Abstract
Aim: To evaluate the accuracy of multi-parametric pelvic magnetic resonance imaging (mpMRI) in diagnosing prostate cancer (PCa) in men with initial biopsy microfocal cancer. Patients and Methods: From January 2012 to July 2014, 40 patients before undergoing repeat transperineal saturation prostate biopsy (SPBx; median, 28 cores) for the presence of a microfocal PCa were submitted to 3.0-Tesla mpMRI. Results: A T1c clinical stage PCa was found in 23 (57.5%) patients submitted to SPBx; mpMRI was positive in 16/40 (40%) cases and in 11 of them a clinically significant PCa was found. On the contrary, the 12 men with negative mpMRI had a quantitative histology suitable for clinically insignificant cancer. Diagnostic accuracy of mpMRI in diagnosing significant PCa was equal to 100%. Conclusion: Multi-parametric pMRI should be suggested in the re-evaluation of microfocal cancer as a selection approach of patients at risk for clinically significant PCa.
The widespread use of serum prostate-specific antigen (PSA) testing, associated with lower PSA threshold and extended biopsy protocols, has led to a marked increase of small, low grade prostate cancer (PCa) that cannot threaten patient's survival. The preoperative prediction (1, 2) of an insignificant PCa (organ confined, less than 0.5 ml cancer without Gleason grade 4 or 5 disease) remains a difficult task because PCa is a multifocal, heterogeneous disease and the employed prostate biopsy technique provides a limited amount of tissue that not necessarily reflects the biology of the disease; therefore, as a consequence, the potential aggressiveness of a small lesion can be underestimated. In the last years, the incidence of biopsy-proven microfocal PCa, characterized by a single positive core (5% or less) of Gleason score (GS) 6 (3-6) has significantly increased with an estimated risk to harbour a clinically significant PCa in about 30% of the cases (6). In the population-based screening study at the Rotterdam section of European Randomized Study on Screening of Prostate Cancer the proportion of focal cancers during the second screening after four years increased from 16% to 29% of all detected cancers (7). On the other hand, to reduce the risk of overtreatment active surveillance protocols (8) have been suggested, but still today the estimated understaging is equal to 30% of the cases. In this light, multiparametric magnetic resonance imaging (mMRI) has demonstrated a good sensitivity to detect only clinically significant PCa missing cancers at risk for indolent disease, especially, in patients submitted to repeat biopsy.
The accuracy of multi-parametric pelvic magnetic resonance imaging (mpMRI) in diagnosing significant PCa in men with initial microfocal biopsy cancer has been prospectively evaluated.
Patients and Methods
From January 2012 to July 2014 among 795 men submitted to prostate biopsy. 40 (5%) aged between 48 and 76 years (median, 62.5 years) underwent repeat saturation biopsy (SPBx; median, 28 cores; range, 24-34 cores) for the presence of a microfocal (Figure 1a) PCa (a single positive core of Gleason score of 6 with a greatest percentage of cancer <5%) diagnosed by extended prostate biopsy (median 18 cores) 6 months before (median; range, 3-9 months). In all cases digital rectal examination was negative and median PSA was equal to 8.9 ng/ml (range, 4.2-15 ng/ml). SPBx was accomplished in a transperineal way with a tru-cut 18 G needle through a GE Logiq 500 PRO ecograph supplied with a biplanar transrectal probe (5-6.5 MHz) under sedation and antibiotic prophylaxis (9).
All patients, who provided a written informed consent, underwent mpMRI 3-10 days before undergoing the SPBx. All examinations were performed using a 3.0 Tesla scanner, (ACHIEVA 3T; Philips Healthcare Best, Amsterdam, the Netherlands) equipped with surface 16 channels phased-array coil placed around the pelvic area with the patient in supine position; multiplanar turbo spin-echo T2-weighted (T2W), axial diffusion weighted imaging (DWI), axial dynamic contrast enhanced (DCE) and spectroscopy were performed for each patient. The mpMRI lesions characterized by a prostate imaging reporting and data system PI-RADS score of 4 and 5 were considered at high risk for the presence of PCa (10). In detail, the criteria (11) for a positive lesion on T2W were the presence of a circumscribed, low signal intensity lesion (hypointense); a positive lesion on DCE was characterized by the presence of foci showing early and intense enhancement and rapid washout after power injection (3.0 ml/s) of gadobutrol 0.1 ml/kg (Gadovist®; Bayer Schering Pharma, Berlin, Germany) followed by a 15 ml saline flush. A positive lesion on spectroscopy was any area where the choline to citrate ratio was 3 or more standard deviations above the mean healthy value. Two radiologists (AF, GP) blinded to pre-imaging clinical parameters evaluated the MRI data separately and independently.
To ensure that histopatological findings matched with mpMRI images (cognitive fusion) the assessment of radiological images and SPBx scheme were performed dividing the prostate into 14 regions as previously reported (12). In the presence of mpMRI lesions suspicious for cancer, 4 targeted TRUS guided-biopsies -in addition to standard SPBx-were performed. A probability (p) level of less than 0.05 was considered statistically significant.
Results
A T1c clinical stage PCa was found in 23 (57.5%) and 11 (27.5%) patients submitted to SPBx and mpMRI targeted-biopsy, respectively; in the remaining cases a normal parenchyma (absence of cancer) was diagnosed. Multi-parametric pMRI was positive (Figure 1b) in 16/40 (40%) patients, in detail, in 11/23 (47.5%) and in 5/15 (33.3%) men with PCa and normal parenchyma, respectively. A total of 64 targeted-biopsies were performed; the median diameter of the suspicious mpMRI lesions was equal to 12 (range=8-15) mm vs. 6 (4-10) mm in the presence vs. absence of PCa (p=0.14), respectively; mpMRI targeted-biopsy found 9 and 2 cancers of the peripheric and anterior zone of the gland, respectively. In 15/23 patients the PCa was found in the same prostatic zone of primary microfocal disease; moreover, mpMRI targeted biopsy diagnosed 2 cancers of the anterior zone that were missed by SPBx.
Clinical parameters, mpMRI and histological biopsy findings of the 23 men with PCa are listed in Table I. All 11 men with PCa and positive mpMRI had a clinically significant cancer (1, 2). On the contrary, in the 12/23 patients with negative mpMRI quantitative histology was suitable for clinically insignificant cancer (median number of positive cores, GPC and GS equal to 1.5, 30% and 6, respectively) or a new microfocus of PCa was found for the second time (6 cases).
Fifteen (65.2%) out of 23 men underwent retropubic radical prostatectomy (RRP): 11 had a positive mpMRI targeted-biopsy combined with clinically significant PCa; moreover, in the 3/4 (75%) patients with negative mpMRI an indolent cancer was found (Table II). The remaining 25 patients were enrolled in an active surveillance protocol.
Clinical, biopsy and mpMRI findings in the 23 patients with prostate cancer (PCa) diagnosed at repeat saturation biopsy.
The diagnostic accuracy of mpMRI targeted-biopsy in diagnosing significant PCa was equal to 100%.
Discussion
Microfocal cancer on prostate biopsy is defined by the presence of a single positive core with a minimal cancer involvement in terms of core length: 0.5-2 mm (3, 4) or in percent of cancer <5% (5, 6). In screen-detected PCa, the overall incidence of PCa that fulfilled the Epstein criteria for indolent cancer in surgical specimen varies from 5.8% to 14% (13). In the sub-group of patients with a preoperative diagnosis of focal cancer this incidence raises from 22-33% up to 60-70% of the specimens, including a 0.8% possibility not to find the cancer at all (vanishing cancer phenomenon) (14-18). Tumor volume and Gleason grading are considered the major determinants of the biological behaviour and clinical outcome of PCa; moreover, the correlation between biopsy findings (19, 20) and overall tumor burden is rather poor and even a single focus of low grade PCa in a biopsy specimen, per se, does not predict the pathologic stage of the disease. Thong et al. (18) reported that 42/192 (22%) patients with biopsy-proven microfocal disease were upgraded and/or upstaged after surgery. In a series of 55 patients with a microfocus of PCa submitted to radical retropubic prostatectomy (RRP) we previously reported in the 27.3 and 14.5% of the cases extraprostatic extent and positive surgical margins, respectively (6).
In the last years, to reduce the risk of overtreatment in the presence of histological biopsy findings predictive of clinically-indolent PCa active surveillance protocols have been introduced in clinical practice; in this respect, mpMRI demonstrated high accuracy in detecting tumors larger than 0.5 ml (21, 22) and delineating clinically significant PCa demonstrating a sensitivity and negative predictive value equal to 85-93 vs. 84-100% (23-26), respectively. Multiparametric MRI has been suggested in the reevaluation of patients with minimal biopsy PCa enrolled in active surveillance protocols; moreover, Delonchamps et al. (27) reported in 391 patients with suspected localized PCa a decreased detection of microfocal cancer performing mpMRI targeted-biopsy. Recently, Ouzzane et al. (28) stated that patients with nonsuspicious mMRI represent a special very low-risk group of men with either no disease or clinically insignificant disease, allowing them to be managed conservatively.
A 67-year-old man with an initial biopsy showing a microfocus of prostate cancer (a) submitted to multi-parametric pelvic MRI (b) that demonstrated a lesion (diameter of 12 mm) suspicious for cancer of the anterior zone of the gland. MRI-targeted biopsy detected 3 cores of Gleason score 6 with a greatest percentage of cancer equal to 50%.
Multi-parametric pelvic magnetic resonance imaging (mpMRI), biopsy and pathological (pT) findings in the 15 patients who underwent radical retropubic prostatectomy.
In our series, the first to our knowledge that evaluated mpMRI accuracy in the reevaluation of microfocal PCa, mpMRI targeted-biopsy found 11/23 cancers characterized by biopsy quantitative histology predictive of clinically significant PCa confirmed in the definitive specimen. On the other hand, the remaining 12 patients with PCa and negative mpMRI were at risk for indolent disease (3/4 patients submitted to surgery had a cancer volume <0.5 ml with a GS of 6). Finally, mpMRI demonstrated a diagnostic accuracy equal to 100% in diagnosing clinically significant PCa in men with initial microfocal disease.
Some limitations and considerations of the present study deserve annotation. Firstly, we do not know the true diagnostic accuracy of mpMRI in PCa diagnosis because the detection rate for cancer was compared only in 15 (37.5%) cases with definitive specimen and in 25 (62.7%) cases with SPBx results. Secondly, we do not know if the false-positive rate (5 cases) of mpMRI was secondary to false-negative SPBx results or biased because an mpMRI imaging/ultrasound fusion-guided biopsy, theoretically more accurate, was not performed. Finally, a greater number of patients should be evaluated.
In conclusion, mpMRI should be suggested in the presence of microfocal cancer as a selection approach of patients at risk for clinically significant PCa.
- Received September 10, 2014.
- Revision received September 24, 2014.
- Accepted September 26, 2014.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
