Random Systematic Versus Directed Ultrasound Guided Transrectal Core Biopsies of the Prostate

0022-5347 /89/1421-007l$02.00/0 THE JOURNAL OF UROLOGY

Vol. 142, July Printed in U.S.A.

Copyright© 1989 by Williams & Wilkins

RANDOM SYSTEMATIC VERSUS DIRECTED ULTRASOUND GUIDED TRANSRECTAL CORE BIOPSIES OF THE PROSTATE KATHRYN K. HODGE, JOHN E. McNEAL, MARTHA K. TERRIS AND THOMAS A. STAMEY* From the Division of Urology, Stanford University Medical Center, Stanford, California

ABSTRACT

Random systematic ultrasound guided transrectal core biopsies of the prostate were compared to directed biopsies of specific hypoechoic defects in 136 men with abnormal prostates on digital rectal examination. Prostate cancer was diagnosed in 83 of 136 patients (62 per cent). In 80 of 83 individuals (94 per cent) the cancer was detected by random systematic biopsies alone. Of 57 men in whom random systematic and directed biopsies were obtained the results of biopsy agreed in 86 per cent, while in 9 per cent random systematic biopsies found cancers missed by directed biopsies and in 5 per cent directed biopsies diagnosed cancers missed by random systematic prostate biopsies. Ultrasound guided random systematic biopsy is simple and easily learned. When combined with additional directed biopsies of the rare hypoechoic areas not included in the pattern of systematic sampling, it provides a highly accurate means to diagnose prostate cancer, minimizing observer and sampling errors. This technique of prostate mapping with 6, 1.5 cm. cores provides valuable additional information on cancer volume, Gleason grade and the potential location of surgically positive margins, all without compromising the operation or the chance for a surgical cure. (J. Ural., 142: 71-75, 1989) In the previous study we established that 90 per cent of all clinical stage B nodules and abnormally firm prostates are accompanied by a hypoechoic defect on ultrasound imaging of the peripheral or central zones of the prostate. 1 Of the stage B, clinical stage C and abnormally firm, nonnodular prostates 75, 100 and 36 per cent, respectively, had prostate cancer by ultrasound guided transrectal core biopsies. We concluded that our high yield of positive biopsies in abnormally palpated prostates is due to the combined technology of a unique transrectal biopsy system guided by ultrasound, and we present evidence that with our technique we were able to make a new diagnosis of cancer in 53 per cent of the patients with previously negative digital biopsies.' Also in the previous study we began to supplement our biopsies, which initially were targeted to the nodule or ultrasound defect, with a few random additional cores to sample some isoechoic areas of the prostate in addition to the targeted area. Once the safety of these multiple biopsies was established' we realized that the prostate could be sampled systematically with 6 carefully positioned biopsies. For example, figure 1 is taken from a plaster cast of an actual 40 gm. prostate removed for prostate cancer. This specimen is representative of the mean weight (41 gm.) of 102 consecutive radical prostatectomies recently reported from our group. 2 A biopsy core of the prostate 1.5 cm. long covers a greater proportion of the distance from the rectal surface to the anterior fibromuscular stroma of the prostate than might be anticipated. This information, plus the observation that biopsies directed anteriorly from the rectal surface offer a better opportunity for systematic sampling than biopsies directed through the perineum, encouraged us to change our technique from that described in the previous study to one of carefully spaced systematic biopsies of the prostate. Because of these considerations we examined the possibility that random systematic biopsies from the apex to the base in the peripheral and central zones of both prostatic lobes may be

a better method to detect posteriorly located cancers than concentrating upon specific hypoechoic defects. METHODS

The 136 patients who had abnormal prostates by digital rectal examination detected when they were in the knee-chest position underwent transrectal biopsy with ultrasound guidance. The spring-driven Biopty gun* using 18 gauge "tru-cut" typet biopsy needles was used in a systematic manner by first taking 1.5 cm. cores from each of 6 standard anatomical sites. These 6 biopsy sites were located approximately 1 cm. apart and were taken from the apex, middle and base of the prostate bilaterally (fig. 2). With respect to the coronal plane, the biopsy sites were oriented in the center of each lobe, equidistant from the midline of the prostate and the lateral border of the lobe. Biopsies were taken with the side-fire sector scanner (Bruel and Kjaer 8537) along the computer generated trajectory that extends from the rectal surface at an angle of 45 degrees (fig. 1). If a suspicious hypoechoic region was located medially or far laterally to this mid lobe parasagittal line of random systematic sampling (fig. 2) additional directed biopsies were taken of the specific defects. Of the 136 patients 57 required these additional biopsies directed at specific hypoechoic defects that did not appear to coincide with any of the 6 standard biopsies. In the remaining 79 patients at least 1 and usually several of the systematic 6 biopsies lay near the center of the hypoechoic defect. The 85 clinical stage B nodules detected on digital rectal examination were classified as stage Bl, B2 or B3.~ A stage Bl nodule, compressible on at least 2 sides, extended for a distance equal to or less than half of 1 prostate lobe. A stage B2 nodule was larger than stage Bl but occupied no more than 1 full lobe. A stage B3 nodule either filled 1 lobe with extension across the midline or represented bilaterally palpable disease. Eight patients with clinical stage C disease had palpable extension above the prostate, presumably into 1 or both seminal vesicles. Transrectal ultrasound imaging was performed with the Bruel and Kjaer 1846 console and 7.0 MHz. transducers 1850, 8537 and 8538 as described previously. 1 Each prostate was

Accepted for publication January 30, 1989. Supported in part by the Richard M. Lucas Cancer Foundation. * Requests for reprints: Division of Urology (S287), Stanford University Medical Center, 300 Pasteur Dr., Stanford, California 943055118.

t C. R. Bard, Inc., Covington, Georgia. 71

72

HODGE AND ASSOCIATES

FIG. 1. Plaster cast of 40 gm. radical prostatectomy specimen. Rectal surface is located posteriorly, apex to right side. Arrowheads indicate 1.5 cm. crypt area for core biopsy. Core (1.5 cm.) would include full thickness of peripheral zone and part of transition zone of prostate. Angle of biopsy, determined by computer generated trajectory of Bruel and Kjaer 8537 sector scanner, is approximately 45 degrees.

random systematic method in 80 (96 per cent), while only in 3 (4 per cent) was cancer detected by biopsy of locations other than the 6 designated random systematic sites. Among the 43 abnormally firm prostates 10 of the 12 cancers (83 per cent) were detected by random systematic biopsy, while 2 (17 per cent) were identified by directed biopsies. Among the 85 clinical stage B nodules (table 1) 62 of the 63 cancers (98 per cent) were detected by random systematic biopsy and only 1 required additional directed biopsies. As expected, all 8 cases of clinical stage C disease were positive with the random systematic biopsy technique. Among the 83-patients with biopsy proved cancer it is noteworthy that, in addition to the biopsy positive hypoechoic segments, the 6 systematic biopsies (fig. 2) detected cancer in nonhypoechoic segments from 7 of 12 (58 per cent) patients with abnormally firm prostates, 32 of 63 (51 per cent) with clinical stage B nodules and 2 of 8 (25 per cent) with clinical stage C disease. Thus, random systematic biopsies offer the pathologist more positive samples with which to judge the overall Gleason grade of the cancer. Although only 3 of 83 biopsy proved cancers (4 per cent) were missed by the random systematic technique, it is interesting to compare the biopsy results in the 57 patients who underwent both techniques because of a hypoechoic defect lying outside the mid lobe parasagittal plane of systematic biopsy (table 2). In 5 patients (9 per cent) random systematic biopsies detected cancers that were missed by the directed biopsy alone (table 2). Directed biopsies revealed cancers in 3 patients (5 per cent) that were not detected by the random systematic technique. DISCUSSION

FIG. 2. Posterior (rectal) surface of 40 gm. prostate seen in figure 1. Apex is at bottom and vesical neck is at top. Three dots over each lobe mark site of 6 random systematic biopsies. Longitudinal distance between center of each dot is 1 cm.

imaged in the transverse and sagittal planes. Pre-biopsy biphosphate enemas and antimicrobial prophylaxis with norfloxacin were used in every case.' All biopsy specimens from each of the 6 areas were processed separately and categorized histologically as described previously. 1 RESULTS

There were 43 patients with an abnormally firm, nonnodular prostate, 85 with discrete, palpable clinical stage B nodules (37 with stage Bl, 28 stage B2 and 20 stage B3 disease) and 8 with clinical stage C disease on digital rectal examination. All 136 patients underwent random systematic biopsy; 57 of the 136 required additional directed biopsies of a hypoechoic defect that was either medial or lateral to the mid lobe parasagittal plane of the systematic biopsies (fig. 2). The mean number of biopsies for all 136 patients was 7.8, with a range of 6 to 11 per patient. Of the 136 prostates 83 (61 per cent) were positive for cancer (table 1). Of these 83 patients cancer was detected by the

The data in tables 1 and 2 show that random systematic core biopsies of the prostate obtained as described detected 62 of 63 cancers that presented as clinical stage B nodules, 10 of 12 that presented as abnormal firmness and all 8 clinical stage C cancers, for an over-all detection rate of 96 per cent (80 of 83). Of the 3 cancers missed by random systematic biopsies 2 were located far laterally at the apex and base, respectively, occupying 2 and 1 mm. core lengths of Gleason grade 3 plus 3. The apical 2 mm. cancer had a 3 to 5 mm. palpable nodule. Four repeat biopsies directed at the 1 mm. area in the patient with cancer at the base of the prostate could not confirm the presence of cancer. The remaining cancer was a 5 mm. core length of Gleason grade 3 plus 4 at the left base of the prostate in a patient who had 1.8 cc of cancer at radical prostatectomy; the random biopsy at the left base contained small, suspicious glands, while the directed biopsy detected the 5 mm. cancer. All 5 tumors detected by random systematic biopsies but not by directed biopsies into the hypoechoic defect were Gleason grade 3 plus 3 and occupied 2 mm. or less of the 1.5 cm. core biopsy. Two cancers were located by biopsy at the apex of the prostate and 2 at the base, and 1 was positive throughout the left lobe (the directed negative biopsy was into a hypoechoic defect at the right base). Only 1 of the 2 patients with a positive apical biopsy had a radical prostatectomy; the cancer volume was 0.92 cc and extended to the apex on histological reconstruction. Of the 2 biopsy positive cancers located at the base of the prostate 1 was a 1 mm. cancer that was re-biopsied with 4 negative cores from this same area. Since 96 per cent of the 83 cancers found in 136 patients were readily identified by random systematic biopsies as described (table 1) and since 5 of 30 cancers (17 per cent) were missed by directed biopsies into hypoechoic defects (table 2), we believe the urologist can adopt the simpler approach of 6 systematic biopsies in the mid lobe parasagittal plane of the prostate (figs. 1 and 2). If a small hypoechoic defect lies close to the midline sagittal plane of the prostate or in a far lateral parasagittal plane additional directed biopsies are advisable. The advantages to this systematic approach of multiple ran-

73

RANDOM SYSTEI\1ATIC ULTRASOUND GUIDED TRANSRECTAL CORE BIOPSIES OF PROSTATE TABLE

1. Ultrasound findings and results of random systematic versus directed ultrasound guided transrectal core biopsies

of the prostate in 136 patients Digital Rectal Examination

No. Pts.

Abnormally firm Stage B nodule Stage C disease Totals

43 85 8 136

No. With Hypoechoic Defects (%) 35 83 8 126

(81) (98) (100) (93)

Random Systematic Biopsies No. Pos. Biopsies(%) 12 63 8 83

(28) (74) (100) (61)

A comparison of the ability of random systematic versus directed biopsies to detect prostate cancer in 57 patients whose hypoechoic defects lay outside the mid lobe parasagittal plane of systematic biopsy

TABLE 2.

Random Systematic Biopsy

Directed Biopsy

Pos. Neg. Pos. Neg. Total

Pos. Neg. Neg. Pos.

No.(%) 22 (39) 27 (47) (9) 5 3 (5) 57 (100)

dom biopsies are equal or perhaps even greater accuracy in detecting cancer compared to the 2 or 3 usual biopsies directed at hypoechoic defects, potential information on the volume of intraprostatic cancer from the spatial distribution of positive biopsies, preoperatively useful information on the possibility of potentially positive margins at the apex and bladder neck if cancer is found at these locations, better representation of the average Gleason grade of the cancer, a substantial simplification of the entire approach to transrectal ultrasonography in the diagnosis of prostate cancer and a technique that is easy to teach, learn and execute. The disadvantages include the possibility of detecting insignificantly small cancers and not knowing how large a cancer can be missed by this technique. Since radical prostatectomies are performed on patients who have only 1 to 3 mm. of cancer on core biopsies, our histological reconstructions of the entire prostate'3 will provide us with more information on the volume sensitivity of these random but systematic biopsies. We have not observed any difficulties in the dissection of Denonvillier's fascia from the longitudinal musculature of the rectum in more than 150 radical prostatectomies after these 6 systematic biopsies using the Biopty gun with an 18 gauge needle. In fact, since the standard 14 gauge Tru-Cut* biopsy needle is nearly twice the diameter of an 18 gauge needle and surely more traumatic from the manual technique of pushing it into the prostate, it is likely that the usual practice of obtaining 2 or 3 Tru-Cut transrectal biopsies produces more scarring than the 6 biopsies described with the :rapid fire, 18 gauge Biopty gun. We are concerned about the accidental detection of small cancers less than 2 or 3 mm. in diameter. For this reason we carefully measure the millimeters of cancer in every biopsy. Biopsies are repeated in the 11 per cent of our patients whose prostates have 3 mm. or less of cancer in only 1 of the 6 areas biopsied systematically, concentrating on the area containing the cancer. Our current technique is to direct 6 additional core biopsies into the precise area of the positive biopsy; 3 of the 6 are directed in such a way as mainly to sample peripheral or central zone tissues, and 3 are obtained from the more anteriorly located transition zone tissues. If repeat biopsies fail to detect further cancer we currently are comfortable, at least with clinical stage B disease or abnormally firm prostates, to follow our patients with serum prostate specific antigen levels and to repeat the biopsies in 1 to 2 years. * Travenol Laboratories, Inc., Deerfield, Illinois.

No. Pos. Biopsies(%) 10 62 8 80

(83) (98) (100) (96)

% Total No. Pts. 23 73 100

59

Ca Detected Only by Directed Biopsy No. Pos. Biopsies(%) 2 1 0 3

(17) (2) (0) (4)

% Total No. Pts. 5 1 0

2

It also must be remembered that this technique applies only to peripheral and central zone cancers. We now know that 50 per cent of all prostate cancers are located spatially in the anterior half of the prostate in either the lateral portions of the peripheral zone or the more medial and anteriorly located transition zone.4 The transition zone contains the nodules of benign prostatic hyperplasia. Since these nodules frequently are hypoechoic relative to peripheral zone tissue and often are placed in the more medial fibers of the preprostatic sphincter,5 which is echopenic, it is difficult if not impossible with current ultrasound technology to detect anteriorly located cancers. Generally, ultrasound is most useful in the detection of cancer in the posterior region of the prostate where the rectal examination is most sensitive. A good example is illustrated by a patient who was excluded from these data. A 62-year-old white lawyer was referred with a palpable linear ridge, 2 to 3 mm. x 2 cm., extending down the middle of the right lobe from the base of an otherwise normal prostate. The prostate specific antigen was 60 ng./ml. by the Yang assay. Several perinea! core biopsies of the right lobe with the patient under general anesthesia had shown no cancer 1 month before referral. The right lobe of the prostate appeared to be hypoechoic from the base to the apex, as did the left base and the middle of the transition zone in the right anterior half of the prostate. Six random systematic biopsies were taken along with 1 biopsy of the right transition zone and both seminal vesicles. All 9 biopsies were free of cancer, showing either atrophy, inflammation (the right transition zone biopsy) or normal glands. Solely because of the strikingly elevated prostate specific antigen the hypoechoic peripheral and central zones of the right lobe were biopsied again at the apex, mid level and base of the prostate. However, because of our concern over anteriorly located cancer as a potential cause of the elevation in prostate specific antigen, we obtained 3 random biopsies from the anteriorly located transition zone on the right side (which again was hypoechoic on ultrasound) and 3 biopsies from the normal-appearing transition zone on the left side. Peripheral and central zone right lobe biopsies again were negative, the right transition zone biopsy showed acute and chronic inflammation, while 2 of the 3 core biopsies from the normal-appearing left transition zone contained a total of 9 mm. of Gleason grade 3 plus 3 cancer, histologically typical of columnar clear cell cancer arising in nodules of benign prostatic hyperplasia.4 At radical prostatectomy the lymph nodes and seminal vesicles were free of cancer but histological reconstruction of the 69 gm. prostate showed 9 cc of anteriorly located cancer well removed from the peripheral and central zones in the posterior third of the prostate. It is clear that while random systematic ultrasound guided transrectal core biopsies are ideal to detect the 68 per cent of cancers arising in the peripheral zone and even the 8 per cent arising from the relatively spared central zone, 6 these biopsies cannot help with the 24 per cent that arise in the benign prostatic hyperplasia nodules of the anteriorly located transition zone. 6 However, a markedly elevated prostate specific antigen, as shown by Cooner and associates, 7 should add to our suspicion of an undiagnosed cancer and should be an integral part of the ultrasound evaluation for prostate cancer. In the presence of negative peripheral zone and central zone random

74

HODGE AND ASSOCIATES

systematic biopsies, a markedly elevated prostate specific antigen should strongly suggest the need for random biopsies of the transition zone. However, in such instances, large amounts of benign prostatic hyperplastic nodules in the transition zone will diminish the statistical chances of a positive biopsy in this area unless the cancer is large. Despite these limitations, we hope that this technique of random systematic biopsies of the prostate will substantially simplify this exciting new diagnostic technology for urologists. With large numbers of patients and more correlations with histological reconstructions of radical prostatectomy specimens, we may be able for the first time to make reasonably accurate preoperative estimations of prostate cancer volume. This would provide important staging and prognostic information that currently is available only after radical excision of the prostate. Dr. John K. Kabalin reviewed the manuscript. REFERENCES 1. Hodge, K. K., McNeal, J. E. and Stamey, T. A.: Ultrasound guided

2.

3. 4.

5. 6.

7.

transrectal core biopsies of the palpably abnormal prostate. J. Urol., 142: 66, 1989. Stamey, T. A., Kabalin, J. N., McNeal, J. E., Johnstone, I. M., Freiha, F., Redwine, E. A. and Yang, N.: Prostate specific antigen in the diagnosis and treatment of adenocarcinoma of the prostate. II. Radical prostatectomy treated patients. J. Urol., 141: 1076, 1989. Stamey, T. A., McNeal, J. E., Freiha, F. S. and Redwine, E.: Morphometric and clinical studies on 68 consecutive radical prostatectomies. J. Urol., 139: 1235, 1988. McNeal, J.E., Price, H., Redwine, E. A., Freiha, F. S. and Stamey, T. A.: Stage A versus stage B adenocarcinoma of the prostate: morphologic comparison and biological significance. J. Urol., 139: 61, 1988. McNeal, J. E.: Origin and evolution of benign prostatic enlargement. Invest. Urol., 15: 340, 1978. McNeal, J. E., Redwine, E. A., Freiha, F. S. and Stamey, T. A.: Zonal distribution of prostatic adenocarcinoma: correlation with histologic patterns and direction of spread. Amer. J. Surg. Path., 12: 897, 1988. Cooner, W. H., Mosley, B. R., Rutherford, C. L., Jr., Beard, J. H., Pond, H. 8., Bass, R. B., Jr. and Terry, W. J.: Clinical application of transrectal ultrasonography and prostate specific antigen in the search for prostate cancer. J. Urol., 139: 758, 1988. EDITORIAL COMMENTS

In the preceding article on ultrasound guided transrectal core biopsies of the palpably abnormal prostate (page 66) the authors have examined carefully the impact of transrectal ultrasound to identify areas of abnormality in the prostate that correspond to biopsy proved prostatic malignancy. They examine a population of 251 men with palpably abnormal prostates and demonstrate a hypoechoic defect identified digitally as abnormal in 90 per cent. It is interesting that only 66 per cent of the palpably abnormal areas were positive for malignancy. They demonstrate further that 156 of 251 patients had palpable nodules, more than 90 per cent of which were hypoechoic by ultrasound. This study should be reviewed carefully by the individual who wishes to use transrectal ultrasound to enhance diagnosis of prostatic malignancy. While nodules that are palpable and clinically suspicious for malignancy usually are hypoechoic and contain prostatic cancer on biopsy, routine screening of prostatic tissue will demonstrate nonpalpable hypoechoic areas that will show no evidence of malignancy despite ultrasonically guided biopsy. Thus, the direct correlation between hypoechogenicity and the presence of malignancy cannot be established with certainty. Finally, the ability to detect malignancy in 94 per cent of the patients proved to have cancer on the basis of digitally directed biopsies and in 53 per cent of the patients with previously negative digitally directed biopsies argues strongly for the use of ultrasonically guided biopsies to direct the needle to areas of specific concern. The authors support the observation that appropriate prophylactic antibiotic coverage can reduce the incidence of biopsy associated infection to an acceptable level. The manuscript should be read with care. Although written by a group enthusiastically committed to the use of

transrectal ultrasound to identify malignancy, it presents a balanced position detailing the advantages of careful use of transrectal ultrasonography to enhance the diagnosis of prostatic malignancy. This article on random systematic versus directed ultrasound guided biopsies, detailing experience with ultrasound guidance to select a course of systematic transrectal biopsies of prostatic substance as opposed to specifically directed core biopsies of the prostate, is of interest but it raises several issues when the details of this study are compared to those of the previous paper. In this article the authors advocate that random systematic biopsies alone may be more effective in detection of malignancies in patients who have a palpably abnormal prostate than are ultrasonically directed biopsies of specific hypoechoic defects. The authors have gone so far as to provide a clay model of a prostate removed at radical prostatectomy to demonstrate the position ing of the biopsies. Although this clay model does demonstrate the parasagittal placement of the biopsy needles, it is difficult for the reader to visualize the course of the needle from the holes that are seen on the surface of the model. One or 2 ultrasound images that show positioning of the needle to correlate with the model would provide significant information to the reader and might provide the reader with the opportunity to biopsy the prostate systematically in this parasagittal area without the use of ultrasound guidance. The authors do make a case that ultrasound guided random systematic biopsy is simple and easily learned, and that it can be used to detect prostatic malignancy with an advantage over directed biopsies. However, they have not established whether digitally guided biopsies in the specific areas identified would provide information of similar impact as does random systematic guided biopsy. Although they may wish to direct attention to this fact in another manuscript, I suspect that this information is available to them and would appreciate their comment in response. The reader also should note that the bulk of the biopsies obtained by the authors, which were positive by random systematic ultrasound guided core biopsy, were obtained in patients with palpable nodules, with less than 25 per cent of the malignancies being detected in patients who had only irregular prostates. Therefore, one must ask how this specific study and the subsequent results that they presented could be biased by patient selection. Finally, the authors have failed to provide sufficient information to confirm their position that random systematic ultrasound guided transrectal core biopsies have the potential to provide quantitative useful information in the preoperative evaluation of men with prostatic malignancy. The concept of sampling of the prostate in a sequential and orderly manner is of value but it may not necessarily require ultrasound guidance to establish such an orderly pattern of biopsy. David F. Paulson Department of Urology Duke University School of Medicine Durham, North Carolina We are indebted to the authors for their determination in attempting to define the role of transrectal ultrasound guided biopsies in the evaluation of men suspected to have prostatic carcinoma. The concept of random systematic biopsy is a novel approach and seems to provide additional information in this select patient population. It is important to remember that all of these men had palpable abnormalities and the implications of this approach in men with palpably normal prostates currently are unknown. The finding of carcinoma in 53 per cent of the men with previously negative digitally guided biopsies is of great interest. This supports the concept that ultrasound guided biopsy should be performed in patients with negative digitally directed biopsies but it also raises the often asked question whether all prostatic nodules should be biopsied under ultrasound guidance or whether the technique should be reserved for those who only have an initially negative digitally directed biopsy. Previous studies have supported both approaches and only additional experience by other investigators will help to answer this question. When such studies are done it would be helpful if the digitally directed biopsy and followup ultrasound guided biopsies are performed by a limited number of investigators, and preferably by a single investigator so as to limit the degree of variability inherently associated with these procedures. The concept of random systematic biopsies is unique and the finding of carcinoma in areas not associated with a palpable abnormality is of significance. When considered further these findings are not unexpected because these patients had palpable tumors and the multicentricity of prostate carcinoma is well known. This concept is supported by the observation that if prostate biopsies are done in men with no