IN 3-D COMPUTER SIMULATED PROSTATE MODELS LATERAL PROSTATE BIOPSYS INCREASE THE DETECTION RATE OF PROSTATE CANCER

John J. Bauer (1,4), Jianchao Zeng (2), James Weir (2), Wei Zhang (3),

Isabell A. Sesterhenn (3), Roger R. Connelly (4), Seong K. Mun (2), Judd W. Moul (4,5)

1- Urology Service, Department of Surgery

Walter Reed Army Medical Center

Washington D.C. 20307-5001

2- Imaging Science and Information Systems Center (ISIS)

Department of Radiology, Georgetown University Medical Center

2115 Wisconsin Avenue, NW, Suite 603, Washington, DC 20007

3- Department of Genitourinary Pathology

Armed Forces Institute of Pathology (AFIP)

Washington D.C. 20306-6000

4- Center for Prostate Disease Research (CPDR)

Department of Surgery,

Uniformed Services University of the Health Sciences

Bethesda, Maryland 20814-4799

5- Reprints/ correspondence:

Judd W. Moul, M.D.

Center for Prostate Disease Research (CPDR)

Dept. of Surgery

Uniformed Services University of the Health Sciences

4301 Jones Bridge Rd.

voice: 301-295-3707

FAX: 301-295-6396

The opinions and assertions contained herein are the private views of the authors and are not to be construed as reflecting the views of the U.S. Army or the Department of Defense.

This research was supported by a grant from the Center for Prostate Disease Research, a program of the Henry M. Jackson Foundation for the Advancement of Military Medicine (Rockville, MD.), funded by the U.S. Army Medical Research and Development Command, and by the Department of Clinical Investigation from the Walter Reed Army Medical Center.

Keywords: 3-D simulation, cancer distribution, computer visualization and simulation, interactive prostate needle biopsy

ABSTRACT

Objectives. Urologists routinely use the systematic sextant needle biopsy technique to detect prostate cancer. However, recent evidence suggests that this technique has a significant sampling error. We developed a novel 3-D computer assisted prostate biopsy simulator based upon whole-mounted step-sectioned radical prostatectomy specimens to compare the diagnostic accuracy of various prostate needle biopsy protocols.

Methods. We obtained digital images of 201 step-sectioned whole-mounted radical prostatectomy specimens. 3-D computer simulation software was developed to accurately depict the anatomy of the prostate and all individual tumor foci. Additional peripheral devices were incorporated in the system to perform interactive prostate biopsies. We obtained 18- biopsies of each prostate model to determine the detection rates of various biopsy protocols.

Results. The 10- and 12- pattern biopsy protocols had a 99.0% detection rate, whereas, the traditional sextant biopsy protocol rate was only 72.6%. The 5-region biopsy protocol had a 90.5% detection rate and the 14- pattern, which includes all the biopsies used in the patterns above only added one additional positive case (99.5%). Transitional zone and seminal vesicle biopsies did not result in a significantly increased detection rate when added to the patterns above. Only one positive model was obtained when the transitional biopsies were added. The lateral sextant pattern revealed a detection rate of 95.5%, whereas, the 4-pattern lateral biopsy protocol had a 93.5% detection rate.

Conclusions. Our results suggest that all the biopsy protocols that use laterally placed biopsies based upon the five region anatomical model are superior to the routinely used sextant prostate biopsy pattern. Lateral biopsies in the mid and apical zones of the gland are the most important.

INTRODUCTION

Transrectal ultrasound (TRUS) guided biopsy is routinely used for diagnosis of prostate cancer. The current biopsy protocol routinely used by urologists is the systematic sextant biopsy (1). However, studies have shown that this protocol results in a positive detection rate of only 20-30% (2, 3). Furthermore, a significant number of patients (20-40%) with elevated PSA have a positive repeat biopsy, suggesting that many patients with prostate cancer are not being diagnosed initially using the sextant protocol (4-7). Recent clinical studies have suggested that the sextant technique may not be optimal and have investigated new biopsy protocols that may yield significantly better results (8-10).

Recently, studies using computer simulation of prostate biopsies have been published suggesting that additional biopsies can increase the sensitivity of the procedure. Both two-dimensional (2-D) and three- dimensional (3-D) computer-based simulation of prostate cancer have been shown to be useful in evaluating existing biopsy protocols (11-14). The objective of our study was to use a large number of 3-D reconstructed radical prostatectomy specimens to determine the detection rate of previously investigated biopsy protocols and, additionally, to determine the frequency of positive biopsies in various regions of the prostate in order to develop an optimized biopsy protocol. Since the new biopsy protocol will be based on statistical analysis of a quantitative database of digitized radical prostatectomy prostate specimens, it may significantly improve the accuracy of prostate cancer detection.

METHODS

Construction of Individual 3-D Computerized Prostate Models

Individual 3-D prostate models are constructed from radical prostatectomy specimens. The prostates are step-sectioned using a deli slicer at 2.25mm intervals and then digitized with a scanning resolution of 1,500 dots per inch. Each digitized image is segmented by a single pathologist (IAS) to identify the key pathological structures including surgical margins, capsule, urethra, seminal vesicles and tumor. The contours of each structure are identified on each slice then stacked to develop the 3-D prostate model. Interpolation between the contours is carried out using a 3-D elastic model-based technique. Two hundred and one 3-D individual prostate models have been constructed using an SGI Onyx 2 Infinite Reality 10000 Workstation (figure 1). An interactive biopsy interface was developed to allow real-time rotation and insertion of the ultrasound probe, and depth of needle placement before the biopsy device was fired (figure 2).

Biopsy Needle Placement

A single urologist (JJB) completed a total of 18 biopsies on each of the 201 prostate models. The various regions of the prostate were sampled a similar manner. Biopsies were performed in the right and left apex, mid and base regions of the prostate models (sextant protocol) half way between the midline and the lateral edge of the prostate. Additional biopsies in the lateral aspect of the right and left apex, mid and base regions were obtained midway between the sextant biopsy locations and the lateral border of the prostate (lateral biopsies). To incorporate a 5-region biopsy pattern similar to Eskew et al (9) biopsies in the midline apex and base were obtained. Both the right and left transition zone and seminal vesicles were also sampled. The tip of the needle was brought up to the capsule of the prostate at a 30-degree angle and then discharged in the interactive mode (Figure 3). The computer then automatically determined if the biopsy was positive for tumor. Several patterns were analyzed as defined in table 1.

Data Analysis

Frequencies of positive biopsies in the various prostate regions were determined. The mean positive biopsy hits were calculated for each biopsy protocol. The needle core biopsy data for each biopsy protocol was analyzed for variance using the McNemar’s test.

RESULTS

Table 2 is a summary of the individual positive biopsy frequencies for each region of the prostate. Of the 201 prostate models, both the 10- and 12- pattern biopsy protocols detected cancer in 199 models for a detection rate of 99.0%. In comparison, the sextant (6-pattern) biopsy protocol only detected 146 models with cancer (72.6%). As noted in table 1, with the addition of the laterally placed biopsies, the 10-, 12-, 14-, 16- pattern and 5-region biopsy protocols have a higher detection rate. The extra biopsies used in the 14- pattern biopsy protocol only added one additional positive model where cancer was detected (99.5%). The 5-region biopsy protocol had a positive detection rate of 90.5% (182/201). The transitional zone and seminal vesicle biopsies added little to the detection rate. One model was detected with cancer solely by the transitional zone biopsy (0.5%, 1/201). The seminal vesicle biopsies where never positive when all other biopsies were negative. The overall positive frequencies for the transitional zone biopsies were 26/201(12.9%) models on the left, 23/201(11.4%) models on the right and for the seminal vesicle biopsies were 2/201(1.0%) models on the left and 3/201(1.5%) models on the right. Interestingly, the lateral sextant 6-pattern and the 4-pattern had a significantly higher detection rate that the traditional sextant pattern (93.5%, 95.5%, 72.6%, respectively).

We analyzed for variance using the McNemar’s test. A 2 x 2 table of two different biopsy patterns was constructed and then compared. For example, comparison of the sextant to lateral sextant biopsy is shown in table 3. In constructing this table, the question of interest is whether the detection rate for sextant pattern (146/201 = 72.6%) and the lateral sextant pattern (192/201 = 95.5%) are the same. The McNemar’s test only analyzes the discordant elements of the 2 x 2 table, these are the data points that are important in determining whether there is a significant difference. In our example, the test statistic is (53-7)²/(53+7) = 46²/60 = 3.267, which is the approximate Chi-square with 1 df (degree of freedom). For Chi-squared (35.3), the p-value is equal to 0.001, so we reject the null hypothesis of equal detection rates. The lateral sextant biopsy patterns detection rate was significantly better than the traditional sextant biopsy. We then compared the 4- pattern biopsy (R/L lateral apex and lateral mid) to the lateral sextant pattern and found the lateral sextant to be marginally better with a p=0.046. Subsequent comparisons were as follows: lateral sextant vs. 10-pattern (p=0.008, in favor of the 10-pattern), 10 vs. 12-pattern (no difference, in favor of 10- since lower number of biopsies), 10 vs. 14-pattern (p=0.317, no significant difference, in favor of the 10-pattern with lower number of biopsies) and 10-pattern vs. 5-region pattern (p=0.001, in favor of the 10- pattern). These results suggest that the 10-pattern biopsy protocol is the optimum pattern for the detection of prostate cancer.

COMMENT

Our study supports the routine use of laterally placed biopsies and suggests the transition zone and seminal vesicle biopsies are rarely required to detect prostate cancer if lateral biopsies are used. The transition zone biopsies were positive in only one model when all other biopsies were negative. There were no models that were positive in only the seminal vesicles. Considering the significant amount of pain that may be associated with transitional zone biopsies, biopsy protocols that include lateral biopsies should decrease the overall patient discomfort of prostate biopsy. When comparing the various patterns, the traditional sextant biopsy pattern revealed a cancer detection rate of 72.6 % in 201 prostate models. By merely shifting these six biopsies laterally one obtains a significantly higher cancer detection rate (95.5% vs 72.6%). In fact, the 4- pattern, a subset of the lateral biopsies which includes the right and left lateral apex and lateral mid region biopsies approximate the detection rates of the biopsy protocols that use 10, 12, 14, or 16 biopsies (93.5% vs 99.0%, 99.0%, 99.5%, %, 100%, respectively). However, our statistical analysis of the various biopsy patterns suggests that the 10-pattern biopsy protocol is the most optimum systematic biopsy protocol. It is also very important that the biopsies are obtained along the posterior and lateral surface of the prostate. The majority of tumors in our radical prostatectomy specimens were near the posterior and lateral capsule. If one merely seats the needle into the prostate before biopsy, a significant number of the needle cores would have been negative.

Current screening tests for prostate cancer include prostate specific antigen (PSA) and digital rectal exam (DRE). The combination of these two tests and a more informed patient and physician population has led to an increased number of prostate needle biopsies. However, with a 20-30% detection rate the accuracy of currently used biopsy techniques needs to be improved. There are also a significant number of prostate cancers that are detected on repeat biopsies. Keetch et al. (5), reported a 24% (104/427) positive repeat biopsy rate in men with persistently elevated PSA after initial negative biopsy. Lui and associates (15), reported a higher repeat positive biopsy rate of 38% (72/187). Lui’s study additionally identified that 28% (53/187) of the repeat biopsies had cancers detected in the peripheral zone and 10% (19/187) in the transitional zone. When Ukimura and associates (16) evaluated 226 men that had undergone repeat biopsies for an elevated PSA, 51 men (26%) were found to have prostate cancer on repeat biopsy. Cancer was found in 17% (33/193) of all men on the first repeat biopsy and 26% (14/54) patients who underwent a second repeat biopsy. Repeat biopsy has potential morbidity and contributes significantly to the costs of detecting prostate cancers. With such a high incidence of positive repeat biopsies many other techniques have been evaluated to determine the patient that would have a higher probability of a repeat positive biopsy. The majority of these indices were used to distinguish between benign prostatic hypertrophy and prostate cancer. Age-referenced PSA, PSA density (PSAD) and PSA velocity (PSAV) are still used to determine if a biopsy is necessary. Currently, the ratio of free-to-total PSA for patients with a PSA value between 4-10 ng/ml is increasingly being used to determine the risk of prostate cancer after an initial negative biopsy (4,17). Morgan et al. showed that a low ratio, despite multiple negative biopsies, demonstrated significant predictive power. A 10% cutoff provided 91% specificity and 86% specificity.

Daneshgari et al (11), developed a 2-D computer simulation of the prostate based on 159 radical prostatectomy specimens. The computer then generated random prostates and tumors. This computer model was used to simulate the sextant biopsy protocol and verify its ability to detect low-volume tumors. Various biases for the angle of biopsy and distribution of cancer foci were incorporated in the model. The simulation showed that only 20.3% of the simulated prostates had a tumor distribution in which sextant biopsy had a 95% probability of tumor detection. In fact, 26.8% of prostates had a distribution that was completely disjointed from the sextant locations. These prior findings show that a significant number of patients who have prostate cancer are not diagnosed at their initial biopsy. Accordingly, improving the predictive value of TRUS guided biopsy by optimizing biopsy protocols will improve its value as a screening and diagnostic tool.

A number of researchers have investigated techniques for improving the accuracy of biopsy protocols; however, several issues remain to be resolved. Eskew et al. introduced a new protocol called the 5-region biopsy in which additional lateral and midline biopsies are added systematically to the traditional sextant biopsy (9). The 5-region biopsy and the traditional sextant biopsy were compared with a total of 119 patients who underwent transrectal ultrasound guided needle biopsy of the prostate. In 48 cancer patients, 17 (35%) were detected only by the additional needles of the 5-region biopsy method within this group. As a result, the new 5-region biopsy method was claimed to improve biopsy results. Eskew’s results are promising, but his study group of 48 patients is small. Therefore, this protocol needs to be validated, and the underlying rationale for using 12 needles instead of some other number should be examined.

Our group found that the 5-region protocol showed a statistically significant advantage over the sextant method based on 89 patients with cancer (13). Chang et al. (10), also showed that lateral biopsies increase the sensitivity of prostate cancer detection. Fourteen percent of 118 patients had prostate cancer detected by only the lateral prostate biopsies. Dietrick et al. conducted a clinical study of 110 men who underwent radical prostatectomy and compared the core cancer length with the volumes of clinically significant and incidental carcinoma (18). For sextant biopsies, Dietrick determined that an optimal biopsy core length of 3 mm or more could reliably detect cancer of clinically significant volume. Vashi et al. developed a statistical model that determined the optimal number of biopsies to achieve a 90 % certainty of detecting various volumes of life threatening cancer (19). This model incorporated the variables of tumor doubling time, age, and prostate volume. Patients with prostate volumes greater than 20 grams and aged 50-69 years required an increased number of biopsies (7-23 biopsies) to detect life threatening prostate cancers. Goto et al. suggested that new biopsy strategies could be developed based on probability maps of cancer distribution within the prostate (9). But issues such as how these maps should be built and how new biopsy protocols could be derived from the maps remain to be investigated.

Recently, the use of 3-D computer simulation has supported the use of lateral biopsies as an essential component in any protocol to increase the detection rate of prostate cancer. Karakiewicz et al, using a 3-D computer assisted analysis of sector biopsies, showed that as they increased the number of biopsies from 4 to 12 and incorporated lateral zone biopsies the detection rates increased (11). This was especially true for the larger volume prostates. Chen et al. and Kaplan et al. both used 3-D computer simulations of prostate biopsy based upon reconstructed radical prostatectomy specimens (12, 13). Both studies showed that the sextant protocol was less sensitive (~ 20%) than biopsy patterns that used an increased number of cores and laterally placed biopsies.

CONCLUSION

The accurate 3-D reconstruction of the radical prostatectomy specimens with spatial anatomy that includes the urethra, ejaculatory ducts, seminal vesicles, capsule, surgical margins and the tumors allows for evaluation of various prostate biopsy protocols. In general it was noted that the majority of the tumors were near the posterior-lateral surface of the prostate. The laterally placed biopsies in the posterior-mid and posterior-apex regions of the gland resulted in the highest positive biopsy frequencies, all near 50%. It is important to note that actually placing the tip of the needle into the prostate before the biopsy is performed results in a higher negative biopsy rate since the tumors are so close to the posterior and lateral capsule.

When we compared the various biopsy patterns our results suggest that the 10-pattern biopsy protocol provides the highest relative detection rate for the number of biopsies performed during a single procedure. When compared to the 10-pattern protocol, the additional biopsies of the 12-, 14-, 16- and 5-region biopsy protocols do not add significantly to the detection rate. The lateral 6-pattern and the 4-pattern biopsies approximate the detection rates of the patterns with higher numbers of biopsies. If clinicians are reluctant to adopt the 10-pattern biopsy protocol because of the extra biopsies, then merely using a sextant pattern with laterally placed locations will approximate the detection rate. Prospective clinical trials that compare these biopsy protocols must be completed to validate these results before any definitive recommendations can be made to replace the currently used traditional sextant biopsy protocol.

REFERENCES

1.) Hodge KK, McNeal, JE, Terris, MK, Stamey, TA: Random systematic versus directed

ultrasound guided trans-rectal core biopsies of the prostate. J. Urol., 142: 71-74, 1989.

2.) Flanigan RC, Catalona WJ, Richie JP, Ahmann FR, Hudson MA, Scardino RC,

DeKernion JB, Ratliff TL, Kavoussi LR, Dalkin BL, Waters WB, MacFarlane MT and

Southwick PC: Accuracy of digital rectal examination and transrectal ultrasonography

in localizing prostate cancer. J. Urol., 152: 1506-1509, 1994.

3.) Van der Kwast TH, Hoedemacker RF, Krause R, Rietenbergen JBW, Kruger AEB and

Schroder FH: Incidence of high grade intraepithelial neoplasia and suspect lesions in

systematic sextant prostate needle biopsies in a population based screening study

(ERSPC). J. Urol., 159(5):180, abstract # 689, 1998.

4.) Catalola WJ, Smith DS, Ratliff TL and Basler JW: Detection of organ-confined prostate

cancer is increased through prostate specific antigen-based screening. JAMMA, 270:

948-954, 1993.

5.) Keetch DW, Catalona WJ and Smith DS: Serial prostatic biopsies in men with

persistently elevated serum prostate specific antigen values. J. Urol., 151:1571-1574,

1994.

6.) Roehrborn CG, Pickens GJ and Sanders JS: Diagnostic yield of repeated transrectal

ultrasound-guided biopsies stratified by histopathologic diagnoses and prostate specific

antigen levels. Urol., 57:347-352, 1996.

7.) Ellis WJ and Brawer MK: Repeat prostate needle biopsy: who needs it? J. Urol.,

153:1496, 1996.

8.) Goto Y, Ohori M, Arakawa A, Kattan MW, Wheeler TM and Scardino PT:

Distinguishing clinically important from unimportant prostate cancers before treatment:

value of systematic biopsies. J. Urol., 156: 1059-1063, 1996.

9.) Eskew AL, Bare RL and McCullough DL: Systematic 5-region prostate biopsy is

superior to sextant method for diagnosing carcinoma of the prostate. J. Urol., 157: 199-

202, 1997.

10.) Chang JJ, Shinohara K, Hovey RM, Bhargava V, Carroll PR and Presti JC: Prospective

evaluation of lateral biopsies of the prostate for cancer detection. J. Urol., 159(5):179,

abstract # 688, 1998.

11.) Daneshgari F, Taylor GD, Miller GJ and Crawford ED: Computer simulation of the

probability of detecting low volume carcinoma of the prostate with six random systematic

core biopsies. Urology, 45: 604-609, 1995.

12.) Chen MF, Troncoso P, Johnston D, Tang K and Babaian JB: Comparison of prostate

biopsy schemes by computer simulation. J. Urol., 159(5):179, abstract # 685, 1998.

13.) Kaplan CR, Lynch JH, Zeng J, Moul JW, Sesterhenn IA and Mun SK: Comparison of

sextant to 5-region biopsy technique using a 3-D computer simulation of actual prostate

specimens. J. Urol., 159(5):179, abstract # 687, 1998.

14.) Karkiewicz PI, Hanely JA and Bazinet M: Three-dimensional computer-assisted analysis

of sector biopsy of the prostate. Urol., 52: 208-212. 1998.

15.) Lui PD, Terris MK, Mcneal JE and Stamey TA: Indications for ultrasound guided

transition zone biopsies in the detection of prostate cancer. J. Urol., 153 1000-1003,

1995.

16.) Ukimura O, Durrani O and Babaian RJ: Role of PSA and its indices in determining the

need for repeat prostate biopsies. Urol., 50:66-72, 1997.

17.) Morgan TO, McLeod DG, Leifer ES, Murphy GP and Moul JW: Prospective use of free

prostate-specific antigen to avoid repeat prostate biopsies in men with elevated total

prostate-specific antigen. Urol., 48:76-80, 1996.

18.) Dietrick DD, McNeal JE and Stamey TA: Core cancer length in ultrasound-guided

systematic sextant biopsies: a preoperative evaluation of prostate cancer volume.

Urol., 45(2): 987-992, 1995.

19.) Vashi AR, Wojno KJ, Gillespie B and Oesterling JE: A model for the number of cores

per prostate biopsy based on patient age and prostate gland volume. J. Urol., 159:920-

924, 1998.

Table 1.

Biopsy Pattern Definitions and Detection Rates

	Detection Rate (%)		# of Biopsies
-----------------	-----------		----------
4-pattern: R/L lateral Apex and lateral Mid		93.5		4
6-Pattern: Sextant = R/L Base + R/L Mid + R/L Apex 6L-Pattern: Lateral Sextant = R/L lateral apex, mid, base		72.6 95.5		6 6
10-Pattern: 6-Pattern + R/L lateral Mid + R/L lateral Apex		99.0		10
12-Pattern: 10-Pattern + R/L lateral Base		99.0		12
14-Pattern: 12-Pattern + Middle Base + Middle Apex		99.0		14
5-Region: 6-Pattern + Middle Base + Middle Apex + R/L lateral Base + 90.5 R/L lateral Apex TZ: R/L transitional zone - 16-Pattern: 14-Pattern + TZ zone biopsies 100.0 SV: R/L seminal vesicle -			12 2 16 2

Table 2:

Needle Biopsy Frequencies by Prostate Regions

Prostate Region	Negative	Positive	% Positive
**LEFT**
Base	152	49	24.4
Mid	111	90	44.8
Apex	133	68	33.8
Lateral Base	126	75	37.7
Lateral Mid	83	118	58.7
Lateral Apex	99	102	50.7
**RIGHT**
Base	160	41	20.4
Mid	130	71	35.3
Apex	140	61	30.9
Lateral Base	139	62	30.8
Lateral Mid	106	95	47.8
Lateral Apex	105	96	47.8
*MIDDLE*
Base	164	37	18.4
Apex	155	46	22.9
Transition Zone
Right	178	23	11.4
Left	175	26	12.9
Seminal Vesicle
Right	198	3	1.5
Left	199	2	1.0

Table 3:

McNemar’s Test: Comparison of Sextant vs. Lateral Sextant Biopsy Protocols

		Lateral Sextant
Sextant Biopsy	NO	YES	TOTAL
NO	2	53	55
YES	7	139	146
TOTAL	9	192	201

NO= negative biopsy, YES= positive biopsy,

McNemar’s test compares the two biopsy patterns ability to detect cancer in a single model, these individual values are then evaluated as part of the entire cohort. Possible combinations are: NO, NO; NO, YES; YES, YES; and YES, NO, where the two discordant pairs are when the two biopsy patterns differed in their ability to detect cancer in a single model. The discordant pairs are the terms used to figure if the two patterns are different.

FIGURES:

(a) (b) (c)

(d) (e)

Figure 1: 3-D Reconstruction of Prostate Models (a) Digitized image of a single slice of a step-sectioned radical prostatectomy specimen (b) Stacked surgical margin contour controls of original slices (c) Surgical margin contour interpolation (d) Complete interpolated model with internal structures (e) Final 3-D reconstructed prostate model.

Figure 2: Virtual Interactive Biopsy Graphical User Interface

(a) (b)

Figure 3: Biopsy Locations (a) Colored dots represent the areas of the prostate biopsied, the pink dots correspond to the traditional sextant locations, transition zone and seminal vesicle biopsies not shown (b) Transparent final prostate model with the angle of biopsy path and position of the needle core when the needle is brought up to the capsule, but not seated in the glad. Mid gland biopsies are positioned at a point where half the volume of prostate is above and below the angle of biopsy path. Apex and Base biopsies are half way between the Mid biopsy and the base and apex of the gland.