One 10-core prostate biopsy is superior to two sets of sextant prostate biopsies

Authors


K.G. Fink, Department of Urology and Andrology, St. Johannsspital, Landeskrankenanstalten, Müllner Hauptstrasse 48, 5020 Salzburg, Austria.
e-mail: k.g.fink@lks.at

Abstract

OBJECTIVES

To compare the efficiency of different transrectal ultrasonography (TRUS)-guided prostate biopsy techniques for detecting prostate cancer.

MATERIALS AND METHODS

In all, 81 prostates from radical prostatectomy were used and two consecutive sets of sextant biopsies and one 10-core biopsy taken in each specimen. The 10-core biopsy consisted of a sextant biopsy and four cores from the far lateral areas of the prostate. To simulate a transrectal biopsy procedure, all biopsies were taken under TRUS guidance.

RESULTS

In the first set of sextant biopsies 44 prostate cancers (54%) were detected and in the second set 51 (63%). Combining both sets of sextant biopsies 57 (70%) of the carcinomas were detected. One set of 10-core biopsies detected 66 (82%) of all prostate cancers. Overall, with the 10-core biopsies 16% more prostate tumours were diagnosed than with two consecutive sets of sextant biopsies. To find the same number of prostate cancers as with the 10-core technique, 14% of patients undergoing sextant biopsy would require a second set and 11% at least a third set of biopsies.

CONCLUSIONS

The 10-core prostate biopsy technique is superior to the commonly used sextant technique and could spare patients unnecessary repeated biopsy. Even after including a second set of sextant biopsies, the total detection rate with these 12 biopsies was inferior to the 10-core technique.

INTRODUCTION

Since Hodge et al.[1] showed that TRUS-guided transrectal sextant prostate biopsy was better than digital guidance, this technique has become a standard procedure for diagnosing prostate cancer. The efforts in early detection have led to more men undergoing this procedure, which has also increased the number of men who are still suspected of having prostate cancer although the first sextant prostate biopsy showed no malignancy. Keetch et al.[2] found prostate cancer in 19% of 427 men who had a second set of sextant prostate biopsies taken, and Roehrborn et al.[3] found a positive biopsy rate of 23% in 123 men after a second set of biopsies. This indicates that significantly many tumours are not diagnosed by the first biopsy.

Levine et al.[4] recommended taking two consecutive sets of TRUS-guided sextant biopsies at one office visit and reported that this would lead to an increase of 30% in prostate cancer detection. However, it is common to take one sextant biopsy and, if the result is negative for cancer and the indication for prostate biopsy persists, to take another set later with the same sextant technique.

Currently the trend is to take more than six cores per biopsy session. Norberg et al.[5] showed that an additional two to four biopsies of the lateral areas of the prostate, added to the routine sextant protocol, increase the sensitivity of cancer detection. These data were confirmed by Bauer et al.[6] using a three-dimensional computer-simulated model. We compared the efficiency of two consecutive sextant biopsies with that of one 10-core prostate biopsy.

MATERIALS AND METHODS

In all, 81 prostates obtained at radical prostatectomy were used and all sets of biopsies taken from each specimen. The samples were taken under TRUS guidance (7.5 MHz biplanar transrectal probe), providing a realistic simulation of a randomized transrectal prostate biopsy procedure (Fig. 1) [7,8]. An 18 G needle was used, with an ‘ultrasound’ tip and depth margins providing a 19 mm cutting length (Manan Medical Products Inc., Illinois, USA), powered by a spring-loaded automatic biopsy gun (High Speed Core Cut System, Biomedizinische-Instrumente Produkte GmbH, Türkenfeld, Germany).

Figure 1.

Simulated prostate biopsy procedure and TRUS view, taking cores from the lateral areas of the prostate.

Sextant biopsies were taken using three probes per side in a paramedian mid-lobular plane. The 10-core biopsy consisted of a sextant biopsy with two extra cores per side, from the far lateral areas of the prostate (Fig. 2). Each patient was assessed with a DRE by an experienced urologist before surgery. The prostate was defined as negative if there were no palpable abnormalities that could be related to cancer. All others were deemed positive, even if the palpable abnormality was not urgently indicative of cancer. The prostate volume was calculated by measuring three diameters in two planes, multiplied by 0.526. Serum PSA was measured using the Tandem R assay (Hybritech Inc., San Diego, CA).

Figure 2.

The location of cores for each biopsy technique.

The volume of prostate cancer was estimated by measuring the area of tumour in the histological slices and multiplying by the thickness of the embedded specimen slices. The tumour volume of each slice was then summed to the find the total tumour volume. Solid and diffuse areas were measured equally and multiple foci of carcinoma measured separately. To allow for tissue shrinkage a correction factor of 1.14 was used.

The results were assessed statistically using commercially available computer software, applying the McNemar unpaired test to evaluate the significance of differences in prostate biopsy results, with P < 0.05 taken to indicate significance.

RESULTS

For diagnosing their disease, 37 patients had had a sextant biopsy, two a sextant and additional lesion-directed biopsy, eight an octant biopsy, one an octant and additional lesion-directed biopsy, 31 a 10-core biopsy, and two an unknown number of cores taken, or with an unknown technique. Thus there was no specific selection of patients.

The mean (range) prostate volume was 40.2 (18–86) mL; 31 prostates were > 40 mL, 19 > 50 mL and eight > 60 mL. The mean (range) PSA value was 10.7 (2.1–31.7) ng/mL, and 27 patients had a PSA level of > 10 ng/mL. The mean (range) tumour volume was 3.02 (0.07–16.89) mL. The mean tumour volume of cancers detected with the first and second sextant biopsy was 3.82 and 3.84 mL, respectively; the mean tumour volume of those additionally detected with the second set was 3.49 mL. The mean tumour volume of the cancers detected with the 10-core biopsy was 3.48 mL. The different sets of biopsies did not detect significantly different tumour volumes.

With the first sextant biopsy, 44 prostate cancers (54%) were detected; for the second set of biopsies, again by sextant technique, 51 were found (63%), so with both sets together 57 (70%) of the tumours were detected. Using the 10-core technique including lateral areas, 66 prostate cancers were detected (82% detection rate; Table 1). Therefore, with one 10-core biopsy, 16% more tumours were found than with two sets of sextant biopsies (P = 0035). In an analysis of subgroups these results were not statistically significantly influenced by different DRE findings, different PSA levels or prostate volume.

Table 1.  The cancer detection rate (n) with two sets of sextant biopsies and one 10-core biopsy
Two sextant10-core biopsyTotal
negativepositive
negative121224
positive  35457
total156681

DISCUSSION

As most cancers are impalpable or invisible on TRUS, systematic biopsies must be taken and the TRUS-guided technique has become the most common procedure for diagnosing prostate cancer. When there is a suspicion of carcinoma, the prostate biopsy should be repeated if the first samples showed no malignancy [9]. Stroumbakis et al.[10] and Rabbani et al.[11] found, respectively, 80% and 77% of prostate cancers on re-biopsy of patients with confirmed prostate cancer, using the sextant technique. More recently Shinghal et al.[12] took TRUS-guided sextant biopsies in 20 patients before radical prostatectomy and found a sensitivity for the method of 65%; Terris et al.[13] reported a sensitivity of 60% for the sextant technique using the same approach. The present results gave a prostate cancer detection rate of 59% (the average cancer detection of 162 sextant biopsies in 81 prostate specimens). Stamey et al.[14] evaluated the annual changes in clinical and histological findings in specimens from radical prostatectomy, and showed that T1c cancer increased from 10% in 1988 to 73% in 1996, and organ-confined cancer from 40% to 75%, respectively. There is an apparent shift to smaller cancer volumes and this explains these differences. As the sensitivity of the sextant biopsy in the present population is low, it must be concluded that it is no longer sufficient for detecting cancer.

In the present simulation a second sextant biopsy found 13 additional tumours (29% more cancers), but intuitively it would seem reasonable that greater positional coverage should result in better detection. That a consecutive set of biopsies using the same technique as the first detects additional tumours is only because the biopsies are not taken perfectly, and that the cores are taken unintentionally from slightly different areas the second time. More cancers would be expected to be detected by taking the samples intentionally from areas that have not been biopsied the first time. The present results show that with four cores taken from areas untouched by the sextant biopsy, more cancers are detected than with six cores from the area previously sampled by the first six cores.

Whether a simulation really reflects in vivo conditions is an arguable point. As the data obtained in the present ex vivo simulation for one or two sextant biopsies were similar to those obtained in vivo[4,12,13] we consider that the simulation provides realistic results. However, we cannot discount the possibility that this model is not entirely compatible with an in vivo biopsy procedure. A simulation has several advantages; it is possible to take more biopsies than a patient would tolerate, and the sensitivity can be determined as there is cancer in all prostates. Usually cancer can be found by prostate biopsy in 25% of men undergoing this procedure [15]. To analyse the same number of cancers as used here it would be necessary to take biopsies in four times as many patients. Another advantage is that detailed information is available about the tumours not detected [8].

One 10-core biopsy detected 66 cancers; for equal cancer detection, 13 patients undergoing sextant biopsy would require a second set, and as two sets of sextant biopsies are inferior to one set using the 10-core technique, nine patients would need at least a third set of biopsies. This not only affects the patient, as a prostate biopsy implies discomfort and a risk of side-effects, but also affects the tumour, as usually there is a delay between taking the biopsies, and the tumour might grow in the meantime. A prostate biopsy technique that can spare the patient a second or even third set is also more cost-effective, which is increasingly important.

The results of Norberg et al.[5] were confirmed by Chang et al.[16], who compared various biopsy schemes, and Bauer et al.[6], who reported that the 10-core technique used in this simulation provided the highest relative detection rate for the number of biopsies taken in one procedure. Likewise, in the present study, this technique was better than two sets of the most common biopsy procedure.

In conclusion, sextant biopsy can no longer be recommended; the 10-core method is far better and can spare patients an unnecessary repeated biopsy. Even with a second set of sextant biopsies the total cancer detection rate of the 12 biopsies was inferior to the 10-core technique, indicating that apart from the total number of biopsies, the three-dimensional distribution of the sampling sites in the prostate, and therefore the biopsy technique, is an important factor in cancer detection.

ACKNOWLEDGEMENTS

We thank the employees from the department of pathology for their support and kindness.

Ancillary