermittent thin and irregular posterior peripheral margin, classified according to our criteria as type b: capsular involvement indefinite. (c) Nodule classified type c with a vessel perforating the posterior margin; extraprostatic spread present.

5 J.L. Sauvain et al. / European Urology 44 (2003) Fig. 4. Patient with normal DRE (digital rectal examination) and a PSA of 11 who underwent 6 initial negative randomized biopsies. (a) Deformation of the basal portion of the prostate that raises the left seminal vesicle but having normal echo-structure and identical to the rest of the gland. (b and c) Pathologic hypervascularization on Power Doppler and 3D imaging. (d) Positive echo-guided biopsy (Gleason score 7). into an inferior vesicular artery going to the inferior surface of the bladder and the internal prostate, and a prostate artery vascularizing the peripheral prostate by way of its pedicular, capsular, and parenchymal divisions. Accessorially the middle hemorrhoidal and internal pudendal artery also participate in this blood supply. With PDS and 3D PDS, in normal subjects (Fig. 1)at the level of the peripheral prostate and at the lateral edges, capsular arteries radially supplying the parenchymal branches were found in all the patients: on coronal section they were symmetrically distributed with a weak Doppler signal. There were no secondary division branches observed at the level of the intraprostatic parenchymal vessels. In 26 of the 41 referring subjects, a 1 to 2 mm thick posterior peripheral margin was identified that contained vessels of diverse origin: sagittally, medially, and laterally arranged vessels having their origin in the inferior vesical net; others having a periurethral origin and finally, others originating at the lateral pedicular level. On 3D PDS reconstruction, this group of vessels forms a marginal system that delimits the posterior surface of the prostate. In the normal state there were no apparent anastomoses visualized on Power Doppler Sonography appearing between the vessels forming the marginal system and the intra-parenchymal or extraprostate vessels. Near the postero lateral border, identification of neurovascular bundles in this referring group were suggested on coronal scans on both sides of the prostate like vessels giving branches to the posterior peripheral margin Patients with suspicion of cancer (n ¼ 282) The criteria of analysis were: increase in number of intra-lesion vessels, disoriented vessels or verticalized vessels in peripheral gland, asymmetrical blood flow, mass effect on the intraprostatic perilesional vessels and vessels in peripheral margins. In the series including 282 patients in which a cancer was found in 157 patients (55.6%) (Table 1): 145 (92.3%) cancers showed blood flow abnormalities, 118 of which were hypoechoic, 19 isoechoic and 20 hyperechoic or heterogeneous. In 12 cancers blood flow was absent (4 hypoechoic and 8 isoechoic). Among the hypoechoic lesions with gray scale US 27 were subtle being only visualized after PDS showed anomalous blood flow.

6 26 J.L. Sauvain et al. / European Urology 44 (2003) cancers (median PSA level ¼ 22 ng/ml) were diagnosed with initial sextant biopsies (31 cancers with less than 3 positive sextants and 21 with one positive sextant). In 72 patients keeping an elevated serum PSA level (median PSA level ¼ 8.4 ng/ml) 41 cancers were diagnosed with repeat biopsies with Power Doppler (Fig. 4) six months after negative initial biopsies and corresponding to 26% of initially misdiagnosed cancers. Among the 49 cancers with PSA level range 4 10 ng/ml, the number of misdiagnosed cancers was 41%. In this series 24 cancers (58%) had less than 3 positive sextants and 14 (34%) only one positive sextant. In 26 cancers (63%), positive biopsies corresponded with PDS echographic findings. In 12 other cancers (29%) some of the positive biopsies corresponded to the hypervascular area, but other positive biopsies were found in hypovascular lesions in other parts of the prostate. Positive biopsies were present in six isoechoic and hypervascular cancers but also in three hypovascular cancers misdiagnosed (7.3%) with PDS. When a lesion was identified with gray scale US, if an increased blood flow was present the risk of having positive core biopsies was 81%, conversely if the echographic lesion was hypovascular the risk was 14%. In the 125 patients with elevated PSA level and where no cancer was found on biopsies, no pathological findings were present in 16 patients, adenoma was found in 93 cases associated with prostatitis lesions in 33 cases. Isolated prostatitis was present in 16 patients. Among patients with prostatitis, hypoechoic lesions with increased blood flow were found in 35 cases and explain our false positive results. This study of 282 patients suspected of cancer allowed us to evaluate the appropriateness of Power Doppler in the diagnoses of cancers (Table 3): sensitivity was increased to 92.4%, specificity to 72% (versus 87.9% and 57.6% for sonography alone). Table 3 Comparison of gray scale US and Power Doppler in a series of 282 patients with an elevated serum PSA >4 ng/ml: 157 cancers were diagnosed (56%) Gray scale US True positive True negative False positive False negative Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy Power Doppler ThePPVwas80.6%andtheNPVwas88.2%(Table 3). Gray scale US and PDS significantly enable prediction of cancer (gray scale US: odd ratio ¼ 9:87 p<0:0001 and PDS: odd ratio ¼ 31:07 p < 0:0001). With the Mac Nemar test there was no significant difference between US gray scale and PDS in diagnoses of cancer (p ¼ 0:12), but PDS was significantly higher than gray scale US to eliminate a cancerous lesion ( p ¼ 0:0013). In the group of 72 patients with repeat biopsies sensitivity, specificity, PPV and NPV were respectively 92.7%, 80.6%, 86.4% and 89.3% for PDS (versus 78%, 74.4%, 80%, and 71.9% for US gray scale) 4. Prospective series: evaluation of extracapsular extension The three vascular types a, b, c (Fig. 3)defined in the methods were prospectively evaluated in 63 patients The results are presented in Table 2: 27 lesions were classified as type a without signs of extraprostatic spread. In 24 cases there was no capsular involvement, but in 3 patients there was lateral extraprostatic spread. In this group 3 cancers had no measurable blood flow and no extraprostatic extension and 2 cancers were insignificant. In this group no seminal invasion was noted. 18 lesions were classified type b: Extraprostatic tumor spread was present in 8 cases and absent in 10 cases. 18 lesions were classified type c and exhibited signs of extraprostatic spread on PDS: in 16 instances there was capsular involvement but in 2 cases the capsule was unscathed. In this group 33% of the patients had a PSA level greater than 20 ng/ml and 50% a tumor grading Gleason score elevated (4 or 5). 84% of the tumor had more than 50% of positive sextant. The risk of extracapsular extension in our cancer group with vascular type c was 88% and only 12% in vascular type a (44% in vascular type b). The presence or the absence of vessels crossing the capsule to determine an extraprostatic tumor spread was a significant sign (odd ratio ¼ 24:93 and p < 0:0001) (sensitivity ¼ 59:3%, specificity ¼ 94:4%, PPV ¼ 88:9%, NPV ¼ 75:6% and accuracy ¼ 79:3%). In the group of 27 patients with a vascular type a, 6 patients (22%) had a high risk of showing an extraprostatic spread with a PSA level greater than 15 ng/ml, Gleason score greater than or equal to 7 and 50% more of positive sextants after biopsies; an extraprostaic spread was found in only one patient.

7 J.L. Sauvain et al. / European Urology 44 (2003) Conversely in the group of 18 patients with a vascular type c in the 5 patients (27%) with a PSA level less than 10 ng/ml, an extracapsular involvement was present. 5. Discussion The vascular anatomy of normal prostate as displayed by PDS is highly sensitive in depicting blood flow, the number, course and continuity of vessels. These data allow comparison of the vascular anatomy of the normal prostate with diseases such as prostate cancer [17,18]. 3D PDS allows a spatial study of intraprostatic vessels and improves detection of abnormal vascular density in isoechoic cancer or tumor mass effect in lesions where no blood flow is measurable. PDS facilitates the detection of subtle abnormal echographic lesions not initially targeted but discovered at a later date. In our series, 6 localized cancers (3.8% of all cancers)without any modification on gray scale sonography were controlled by echo-guided biopsy and thanks to Power Doppler. Initial studies performed with color Doppler imaging searched for areas of increased blood flow or asymmetry in color flow signal intensity between two sides in the prostate but did not analyze modification of vessels in space as is performed with 3D PDS. The issues of this study were extremely positive (overall sensitivity of PDS ¼ 92:4% and specificity ¼ 72%) and perhaps its seems every tumor can be detected with PDS contrary to experience of other investigators. Several remarks must moderate our results: First, to evaluate the sensitivity and the specificity of initial PDS performed before biopsies in the set of 282 patients with suspected cancer and to calculate the number of true negative and true positive explorations a pathological correlation in each case was necessary. Admittedly approximately about 20% to 30% of cancers were misdiagnosed with first sextant random biopsies (21). So repeat biopsies are indicated in patients with still abnormal PSA level results six months after the first set of biopsies. The fact we performed a second biopsy session increased the sensitivity if only first sextant biopsies with gray scale US had been used. Furthermore the fact the investigator knew the PSA was still above 4 after six months also influenced the chance of finding a cancer. We can also discuss that only positive result after biopsy must be consider and theoretically after the two set of biopsy we don t have the certitude a cancer was not present in patients with negative biopsy. The second remark concern the 29% of cancer diagnosed with PDS in repeat set of biopsies: positive biopsies were found in hypervascular area visualised with PDS but also in normal echographic area where no blood flow was measurable. A true positive result was recorded for patients in whom a single cancer was present in the same location as PDS depicted. If additional biopsies in same patient were positive in normal area with PDS, the case was still considered like a true positive result: this fact can also falsely elevated the sensitivity of PDS in initial diagnosis but explain the limits of PDS to screen prostate cancer. The third remark concern conditions who permitted to realize this study: this work was performed by the same team during about for years between January 1997 and October 2001 with the same protocol and the same investigators who get progressively experiences of PDS in Prostatic cancer. The inter- and intra-observer variability was not studied. The quality of machine being used can also modified final results. Immunohistological studies confirm the existence of vascularization on a par with prostate cancers thanks to the measurement of microvessel density [24,25] in terms of healthy tissue and tumor tissue. Microvessels density is increased in tumor tissue compared to normal prostate tissue and this microvessel density is more significant at the center of the tumor than in its periphery which explains the absence of necrosis. An increased microvessel density is an element of unfavorable prognosis and would correlate well with the Gleason score and tumor extension [26]. Some authors [20] performed contrast 3D PDS with contrast agent and concluded that contrast ultrasonography is a minimally invasive imaging method which has the potential to visualize lesion with increased microvessels density and could be used in detection of cancer. A blood flow is present in 85 to 90% of all cancers with color Doppler US and color Doppler has been shown to be an important adjunct to gray scale ultrasonography, improving the sensitivity of isoechoic cancer detection [17,27]. So in our study 27 lesions were isoechoic (17% of all lesions) and in 19 cases PDS detected an increased blood flow. If we added to these lesions the 27 cases where gray scale abnormality was subtle and identified only after PDS, 34% of cancerous lesion were not visible on gray scale US alone. The subtle hypoechoic lesions are included in the hypoechoic lesions of our series and this fact can also explain the weak number of isoechoic cancers in our final results. Our study concerns a heterogeneous population of cancers where 92.3% of which presented abnormal blood flow with Power Doppler. In 7.8% of the other cancers, there was no measurable blood flow with 3D

8 28 J.L. Sauvain et al. / European Urology 44 (2003) PDS according to small lesions or lesions with a very low flow. Perhaps in these cases contrast agent would be interesting to detect microvessels. Contrast enhancement color Doppler was used by some authors to target biopsy and detected as many cancers as systematic biopsies with less than half the number of biopsy cores [28]. However color Doppler and Power Doppler, 3D PDS with or without contrast agent did not reveal prostatic cancer with sufficient accuracy to avoid sextant or extended biopsies. Sextant biopsies alone do not provide an adequate tissue sample for accurate detection of disease in the prostate gland and are sub optimal in patients with PSA range of 4 10 ng/ml [29] and extensive biopsy protocols are suggested [30,31]. 26.1% of all the cancers in our study were misdiagnosed by the first sextant biopsies without Doppler and 41% in patients with a PSA level range 4 10 ng/ml. Prevalence of a pathology is very important data to estimate the predictive value: Brawer in a large series of patients with a serum PSA level greater than 4 ng/ml estimated the predictive positive value was about 30% [21]. Although in our study the prevalence of cancer was elevated to 55.6% and the NPVof PDS was 88.2%: a negative Doppler echo examination can provide reassurance specially for patients whose PSA values are ranged 4 10 ng/ml when all studies, including repeat biopsies are negative. On this point our study confirms the positions of other authors [27]. The sole risk undertaken by these patients would be that of being a patient with a cancer with no measurable blood flow (11/157 cases or 7.6% in our series) but the central question that remains is whether radical treatments can improve survival and quality of life in small growing lesions [1]. All the more, several studies confirm there to be a correlation between blood supply and Gleason score [27,32,33] and in the case of avascular cancer the probability of misdiagnosing an aggressive tumor is low [25]. Prostate cancer was found [31] in 54% of 74 cases suspected by Niemann performing extended sector biopsy: 25% of these cancers were detected in additional regions only. In our series similar results were observed and a cancer was found in 56% of patients, 26% of which were misdiagnosed with the first set of sextant biopsies. 3D-PDS can contribute to the decrease of the number of biopsy cores and can be proposed for initial biopsy in patients with a PSA level range 4 10 ng/ml where sextant biopsies are sub optimal (41% of cancer with PSA level range 4 10 ng/ml in our series were misdiagnosed with the first set of sextant of biopsies). PDS and contrast enhancement color Doppler seem to give similar results. On this point we agree with authors [28] practicing contrast enhancement color Doppler targeted biopsies and detected as many cancers as systematic biopsies with fewer than half the number of biopsy cores. Clinical examination by rectal palpation is inadequate and underestimates the developmental stage [4]; 40 to 50% of the patients operated on for a T1 or T2 stage cancer have signs of extraprostate extension on the prostatectomy specimen (42.8% in our series). With respect to diagnosis of invasion of the apex, where the capsule and perirectal fat are absent, imaging diagnoses are still more delicate and study of intraprostate invasion of seminal vesicles can be made only on anatomical-pathological specimen. Isolated endorectal echography has limited interest because it has low sensitivity, the deformation of the capsule being the sign of capsule invasion, rupture of the periprostatic fatty border, penetration of the neuro-vascular bundles. In normal subjects a capsule is sometimes identified on the echography in the form of a small hypoechogenic band [34], but this is not a histological reality; it is a pseudocapsule containing connective tissue, smooth muscle fibers and vasculo-neural elements 1 to 2 mm in thickness separating the prostate tissue from the perirectal fat. This capsule is thinner at the base and nonexistent at the apex [35]. The irregular appearance or the absence of this ultrasound capsule in the event of cancer would be in favor of capsular spread, but this sign overestimates the extension of many cancers [36]. Tumor vascularization is central, its extension is centrifugal and the posterior peripheral margin with vascular structures are initially obliterated by the tumor mass effect (Fig. 2): there are no vessels behind the tumor and the posterior peripheral margin presents regular borders (type a). When the tumor is growing it thins out and becomes irregular (type b). Tumor vascular elements can thus perforate the capsule and anastomize with the extraprostate vessels (type c). This theoretical aspect cannot, however, be applied to the apex and the base which are deprived of capsule that can be analyzed and that are at the base of false negatives results. In type b cases where the investigation of extraprostatic spread is undetermined, the risk is increased and calculated at 44%. This vascular type is encountered in 28% of the cases in our series and could need other imaging techniques or biopsies of periprostatic spaces. Each cancer is different and it is difficult to apply a statistical risk to an isolated patient. Partin s and D Amico s tables allow evaluation from PSA level, grading Gleason score and the number of positive

9 J.L. Sauvain et al. / European Urology 44 (2003) sextants patients with either a low risk or high risk of extraprostatic spread but never provide a certitude. When a patient is included, with these tables, in a group with a high risk of extracapsular involvement, in case of vascular type a observed with PDS, a surgical abstention can be questionned; conversely in a patient included in a group with a low risk of extracapsular spread but a vascular type c with PDS, treatment such as brachytherapy is perhaps not the best choice. combining a first set of sextant biopsies and targeted areas presenting abnormal blood flow, the use of PDS can increase in cancer detection with an optimized number of biopsy cores. The risk of extracapsular involvement can be evaluated by the presence of vessels perforating the capsule. On the other hand a mass effect exerted by a hypo or hyper vascular lesion on a regular peripheral margin where no blood flow is found behind this lesion can suggest a localized intraprostatic cancer. 6. Conclusion Power Doppler and three-dimensional vascular imaging improve the accuracy of echographic imaging in the diagnosis of cancer. In patients suspected with cancer with serum PSA level range 4 10 ng/ml Acknowledgements We thank Robert Pepin and Claire Blandin for the translation. We also thank all the staff of the department of Urology for its participation and the DIM division for the statistical analyses. References [1] Neal DE, Donovan JL. Prostate cancer: to screen or not to screen? Lancet Oncol 2000;1(1): [2] Masai M, Ito H, Kotake T, Nagao K. Serum prostate specific antigen and prostate specific antigen density in patients receiving radical prostatectomy. Int J Urol 1996;3: [3] Stamey TA, Yang N, Hay AR, McNeal JE. Prostate specific antigen as a serum marker for adenocarcinoma of the prostate. New Engl J Med 1987;317: [4] Bostwick DG. Staging Prostate cancer-1997: Current Methods and limitation. Eur Urol 1997;32:2 14. [5] Littrup P, Sparschu R. Transrectal ultrasound and prostate cancer risks: the tailored prostate biopsy. Cancer 1995;75(Suppl): [6] Resnick NI, Willard JW, Boyce WH. Ultrasonic evaluation of the prostatic nodule. J Urol 1981;120:86 9. [7] Rifkin M, Sudakoff G, Alexander A. Prostate: techniques, results and potential applications of color Doppler US scanning. Radiology 1993;186: [8] Aarnink RG, Beerlage HP, de la Rosette JJ, Debruyne FM, Wijkstra. Transrectal ultrasound of the prostate: innovations and future applications. J Urol 1998;159: [9] Brendler CB. Characteristics of prostate cancer found with early detection regimens. Urology 1995;46(3 Suppl A):71 6. [10] Kelly I, Lees W, Rickards D. Prostate cancer and the role of color Doppler US. Radiology 1993;189: [11] Newmann JS, Bree RL, Rubin JM. Prostate cancer: diagnosis with color Doppler sonography with histological correlation of each biopsy site. Radiology 1995;195(1): [12] Bree RL. The role of color Doppler and staging biopsies in prostate cancer detection. Urology 1997;49:31 4. [13] Lavoipierre AM, Snow RM, Frydenberg M, Gunter D, Reisner G, Royce PL, et al. Prostatic cancer: role of color Doppler imaging in transrectal sonography. Am J Roentgenol 1998;171(1): [14] Cornud F, Hamida K, Flam T, Helenon O, Chretien Y, Thiounn N, et al. Endorectal color Doppler Sonography and endorectal MR, imaging feature of non palpable prostate cancer: correlation with radical prostatectomy findings. Am J Roentgenol 2000;175: [15] Halpern EJ, Strup SE. Using gray-scale and color and power Doppler sonography to detect prostatic cancer. Am J Roentgenol 2000;174: [16] Sauvain JL, Palascak P, Bremon JM. Power Doppler ultrasonography and hypoechoic nodules of peripheral prostate: prospective and limitations. J Radiol 1997;78: [17] Sauvain JL, Palascak P, Bourscheid D, Bloqueau P, Bremon JM, Jung L, et al. Power Doppler and 3D vascular sonography of intra prostatic blood supply: assessment criteria and value for diagnosis and staging of prostate cancer. Prog Urol 2000;10: [18] Leventis AK, Shariat SF, Utsunomiya T, Slavim KT. Characteristics of normal prostate vascular anatomy as displayed by power Doppler. Prostate 2001;46(4): [19] Rubin JM, Bude RO, Carson PL, Bree RL, Adler RS. Power Doppler US: a potential useful alternative to mean frequency based color Doppler US. Radiology 1994;190: [20] Sedelar JP, Van Leenders GL, Hulsbergen-van De Kaa CA, Van Der Poel HG, Van Der Laak JA, Debruyne FM, et al. Microvessels density: correlation between contrast ultrasonography and histology of prostate cancer. Eur Urol 2001;40(3): [21] Brawer MK, Beatie J, Wener MH, Vessella RL, Preston SD, Lange PH. Screening for prostatic carcinoma with prostate specific antigen: results of the second year. J Urol 1993;150: [22] Epstein JL, Carmichael MJ, Pizov G, Walsh PC. Influence of capsular penetration on progression following radical prostatectomy: a study of 196 cases with long term follow up. J Urol 1993;150: [23] Rosen MA, Goldstone L, Lapin S, Wheeler T, Scardino PT. Frequency and location of extracapsular extension and positive surgical margins in radical prostatectomy specimen. J Urol 1992; 148: [24] Furosato K, Walkui S, Sasaki H. Tumor angiogenesis in latent prostatic carcinoma. Br J of Cancer 1994;70: [25] Brawer MK, Deering RE, Brown M. Predictors of pathologic stage in prostatic carcinoma: the role of neovascularity. Cancer 1994;73: [26] Weidner N, Caroll PR, Flax J, Folkmann J. Tumor angiogenesis correlates with metastasis in invasive prostate carcinoma. Am J Sug Pathol 143: [27] Cornud F, Belin X, Piron D, Chretien Y, Flam T, Casanova JM, et al. Color Doppler guided prostate biopsies in 591 patients with an

10 30 J.L. Sauvain et al. / European Urology 44 (2003) elevated serum PSA level: impact on Gleason score for non palpable lesions. Urology 1997;49(5): [28] Frauscher F, Klauser A, Volgger H, Halpern EJ, Pallwein L, Steiner H, et al. Comparison of contrast enhanced color Doppler targeted biopsy with conventional systematic biopsy: impact on prostate cancer detection. J Urol 2002;167(4): [29] Boccon-Gibod L. Rising PSA with a negative biopsy. Eur Urol 2001;40(2):3 8. [30] Djavan B, Remzi M, Schulmann CC, Marberger M, Zlotta AR. Repeat prostate biopsy: who, how and when? A review. Eur Urol 2002;42(2): [31] Niemann TH, Bahnson RR. Extended sector biopsy for detection of carcinoma of the prostate. Urol Oncol 2001;6(3):91 3. [32] Gohji K, Morisue K, Kizaki T, Fujii A. Correlation of transrectal ultrasound imaging and the results of systematic biopsy with pathological examination of radical prostatectomy specimens. Br J Urol 1995;75: [33] Ismail M, Petersen RO, Alexander AA, Newschaffer C, Gomella LG. Color Doppler Imaging in predicting the biologic behavior of prostate cancer: correlation with disease-free survival. Urology 1997;50(6): [34] Shinohara K, Wheeler T, Scardino PT. The appearance of prostate cancer on transrectal ultrasonography: correlation of imaging and pathological examinations. J Urol 1989;142: [35] Ayala AG, Ro JY, Babaian R, Troncoso P, Grignon DJ. The prostatic capsule: does it exist? Its importance in the staging and treatment of prostatic carcinoma. Am J Surg Pathol 1989;13:21 7. [36] Young MP, Jones DR, Grifiths GJ, Peeling WB, Roberts EE, Parkinson MC. Prostatic capsule: a comparative study of histological and ultrasonic appearances. Eur Urol 1993;24: Editorial Comment J.J.M.C.H. de la Rosette, AMC University Hospital, Amsterdam, The Netherlands New ultrasound imaging modalities are abandoning the strict imaging of prostatic tissue, and are focussing on imaging the vascularity of the prostate, mostly in combination with prostate tissue imaging. Two-dimensional and 3-dimensional ultrasound techniques are combined with Doppler modalities and are constantly modified, to try and detect the small blood vessels. In addition the use of contrast agents, administered to act like an intravascular ultrasound reflector, are studied to evaluate their value in the enhancement of the acoustic properties of the blood flow [1]. Probably the most exciting feature of these newer ultrasound studies of the prostate could be the changing of the prostate biopsy protocol by adding lesion directed biopsies towards abnormalities seen on ultrasound [2]. At present more and younger men consult the urologist without complaints but with a slightly moderately elevated PSA value. To confirm or rule out (early) prostate carcinoma extensive prostate biopsy protocols have been proposed, including up to 18 prostate biopsies per biopsy session [3]. However, the extension of the biopsy protocol increases the patient discomfort of the investigation up to levels that it might be almost impossible to perform the investigation on an outpatient basis. Once a prostate carcinoma is identified, one would like to properly study the stage of the disease. Almost all new ultrasound scanners have Power Doppler, or similar sensitive Doppler techniques and it is expected that in the future most scanners will have 3D modalities implemented. The introduction of these newer applications has, however, also its considerations. The most important restriction is the time needed for the investigation, especially when used in combination with contrast agents, as well as the evaluation of the investigation. Time is a scarce asset in the modern medical practice and should be used with utmost circumspection. Therefore the technique of Contrast Enhanced 3D-Power Doppler ultrasound guided prostate biopsies should maybe only be reserved for patients at risk of prostate cancer, but with no abnormalities on DRE or on greyscale/doppler ultrasound. Using Contrast Enhanced 3D-Power Doppler ultrasound guided biopsies, however, could be a major step forward in reducing the total number of prostate biopsy sessions. The investigations described above are static investigations: the main purpose was to image the vascular architecture, and possibly distinguish areas with asymmetry and/or focal increase of blood vessels. Another possibility of contrast ultrasound that could be exploited is the assessment of hemodynamics within the prostate. Tumour growth induces alterations in the vascularity resulting in changes in the hemodynamics. Even small tumours do alter the hemodynamics, while the vascular architecture has not yet changed. These changes in hemodynamics could be detected using Dynamic Contrast Enhanced ultrasound investigations. Instead of creating a 3D image of the prostate vasculature architecture, the blood circulation in one single plane of the prostate is imaged using Doppler ultrasound [4]. The addition of Dynamic Contrast Enhanced Power Doppler ultrasound investigations to static investigations could improve the detection of prostate cancer even further, possibly detecting more small tumours. In conclusion further improvement of ultrasound scanners and probes, with the addition of new techniques for optimal exploitation of the physical properties of contrast agents will be the exciting future of imaging the prostate and making the proper diagnosis and staging.

11 J.L. Sauvain et al. / European Urology 44 (2003) References [1] Sedelaar JPM, Vijverberg PLM, de Reijke TM, de la Rosette JJMCH, Kil PJM, Braeckinan JG, et al. Transrectal ultrasound in diagnosis of prostate cancer: state of the art and future perspectives. Eur Urol 2001;40:275. [2] Frauscher F, Klauser A, Volgger H, Halpern EJ, Pallwein L, Steiner H, et al. Comparison of contrast enhanced color Doppler targeted biopsy with conventional systematic biopsy: impact on prostate cancer detection. J Urol 2002;167(4): [3] Djavan B, Remzi M, Schulman CC, Marberger M, Zlotta AR. Repeat prostate biopsy: who, how and when? A review Eur Urol 2002; 42: [4] Goossen TEB, de la Rosette JJMCH, Hulsbergen-van de Kaa CA, van Leenders GJLH, Wijkstra H. The value of Dynamic Contrast Enhanced Power Doppler Ultrasound imaging of the localisation of prostate. Cancer Eur Urol 2002;43: