Les biopsies systématiques sont-elles nécessaires dans le même temps que les biopsies ciblées chez les patients ayant des biopsies de prostate initialement négatives ?

25 janvier 2018

Auteurs : S. Albisinni, F. Aoun, A. Noel, E. El Rassy, M. Lemort, M. Paesmans, R. van Velthoven, T. Roumeguère, A. Peltier
Référence : Prog Urol, 2018, 1, 28, 18-24




 




Introduction


Prostate cancer (PCa) is the most frequent male non-cutaneous malignancy across Europe [1]. After multiple years of overly aggressive treatment even in cases of low-risk, indolent cancer, today the urologic community is driving its attention to the identification and treatment of clinically significant PCa, advising to reduce detection and practice less morbid therapeutic options for indolent prostate PCa, in the light of the results of recent large, prospective studies [2]. In this setting, multiparametric MRI (mp-MRI) has emerged as a promising tool to discriminate indolent from clinically significant PCa, and recent trials [3] have reported intriguing findings which could reduce the number of biopsies performed across urologic departments, as well as its associated morbidity.


MRI-guided targeted prostate biopsies have emerged as a promising technique to adequately sample the index lesion detected on mp-MRI, identifying more men with significant disease as opposed to systematic biopsies. Several studies have demonstrated more homogeneous mapping and precise targeting of the index lesion in contemporary series [4, 5]. Moreover, MRI-targeted biopsies allow a more accurate diagnosis of anterior and apical tumors [6]. As such, in patients in whom initial prostate biopsies yielded negative results and presenting nonetheless a high suspicion for neoplasia, current guidelines advise to perform mp-MRI and MRI-guided biopsies [7]. However, in this scenario, it remains poorly understood whether men should undergo purely MRI-targeted biopsies, or if a novel systematic sampling of the prostatic gland should be performed along with MRI-guided targeted biopsies, in order to obtain the maximal detection of clinically significant PCa. Supporter of repeating systematic biopsies advocate the potential added value of such specimens in order to avoid missing clinically significant PCa [8]; on the other hand, some investigators did not find any added value of reperforming standard sampling of the prostate, while this could potentially increase patient discomfort, bleeding, infection rate and eventually lead to indolent cancer overdiagnosis [9]. Aim of the current study was to explore such clinical dilemma in a prospective biopsy cohort at Jules-Bordet Institute, where MRI-targeted biopsies have now been implemented for over 5 years.


Materials and methods


Study population


Between 2012 and 2016, we prospectively enrolled all patients undergoing MRI-guided targeted prostate biopsies in the Jules-Bordet Institute. The study was approved by the ethics committee of our institute (reference number CE2548). The entire cohort includes 433 men. After excluding 192 biopsy-naïve patients, 153 men with known PCa and 14 patients with no identifiable lesion on MRI, we selected 74 men in whom MRI-targeted biopsy was performed due to prior negative biopsy and persistent suspicion of PCa. All the patients included in the present study had only one negative biopsy before inclusion. Patients had undergone a mp-MRI 3.0-T MRI with 4 sequences — triplanar T2-weighted, dynamic contrast-enhanced, diffusion-weighted imaging, and MR spectroscopy either prior to the initial set of biopsies of at least 8 weeks after. Images were reviewed by our central radiologist (ML), who proceeded to the delimitation of the region of interest (ROI), and described the lesion(s) using the PiRADS v2.0 score [10].


Biopsy technique


MRI-guided targeted biopsies were performed by a single operator (AP), performing>100 targeted biopsies/year and with>20 years of experience in prostate biopsy[11]. Patients received fluoroquinolone antibiotic prophylaxis following current guidelines. A transrectal, 3D elastic fusion approach was used to perform the biopsies via the Koelis Urostation [11]. First, an elastic fusion of MRI and ultrasound volumes is performed. Then, with the ROI hidden and not visualized, a 12-14-core systematic scheme is applied, taking 2 biopsies at the base, 2 at the mid and 2 at the apex for each prostatic lobe. Moreover, 2 biopsies are performed in the transitional zone. Once the systematic biopsies are performed, the ROI is finally displayed, and 2 to 4 targeted biopsies are taken in the ROI. This approach was implemented to reduce observer bias, as the operator is less aware of the localization of the ROI while performing the standard, systematic biopsies.


Specimens were labeled separately, fixed in paraffine, and sections were cut at 3μm intervals. If cancer was detected, cancer length was described in mm and histologic grading was accorded following Gleason score by a dedicated uro-pathologist. Clinically significant PCa was defined as any cancer Gleason≥7 and/or cancer length≥6mm on a single biopsy [12]. This threshold allows to detected over 95% of PCa presenting a volume>0.5ml [12].


Statistical analysis


In the present study, each patient was used as its own control, in order to limit selection bias. As such, for each patient we identified systematic biopsies (TRUS B), MRI-guided fusion targeted biopsies (Targeted B) and the combination of the two techniques (TRUS B+Targeted B). McNemar's test was used to compare categoric variables, such as cancer detection rate, number of clinically significant PCa as well as indolent PCa. Univariate logistic regressions were constructed to assess risk factors for the detection of clinically significant PCa. Variables explored included age (continuous), PSA (continuous, log transformed), DRE (categoric), prostate volume (continuous), number of previous biopsies (categoric), high-grade prostatic intraepithelial neoplasia (HG PIN) on previous biopsy (categoric) and a PiRADS score of 4-5 on mp-MRI (categoric). Associations with P -value less than 0.05 were considered statistically significant. All statistical analyses were performed using Stata 13.1 (StataCorp, College Station, TX).


Results


Overall characteristics of the cohort are illustrated in Table 1. Global PCa detection rate in this repeat biopsy cohort was 57% (42/74), of which 69% (29/42) were clinically significant and 31% (13/42) were indolent PCas (Table 2). Detection rate for PCa and for clinically significant PCa was higher in the Combination (TRUS B+Targeted B) group as opposed to the TRUS B group alone (Figure 1). In particular, our combined approach detected 11% (29/42 vs. 21/42) more clinically significant PCa, with a statistically significant difference (P =0.013), as well as an increased overall cancer detection, without increasing the number of indolent PCas detected.


Figure 1
Figure 1. 

Prostate cancer detection rate of systematic TRUS and combination biopsies (TRUS B+Targeted B).




When comparing only biopsy specimens obtained via a standard systematic approach (TRUS B) to those obtained only with a MRI-targeted strategy (targeted B), we did not find any significant difference between the two techniques (Figure 2). Although targeted biopsies did yield an absolute higher number of clinically significant PCa diagnoses (25 vs. 21), this difference was non-significant (P =0.34).


Figure 2
Figure 2. 

Prostate cancer detection rate of systematic TRUS and targeted-only biopsies (Targeted B).




Finally we evaluated the detection difference between a pure, MRI-targeted strategy (Targeted B), and our combination technique (targeted B+TRUS B), in the effort to explore whether a novel set of systematic biopsies should be performed at the moment of repeat targeted biopsy. Indeed, a combination approach was significantly superior to a targeted-only in overall PCa detection (+16.7% overall detection rate, P =0.007) (Figure 3). Although differences in clinically significant PCa detection were statistically non-significant (P =0.13), a combination approach did allow to detect 7 extra clinically significant PCas (+13.8%). The increased detection rate obtained via a combination approach was accompanied by a non-significant (P =0.25) increase in indolent cancer detection, in particular 3 extra indolent PCa (Table 3).


Figure 3
Figure 3. 

Prostate cancer detection rate of targeted-only (Targeted B) and combination biopsies (TRUS B+Targeted B).




On univariate logistic regression (Table 4), the only significant risk factor for the detection of clinically significant prostate cancer was the presence of a PiRADS 4-5 lesion on mp-MRI.


Discussion


Currently, in patients with a high suspicion of PCa after a first set of negative biopsies, guidelines advise to undergo mp-MRI, followed by systematic and targeted repeat biopsies [7]. Logically, it could seem obvious that if cancer is not detected on a first set of systematic biopsies, it is unlikely to be found on a subsequent set during which tissue is sampled from, theoretically, the same glands. Indeed, multiple studies have demonstrated the superiority of an MRI guided approach to repeat systematic biopsy, in order to enhance PCa detection and reduce morbidity [13]. One could imagine that if a patient is to undergo a second, targeted procedure, systematic biopsies could probably be avoided, as prostatic tissue has already been sampled in the first, negative, set of biopsies [8]. Although this reasoning may seem correct on a theoretical basis, we hypothesized that systematic biopsies remain vital in combination with a targeted approach. In the present study, we present our preliminary results finding that indeed a combined approach to repeat biopsy is clinically more accurate and should be pursued, as it allows a significant increase in PCa detection (+16% compared to targeted alone, P =0.007). Complete mapping of PCa, including the index lesion and indolent cancer, may have significant impact of therapeutic strategy, including active surveillance and focal therapy.


Initially, we were quite perplexed by the current findings: indeed in our department we are firm supporters of mp-MRI which we have been performing from>10 years; yet, the reported results support the use of systematic biopsies to targeted ones, which alone are not significantly superior. The reasons for such findings are probably multifactorial: first, our single operator performing all biopsies (AP) has a vast experience in prostate ultrasound and 3D prostatic reconstruction. Using the Koelis Urostation, we are performing 3D reconstructions of the prostate, possibly increasing quality and spatial distribution of the systematic biopsies within the gland. As such, even standard, systematic 12-core biopsies could be more precise compared to those performed by a urologist without such background. Indeed, 3D prostatic mapping has demonstrated its interest in the detection of clinically significant PCa [14, 15]. In fact, our detection rates are higher to those reported by investigators using a 2D targeted system [16]: Sonn et al. reported an overall cancer detection rate of 34% (vs. 56.8% in the current study) and 25% of clinically significant PCa (versus 34%). Second, patient and prostatic movements may determine modifications in the positioning of the ROI, thus reducing accuracy of targeted biopsies; Xu et al. have demonstrated the presence of a modification of up to 4.3mm even with robot-assisted biopsies [17]. However, thanks to the organ-tracking system of Koelis, these modifications are minimal and reduced to 1.2mm between the theoretical and the true target. Finally, mp-MRI may present false positive results in up to 32% cases [18], thus targeted biopsies in falsely positive areas will inevitably yield inaccurate results, rendering repeat systematic biopsies still fundamental.


Similar results have been found by other investigators. A large meta-analysis evaluating 16 studies reported that targeted biopsies had a higher detection rate of clinically significant PCa, and subanalysis confirmed that this difference was statistically significant in the repeat biopsy setting (relative sensitivity 1.62, 95% CI 1.02-2.57) [19]. In a randomized study, Panebianco et al. [20] similarly reported an improved diagnostic yield with MRI-targeted biopsies in the setting of repeat prostate biopsy. Portalez et al., in patients with a previous negative biopsy, did find more positive cores with a targeted strategy compared to random biopsies (36.3% compared with 4.9%, P <0.00001), although the study was not designed to explore such outcome [21]. Similarly, Costa et al. reported a positive biopsy in 92% of men undergoing a targeted template compared to 23% in systematic biopsy, with 77% of tumors detected exclusively in suspicious zones on mp-MRI [22]. Overall, most studies analyzing the repeat biopsy scenario yielded similar results, in that a targeted approach appears to be significantly superior for overall cancer detection and clinically significant PCa [23, 24], though the definition of the latter may vary.


However, to date, few reports have addressed the issue to whether, in addition to targeted biopsies, a set of standard 12-core systematic biopsies should also be performed. Moreover, an expert panel recently reported on the topic, recommending caution in performing solely targeted biopsies in this setting [25]. Costa et al. analyzed 32 men with negative targeted biopsies in spite of highly suspicious MRI, finding intermediate and high-risk PCa in 42% of cases [26].


Some investigators instead advocate against systematic biopsies at the time of repeat targeted procedure, given the low added value of such specimens: Mendhiratta et al. [9] published a series in which targeted-only biopsies detected 92.3% of all significant cancers of their cohort. Similarly, Salami [8] investigated the need for additional template biopsies during repeat targeted biopsy. In 140 men, they reported an overall detection rate of 65%, missing only 3.5% of cancers.


In the current study, the only risk factor for the detection of clinically significant PCa appeared to be a radiological image highly suspicious for PCa (i.e. PiRADS≥4). Current literature supports increasingly the use of mp-MRI, even in the setting of biopsy naïve patients. Recently, the results of the PROMIS trial have been reported by Ahmed et al. [3]: 576 biopsy-naïve men underwent mp-MMRI followed by TRUS and targeted biopsies. The authors demonstrated, in this biopsy naïve cohort, that mp-MRI has an elevated sensitivity for the detection of clinically significant PCa (93%). Performing a screening mp-MRI for patients before TRUS biopsy could reduce by 25% the number of useless, potentially morbid biopsies. Furthermore, the diagnostic precision of mp-MRI is enhanced at increasing percentage of Gleason 4 pattern, given its "solid" nature, and at increasing size of PCa, thus making it a reliable tool to screen for significant cancer and reduce overdiagnosis of indolent PCa.


Our study is not devoid of limitations. First, the study is based on a small sample size, thus larger studies are needed to validate the findings. Second, patients were their own controls, introducing bias. However, this reduces the differences, which may inevitably exist when comparing to randomized or matched pair groups. Third, the definition of clinically significant PCa is still a matter of debate; clearly, setting a different threshold may modify our results in one sense or the other. Ideally, all results should be compared to prostatectomy specimen, in order to be fully able to define the real presence of clinically significant prostate cancer and confirm the localization of the index tumor. As such, our results should be interpreted with caution, given the assumption that if no clinically significant PCa was found on combination biopsy, the patient did not harbor clinically significant PCa.


Conclusions


In this prospective, single center study, patients with prior negative biopsies appeared to benefit from the inclusion of systematic biopsies to targeted cores. Although targeted biopsies seem to determine a superior diagnostic yield, omitting a standardized template at the time of repeat biopsy could miss clinically significant PCas. Larger studies are needed to confirm our results.


Disclosure of interest


The authors declare that they have no competing interest.




Table 1 - General characteristics of the cohort.
Age (years) median (IQR)  65 (62-69) 
PSA (ng/ml) median (IQR)  9.27 (6.84-13.4) 
DRE   
Normal, n (%)  59 (80) 
Abnormal, n (%)  15 (20) 
Prostate volume (ml) median (IQR)  45 (36-60) 
Time from last negative biopsy (mo) median (IQR)  27 (12-48) 
Number of MRI detected lesions per patient (PiRADS>3) median (IQR)   
Histology of prior biopsy  1 (1-2) 
BPH, n (%)  37 (50) 
Chronic prostatitis, n (%)  21 (28) 
HG PIN, n (%)  13 (18) 
ASAP, n (%)  3 (4) 





Table 2 - Detection rates according to biopsy specimen analyzed.
  Overall PCa detection  Indolent PCa  Clinically significant PCa 
TRUS B only  44.6% (33/74)  16.2% (12/74)  28.4% (21/74) 
Targeted B only  47.3% (35/74)  13.5% (10/74)  33.8% (25/74) 
Combination  56.8% (42/74)  17.6% (13/74)  39.2% (29/74) 



Légende :
TRUS B: randomized TRUS biopsies only; Targeted B: MRI guided fusion targeted biopsies only; Combination: TRUS B+Targeted B.



Table 3 - Comparing detection rates across biopsy techniques.
  Detection rate, %  Difference across techniques, % (95% CI)  P  
Overall cancer (n =42)       
TRUS B vs. targeted B  78.6 vs. 83.3  4.7 (0-11.1)  0.8 
Combination vs. targeted B  100 vs. 83.3  16.7 (9-33.8)  0.007 
Clinically significant PCa (n =29)       
TRUS B vs. targeted B  72.4 vs. 86.2  13.8 (1.2-33.8)  0.39 
Combination vs. targeted B  100 vs. 86.2  13.8 (1.2-26.4)  0.13 
Indolent prostate cancer (n =13)       
TRUS B vs. targeted B  92.3 vs. 76.9  15.4 (0-35)  0.79 
Combination vs. targeted B  100 vs. 76.9  23.1 (0.2-46)  0.25 



Légende :
TRUS B: ranbdomized TRUS biopsies only; Targeted B: MRI guided fusion targeted biopsies only; Combination: TRUS B+Targeted B.



Table 4 - Univariate logistic regression analyzing risk factors for clinically significant prostate cancer.
  Univariate 
  OR  95% CI  P  
Age  1.05  0.97-1.13  0.27 
PSA  1.04  0.96-1.12  0.36 
DRE  2.82  0.89-9.0  0.08 
Prostate volume  0.98  0.96-1.01  0.22 
HG PIN on previous Bx  0.85  0.23-3.08  0.8 
Number of previous Bx  1.23  0.47-3.24  0.67 
PiRADS 4-5  28.7  6-138.4  <0.001 




References



Center M.M., Jemal A., Lortet-Tieulent J., Ward E., Ferlay J., Brawley O., et al. International variation in prostate cancer incidence and mortality rates Eur Urol 2012 ;  61 (6) : 1079-1092 [cross-ref]
Hamdy F.C., Donovan J.L., Lane J.A., Mason M., Metcalfe C., Holding P., et al. 10-year outcomes after monitoring, surgery, or radiotherapy for localized prostate cancer N Engl J Med 2016 ;  375 (15) : 1415-1424 [cross-ref]
Ahmed H.U., El-Shater Bosaily A., Brown L.C., Gabe R., Kaplan R., Parmar M.K., et al. Diagnostic accuracy of multi-parametric MRI and TRUS biopsy in prostate cancer (PROMIS): a paired validating confirmatory study Lancet 2017 ;  389 (10071) : 815-822 [cross-ref]
Volkin D., Turkbey B., Hoang A.N., Rais-Bahrami S., Yerram N., Walton-Diaz A., et al. Multiparametric magnetic resonance imaging (MRI) and subsequent MRI/ultrasonography fusion-guided biopsy increase the detection of anteriorly located prostate cancers BJU Int 2014 ;  114 (6b) : E43-E49
Cerantola Y., Haberer E., Torres J., Alameldin M., Aronson S., Levental M., et al. Accuracy of cognitive MRI-targeted biopsy in hitting prostate cancer-positive regions of interest World J Urol 2016 ;  34 (1) : 75-82 [cross-ref]
Kim M., Choi S.-K., Park M., Shim M., Song C., Jeong I.G., et al. Characteristics of anteriorly located prostate cancer and the usefulness of multiparametric magnetic resonance imaging for diagnosis J Urol 2016 ;  196 (2) : 367-373 [cross-ref]
Mottet N., Bellmunt J., Bolla M., Briers E., Cumberbatch M.G., De Santis M., et al. EAU-ESTRO-SIOG guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent Eur Urol 2017 ;  71 (4) : 618-629 [cross-ref]
Salami S.S., Ben-Levi E., Yaskiv O., Ryniker L., Turkbey B., Kavoussi L.R., et al. In patients with a previous negative prostate biopsy and a suspicious lesion on magnetic resonance imaging, is a 12-core biopsy still necessary in addition to a targeted biopsy? BJU Int 2015 ;  115 (4) : 562-570 [cross-ref]
Mendhiratta N., Meng X., Rosenkrantz A.B., Wysock J.S., Fenstermaker M., Huang R., et al. Prebiopsy MRI and MRI-ultrasound fusion-targeted prostate biopsy in men with previous negative biopsies: impact on repeat biopsy strategies Urology 2015 ;  86 (6) : 1192-1198 [inter-ref]
Barentsz J.O., Weinreb J.C., Verma S., Thoeny H.C., Tempany C.M., Shtern F., et al. Synopsis of the PI-RADS v2 guidelines for multiparametric prostate magnetic resonance imaging and recommendations for use Eur Urol 2016 ;  69 (1) : 41-49 [cross-ref]
Peltier A., Aoun F., Lemort M., Kwizera F., Paesmans M., Van Velthoven R. MRI-targeted biopsies versus systematic transrectal ultrasound guided biopsies for the diagnosis of localized prostate cancer in biopsy naïve men Biomed Res Int 2015 ;  2015 : 571708
Ahmed H.U., Hu Y., Carter T., Arumainayagam N., Lecornet E., Freeman A., et al. Characterizing clinically significant prostate cancer using template prostate mapping biopsy J Urol 2011 ;  186 (2) : 458-464 [cross-ref]
Siddiqui M.M., Rais-Bahrami S., Turkbey B., George A.K., Rothwax J., Shakir N., et al. Comparison of MR/ultrasound fusion-guided biopsy with ultrasound-guided biopsy for the diagnosis of prostate cancer JAMA 2015 ;  313 (4) : 390-397 [cross-ref]
Klein J., De Gorski A., Iselin C. [3D ultrasound guidance for prostate biopsy with MRI-image fusion] Rev Med Suisse 2013 ;  9 (409) : 2275-2278
Peltier A., Aoun F., Albisinni S., Marcelis Q., Ledinh D., Paesmans M., et al. Results of a comparative analysis of magnetic resonance imaging-targeted versus three-dimensional transrectal ultrasound prostate biopsies: size does matter Scand J Urol 2016 ;  50 (3) : 144-148 [cross-ref]
Sonn G.A., Chang E., Natarajan S., Margolis D.J., Macairan M., Lieu P., et al. Value of targeted prostate biopsy using magnetic resonance-ultrasound fusion in men with prior negative biopsy and elevated prostate-specific antigen Eur Urol 2014 ;  65 (4) : 809-815 [cross-ref]
Xu H., Lasso A., Guion P., Krieger A., Kaushal A., Singh A.K., et al. Accuracy analysis in MRI-guided robotic prostate biopsy Int J Comput Assist Radiol Surg 2013 ;  8 (6) : 937-944 [cross-ref]
Bratan F., Niaf E., Melodelima C., Chesnais A.L., Souchon R., Mège-Lechevallier F., et al. Influence of imaging and histological factors on prostate cancer detection and localisation on multiparametric MRI: a prospective study Eur Radiol 2013 ;  23 (7) : 2019-2029 [cross-ref]
Schoots I.G., Roobol M.J., Nieboer D., Bangma C.H., Steyerberg E.W., Hunink M.G.M. Magnetic resonance imaging-targeted biopsy may enhance the diagnostic accuracy of significant prostate cancer detection compared to standard transrectal ultrasound-guided biopsy: a systematic review and meta-analysis Eur Urol 2015 ;  68 (3) : 438-450 [cross-ref]
Panebianco V., Barchetti F., Sciarra A., Ciardi A., Indino E.L., Papalia R., et al. Multiparametric magnetic resonance imaging vs. standard care in men being evaluated for prostate cancer: a randomized study Urol Oncol 2015 ;  33 (1) : 17e1-17e7
Portalez D., Mozer P., Cornud F., Renard-Penna R., Misrai V., Thoulouzan M., et al. Validation of the European Society of Urogenital Radiology scoring system for prostate cancer diagnosis on multiparametric magnetic resonance imaging in a cohort of repeat biopsy patients Eur Urol 2012 ;  62 (6) : 986-996 [cross-ref]
Costa D.N., Bloch B.N., Yao D.F., Sanda M.G., Ngo L., Genega E.M., et al. Diagnosis of relevant prostate cancer using supplementary cores from magnetic resonance imaging-prompted areas following multiple failed biopsies Magn Reson Imaging 2013 ;  31 (6) : 947-952 [cross-ref]
Durmus T., Stephan C., Grigoryev M., Diederichs G., Saleh M., Slowinski T., et al. [Detection of prostate cancer by real-time MR/ultrasound fusion-guided biopsy: 3T MRI and state of the art sonography] ROFO Fortschr Geb Rontgenstr Nuklearmed 2013 ;  185 (5) : 428-433
Puech P., Rouvière O., Renard-Penna R., Villers A., Devos P., Colombel M., et al. Prostate cancer diagnosis: multiparametric MR-targeted biopsy with cognitive and transrectal US-MR fusion guidance versus systematic biopsy — prospective multicenter study Radiology 2013 ;  268 (2) : 461-469 [cross-ref]
Rosenkrantz A.B., Verma S., Choyke P., Eberhardt S.C., Eggener S.E., Gaitonde K., et al. Prostate magnetic resonance imaging and magnetic resonance imaging targeted biopsy in patients with a prior negative biopsy: a consensus statement by AUA and SAR J Urol 2016 ;  196 (6) : 1613-1618 [cross-ref]
Costa D.N., Kay F.U., Pedrosa I., Kolski L., Lotan Y., Roehrborn C.G., et al. An initial negative round of targeted biopsies in men with highly suspicious multiparametric magnetic resonance findings does not exclude clinically significant prostate cancer-Preliminary experience Urol Oncol 2017 ;  35 (4) : 14910.1016/j.urolonc.2016.11.006[e15-149.e21 Epub 2016 Dec 9].






© 2017 
Elsevier Masson SAS. Tous droits réservés.