The Appearance of Prostate Cancer on Transrectal Ultrasonography: Correlation of Imaging and Pathological Examinations

0022-534 7/89/1421-0076$02.00/0 Vol. 142, July

THE JOURNAL OF UROLOGY

Printed in U.S.A.

Copyright© 1989 by Williams & Wilkins

THE APPEARANCE OF PROSTATE CANCER ON TRANSRECTAL ULTRASONOGRAPHY: CORRELATION OF IMAGING AND PATHOLOGICAL EXAMINATIONS KATSUTO SHINOHARA, THOMAS M. WHEELER AND PETER T. SCARDINO From the Scott Department of Urology and Department of Pathology, Baylor College of Medicine and The Methodist Hospital, Houston, Texas

ABSTRACT

Preoperative transrectal ultrasonograms in 70 patients who underwent radical prostatectomy were compared retrospectively to the histological features of whole mount sections of the surgical specimens. In 42 cases (60 per cent) the tumor could be appreciated as a hypoechoic area on the sonogram. In only 1 case did the tumor appear as a hyperechoic area. In the other 27 cases (39 per cent) the tumor was isoechoic and could not be distinguished clearly from the surrounding tissue. The smallest tumor visualized sonographically measured 4.4 mm. in diameter in the surgical specimen. There was a statistically significant correlation among tumor echogenicity, the actual tumor size and the Gleason grade. Although there was a linear correlation between the size (maximum diameter) of the tumor measured by ultrasound and that determined from the histological sections, ultrasonography generally underestimated the size of the tumor, with the maximum diameter measured sonographically being approximately 4.8 mm. smaller than the diameter measured in the whole mount sections. Of 25 nonpalpable (stage A) tumors 9 (36 per cent) were visualized on ultrasound, while of 45 palpable (stage B) tumors 11 (21 per cent) were not. Although most clinically recognized tumors can be appreciated sonographically, the tumors that are visualized tend to be larger, less well differentiated and palpable. These features favor the use of sonography to stage and monitor established prostate cancers, and suggest that tumors detected by sonography in an early detection program are likely to be clinically important cancers. (J. Ural., 142: 76-82, 1989) Transrectal ultrasonography has received increasing attention recently because of its potential for the early detection of prostate cancer. Within the last 2 years several investigators have reported detection of small prostate cancers with ultrasonography.1· 2 Previously, the sonographic diagnosis of prostate cancer was based upon the indirect effect of the disease, such as deformity or asymmetry of the prostate gland, or disruption of the capsular echo. 3·4 However, the development of gray scale ultrasound equipment made it possible to visualize the internal echo patterns of the prostate and to study the echogenic characteristics of prostate cancer directly. 5 Most early studies reported that cancer appeared as a hyperechoic focus 6 · 10 and a few suggested that hypoechoic areas were associated with cancer.11·12 Confusion about the appearance of cancer arose in the absence of information correlating ultrasonograms with histological features of the prostate. In 1985 Lee and associates reported that the presence of a hypoechoic area in the peripheral zone was a reliable criterion for the diagnosis of prostate cancer. 13 In an early study correlating ultrasonograms with histology in radical prostatectomy specimens, Diihnert and associates reported that 54 per cent of the tumors were echopenic, 22 per cent slightly hypoechoic and 24 per cent isoechoic.14 The extent to which trnnsrectal ultrasonography adds to the ability of digital rectal examination to detect prostate cancer remains open to question. Although some malignant lesions can be visualized with ultrasonography not all tumors are detected. The proportion that are visualized and the number of these that are palpable remain unclear. Moreover, the reason some tumors are visualized and others are not has not been established. We conducted a retrospective study correlating preoperative

transrectal ultrasonograms with the histological features of 70 consecutive step-sectioned whole mount radical prostatectomy specimens to determine the characteristic sonographic features of prostate cancer, and to analyze the relationship among the ultrasonographic appearance of the tumor and the actual size of the tumor, its histological grade and the results of a digital rectal examination of the prostate. MATERIALS AND METHODS

Between February 1984 and August 1987, 70 patients with clinically localized prostate cancer had transrectal ultrasonography of the prostate before radical retropubic prostatectomy. Patient age ranged from 48 to 78 years, with a mean age of 62.9 years. Of the patients 8 had clinical stage Al, 17 stage A2, 10 stage BlN, 23 stage Bl, 10 stage B2 and 2 stage C tumors. Staging was based on digital rectal examination regardless of the results of ultrasonography and the results of the rectal examination were recorded on a designated form in an effort to establish an objective data base. The Johns Hopkins criteria were used to classify stages Al and A2 tumors. 15 If not more than 5 per cent of the surface area was involved by tumor and the Gleason grade was less than 8 the tumor was considered stage Al. Palpable tumors 1.5 cm. in diameter or smaller were considered stage BlN. Tumors larger than 1.5 cm. in diameter but confined within 1 lobe were classified as stage Bl. Tumors involving both lobes were classified as stage B2. Stage C tumors extended outside of the prostate into the lateral sulci or seminal vesicles. Transrectal ultrasonography. Transrectal ultrasonography was performed with an Aloka 502 scanner (26 patients) or an Aloka 520 scanner (44). Both machines were equipped with a chair-mounted 5 MHz. radial transrectal probe and produced comparable images of the prostate. Continuous scanning was performed from the apex of the prostate to the tip of the seminal vesicles. Transverse images of the prostate and seminal

Accepted for publication January 30, 1989. Supported in part by Grant CA28500-06 from the National Cancer Institute, United States Public Health Service. 76

77

APPEARANCE OF PROSTATE CANCER ON TRANSRECTAL ULTRASONOGRAPHY

FIG. 1. A, representative images achieved with transrectal ultrasonography before radical prostatectomy. Transverse images are taken from apex of prostate (left upper section) to seminal vesicles (right lower section) at 5 mm. intervals. B, hypoechoic area in right peripheral zone in mid prostate (left lower section in part A) corresponds to tumor drawn in black on tumor map traced from whole mount sections of radical prostatectomy specimen. Prostate was sectioned transversely at 5 mm. intervals from apex (P2) to base (P7) to correspond to ultrasound. Seminal vesicles are seen in PS to PIO.

vesicles were recorded at 5 mm. intervals on Polaroid film (fig. 1, A) and additional films were made of areas of interest. Preoperative sonography was performed at least 4 to 6 weeks after a needle biopsy or transurethral resection of the prostate so that any artifacts introduced by these procedures would be minimized. In patients with stage A cancer only sonograms performed after transurethral resection of the prostate were used for this analysis. Pathological examinations. Surgical specimens were fixed and whole mount step sections were cut transversely at 5 mm. intervals from the apex of the prostate to the tips of the seminal vesicles in a pattern corresponding to the standard preoperative ultrasound images. Each section was carefully examined under a microscope by 1 of us (T. M. W.), and all areas of tumor and severe ductal dysplasia (carcinoma in situ) were traced separately on paper to create a map (fig. 1, B )o The grade of tumor also was recorded using the Gleason grading system. Distributions of clinical stage and Gleason score in the 70 patients are shown in table 1. Retrospective correlation study. By comparing the ultrasonograms with the tumor maps we determined the echogenic pattern of prostate cancer in each section. The echogenicity was classified as hypoechoic-tumor less echogenic than surrounding tissue, isoechoic-tumor could not be distinguished from the surrounding tissue and hyperechoic-tumor more echogenic than surrounding tissue. The maximum diameter of the largest single focus of tumor was measured on the ultrasound image, if the tumor was visualized, and on the histological tumor map. The actual tumor size was obtained by multiplying the histological size of the tumor by a factor of 1.1 to compensate for shrinkage of the specimen during processing. Statistical analysis. Correlations between the echogenic pattern (hypoechoic versus isoechoic) and the Gleason score (primary + secondary grade) and pathological size (maximum diameter) of the tumor were tested for significance with the t

TABLE

1. Distribution of clinical stage and Gleason score in 70

patients Gleason Score Clinical Stage

Totals 4

Al A2 BIN Bl B2 C Totals

1 5 2

8

5 5 7 2 5 3

22

6

7

9

10

2

8 3

3 10 3 1 19

3 8 3

1 18

2

17 10 23 10 2 70

test with 1 degree of freedom. Since only l tumor was hyperechoic it was excluded from the statistical analysis. RESULTS

Echogenicity. In 42 prostates (60 per cent) the principal or index tumor was visualized as a hypoechoic area on transrectal ultrasonography. In only 1 prostate (1 per cent) did the tumor show regions of high echogenicity interspersed with smaller hypoechoic areas. This tumor was a ductal-type adenocarcinoma with comedo-type tumor nests containing calcification and necrotic debris. In the other 27 prostates (39 per cent) no tumor was visualized on ultrasonography. Both 5 MHz. Aloka scanners used gave comparable results. Of the 42 prostates in which the principal tumor could be seen as a hypoechoic area 36 (86 per cent) had at least 1 other area of tumor within the prostate. However, these smaller accessory tumors were visualized in only 4 instances (11 per cent). Nearly 90 per cent of the prostates contained areas of severe ductal dysplasia (carcinoma in situ) of varying size but none of these areas was visualized on sonography. Tumor grade and echogenicity. Table 2 shows the Gleason score of the principal tumor in 70 patients. Most of the hypoechoic tumors were moderately or poorly differentiated (mean

78

SHINOHARA, WHEELER AND SCARDINO

Gleason score 6.2), while the isoechoic tumors tended to be well or moderately differentiated (mean Gleason score 5.2). Overall, only 13 per cent of the Gleason scores 2 to 4 (well differentiated) tumors were visualized, compared to 80 per cent of the Gleason scores 5 to 7 (moderately differentiated) tumors and 100 per cent of the Gleason scores 8 to 10 (poorly differentiated ) tumors. There was a statistically significant correlation between Gleason score and tumor echogenicity (hypoechoic versus isoechoic, t test, p <0.0002). Tumor size and echogenicity. The size (maximum diameter) of the principal tumor in each case was correlated with the echogenic pattern (table 3). There was a statistically significant correlation between tumor size and echogenicity (hypoechoic versus isoechoic, t test, p <0.0004). Only 18 per cent (3 of 17) of the tumors 10 mm. in diameter or less were visualized, compared to 77 per cent (40 of 52) of those more than 10 mm. in diameter. The smallest tumor visualized on transrectal ultra sonography was 4.4 mm. in maximum diameter. The largest tumor that was not visualized was 49.5 mm. in diameter. Tumor size measured histologically (actual size) and sonographically. Figure 2 shows the relationship between the maximum diameter of the tumor measured by ultrasound and that determined from the histological section. In most cases the tumor appeared to be smaller on ultrasonography than the actual size but there was a linear and highly significant correlation between these 2 measurements (r = 0.86, p <0.001), described by the equation y (actual size) = 0.97 X (size on ultrasound) + 4.8. Interpreting this equation we find that the actual size of the tumor is approximately 4.8 mm. larger than its size on ultrasound. Clinical stage (palpability) and echogenicity. Although residual invasive cancer was present in the radical prostatectomy specimens in 7 of the 8 patients with stage Al tumors, in only 1 could we recognize a hypoechoic area (table 4). However, in 8 of 17 patients (47 per cent) with stage A2 cancer an area of tumor could be visualized on transrectal ultrasonography (fig. 3). Stage BlN tumors usually were visible ultrasonographically with 8 of 10 (80 per cent) visualized as a hypoechoic area (fig. 4). In large palpable tumors (stages Bl, B2 and C) 70, 80 and 100 per cent, respectively, were visualized (fig. 5). Thus, 9 of 25 (36 per cent) nonpalpable tumors (stages Al and A2) were detectable with ultrasound, whereas 11 of 45 palpable tumors (21 per cent) were not detectable (table 4). In 2 stage BlN tumors the principal tumor actually was in the lobe contralateral to the palpable nodule. Therefore, the tumor actually was palpable in 43 of 70 cases (61 per cent), which, interestingly, is the same number that were detectable sonographically (42 hypoechoic and 1 hyperechoic). TABLE

In patients with stage A tumors sonograms were obtained after transurethral resection of the prostate. The residual cancer visualized sonographically was located anteriorly in the transition zone in some patients but posteriorly in the peripheral zone in others (fig. 3). Palpable, stage B and C tumors almost always were visualized in the peripheral zone (fig. 4) with some extending anteriorly into the transition zone (fig. 5). DISCUSSION

The echogenic characteristics of prostate cancer have been described in several recent studies comparing sonographic findings with histological examinations of the prostate. 14• 16 How-

50

40

E

EQ)

N

30

"cii

0

1,1

Y=4.8031 +0.968X, R=0.86

1,1

E

.a

20

cii

::::,

P < 0.001

ffl

t5

<(

10 1,1

0 0

10

20

30

40

50

Tumor size on ultrasound (mm)

FIG. 2. Correlation between size (maximum diameter) of tumor on ultrasonography and on whole mount sections. If tumor was visualized (hypoechoic or hyperechoic) there was linear and highly significant correlation between sizes (r = 0.86, p <0.001) described by equation (actual size)= 0.97 X (size on ultrasound) + 4.8.

TABLE

4. Echogenic pattern as a function of the clinical stage (based on digital rectal examination) Clinical Stage*

Echogenicity

Al

A2

BlN

Bl

Hypoechoic 1 (13) 8 (47) 8 (80) 16 (70) 7 7 (87) 9 (53) 2 (20) 7 (30) 2 Isoechoic 1 Hyperechoic 8 (100) 17 (100) 10 (100) 23 (100) 10 Totals

Total B2

C

No. (%)

(70) 2 (100) 42 (60) (20) 27 (39) (10) 1 (1) (100) 2 (100) 70 (100)

* Number of prostates (per cent).

2. Echogenicity as a function of the grade of tumor (hypoechoic versus isoechoic, t test, p <0.0002) Gleason Score

TABLE

Total No.(%)

Echogenicity

4 No.(%)

5 No.(%)

6 No.(%)

7 No.(%)

9 No.(%)

10 No.(%)

Hypoechoic Isoechoic Hyperechoic Totals

1 (13) 7 (87)

10 (45) 12 (55)

15 (79) 4 (21)

14 (78) 4 (22)

1 (100)

1 (50)

42 (60) 27 (39)

..l._l!2

1 (100)

1 (50) 2 (100)

8 (100)

22 (100)

19 (100)

18 (100)

70 (100)

3. Echogenic pattern as a function of the actual size of tumor measured in the whole mount sections of the prostate (hypoechoic versus isoechoic, t test, p <0.0004) Size (greatest diameter in mm.)* Echogenicity Hypoechoic Isoechoic Hyperechoic Totals

Ot

Total No.(%)

No.(%)

0.1-5.0 No.(%)

5.1-10.0 No.(%)

10.1-15.0 No.(%)

>15.1 No.(%)

1 (100)

1 (25) 3 (75)

2 (15) 11 (85)

8 (57) 6 (43)

31 (82) 6 (16)

42 (60) 27 (39)

-1....._ill

_l___ill

1 (100)

4 (100)

13 (100)

14 (100)

38 (100)

70 (100)

* The measured size was multiplied by 1.1 to correct for shrinkage during processing. t One patient with a stage Al tumor had no residual cancer in the radical prostatectomy specimen.

APPEARANCE OF PROSTATE CANCER ON TRANSRECTAL ULTRASONOGRAPHY

79

Ill Invasive cancer ~

Missing part 0 Carcinoma in situ

Invasive cancer FIG. 3. Stage A2 prostate cancer after transurethral prostatectomy. A, broad hypoechoic area anterior to urethra is visualized clearly (arrows). B, large solid sheet of poorly differentiated cancer is visualized anteriorly but smaller posterior foci are not. This is typical pattern of stage A prostate cancer. 19 C and D, ultrasound after transurethral prostatectomy shows no abnormality anteriorly but hypoechoic focus of residual cancer on right side in peripheral zone posteriorly (arrow) where largest focus of residual cancer was found in radical prostatectomy specimen.

11 Invasive cancer CJ] Carcinoma in situ FIG. 4. Stage Bl nodule. A, ultrasound of typical small stage Bl tumor. Hypoechoic area in right peripheral zone posteriorly is identified readily (arrow). B, in whole mount histological section tumor is larger than it appears on ultrasound but is in same location. At periphery (arrow) tumor appears indistinct as it infiltrates normal gland structure; this area is not well seen on sonogram. H & E, reduced from X3. C, as tumor map illustrates, there were multiple accessory foci of cancer that were not visualized on sonogram and are difficult to appreciate on whole mount section at low power. If tumor cannot be appreciated at low power on whole mount section it generally will not be visualized on ultrasound.

ever, these authors did not investigate the reason for the varied echo patterns of prostate cancer. The finely stippled internal echo pattern of the prostate results from reflections of the sound wave at the interface between prostatic stroma and the fluid-filled acinar lumen. Thus, the normal prostate gland, especially in the peripheral and central zones, shows a relatively homogeneous echogenic texture corresponding to the uniform distribution of glands in these areas. 17 Prostate cancer is composed of a small gland-forming or even nongland-forming mass of cells and is different in structure from the normal prostate gland. Cancer often destroys the normal glandular structure and replaces it with a packed mass of cells containing smaller glands (fig. 4, B). The malignant tissue contains few sonographically detectable interfaces and,

therefore, appears hypoechoic (less echogenic) compared to the adjacent normal tissue (fig. 4, A). However, tumor structure varies. Some tumors are composed entirely of a small gland-forming mass of cells, whereas others form relatively large glandular structures (fig. 5, D) or even infiltrate among normal glands (fig. 6, E). The latter tumors may appear similar to normal prostatic tissue sonographically because they contain many echo-reflective interfaces (fig. 6, A). Higher grade tumors tend to be large and medullary, while lower grade tumors tend to form glands or to mix with normal glandular structures. Consequently, higher grade tumors are more likely to be visualized as hypoechoic areas on ultrasonography (table 2). In addition to grade, the size of a tumor nodule proved to be

80

SHINOHARA, WHEELER AND SCARDINO

• Invasive cancer D Carcinoma in situ C

FIG. 5. Stage Bl disease. A and B, tumor can be seen as low echogenic area replacing most of left lobe (arrows) but borders are indistinct. C, on whole mount section tumor (arrows) is difficult to see at low power because of its similarity to normal gland structure. H & E, reduced from X3. D, at high power tumor consists of large gland-forming mass of cells with Gleason score of 6. H & E, reduced from X315.

a critical determinant of visualization in our series. No previous reports have correlated the echogenicity of prostate cancer with the size of the corresponding lesion pathologically. We were unable to visualize, even in retrospect, any tumor smaller than 4.4 mm. in maximum diameter and failed to visualize 89 per cent of the small incidental foci of cancer present within the pathological specimen (figs. 4 and 5). Even large tumors up to 49.5 mm. could not be identified sonographically in 16 per cent of the patients (table 3), perhaps because the prostate was largely replaced by tumor that eliminated the areas of contrast with normal tissue. When a tumor was visualized it usually appeared to be smaller than its actual size (fig. 2). In the periphery of a cancer malignant cells often invade between but do not completely replace normal glandular structures (fig. 4). Thus, the echo pattern of the margins of a tumor may remain isoechoic. In our experience transrectal ultrasonography usually visualizes the core of the tumor rather than the entire lesion. However, visualization of a tumor within the prostate is dependent not only on the size and architectural structure (grade and pattern of invasion) of the tumor itself but also on its environment. A tumor within the transition zone may be difficult to appreciate because of the more heterogeneous echo pattern of the glandular and fibrous adenoma or the normally hypoechoic periurethral fibromuscular tissue (fig. 4). Even within the peripheral zone a small tumor may be difficult to distinguish from normal structures, such as the ejaculatory duct complex or a thickened fibrous prostatic capsule, which appear hypoechoic. Indirect signs, such as thickening or asymmetry of the peripheral zone, asymmetry of the prostate or distortion of normal anatomical structures, such as the ejaculatory ducts, may provide clues to the astute ultrasonographer that a tumor is present. As we gained experience correlating ultrasonograms with whole mount sections of the prostate, we found that generally if the tumor cannot be appreciated by the naked eye when the

histological section is held up to the light it will not be visualized sonographically (figs. 4 and 6). While 79 per cent of the palpable tumors in our series were visualized, only 36 per cent of the nonpalpable tumors could be identified sonographically. All nonpalpable tumors were clinical stage A cancers discovered incidentally at transurethral resection of the prostate. We reported previously that stage A cancer tends to be located anterior to the mid point of the urethra, often is multifocal and diffuse, and frequently is of low grade (especially stage Al). 18• 19 Others have reported similar findings independently. 20 •21 Our ultrasonograms were performed after transurethral resection of the prostate. Although residual cancer was present in every radical prostatectomy specimen except 1 (which had foci of carcinoma in situ), the volume of residual cancer usually was small. Based on our analysis of the characteristics that make a tumor visible sonographically, one would expect that stage A tumors would be difficult to appreciate on ultrasound. Residual tumor in the anterior transition zone was visualized only when a large focus of moderate to high grade tumor was present (fig. 3, A). However, visualization of foci of residual tumor in the peripheral zone posteriorly was more common (fig. 3, C). Identification of stage A cancer before transurethral resection of the prostate may even be more difficult because of the distortion and heterogeneous echo pattern caused by benign prostatic hyperplasia. Lee and associates reported a series in which hypoechoic lesions suspicious for cancer were biopsied using ultrasound guidance. Cancer was found in approximately half of the biopsies but an additional 10 per cent showed dysplasia. 22 Severe dysplasia (carcinoma in situ) is a malignant transformation of the epithelial cells in the prostatic acini but it does not destroy normal glandular structure. Carcinoma in situ maintains a structure that is architecturally similar to the normal prostate gland. In fact, in our experience none of the areas of carcinoma in situ, present in more than 85 per cent of the radical prostatectomy specimens, could be distinguished from normal pros-

APPEARANCE OF PROSTATE CANCER ON TRANSRECTAL ULTRASONOGRAPHY

81

II! Invasive cancer D Carcinoma in situ

FIG. 6. A and B, stage Bl tumor is visualized in right lobe (arrow) but nonpalpable tumor in left lobe (arrowhead) is not apparent, although both are about same size on tumor map (B). C, on whole mount section tumor on right side (arrow) is easy to appreciate but mirror image tumor on left side (arrowhead) is not. H & E, reduced from X3. D, at high power tumor in right lobe consists of small-gland forming mass of cells. H & E, reduced from X315. E, tumor in left lobe has same Gleason grade but infiltrates among normal glands. H & E, reduced from X315.

tate tissue ultrasonographically (figs. 3 and 5). Foci of carcinoma in situ usually are multiple and often are adjacent to a cancer. 18 Carcinoma in situ found in a biopsy specimen may herald the presence of a nearby cancer. Such patients should be followed closely if a repeat biopsy is not performed. We found that 60 per cent of the clinically diagnosed prostate cancers were hypoechoic and only 1 per cent was hyperechoic. This unusual tumor actually had a mixed echo pattern resulting from hypoechoic areas of cancer intermingled with hyperechoic foci of calcification within necrotic tumor nests. Earlier reports that prostate cancer typically is hyperechoic or mixed in its echo pattern appear to be unfounded6 · 10 and arose in the absence of ultrasonographic-pathological correlations in which the entire prostate was available for examination. Because ours was a retrospective study based on radical prostatectomy specimens removed from patients with a clinically (not sonographically) diagnosed cancer, the exact location of each tumor was known and the optimal sensitivity of ultrasonography to visualize clinically recognized prostate cancer could be evaluated. Nevertheless, our series does not include patients initially diagnosed by transrectal ultrasonography and, therefore, it may underestimate the ability of this imaging

technique to detect some nonpalpable cancers in the peripheral zone posteriorly. Our study also may be limited by the ultrasound equipment used. Had we used probes with a higher frequency (7 to 7.5 MHz.) and biplanar capability, as we do currently, we might have been able to identify more tumors. The value of transrectal ultrasonography in screening for prostate cancer remains controversial. In 1985 Watanabe and associates reported screening of 5,770 men with transrectal ultrasonography, with a sensitivity of 97.1 per cent and specificity of 80.4 per cent, and they concluded that this procedure is useful for screening. 23 On the other hand, Chodak and associates reported a sensitivity of 86 per cent but a specificity of only 41 per cent in screening 216 men. 24 However, both of these studies were conducted without the knowledge of recently developed ultrasonographic criteria for the diagnosis of cancer, including the importance of a small hypoechoic area. Recently, Lee and associates reported detection by ultrasound-guided biopsy of 23 tumors that were either nonpalpable or equivocal on digital rectal examination. 22 Cooner and associates conducted a prospective screening study in 238 men and found prostate cancer in 13 per cent of this population, including 12 patients with nonpalpable tumors. 2

82

SHINOHARA, WHEELER AND SCARDINO

Our correlations of ultrasonography with whole organ mapping of prostate cancers showed that tumors visualized sonographically tended to be larger (more than 4.4 mm.), of higher grade and palpable. Consequently, tumors detected with current technology in an ultrasound screening program are likely to be clinically important rather than latent or incidental carcinomas. In our series 36 per cent of the nonpalpable tumors were visualized with ultrasonography but 21 per cent of the palpable tumors were not visualized. The population of ultrasonically detectable tumors and palpable tumors overlap (75 per cent of our patients) but some tumors were detectable with 1 method and not the other. A combination of these 2 techniques, perhaps in association with serum markers, such as prostate specific antigen, should enhance our ability to detect prostate cancer earlier.

11. 12. 13.

14.

15.

Ms. Carolyn Schum provided editorial assistance. 16. REFERENCES

1. Lee, F., Gray, J. M., McLeary, R. D., Lee, F., Jr., McHugh, T. A., Solomon, M. H., Kumasaka, G. H., Straub, W. H., Borlaza, G. S. and Murphy, G. P.: Prostatic evaluation by transrectal sonography: criteria for diagnosis of early carcinoma. Radiology, 158: 91, 1986. 2. Cooner, W. H., Eggers, C. W. and Lichtenstein, P.: Prostate cancer: new hope for early diagnosis. Alabama Med., 56: 13, 1987. 3. Watanabe, H.: Prostate and seminal vesicles. In: Diagnostic Ultrasound in Urology and Nephrology. Edited by H. Watanabe, J. N. Holmes, H. H. Holm and B. B. Goldberg. Tokyo: lgaku-Shoin, pp. 130-134, 1981. 4. Gammelgaard, J. and Holm, H. H.: Transurethral and transrectal ultrasonic scanning in urology. J. Urol., 124: 863, 1980. 5. Harada, K., Tanahashi, Y., lgari, D., Numata, I. and Orikasa, S.: Clinical evaluation of inside echo patterns in gray scale prostatic echography. J. Urol., 124: 216, 1980. 6. Boyce, W. H., McKinney, W. M., Resnick, M. I. and Willard, J. W.: Ultrasonography as an aid in the diagnosis and management of surgical diseases of the pelvis: special emphasis on the genitourinary system. Ann. Surg., 184: 477, 1976. 7. Peeling, W. B., Griffiths, G. J., Evans, K. T. and Roberts, E. E.: Diagnosis and staging of prostatic cancer by transrectal ultrasonography: a preliminary study. Brit. J. Urol., 51: 565, 1979. 8. Rifkin, M. D., Kurtz, A. B., Choi, H. Y. and Goldberg, B. B.: Endoscopic ultrasonic evaluation of the prostate using a transrectal probe: prospective evaluation and acoustic characterization. Radiology, 149: 265, 1983. 9. Resnick, M. I., Willard, J. W. and Boyce, W. H.: Ultrasonic evaluation of the prostatic nodule. J. Urol., 120: 86, 1978. 10. Abu-Yousef, M. M. and Narayana, A .. S.: Prostatic carcinoma:

17. 18. 19.

20.

21.

22.

23.

24.

detection and staging using suprapubic US. Radiology, 156: 175, 1985. Egender, G., Rapf, C., Feichtinger, I., Mikuz, G., Bartsch, G. and Frommhold, H.: Vergleichende histopathologische und sonomorphologische Prostatauntersuchungen. ROFO, 140: 60, 1984. Frentzel-Beyme, B., Aurich, B. and Dakopoulous, A.: Transrectal prostate ultrasonography in early cancer recognition. Computertomographie, 3: 153, 1983. Lee, F., Gray, J. M., McLeary, R. D., Meadows, T. R., Kumasaka, G. H., Borlaza, G. S., Straub, W. L., Lee, F., Jr., Solomon, M. H., McHugh, T. A. and Wolf, R. M.: Transrectal ultrasound in the diagnosis of prostate cancer: location, echogenicity, histopathology, and staging. Prostate, 7: 117, 1985. Diihnert, W. F., Hamper, U. M., Eggleston, J.C., Walsh, P. C. and Sanders, R. C.: Prostatic evaluation by transrectal sonography with histopathologic correlation: the echopenic appearance of early carcinoma. Radiology, 158: 97, 1986. Epstein, J. I., Paull, G., Eggleston, J.C. and Walsh, P. C.: Prognosis of untreated stage Al prostatic carcinoma: a study of 94 cases with extended followup. J. Urol., 136: 837, 1986. Griffiths, G. J., Clements, R., Jones, D. R., Roberts, E. E., Peeling, W. B. and Evans, K. T.: The ultrasound appearances ofprostatic cancer with histological correlation. Clin. Rad., 38: 219, 1987. McNeal, J. E.: Normal histology of the prostate, Amer. J. Surg. Path., 12: 619, 1988. Wheeler, T. M., Scardino, P. T. and Cantini, M.: Whole organ mapping of radical prostatectomy specimens. J. Urol., part 2, 137: 224A, abstract 483, 1987. Neerhut, G. J., Wheeler, T. M., Dunn, J. K. and Scardino, P. T.: Residual tumor after TUR: pathologic features of stage A prostate cancer in the transurethral and radical prostatectomy specimens. J. Urol., part 2, 139: 315A, abstract 612, 1988. Epstein, J. I., Oesterling, J.E. and Walsh, P. C.: The volume and anatomical location of residual tumor in radical prostatectomy specimens removed for stage Al prostate cancer. J. Urol., 139: 975, 1988. McNeal, J. E., Price, H. M., Redwine, E. A., Freiha, F. S. and Stamey, T. A.: Stage A versus stage B adenocarcinoma of the prostate: morphological comparison and biological significance. J. Urol., 139: 61, 1988. Lee, F., Littrup, P. J., McLeary, R. D. Kumasaka, G. H., Borlaza, G. S., McHugh, T. A., Soiderer, M. H. and Roi, L. D.: Needle aspiration and core biopsy of prostate cancer: comparative evaluation with biplanar transrectal US guidance. Radiology, 163: 515, 1987. Watanabe, H., Ohe, H., Saitoh, M., Itakura, Y., Inaba, T., Miyashita, H., Ohnishi, K. and Abe, M.: A survey on 10,000 examinations by transrectal ultrasonotomography in our clinic. Jap. J. Med. Ultrason., suppl. 1, 12: 921, 1985. Chodak, G. W., Wald, V., Parmer, E., Watanabe, H., Ohe, H. and Saitoh, M.: Comparison of digital examination and transrectal ultrasonography for the diagnosis of prostatic cancer. J. Urol., 135: 951, 1986.