Saturation biopsy for detecting and characterizing prostate cancer

Authors

  • J. Stephen Jones

    1. Glickman Urological Institute and Cleveland Clinic Lerner College of Medicine at Case Western Reserve University, Cleveland, OH, USA
    Search for more papers by this author

J. Stephen Jones, Glickman Urological Institute, Cleveland Clinic, 9500 Euclid Ave St. A100, Cleveland, OH 44195 USA.
e-mail: jonesS7@ccf.org

Abbreviations
SB

saturation biopsy

PZ

peripheral zone.

INTRODUCTION

Systematic biopsy became the standard for diagnosing prostate cancer after the landmark report of Hodge et al. in 1989 [1]. Their scheme included six evenly distributed parasagittal cores, a number chosen based on Stamey’s observation that six cores detected most of the large palpable tumours seen before the use of PSA testing (personal communication). Subsequent studies after the start of PSA testing showed a high false-negative rate. Additional sampling of the lateral and apical peripheral zone (PZ) increases accuracy [2,3], but misses many tumours even with extended biopsies [4]. As a result, ‘saturation’ biopsy (SB) has been explored, resulting in substantial cancer detection rates even after many negative biopsies [5,6]. This was initially done in the operating room under general or regional anaesthesia, but recent reports suggest that office-based SB is as well tolerated, with a lower risk of urinary retention [7].

DEFINITION AND RATIONALE

Stewart et al.[6] coined the term SB to describe the technique that they had developed almost simultaneously with Borboroglu et al.[5]. The studies obtained a mean of 23 and 22.5 cores, respectively, setting the de facto threshold of 22–24. It was shown that the additional value of obtaining >20 cores might be limited, so most contemporary protocols use this threshold.

Many urologists were concerned about the morbidity that such an aggressive strategy would entail when Hodge and Stamey proposed the ‘random’ sextant biopsy after testing in 136 patients. Such concerns were dismissed when broad experience showed few complications, and their six-core technique quickly became the standard of care. Patients usually accepted the pain of the few needle punctures, based on the inability to anaesthetise the prostate during that era. The advent of periprostatic block [7,8] allowed even more aggressive 8–12-core strategies previously considered unthinkable [9].

Many patients continue to have findings that create a suspicion of undiagnosed malignancy even after repeated negative biopsy. Almost a quarter of cancers are identified after an initial negative biopsy [10]. However, after two negative biopsies, a repeat sextant or even extended biopsy uncommonly identifies significant malignancy [11].

The 17th century poet, John Dryden, defined insanity as doing the same thing over and over again while expecting different results. The concept applies to repeat biopsy; if no cancer is identified in the parasagittal cores once, twice or more, it probably means that cancer does not exist in that location. As a result, SB often involves a different scheme, both numerically and geometrically, to identify cancer where it might previously have been overlooked. By contrast, with repeat routine extended biopsy, SB has been shown in several series to identify a substantial risk of high-grade cancers even after several routine biopsies.

APPROACHES

The original reports involved transrectal biopsy in the operating room. The use of a brachytherapy grid to map the prostate through a transperineal approach has also been described [12]. This approach requires anaesthetic-based operating room use, and so is associated with increased cost, time and anaesthetic risk.

The original report of office-based SB used a 24-core transrectal template, with cores concentrated laterally and apically, based on the preponderance of cancer in these locations. The apex is especially important during repeat biopsy after extended schemes, as it is the most common site of unique site-specific positive cores [13]. This sector is composed entirely of PZ tissue, so these cores are similar to lateral cores in the remainder of the gland. In site-specific studies, the lateral sectors are the primary site of essentially all tumours in patients having a repeat biopsy. As a result, sampling from the medial sector (midgland and base) appears to add less value, so a 20-core template for patients undergoing repeat biopsy has been proposed (Fig. 1) [14].

Figure 1.

Reduced sampling in the medial sectors is justified by the finding that parasagittal sectors (light shading) are rarely the site of unique positive cores in the absence of positive lateral cores (dark shading) in a repeat biopsy population.

To avoid repeat biopsy of the same site, the examiner should be observant of the hyperechoic linear echoes on TRUS that occur due to the interface of tissue with blood and the needle tract. With experience, it is possible to determine the site of biopsies taken earlier in the same session, to evenly distribute the sampling.

Areas likely to harbour cancer undiagnosed during previous biopsy sessions should be the focus of repeat biopsy whether SB or standard biopsy is used. Sampling at the vital apical level is often limited by the painful nature of apical biopsy. Insufficient sampling based on this pain might contribute to the apex often being the site of cancer overlooked during routine biopsy [14]. The pain of apical biopsy can be avoided by using the rectal sensation test to bypass anal pain fibres [15].

Side-fire ultrasound probes primarily target the posterior tissue of the PZ, which is indeed the most common site of malignancy. However, if initial biopsy in this area fails to identify cancer, it is mandatory that the ‘anterior horn’ of apical tissue be sampled, as shown in Figs 2 and 3.

Figure 2.

Transition zone (TZ) tissue is unlikely to harbour cancer, so apical biopsy is often positive, based on its entirety being comprised of PZ tissue. The ‘anterior horn’ of PZ tissue can be seen to the right of the more heterogeneous TZ tissue.

Figure 3.

The needle has been placed caudal (to the right) to the transition zone (TZ) in the ‘anterior horn’ of PZ tissue that wraps around the TZ. These anterior biopsies are the most likely place for cancer to have been undiagnosed during previous biopsy sessions.

RESULTS

Operating room-based SB was reported to identify cancer in 30–34% of cases after repeated negative sextant biopsies. Compared to sextant biopsy, there is a less steep decline in positivity with the number of previous negative biopsies [7,16].

Rabets et al.[17] reported a positive rate of 29% when using SB in the office setting. Importantly, most of these patients had had previous extended-core biopsy instead of the traditional sextant biopsy in previous reports. Seven of 11 patients who had had one negative sextant biopsy were found to have cancer, emphasising its false-negative rate· Walz et al.[16], using SB in the office, identified cancer in 41% of 161 patients who had had at least two previous negative biopsies.

A French study [18] reported 303 patients biopsied with 21 cores under periprostatic block. A second biopsy was positive in 25.6%, whereas only one of 37 patients who had had more than one previous biopsy was positive. Again, medial cores were of little additional value. Interestingly, their patients tolerated this protocol even with no periprostatic block. A subsequent report from this group in 650 patients found that SB improved the ability to predict pathological T3 tumours and positive surgical margins in the 150 who had a prostatectomy [19].

A report of transperineal biopsy using a brachytherapy grid, with a mean of 21 cores, identified cancer in 37% of patients; of these, 81% had their third or more biopsy. By contrast with the other publications, this group reported 46% of positive biopsies were in the transition zone . Biopsy via a brachytherapy grid has been the subject of other recent series, but has shown similar detection rates but higher rates of urinary retention [20–22].

By contrast, Fleshner and Klotz [23] found only 13.5% positive results in 37 patients who had had several previous biopsies. Their study is sometimes misinterpreted as disproving the benefit of SB; on the contrary, it shows that men who have had several (three to six) negative biopsies are unlikely to have significant cancer identified using SB or any biopsy. Their report does not address SB as the first repeat biopsy. All cancers were identified in the PZ.

Complication rates for office-based SB were reported to be no higher than for sextant biopsy, consistent with previous reports showing that taking more cores has little if any impact on morbidity. However, when done under anaesthesia, the risk of urinary retention was reported to be 3.1–19%, compared to the 0–1.9% risk reported in series in the clinic with local anaesthesia alone. Retention rates of ≥ 10% were reported in five of eight series with biopsies taken from patients under i.v. sedation, general or regional anaesthesia, supporting the simpler office-based SB using periprostatic block as the preference. The results and complications of SN in peer-reviewed reports are shown in Table 1[5,6,12,16–25].

Table 1.  The results and complications of SB in peer-reviewed studies
RefRouteN ptsYieldCores prev/bxN pts 1ry bx1ry yieldN pts 2nd bxYield 2nd bxN > 2 prev bxYield > 2 prev bxCores SB, nAxSettingCI PCURPr/sepSBl
  • *

    Includes previous patient population of the investigators’ initial report; setting NR but not under anaesthetic, intrarectal lidocaine used initially, PPB for later patients;

  • does not include 3/35 with catheter placed empirically, bringing total to 25.8%, which authors attribute to transurethral cores;

  • number having 2nd vs >2 biopsies not specified;

  • ¶six-core followed by at least two eight-core biopsies;

  • §

    mean 17.4 previous cores but cores/session NR;

  • °

    1.6% complete retention, 1.6%‘difficult urination’ described as adverse event. CI PC, clinically insignificant prostate cancer; UR, urinary retention; Pr/sep, prostatitis or sepsis; SBl, severe bleeding; IR, intrarectal; IVS, intravenous sedation; TR, transrectal; TP, transperineal; NR, not reported; NA, not applicable; OR, operating room; Ax, anaesthetic; bx, biopsy; PPB, periprostatic local anaesthetic block; gen, general; (r) rectal; (h) haematuria; (n), median; {n}, mean.

[19]*TR650NRNRNRNRNRNRNRNRNRIR and/or  PPBClinicNRNRNRNR
[18]TR30331.3NR18839.37825.6 37 2.721PPBClinicNR 1.90.60.3 (r)
[17]TR 11629mixed0NA7033 4620–2420–24 {22.8}PPBClinic 0 000.9 (r)
[16]TR161418+0NA 0NA 1614118+IVSClinic15.6 1.21.2NR
[24]TR13944.6NA13944.6 0NA  0NA24PPBClinic15.8 002.2 (r)
[25]TR 352060NA 0NA 352014–28 (21)genOR 017.100
[6]TR2243460NA11236 1124214–45 {23}genOR14.3 4.50.45% (h)
[5]TR 573060NA5730NRNR{22.5}IVSClinic 710.501.7 (r)
[23]TR 3713.5mixed0NA 0NA 3713.532–38genORNR1900
[12]TP21037NR§0NA40NR170NR{21}genOR 0 1100
[20]TP 603880NA 0NA 6038{24}genOR  3.301.6 (h)
[21]TP12822.780NA7018.6 5827.622spinalORNR 3.1°0.80
[22]TP18038(12)0NA9344 9332(41)genORNR10NRNR

SB AS AN INITIAL BIOPSY STRATEGY

By contrast with the evidence that cancer detection is increased with SB as a repeat biopsy strategy, its use as an initial biopsy appears to be limited. In 139 patients undergoing initial biopsy, 24 cores yielded no higher detection rate than 10 cores (44.6% vs 51.7%, P > 0.9). Analysis by PSA level failed to show any benefit to SB for any degree of PSA elevation. Men with a PSA level of 2.5–9.9 ng/mL were found to have cancer in 53 of 122 (43.4%) of SBs and 26 of 58 (44.8%) 10-core biopsies [24].

In the French report of 188 patients having a 21-core strategy as their initial biopsy, the detection rate was 39.3%, or similar to 10–12-core biopsy schemes [19]. Both these reports are consistent with the ≈ 40% detection rate for all schemes using ≥ 12 cores for the initial diagnosis, so it appears that SB offers no advantage when used as an initial biopsy strategy.

MANAGEMENT AFTER A NEGATIVE SB

Several groups showed that more than one repeat routine biopsy is unlikely to identify significant cancer, presumably because cancers in the areas routinely biopsied are usually identified during the first or second biopsies. By contrast, the above studies show that SB is still often positive after several routine negative biopsies. The findings of prostatic intraepithelial neoplasia or atypical small acinar proliferation must be addressed separately.

Nevertheless, highly selected patients are occasionally identified with cancer after a negative SB. de la Taille et al.[18] found cancer in 14.8% of 27 men whose PSA level increased substantially after SB, although they did not report if the first SB was an initial or repeat biopsy. Pryor and Schellhammer [25] found cancer in five of seven highly selected patients. We have rarely identified cancer during the follow-up; the few patients with cancer have uniformly had clinically insignificant disease and not one case to date has been identified as metastatic. Therefore, the threshold to repeat SB appears to be high, and is usually only triggered by the diagnosis of atypical small acinar proliferation or a substantial increase in PSA level.

THE RISK OF CLINICALLY INSIGNIFICANT CANCER

After random sextant biopsy became common, concern that it would primarily identify clinically insignificant cancer was shown to be unfounded [26]. The CAPSure database shows that taking more cores improves cancer detection, and does not appear to increase the risk of detecting clinically insignificant cancer [27].

Patients with low-risk cancer will theoretically be less likely to be treated with surgical intervention, so prostatectomy studies might underestimate the risk of clinically insignificant cancer. Mindful of this, such studies have found that this risk is ≥ 16% with routine biopsy [28]. There is no universally accepted definition of ‘clinically insignificant’, but Carter et al.[29] concluded that tumours were ‘insignificant’ (confined tumour <0.2 cm3 with a Gleason score of <7) in 17% and ‘minimal’ (confined tumour 0.2–<0.5 cm3 with a Gleason score of <7) in 12% of 240 men who had prostatectomy, suggesting that over a quarter of men might be ‘over-diagnosed’ using traditional biopsy.

Drawing the same conclusion for cancer identified during SB is not automatically intuitive, because the cancers might be different than those identified using traditional templates. The original SB studies had conflicting findings; Borboroglu et al.[5] found that 12 of 13 patients after prostatectomy had clinically significant cancer. However, Stewart et al.[6] found that the risk of incidental cancer increased from 11.1% to 15.4% to 22.2% in patients with one, two and three or more previous sextant biopsies, respectively; 30.6% of the tumours were <0.5 mL. Of four patients who had a prostatectomy in the series by Rabets et al.[17], all were deemed clinically significant, with volumes of >0.5 mL, and three of the four were Gleason 7. Of patients who had prostatectomy after perineal SB, all had tumours of >0.5 mL and 90% had stage pT2 disease [13]. In the small series reported by Fleshner and Klotz [23], the one patient who had had several (three to six) previous negative biopsies, and had radical prostatectomy, had extensive Gleason 7 cancer and multiple positive margins.

Clinically insignificant cancer was identified in three of 19 patients (16%) undergoing prostatectomy after initial SB [25], which is comparable to the risk during sextant biopsy.

The concern of over-detection must be weighed against the risk of missing clinically significant malignancy. Balancing these risks presumably involves the concept espoused by Carroll [30], i.e. ‘unlinking detection and treatment, as they are separate processes.’ Many men with low-volume, low-grade disease might be managed expectantly, and treatment can be safely avoided or delayed until the time that clinical indicators such as sequential PSA levels, repeat biopsy, or physical examination indicate the need for delayed intervention [31].

SB AS PART OF AN ACTIVE SURVEILLANCE PROTOCOL

Although initial efforts focused on cancer detection, SB for quantification and qualification of prostate cancer has also been the subject of investigation. Epstein et al.[28] took SBs ex vivo, using a plastic template (mean 44 cores), in 103 radical prostatectomy specimens from patients that they determined would have been candidates for active surveillance based on favourable pathological features. This scheme predicted insignificant cancer if the Gleason score was <7 with three or fewer positive cores, or if the maximum length of cancer in one core was <4.5 mm and the total for all cores was <5.5 mm. These criteria for clinically significant cancer had a sensitivity of 71.9% and specificity of 95.5%. An alternative considering only every other core was assessed. Half as many cores (mean 22) predicted clinically insignificant cancer if the Gleason score was <7 and the total cancer was <2.75 mm, or if the total in all cores was <9.5 mm. This approach yielded the same sensitivity but an increased specificity of 97.1%, suggesting the 22-core template might be preferred to a more extensive scheme. This notably approximates the number used by most of the referenced groups in clinical practice.

SB under periprostatic block in 35 patients with microfocal cancer [32] was reported to determine the patients that the authors felt could be observed. The authors proceeded to treatment in the 70% with multiple positive cores on SB after initial biopsy showing one microfocus. We have used a similar protocol to follow all patients on active surveillance since 2002. No patient has developed metastatic disease, but the follow-up is presently ≤ 5 years. To date, in almost 100 patients, an increase in grade or volume has led to the recommendation to pursue curative therapy in only a third of cases (Abouassaly R, Jones JS, in preparation).

CONCLUSION

The complications and risk of diagnosing clinically insignificant cancer using SB after a previous negative biopsy are reported to be no higher than with routine sextant or extended-core biopsy, unless general or regional anaesthesia is used, whereas the detection of clinically significant cancer is higher. Although initial investigators used regional or general anaesthesia, the periprostatic block has allowed several authors to now report its routine use in the office setting. This appears to overcome the increased risk of urinary retention related to systemic anaesthesia. SB does not seem to provide additional value during the initial biopsy, so should be reserved for repeat biopsy. The role and appropriate number of cores for SB continue to be defined, but a threshold of 20 cores, with emphasis on the lateral areas and apex, is supported by published reports.

ACKNOWLEDGEMENTS

I thank Drs Brian J. Moran and Jochen Walz for providing additional information on their work to strengthen this review.

CONFLICT OF INTEREST

None declared. Source of funding: Glickman Urological Institute, Cleveland Clinic.

Ancillary