Paradigm Shift in Prostate Cancer Diagnosis: Pre-Biopsy Prostate Magnetic Resonance Imaging and Targeted Biopsy.

With regard to the indolent clinical characteristics of prostate cancer (PCa), the more selective detection of clinically significant PCa (CSC) has been emphasized in its diagnosis and management. Magnetic resonance imaging (MRI) has advanced technically, and recent international cooperation has provided a standardized imaging and reporting system for prostate MRI. Accordingly, prostate MRI has recently been investigated and utilized as a triage tool before biopsy to guide tissue sampling to increase the detection rate of CSC beyond the staging tool for patients in whom PCa was already confirmed on conventional systematic biopsy. Radiologists must understand the current paradigm shift for better PCa diagnosis and management. This article reviewed the recent literature, demonstrating the diagnostic value of pre-biopsy prostate MRI with targeted biopsy and discussed unsolved issues regarding the paradigm shift in the diagnosis of PCa.

INTRODUCTION

Prostate cancer (PCa) is the second-most prevalent cancer and the fifth leading cause of cancer-specific mortality in male, worldwide [1]. In the Republic of Korea, the incidence of PCa has rapidly increased over the last couple of decades, and it is the fourth-most prevalent cancer and seventh leading cause of cancer-specific mortality in male [2]. Considering the global prevalence in developed countries and the aging tendency in Korean demographics, PCa is expected to become more prevalent in the near future. In PCa epidemiology, it is remarkable that the cancer-specific mortality rate is relatively low compared with the incidence rate. This is because PCa predominantly occurs in elderly male who are at high risk of developing aging-related comorbidities. Furthermore, the introduction of serum prostate-specific antigen (PSA)-based screening tests has contributed to an increase in low-risk PCa detection associated with excellent prognosis [3].

PCa is clinically suspected based on the results of digital rectal examination (DRE) and/or an elevated serum PSA level. Although abnormal DRE results are an indication for biopsy, PSA level is a more sensitive indicator of PCa than either DRE or transrectal ultrasound (TRUS) results [4]. The standard method for pathological diagnosis is TRUS-guided 10–12-core systematic biopsy, in male with clinical suspicion of PCa (Fig. 1A). However, conventional screening systems based on PSA levels and diagnostic strategies using systematic biopsy have several limitations. First, some PCa can be missed in relation to tumor size and location because TRUS is usually confined to the role of anatomical guidance, owing to its low capability of detecting PCa [5]. Second, the sampling power may decrease as prostate volume increases in association with the underlying benign prostatic hypertrophy. Finally, overdiagnosis of low-risk PCa is an important issue because serum PSA levels are not specific to clinically significant PCa (CSC). Low-risk PCa, characterized by an early stage and a low pathologic grade, has a substantially excellent prognosis [67]. Therefore, active management can be clinically doubtful for silent PCa, especially in elderly male with a short life expectancy; this is because radical prostatectomy, which is the standard therapeutic option for localized PCa, yields morbidity [8]. The conventional diagnostic pathway has been reported to be ineffective in the selective detection of CSC, despite its diagnostic ability to detect overall PCa. Several factors, such as volume, pathologic grade, and local extent of the index lesion, should be actively managed owing to its aggressiveness and poor prognosis [9]. CSC was defined based on surgical specimen findings as follows: tumor volume ≥ 0.5 cm3, Gleason score > 6, or presence of extraprostatic extension [10]. Among the variable criteria for CSC, tumors with group 2 International Society of Urological Pathology (ISUP) grade (i.e., Gleason score 3 + 4) or higher constitute the most common and important criteria for CSC in both biopsy and prostatectomy specimens [11].

Fig. 1

Diagnostic pathway for prostate cancer.

A, B. Conventional (A) and new MRI-based diagnostic pathway (B) for PCa detection. Dashed arrow means controversial pathway with weak level of evidence. DRE = digital rectal examination, PCa = prostate cancer, PI-RADS = Prostate Imaging Reporting and Data System, PSA = prostate-specific antigen, TRUS = transrectal ultrasonography

Over the last twenty years, prostate magnetic resonance imaging (MRI) has advanced technologically and has been widely investigated in PCa detection, localization, and characterization. With increasing knowledge on prostate MRI, to compensate for the limitations of conventional diagnostic strategies, several researchers have suggested the potential of prostate MRI before biopsy in PCa diagnosis. However, heterogeneity in imaging protocols and interpretive methods has been recognized as an obstacle in utilizing prostate MRI beyond cancer staging. Recently, global collaboration has attempted to standardize the protocol and interpretation of multiparametric prostate MRI (mpMRI), and efforts have brought about promising results, such as the formation of international guidelines termed the Prostate Imaging Reporting and Data System (PI-RADS) [12]. Positive results may induce a change in PCa diagnosis using prostate MRI before biopsy to identify specific target lesions or even to determine whether to perform a biopsy (Fig. 1B). The current paradigm shift in PCa diagnosis aims at the following points: improvement in CSC detection, reduction in the number of unnecessary biopsies or biopsy cores, and prevention of the over-detection of clinically insignificant PCa.

The purpose of this review is to introduce the achievements of current investigations in association with pre-biopsy MRI and MRI-targeted biopsy, for PCa diagnosis and providing insight into the unsolved issues of this paradigm shift.

Pre-Biopsy MRI and MRI-Targeted Biopsy in Patients with Prior Negative Biopsy

Persistent or even increased PSA levels are clinically dilemmatic in patients with prior negative biopsy results. Conventionally, a repetitive systematic TRUS-guided biopsy is the only diagnostic approach. However, repetitive systematic biopsies yield a decreasing PCa detection rate at each subsequent sampling. In a previous study of 2526 patients, the cancer detection rates of serial systematic biopsies after initial biopsy were 17%, 14%, 11%, and 7%, respectively [13]. Similarly, in another study, the detection rates of PCa on the first and second biopsies in 1051 male with prior negative biopsy were 10% and 5%, respectively [14]. Therefore, previous guidelines recommend at least a single session of TRUS-guided biopsy for patients with an initial negative biopsy result. To improve the detection rate of PCa in repetitive biopsies, a study performed saturation biopsies with a markedly increased number of cores [15]. In this study, the PCa detection rate was 34% on the first repeat biopsy in male with a prior negative biopsy. However, limitation of saturation biopsy is the necessity for general anesthesia or conscious sedation, which is not mandatory in conventional biopsy. Furthermore, the increased detection rate of PCa is primarily attributed to the increased detection of clinically insignificant PCa.

Although prostate MRI has been used for staging pathologically confirmed PCa on TRUS-guided systematic biopsy, technical development and accumulated data on prostate MRI have enabled the utilization of MRI before re-biopsy, to potentially solve the limitations of both repetitive systematic biopsy and saturation biopsy. Hambrock et al. [16] reported a superior PCa detection rate of MRI with targeted biopsy compared with systematic TRUS-guided biopsy in male with a prior negative biopsy. In the study by Portalez et al. [17], more targeted cores were found to be PCa than random systematic cores (36.3% vs. 4.9%) in patients with prior negative biopsy. In another study, the positive biopsy yield was higher in MRI-prompted biopsies than in systematic samplings (92% vs. 23%), and 77% of tumors were exclusively detected in MRI-prompted zones [18]. Furthermore, the authors demonstrated that the anterior and apical regions contained most of the tumors that were missed by prior systematic TRUS-guided biopsy. In a study by Sonn et al. [19], more CSC was detected on targeted biopsy than on systematic biopsy, and the degree of suspicion on MRI was the most powerful predictor of CSC in male with prior negative biopsy. In a systematic review and meta-analysis of the abovementioned studies, MRI-targeted biopsy improved both overall PCa and CSC detection rates (relative sensitivity, 1.62 and 1.22, respectively) compared with systematic TRUS-guided biopsy in male with prior negative biopsy [20]. The added value of MRI-targeted biopsy is related to the tumor location, where contact between biopsy needles and PCa can be easily avoided during systematic biopsy. According to accumulating data, recent international guidelines recommend pre-biopsy MRI and targeted biopsy in patients with prior negative biopsy [21222324]. Therefore, pre-biopsy MRI should be considered in repeat biopsy cases if both quality-controlled prostate MRI and experienced operators are available for targeted biopsies.

Pre-Biopsy MRI and MRI-Targeted Biopsy in Biopsy-Naïve Patients

In biopsy-naïve patients with clinically suspicious PCa, there has been growing interest in adopting pre-biopsy MRI and MRI-targeted biopsy (Table 1). Panebianco et al. [25] reported the results of pre-biopsy MRI and targeted biopsy in a randomized prospective analysis of 1140 male who were initially evaluated for PCa. In their study, the proportion of male with an overall PCa diagnosis was higher in those randomized to the MRI-first strategy than in those randomized to the standard TRUS-guided biopsy. However, another prospective randomized study by Tonttila et al. [26] did not find a significant difference between the pre-biopsy MRI group and the standard TRUS-guided biopsy group among 113 biopsy-naïve patients, although the pre-biopsy MRI group showed a slightly higher detection rate for both overall PCa and CSC than the standard TRUS-guided biopsy group. Similarly, there was no significant difference in the detection rates of the two biopsy strategies, for overall PCa and CSC, in a study by Baco et al. [27]. However, the authors emphasized the utility of MRI-targeted biopsy because the majority of CSCs (87%) were detected by targeted biopsy. In a more recently published study by Porpiglia et al. [28], the diagnostic pathway using pre-biopsy MRI was stated to be superior to the standard pathway, in detecting both overall PCa and CSC. This topic was further investigated in multicenter-based studies, such as in a prospective study including 626 biopsy-naïve men by van der Leest et al. [29], where the MRI pathway (i.e., pre-biopsy MRI with MRI-targeted biopsy) showed an identical detection rate for CSC as the standard pathway (25% vs. 23%). The analysis showed that the MRI pathway enabled biopsy avoidance in 49% of the enrolled patients, at the cost of missing CSC in 4% [30]. In another prospective study including 576 male without a previous biopsy by Ahmed et al. [30] (PROMIS trial), MRI-targeted biopsy was more sensitive and less specific in detecting CSC (sensitivity, 93%; specificity, 41%) than TRUS-guided biopsy (sensitivity, 48%; specificity, 96%). Triage using MRI allowed 27% of the patients to avoid biopsy. Furthermore, a recent study including 500 biopsy-naïve male (PRECISION trial) reported similar results [31], as MRI-targeted biopsy was superior to standard TRUS-guided biopsy (38% vs. 26%), and fewer patients were diagnosed with clinically insignificant PCa in the MRI pathway than in the standard pathway (adjusted difference, -13%). Another prospective study (MRI-FIRST trial) in 251 biopsy-naïve patients demonstrated that targeted biopsy was similar to systematic biopsy and added value to systematic biopsy in detecting CSC [32]. In summary, recent data from large and high-quality prospective studies consistently demonstrated the superiority of pre-biopsy MRI with MRI-targeted biopsy over standard TRUS-guided biopsy in detecting CSC (Fig. 1B), which potentially reduced unnecessary biopsies in biopsy-naïve patients with clinical suspicion of PCa. These results may support the current paradigm shift in the diagnostic strategies for PCa.

Table 1

Comparison of Diagnostic Performance between MRI-Targeted and Standard TRUS-Guided Biopsy in the Biopsy Naïve Patient

Study (Year)	Study Design	MRI Interpretation	Population (Male)	Cancer Detection Rate (%)
				MRI-Targeted Biopsy		Standard Biopsy
				Overall PCa	CSC	Overall PCa	CSC
Panebianco et al.* (2015)[25]	Prospective	PI-RADS v1	1140	73	N/A	38	N/A
Tonttila et al.* (2016) [26]	Prospective	Likert scale	113	64	55	57	45
Baco et al.* (2016) [27]	Prospective	PI-RADS v1	175	51	44	48	49
Porpiglia et al. (2017) [28]	Prospective	PI-RADS v1	212	51	44	30	18
Kasivisvanathan et al. (2018) [31]	Prospective	PI-RADS v2	500	47	38	48	26
van der Leest et al.* (2019) [29]	Prospective	PI-RADS v2	626	39	25	48	23
Rouviere et al.* (2019) [32]	Prospective	Likert scale	251	64	3	52	30

*These studies included systematic biopsies in the MRI-targeted biopsy process. CSC = clinically significant cancer, MRI = magnetic resonance imaging, N/A = not applicable, PCa = prostate cancer, PI-RADS = Prostate Imaging Reporting and Data System, TRUS = transrectal ultrasonography

Interpretation of Pre-Biopsy MRI and Indication of MRI-Targeted Biopsy

Precise and standardized interpretation of prostate MRI is essential for utilizing MRI as a triage system in the initial assessment of patients with clinically suspected PCa. In 2014, PI-RADS was initially proposed by the European Society of Urogenital Radiology (ESUR); in 2015, it was updated to its second version by the ESUR and American Urologic Association (AUA) [1233]. In the updated version, the PI-RADS was further simplified to improve CSC detection. This system defines each category according to the probability of CSC. PI-RADS category 4 or 5 was assigned if CSC was likely or highly likely to be present, and category 3 if a lesion was equivalent to the probability of CSC. The guidelines distinctly described that biopsy should be considered for category 4 or 5, but not for category 1 or 2. For category 3, the PI-RADS ambiguously described that biopsy may or may not be appropriate, depending on non-imaging factors. This is because the PI-RADS was developed and modified based on the consensus of the expert committee; therefore, it encourages researchers to validate these guidelines.

Many studies have reported either biopsy or surgical pathological findings in each of the PI-RADS version 2 categories. A study reported that CSC detection rates on MRI-targeted biopsy were 44%–49% for category 4 lesions, and 72%–74% for category 5 lesions [34]. In this study, 11% of the category 3 lesions were CSCs. In the PRECISION trial, CSCs were identified in 12% of category 3 lesions, 60% of category 4 lesions, and 83% of category 5 lesions. In a prospective study by van der Leest et al. [29], CSCs were identified in 15%–18% of category 3 lesions, 25%–40% of category 4 lesions, and 68%–70% of category 5 lesions. In a multicenter study of 3349 male from 26 different centers, Westphalen et al. [35] reported that the PCa detection rates were 15% for category 3 lesions, 39% for category 4 lesions, and 72% for category 5 lesions; moreover, the overall PCa detection rates were 35% for PI-RADS scores greater than or equal to 3 and 49% for PI-RADS scores greater than or equal to 4.

Considering the reported data, category 4 or 5 lesions should be targeted during biopsy because of the high probability of CSC, as recommended by the PI-RADS. However, routine inclusion of category 3 lesions in targeted biopsy may still be controversial because both the number of biopsy avoidances and the detection rate of CSC can be increased by omitting biopsies for category 3 lesions. If category 3 lesions were excluded from targeted biopsy, biopsy avoidance would increase from 28% to 48%, with a higher detection rate of CSC (from 28% to 71%) in the PRECISION trial. Similarly, biopsy avoidance increased from 49% to 56%, with an increased detection rate of CSC (from 25% to 55%) in the study by van der Leest et al. [29]. These findings are associated with a relatively low rate of CSC detection in category 3 compared with that in category 4 or 5. However, the absolute number of missed CSCs would increase if targeted biopsy was omitted for category 3 lesions. Table 2 summarizes the literature reporting PCa detection rates in patients with a PI-RADS version 2 score of 3 on prostate MRI. Tan et al. [36] reported that 3 of 31 (9.7%) PI-RADS category 3 lesions were confirmed as CSCs in their analysis. In a retrospective analysis by Sheridan et al. [37], 19 of 111 (17.1%) PI-RADS category 3 lesions were CSCs on MRI-TRUS fusion biopsy. In another study, 26 of 156 patients (16.7%) with PI-RADS category 3 lesions showed CSCs on targeted biopsy [38]. Therefore, a targeted biopsy is required to increase the absolute number and sensitivity of CSC detection, even though the detection rate and specificity may decrease. Almost all recent multicenter prospective trials have included equivocal lesions (i.e., score 3 on a Likert scale or category 3 in the PI-RADS) in MRI-targeted biopsy to prevent under-diagnosis of PCa [293031323940].

Table 2

PCa Detection Rate in the Subjects with PI-RADS v2 Score 3 on MRI

Study (Year)	Study Design	MRI Scanner	Subject No.	Subject No. with Score 3 (%)	Cancer Detection Rate (%)
					Overall PCa	CSC
Tan et al. (2017) [36]	Retrospective	3T	134 lesions	31 lesions (23)	19	10
Venderink et al.* (2018) [38]	Retrospective	3T	1057 male	156 male (15)	35	17
Rosenkrantz et al. (2017) [34]	Retrospective	3T	343 lesions	79 lesions (23)	28	11
Kasivisvanathan et al. (2018) [31]	Prospective	1.5 or 3T	252 male	51 male (20)	33	12
Sheridan et al. (2018) [37]	Retrospective	3T	474 lesions	111 lesions (23)	27	17
van der Leest et al. (2019) [29]	Prospective	3T	317 male	40 male (13)	35	18
Wegelin et al. (2019) [39]	Prospective	3T	665 male	64 male (10)	25	17
Westphalen et al. (2020) [35]	Retrospective	1.5 or 3T	5082 lesions	1490 lesions (29)	30	15

*This study utilized both PI-RADS v1 and v2. CSC = clinically significant cancer, MRI = magnetic resonance imaging, PCa = prostate cancer, PI-RADS = Prostate Imaging Reporting and Data System

In the recently modified PI-RADS version 2.1, there were some changes, especially in the definitions of Categories 2 and 3 in the transition zone [41]. Rosenkrantz et al. [34] reported a relatively wide discrepancy in the frequency of scoring category 3 in PI-RADS version 2 among radiologists. This tendency resulted in discrepancies in the overall PCa and CSC detection rates for category 3 lesions. On the basis of these findings, the authors proposed several adjustments to PI-RADS version 2 for more concordant and better interpretations. Several recent studies reported slightly improved diagnostic performance of PI-RADS version 2.1 compared with the previous version, in both transitional and peripheral zone cancer [424344]. The recent change in the PI-RADS might influence the frequency of scoring category 3 in MRI interpretation; accordingly, the results of targeted biopsy might be slightly different from the reported range according to the previous PI-RADS version.

In summary, a targeted biopsy should be performed for lesions of category 3 or higher on mpMRI, according to the latest PI-RADS version. Furthermore, more data should be collected and analyzed in future PI-RADS versions to validate the effectiveness and appropriateness of MRI-targeted biopsies for category 3 lesions.

Techniques for MRI-Targeted Biopsy

There are three different technical strategies for MRI-targeted biopsies: in-bore MRI biopsy, MRI-TRUS fusion biopsy, and cognitive registration TRUS biopsy [45]. In-bore MRI biopsy is the first technique developed for targeted prostate biopsy under MRI guidance; it allows direct and precise sampling of suspicious lesions on MRI [46]. Several studies have reported PCa detection rates ranging from 15% to 52% by adopting this technique in patients with prior negative systematic biopsy [4647484950515253]. However, this technique requires specialized MRI-compatible equipment. Furthermore, systematic biopsy is not affordable because each sampling takes a substantial amount of time compared to TRUS biopsy; and is therefore costly. Conversely, cognitive registration for TRUS biopsy requires no additional software or equipment. The operator reviews the lesion and anatomy of the prostate gland on MRI and then estimates the target using real-time TRUS imaging. Both targeted and systematic biopsies can be sequentially performed under TRUS guidance. Therefore, this biopsy technique has advantages over in-bore MRI biopsies in terms of time and cost. It has been proven to be a better technique for detecting CSC than non-targeted systematic biopsy [545556575859606162]. However, a disadvantage is that the processes of cognitive fusion and visual registration are operator-dependent. Furthermore, visual discrepancies between parallel axial images and fanwise-acquired TRUS images may result in incorrect registration, especially for lesions located in the far apex or base of the prostate gland [63]. Instead of cognitive registration, MRI-TRUS fusion biopsy utilizes software-based platforms for fusion during biopsy to minimize operator errors. Therefore, MRI-TRUS fusion biopsy yields a moderate position regarding cost, procedure time, and technical availability, compared with the other biopsy techniques.

Several studies have reported the diagnostic performance of each MRI-targeted biopsy method; however, only a few studies have directly compared the results of each technique (Table 3). Initial studies compared the diagnostic performance between MRI-TRUS fusion biopsy and cognitive registration TRUS biopsy, but could not demonstrate a significant superiority of MRI-TRUS fusion biopsy in detecting PCa [646566]. Arsov et al. [67] found no significant difference between in-bore MRI biopsy and MRI-TRUS fusion biopsy in detecting both overall PCa (37% vs. 39%) and CSC (29% vs. 32%). Yaxley et al. [68] also reported no advantage of in-bore MRI biopsy over cognitive registration TRUS biopsy in detecting overall PCa and CSC. In a prospective trial by Hamid et al. [69], neither the overall PCa nor CSC detection rates were significantly different between cognitive registration and MRI-TRUS fusion techniques. However, Kaufmann et al. [70] found a significant advantage of in-bore MRI or MRI-TRUS fusion biopsy over cognitive registration for overall PCa detection, although they also failed to find superiority of any technique in detecting CSC. Similarly, a recent meta-analysis reported that in-bore MRI biopsy showed superior diagnostic performance in overall PCa detection compared with cognitive registration TRUS biopsy [71]. However, MRI-TRUS fusion biopsy showed a similar performance to in-bore MRI biopsy in detecting overall PCa and CSC, and there was no significant difference between any single biopsy technique, in detecting CSC. According to a multicenter randomized controlled trial (FUTURE trial), the detection rates of overall PCa and CSC were not significantly different among the three techniques in a repeat biopsy setting in patients with prior negative biopsies [39]. In the trial, the detection rates of PCa and CSC respectively were 55% and 33% by in-bore MRI biopsy, 49% and 34% by MRI-TRUS fusion biopsy, and 44% and 33% by cognitive registration TRUS biopsy (all p > 0.05). These results suggest that the software or equipment for MRI-TRUS fusion or in-bore MRI biopsy is not mandatory for MRI-targeted biopsy, as cognitive registration TRUS biopsy has shown a similar diagnostic performance, especially in detecting CSC. Nonetheless, we must perceive a potential bias in the results because the majority of procedures in the literature might be performed by experienced operators. The outcome of cognitive registration for TRUS biopsy can be influenced by the skill and experience of the operator. Therefore, the use of in-bore MRI and MRI-TRUS fusion biopsy is recommended if they are available, as they may enable a more standardized and uniform fusion process than cognitive registration TRUS biopsy, especially if the lesions are small or invisible on TRUS.

Table 3

Comparison of Diagnostic Performance among MRI-Targeted Biopsy Techniques

Study (Year)	Study Design	MRI Interpretation	Subject No.	Cancer Detection Rate (%)
				Cognitive Registration		MRI-TRUS Fusion		In-Bore MRI
				Overall Pca	CSC	Overall PCa	CSC	Overall PCa	CSC
Puech et al. (2013) [64]	Prospective	Likert scale	79 lesions	47	N/A	53	N/A	N/A	N/A
Wysock et al. (2014) [65]	Prospective	PI-RADS v1	172 lesions	27	15	32	20	N/A	N/A
Arsov et al. (2015) [67]	Prospective	PI-RADS v1	201 male	N/A	N/A	39	32	37	29
Lee et al. (2016) [66]	Prospective	Likert scale	396 lesions	33	23	37	21	N/A	N/A
Yaxley et al. (2017) [68]	Retrospective	PI-RADS v1	595 lesions	75	68	N/A	N/A	74	66
Kaufmann et al. (2018) [70]	Retrospective	PI-RADS v2	156 male	29	24	52	36	51	40
Hamid et al. (2019) [69]	Prospective	PI-RADS v1	129 male	66	53	69	53	N/A	N/A
Wegelin et al. (2019) [39]	Prospective	PI-RADS v2	665 male	44	33	49	34	55	33

CSC = clinically significant cancer, MRI = magnetic resonance imaging, N/A = not applicable, PCa = prostate cancer, PI-RADS = Prostate Imaging Reporting and Data System, TRUS = transrectal ultrasonography

Optimal Number of Biopsy Cores Per Lesion during MRI-Targeted Biopsy

The ideal number of targeted biopsies per lesion was not determined because of a lack of accumulated data. A recent study reported that increasing the number of biopsy core samples from one to three and three to five per target lesion increased the detection rate of CSC by 6.4% and 2.4%, respectively [72]. The authors also described that increasing the number of samples to more than five per lesion would be ineffective because it would diminish the incremental detection rate of CSC. These results can be explained in terms of the characteristics of the Gleason score for PCa grading. The final Gleason score is calculated as the sum of the first half of the score, based on the dominant morphological pattern, and the second half, based on the non-dominant pattern of the highest grade [7374]. Accordingly, undersampling can lead to underestimation of the Gleason score, and this phenomenon occurs more frequently in low-volume tumors [75]. Several studies have demonstrated that upto 60% of clinically insignificant PCa determined on biopsy changed to CSC among prostatectomy specimens [767778]. Therefore, multiple cores obtained from a target lesion may lead to the detection of CSC, which could be falsely determined on a single biopsy core, as clinically insignificant PCa. This is especially crucial in determining the eligibility for active surveillance (AS) because a Gleason score of 6 on biopsy is one of the most common and important inclusion criteria for AS [7980]. While it is reasonable to obtain multiple cores during targeted biopsies for the detection of CSC, there may be a risk of oversampling or increased detection of clinically insignificant cancer. The current consensus statement of the AUA and Society of Abdominal Radiology also recommends at least two cores per target lesion [81]. Nevertheless, the operators cannot neglect the increasing cost and potential complication rate associated with the number of biopsy cores. Therefore, the definitive number of biopsy cores should be determined by each operator during biopsy, considering individual confidence in targeting and lesion characteristics, such as size, location, and visibility during biopsy.

Necessity of Routine Systematic Biopsy in Conjunction with MRI-Targeted Biopsy

Although the PRECISION trial demonstrated the superiority of pre-biopsy MRI with or without the targeted biopsy pathway, the results regarding whether systematic biopsy should be performed in conjunction with targeted biopsy remain unclear. A recent prospective multicenter study (MRI-FIRST trial) compared CSC detection rates between targeted, systematic, and targeted systematic biopsies [32]. The CSC detection rate in the biopsy-naïve cohort was higher in the combined biopsy group (37.5% for ISUP grade ≥ 2; 21.1% for ISUP grade ≥ 3) than in either the systematic (29.9% for ISUP grade ≥ 2; 15.1% for ISUP grade ≥ 3) or targeted biopsy groups (32.3% for ISUP grade ≥ 2; 19.9% for ISUP grade ≥ 3). In a recent study by Kim et al. [82], combined targeted and systematic biopsy yielded an increased detection rate of 5.6%, compared to targeted biopsy alone. In another prospective study, the underdetection rate of CSC was higher in targeted biopsy only (9%) than in the combination of targeted and systematic biopsies (2%) [29]. In the repeated biopsy cohort, combined targeted and systematic biopsies increased the detection rate of ISUP grade ≥ 2 and ≥ 3 PCa by approximately 40% and 50%, respectively [83]. Therefore, targeted biopsy should be accompanied by systematic biopsy to increase the detection rate of CSC for the initial assessment of biopsy-naïve patients as well as repeated biopsy patients. The disadvantages of systematic biopsy include the cost and potential increase in biopsy-related complications. However, there are no results from large prospective trials that directly compare the complication rates between targeted biopsy only and targeted biopsy with systematic biopsy. In a systematic review of prostate biopsy complications, more biopsy cores were somewhat related to minor complications associated with pain, bleeding, infection, hematospermia, and erectile dysfunction, although there were substantial controversies among the results of these studies [84]. However, no previous studies have demonstrated a definite relationship between the number of biopsy cores and fatal complications. Therefore, the addition of systematic biopsy may not inflict a significant disadvantage in the management of patients, considering the diagnostic benefit.

Systematic Biopsy in Patients without Any Target Lesion on MRI

Several studies have reported that some PCa can be missed on MRI by analyzing preoperative prostate MRI with surgical pathological data or systematic TRUS biopsy. According to a systematic review and meta-analysis, the median negative predictive value (NPV) of mpMRI was 82.4% for overall PCa, and 88.1% for CSC in 48 studies (median disease prevalence rate, 50.4% and 32.9% for overall PCa and CSC, respectively) [85]. In conclusion, this study emphasized the variation in the NPV of MRI depending on the prevalence of PCa, definition of CSC, and interpretation of positive MRI findings (i.e., Likert scale or PI-RADS version 1). In a recent retrospective study that adopted PI-RADS version 2 for MRI interpretation by Kim et al. [86], cancer-negative findings on pre-biopsy MRI yielded a missed detection rate of 12.6% for PCa, including 3.9% for CSC (disease prevalence rate, 25% and 8.9% for overall PCa and CSC, respectively). In the PROMIS trial, 158 of 576 (27.4%) biopsy-naïve patients showed no target lesion on MRI, of which 17 (10.8%) had CSC on template prostate mapping biopsy (disease prevalence rate, 71% and 40% for overall PCa and CSC, respectively). Although the NPV and false-negative value of MRI are variable depending on the study design, omitting a biopsy on the basis of negative MRI findings may result in missed PCa, including CSC. This is because small-volume PCa, especially of less than 1.0 cm3, can be invisible on mpMRI [878889].

Performing fewer biopsies may have the advantage of avoiding cost- and procedure-related problems, at the expense of missing cancer. Therefore, it is difficult to determine the diagnostic risk and economic benefits of MRI for determining prostate biopsy. The following issues should be thoroughly considered to conclude whether omitting a biopsy based on negative MRI findings is clinically justifiable. First, the exact epidemiology of PCa should be understood at institutional and national levels. Furthermore, a cost-effectiveness analysis should be conducted to assess the economic benefit of omitting a biopsy and economic loss for MRI under the national medical environment. Faria et al. [90] attempted to optimize PCa diagnosis in terms of effectiveness and cost-effectiveness, according to the PROMIS trial. They concluded that the MRI-first strategy was effective and cost-effective for CSC diagnosis under the circumstances of the UK National Health Service. Second, the interpretation of MRI findings should be standardized and quality-controlled. The results of published studies may be acquired from imaging and biopsy data handled by experienced radiologists or urologists. False-negative MRI findings can result from MRI reading errors in addition to the technical limitations of mpMRI [91]. Therefore, radiologists should use the most recently updated version of PI-RADS. Furthermore, quality control may be mandatory in terms of the imaging protocols and equipment for mpMRI. Finally, patient stratification can be useful in determining candidates who are more eligible to skip biopsy based on negative MRI findings. Panebianco et al. [92] concluded that systematic biopsy should be recommended in younger patients with high or increasing PSA levels despite negative MRI findings. Omitting a biopsy can be relatively effective in low-risk patients, in whom the NPV of MRI may be high owing to the low prevalence of PCa. Meanwhile, systematic biopsy may be necessary in high-risk patients because omitting a biopsy may be at the expense of substantial CSC under-detection.

In patients with negative MRI findings, serum tumor markers can be useful as supplementary indicators for active monitoring or intervention. Washino et al. [93] reported a threshold PSA density of < 0.15 ng/mL2, which may avoid unnecessary biopsy in conjunction with a PI-RADS version 2 score of ≤ 3 on MRI. In addition to PSA density, the prostate health index outperformed PSA, free PSA, and free-to-total PSA levels in predicting PCa; accordingly, the results suggest its potential as a biomarker to triage patients with negative MRI findings [94]. However, the levels of tumor markers used to stratify patients can be affected by the methodology of MRI interpretation and the definition of negative MRI findings. Therefore, further data are needed to determine the threshold values of these novel biomarkers to stratify patients with negative MRI findings, which can be properly interpreted according to the latest version of the PI-RADS.

Quality Control for Prostate MRI and Imaging Interpretation

Accurate and standardized imaging interpretations based on quality-controlled mpMRI are preconditions for using MRI as a new diagnostic strategy for PCa. Although the PI-RADS has standardized the process of prostate imaging and interpretation, subjectivity in imaging interpretation owing to the intrinsic limitations of the system still remains. Furthermore, the system does not guarantee the quality of the acquired images or the performance of individual radiologists in practice. The ESUR and the European Urological Association Section of Urologic Imaging recently provided a consensus statement on recommendations for controlling image quality and interpretation performance [95]. Furthermore, a new scoring system called the Prostate Imaging Quality was suggested according to the PRECISION trial [96]. Although these attempts are currently incipient, the accumulation of consensus statements and guidelines for quality standards may impel preparation for national or international certifications in the near future. Radiologists need to consider not only the technical aspects of PI-RADS, but also the efforts for quality control in prostate imaging and interpretation.

CONCLUSION

Pre-biopsy MRI with subsequent targeted biopsy has added value in diagnosing CSC in both biopsy-naïve patients and those with prior negative biopsy results. The accumulated data seems to be sufficient for a paradigm shift in diagnosing PCa because recent prospective studies have consistently demonstrated the superiority of the MRI-first strategy over the conventional diagnostic pathway. Cognitive registration TRUS biopsy is the most cost-effective method for targeted biopsy without significant limitations in CSC detection rate, although in-bore MRI or MRI-TRUS fusion biopsy is recommended if available. During targeted biopsy, systematic biopsy seems to be necessary in both biopsy-naïve and repeated biopsy patients, especially in those at high risk for CSC. However, whether a systematic biopsy can be omitted in patients without a target lesion on MRI remains controversial. Risk stratification and a stepwise strategy can be effective, although further data are necessary to address this issue. Quality control of imaging and interpretation is an important precondition for these above issues.