Study characteristics

Patient sampling

Aim of the study: to assess the performance of mpMRI in men with prior‐negative SBx

Type of study: retrospective cohort

Selection: unclearly reported

Enrolled/eligible: 54/58

Inclusion period: not reported, but before April 2013

Patient characteristics and setting

Inclusion criteria: men who had ≥ 1 negative SBx and underwent mpMRI (index test) followed by TTMB (reference standard). All men included in the study had either increasing or persistently high PSA level

Exclusion criteria: 4 men were excluded from the study as they received limited TTMB (< 20 cores were taken)

Setting: London, UK. University hospital

Age: median 64 years (range 39‐75)

PSA: 10 ng/mL (range 2‐23)

Prostate volume: 53 mL (range 19‐136)

Previous number of negative Bx: 33 men had 1, 16 had 2, 5 had 3

Index tests

Index tests: MRI only, with an MRI‐score 1‐5 with threshold ≥ 3 for positivity. A 1.5 Tesla (Philips Achiva) and 3.0 Tesla (Siemens Avanto) MRI machine, with T2, DWI and DCE sequences were used. Index test performed first, then the reference test.

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4, GS ≥ 4+3 and others. Pathology grading before ISUP 2005: GS was based upon most frequent pattern instead of highest grade detected.

Reference standard: systematic TTMB with the use of a brachytherapy grid under general anaesthesia, as described by Barzell. Basal and apical cores were obtained routinely, and the minimum number of samples was 20. MRI results were available during TTMB.

Flow and timing

All men underwent the same reference test. No men were excluded from analysis

Comparative

Notes

Study authors provided additional data

Although MRI‐TBx were taken in a subset of 15 men, their results are not reported nor are they taken into account in our analysis.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

No

Did the study avoid inappropriate exclusions?

Unclear

High

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

No

High

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to test diagnostic accuracy of mpMRI and SBx against a reference test, TTMB

Type of study: multicentre, paired‐cohort, prospective study

Selection: consecutive

Enrolled/eligible: 576/740

Inclusion period: May 2012‐November 2015

Patient characteristics and setting

Inclusion criteria: PSA ≤ 15 ng/mL within previous 3 months, organ confined disease on DRE

Exclusion criteria: previous history of PBx, prostate surgery or treatment for PCa (interventions for benign prostatic hyperplasia/bladder outflow obstruction were accepted. Evidence of a urinary tract infection or history of acute prostatitis within the last 3 months. Contraindication to MRI (e.g. claustrophobia, pacemaker, estimated GFR = 50). Previous history of hip replacement surgery, metallic hip replacement or extensive pelvic orthopedic metal work. Treated using 5‐alpha‐reductase inhibitors at time of registration or during the prior 6 months.

Setting: London, UK. University and peripheral hospitals

Age: mean 63.4 years (SD 7.6)

PSA: mean 7.1 ng/mL (SD 2.9)

Prostate volume: not reported

Index tests

Index test 1: MRI only: at multiple sites, 1.5 Tesla MRI scanners (T1, T2, DWI and DCE sequences) were used. Radiologists were provided with clinical details. A 5‐point Likert radiology reporting scale was used, with score of ≥ 3 designated a suspicious scan. Radiologist had undergone additional centralised training.

Index test 2: 10–12 core transrectal SBx.

Participants and physicians remained blinded to the mpMRI images and report. Participants first underwent the TTMB, followed by the SBx

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4, GS ≥ 4+3 and others

Reference standard: TTMB, blinded for MRI report, with cores taken from every hole in the 5‐mm sampling frame.

Flow and timing

All men underwent the same reference test.

Participants left the study for various reasons: 4 were ineligible, 2 were unblinded, 69 had large prostates (> 100 mL), 5 had T4 or nodal disease, 21 had clinical reasons, 52 did not want to proceed, 11 had other reasons

Comparative

Notes

Study authors provided additional data. Number of TTMB cores not reported but estimated

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Yes

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

No

High

Study characteristics

Patient sampling

Aim of the study: to assess the potential of a risk‐based strategy including MRI to selectively identify men aged ≥ 70 years with high‐grade PCa

Type of study: prospective, 2‐arm, PSA‐screening study: 179 men received 6 core SBx only; 158 received MRI+/‐MRI‐TBx and SBx

Selection: consecutive selection based on invitation to participate in a population‐based PSA screening trial

Enrolled/eligible: 337/406 (69 participants refused Bx)

In the current analysis, only the 158 men in the group receiving MRI and MRI‐TBx are included, of which 85 had a prior‐negative Bx and 74 were Bx‐naïve

Inclusion period: Octobr 2013‐April 2016

Patient characteristics and setting

Inclusion criteria: PSA ≥ 3.0 ng/mL

Exclusion criteria: none

Setting: PSA‐screening study. Rotterdam, the Netherlands. University hospital

Age: median 73.1 years (IQR 72.4‐73.8)*

PSA: median 4.2 ng/mL (IQR 3.4–5.8)*

Prostate volume: median 52.9 (IQR 36.8‐70.9)*

DRE positive: 14 participants*

*of the 158 prior‐negative‐ and Bx‐naïve participants taken together

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (Discovery MR750, General Electric Healthcare) was used, with T2, DWI, and DCE sequences. PI‐RADS version 2 was used, with score 1‐5 and score ≥ 3 for positivity. The Koelis Urostation was used for software fused transrectal MRI‐TBx from all MRI‐positive lesions

Index test 2: transrectal extended sextant SBx were taken, blinded for MRI results, before taking the MRI‐TBx

Target condition and reference standard(s)

No reference standard is used in this agreement analyses (MRI‐pathway vs SBx) study, therefore the reference standard domain is not applicable and disregarded

Flow and timing

All participants underwent the same reference test.

During the study, 69 participants refused Bx.

Comparative

Notes

Only the 158 participants in the group receiving MRI and MRI‐TBx are included in the current analysis; the 179 participants with sextant Bx only are excluded from our analysis.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

High

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

No

High

Study characteristics

Patient sampling

Aim of the study: to compare the PCa detection rate of SBx and mpMRI‐TBx

Type of study: prospective cohort

Selection: unclear

Enrolled/eligible: 206/213

Inclusion period: September 2012‐September 2013

Patient characteristics and setting

Inclusion criteria: ≥ 1 prior‐negative SBx session (10–12 cores) and a persistent clinical suspicion of PCa (elevated PSA, an abnormal DRE, or a previous abnormal TRUS image) that warranted a repeat SBx

Exclusion criteria: a prior PCa diagnosis, prior prostate mpMRI, or presence of general contraindications for MRI

Setting: Herlev, Denmark. University hospital

Age: median 65 years (IQR 58‐68)

PSA: median 12.8 ng/mL (IQR 8.9‐19.6)

Prostate volume: not reported. Instead, PSA‐density: median 0.20 (IQR 0.13‐0.29)

DRE positive: 18 men

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI (Ingenia, Philips) was used, with T2, DWI and DCE sequencing. PI‐RADS version 1 with a Likert 1‐5 score and threshold for positivity of ≥ 2 was used*. Software fusion (Hitachi Ltd, HI‐RVS‐system) MRI‐TBx were taken of all MRI‐positive lesions

Index test 2: a 10‐core transrectal SBx. Abnormalities on TRUS were sampled using the standard core for the relevant segment. This was performed first and blinded for MRI results, afterwards the MRI‐TBx were taken

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded

Flow and timing

All participants underwent the same reference test.

During the study, 7 participants were excluded because of technical problems or claustrophobia

Comparative

Notes

*Although MRI‐TBx scores 2‐5 were taken, the results for a threshold of ≥ 3 can be distinguished. In our analysis, therefore, we used the threshold of ≥ 3 for a positive MRI and MRI‐TBx.

We contacted study authors but they could not provide additional data. Other comparisons were not possible to the lack of detailed data.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to assess the diagnostic accuracy and negative predictive value of a novel bpMRI method in Bx‐naïve men in detecting and ruling out significant PCa.

Type of study: prospective, single‐institutional, paired diagnostic study

Selection: consecutive selection

Enrolled/eligible: 1020/1063 (43 participants were excluded for various reasons)

Inclusion period: November 2015–June 2017

Patient characteristics and setting

Inclusion criteria: Bx‐naïve men with a clinical suspicion of PCa (PSA ≥ 4 ng/mL and/or abnormal DRE results)

Exclusion criteria: prior PBx, evidence of acute urinary tract infections, acute prostatitis, general contraindications for MRI, and prior hip replacement surgery or other metallic implants in the pelvic area

Setting: Herlev, Denmark, University Hospital

Age: median 67 years (IQR 61‐71)

PSA: median 8 ng/mL (IQR 5.7‐13)

Prostate volume: median 53 mL (IQR 40‐72)

DRE positive: 377/1020 (37%) participants

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (Philips) with a pelvic‐phased‐array coil was used with T2, DWI sequences, without DCE (bpMRI). An in‐house modified PI‐RADS version 2 was used with score 1‐5 and score ≥ 3 for positivity. Transrectal software fused MRI‐TBx were performed from all MRI‐positive lesions, using Hitachi (n = 877) and Invivo (n = 143) systems.

Index test 2: transrectal 10‐core SBx were taken before the MRI‐TBx, the performers were blinded for the MRI results

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same type of tests. The minimal exclusions were sufficiently explained not leading to relevant bias.

Comparative

Notes

Study authors provided additional data.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the diagnostic efficacy of cognitive mpMRI‐TBx compared to SBx in Bx‐naïve men

Type of study: prospective single‐centre cohort study

Selection: consecutive

Enrolled/eligible: 168/168

Inclusion period: July 2011‐July 2014

Patient characteristics and setting

Inclusion criteria: Bx‐naïve men, with a clinical suspicion of PCa because of elevated PSA levels and/or an abnormal DRE

Setting: Madrid, Spain. University Hospital

Age: mean 61.4 years (± 7.6)

PSA: mean 8.3 ng/mL (± 6.1)

Prostate volume: mean 48.9 mL (± 6.7)

Index tests

Index test 1: MRI‐pathway: mpMRI was performed with a 1.5 Tesla machine (Achieva, Philips Healtcare, Best, the Netherlands) with surface coil, using T1, T2 and DWI. PI‐RADS version 1 was used to assess the MRI by 2 readers independently, with a 1‐5 score and score ≥ 3 for positivity. All PI‐RADS ≥ 3 lesions were targeted cognitively with 2 MRI‐TBx cores.

Index test 2: all men underwent transrectal SBx based on the Vienna nomogram (8‐19 biopsy cores depended on age and prostate volume), before MRI‐TBx were taken, by the same urologist. Blinding of MRI results during SBx was not reported

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded

Flow and timing

All participants underwent the same type of tests

No participants were excluded

Comparative

Notes

Study authors provided additional data

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Unclear

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Unclear

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to investigate the overall and clinically significant PCa detection rates of MRI‐TBx and SBx in prior‐negative Bx men

Type of study: retrospective study

Selection: consecutive selection, but performance of MRI according to the physicians’ clinical considerations

Enrolled/eligible: 185/185 (65 men underwent MRI and Bx, 120 men underwent only Bx without prior MRI)

Inclusion period: March 2012–December 2014

Patient characteristics and setting

Inclusion criteria: men with prior‐negative Bx, persistently elevated serum PSA level and normal DRE

Exclusion criteria: positive DRE

Setting: Taichung, Taiwan. University hospital

Age: median 64 years (IQR 60.3‐67.8)

PSA: median 10.9 ng/mL (IQR 7.2‐14.7)

Prostate volume: median 48 mL (IQR 33.5‐62.5)

DRE positive: none

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (Signa HDx, General Electric Healthcare) was used with T2, DWI, and DCE sequences. PI‐RADS version 1 was converted to PI‐RADS version 2, with score 1‐5 and score ≥ 3 for positivity. Transrectal cognitive MRI‐TBx were performed from all MRI‐positive lesions with ≥ 2 cores

Index test 2: transrectal SBx were taken with ≥ 16 cores from the peripheral zone and transitional zone, after the MRI‐TBx were taken by the same operator.

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same reference test. No participants were excluded.

Comparative

Notes

The 120 participants in the control group who underwent only SBx without prior MRI were excluded from our analysis. Study authors provided additional data.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

No

Did the study avoid inappropriate exclusions?

No

High

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

No

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

High

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to determine the detection rate of 3‐Tesla MRI and MRI‐TBx compared to SBx

Type of study: prospective cohort

Selection: consecutive selection of participants who presented with a suspicion of PCa

Enrolled/eligible: 420/429

Inclusion period: June 2008‐December 2013

Patient characteristics and setting

Inclusion criteria: abnormal DRE findings and/or persistently elevated PSA levels

Exclusion criteria: not reported

Setting: Shanghai, China. University hospital

Age: median 67 years (range 45‐91)

PSA: median 9.7 ng/mL (range 2.4‐35.7)

Prostate volume: median 44.8 mL (range 21.2‐83.2)

DRE positive: 52 participants

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (Philips Achieva) was used, with T1, T2, T2 spectral presaturation attenuated inversion recovery (SPAIR) and DWI sequences. An in‐house MRI score 1‐5 with threshold ≥ 3 for positivity were used. Cognitive transperineal MRI‐TBx were performed from all MRI‐positive lesions.

Index test 2: a 10‐core fan‐shaped transperineal SBx from the peripheral zone with 2‐cores from transition zone was performed, blinded for MRI results, before taking the MRI‐TBx

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same reference test

Except for the 9 excluded participants (DWI artifacts due to movement of the participant) all participants were included in the analysis.

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

High

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the clinical benefit of MRI‐TBx over SBx between first‐time and repeat SBx patients with prior (ASAP)

Type of study: prospective cohort

Selection: unclear

Enrolled/eligible: 100/unclear (50 participants with prior‐negative Bx, 50 Bx‐naïve men)

Inclusion period: September 2011‐March 2014

Patient characteristics and setting

Inclusion criteria: PSA 2‒20 ng/L or DRE abnormalities. Bx‐naïve or ≥ 1 prior Bx with ASAP and ongoing clinical concern for malignancy

Exclusion criteria: known PCa diagnosis, previous prostate MRI or contraindication to MRI or SBx

Setting: Ontario, Canada. University hospital

Age*: mean (SD) 59.4 (7.7); 61.9 (6.5)

PSA*: mean (SD) 6.0 ng/mL (3.5); 7.9 (3.9)

Prostate volume*: mean (SD) 38 g (18); 56 (27)

*for Bx‐naïve men; and previous ASAP men, respectively

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (GE Healthcare) was used, with T2, DWI and DCE sequences. A binary MRI suspicion score was used with a low threshold set to initiate software fusion (Artemis system) transrectal MRI‐TBx from all MRI‐positive lesions.

Index test 2: a standard 12‐core transrectal SBx was performed in Bx‐naïve men, 2 additional cores were taken from the transition zone in previous ASAP men. No blinding of MRI results is reported. MRI‐TBx were taken prior to SBx

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants (with the same indication) underwent the same type of tests

All participants were included in the analysis

Comparative

Notes

Retrospectively, MRI was reassessed using PI‐RADS version 2. The presented Bx data, however, are based on the prospective binary MRI‐score

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Unclear

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Unclear

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

High

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to assess the value of MRI and MRI‐TBx added to standard SBx for detecting clinical relevant PCa

Type of study: retrospective analysis

Selection: retrospective selection of participants meeting inclusion criteria

Enrolled/eligible: 38/1053 (of the 1053 participants who had had an MRI, 38 participants met the inclusion criteria)

Inclusion period: August 2003‐August 2008

Patient characteristics and setting

Inclusion criteria: men with ≥ 2 prior‐negative biopsies who underwent MRI and subsequent SBx complemented with MRI‐TBx of MRI‐suspicious lesions. All men were referred for MRI because of PSA > 4 ng/mL, PSA velocity > 0.75 ng/mL/year or equivocal histopathology from previous Bx

Setting: Boston, USA. University hospital

Age: mean 64 (range 48‐77)

PSA: mean 14.4 (range 1.8‐33.1)

Prostate volume: not reported

Index tests

Index test 1 : MRI‐pathway: a 3 Tesla MRI machine (Genesis Signa LX Excite, GE Healthcare) was used, with T1, T2 and DCE sequences. An in‐house MRI Likert 1‐5 scale was used, grouping score 1‐3 negative and 4‐5 positive. Cognitive MRI‐TBx from MRI‐positive lesions were taken, depending on judgement of urologist

Index test 2: transrectal SBx was performed. Sequence of tests and number of cores were dependent on judgement of urologist. A total median of 19 (range 8‐28) cores (MRI‐TBx + SBx) were taken. MRI results were known at time of SBx performance

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same type of tests. All participants were included in the analysis

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

No

Did the study avoid inappropriate exclusions?

Yes

High

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

No

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

High

High

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Unclear

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Unclear

High

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate whether adding 1.5 T magnetic field mpMRI‐TBx improves PCa detection in men undergoing blind 24‐core saturation PBx

Type of study: prospective collected data

Selection: consecutive selection

Enrolled/eligible: 123/123

Inclusion period: January 2013–December 2016

Patient characteristics and setting

Inclusion criteria: men who had already undergone a first 10/12‐core PBx with suspected PCa due to an increased PSA level and/or positive DRE

Exclusion criteria: > 1 set of 10/12‐core Bx, TURP or other lower urinary tract endoscopic procedures

Setting: Padua, Italy. University hospital

Age: median 62 years (IQR 57‐68)

PSA: median 6.27 ng/mL (IQR 4.75‐8.9)

Prostate volume: mean 54.59 mL (range 20‐149)

DRE positive: 8.9% (11/123) of the participants

Index tests

Index tests: MRI only + MRI‐TBx + MRI‐pathway: a 1.5 Tesla MRI machine was used with T2 and DWI sequences. PI‐RADS version 1 was used with score 1‐5 and score ≥ 3 for positivity. Transrectal cognitive MRI‐TBx were performed from all MRI‐positive lesions

Target condition and reference standard(s)

Target conditions: GS 3+3 = 6, GS ≥ 3+3, GS ≥ 3+4

Reference standard: transrectal 24‐core saturation Bx including 8 anterior biopsies, blinded for MRI results. When a suspicious lesion was present, the operator performed first the MRI‐TBx and then the saturation biopsies (unblinded)

Flow and timing

All participants underwent the same reference test. No participants were excluded

Comparative

Notes

Study authors provided additional data

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐TBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to compare the accuracy of visual MRI‐TBx versus software MRI‐TBx using a rigid or elastic approach

Type of study: prospective cohort

Selection: consecutive selection, divided into 3 groups: the first 127 participants received visual MRI‐TBx, the next 131 participants had the rigid fusion MRI‐TBx and the last 133 participants had the elastic fusion MRI‐TBx

Enrolled/eligible: 391/391

Inclusion period: January 2011‐March 2012

Patient characteristics and setting

Inclusion criteria: PSA > 4 ng/mL, and/or suspicious DRE and no previous PBx

Exclusion criteria: none

Setting: Paris, France. University hospital

Age*: mean 62.7 years (SD 7.4); 64.5 (7.9); 64.6 (6.7)

PSA*: mean (SD) 8.1 ng/mL (3.7); 9.0 (3.9); 8.3 (4.1)

Prostate volume* (SD): 53 mL (25); 58.3 (28.6); 55.7 (35.1)

DRE positive*: 20; 16; 16

*For the 3 groups, respectively: visual‐; elastic‐; rigid fusion

Index tests

Index test 1: MRI‐pathway: a 1.5 Tesla MRI machine was used, with T2, DWI and DCE sequences. An in‐house MRI‐score: 0‐4 score in transitional zone and 0‐10 in peripheral zone were used, with threshold ≥ 2 and ≥ 6 for positivity, respectively. Either cognitive MRI‐TBx or software fusion MRI‐TBx (Koelis, elastic MRI‐TRUS image registration System; Esaote, rigid navigation system) were taken from all positive lesions.

Index test 2: 10‐12 core transrectal SBx was performed first. Blinding for MRI results was not reported. Subsequently, MRI‐TBx were taken of suspicious lesions

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same type of tests. All participants were included in the analysis.

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Unclear

Unclear

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

High

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to analyse the negative predictive value of MRI and PSA density to rule out significant PCa

Type of study: prospective cohort

Selection: consecutive selection of men with a suspicion of PCa (PSA >4.0 ng/ml and/or suspicious digital rectal examination (DRE)) who were either biopsy‐naïve or after previous negative biopsy.

Enrolled/eligible: 1040/1040 (597 Bx‐naïve + 443 prior‐negative Bx men)

Inclusion period: October 2012‐December 2015

Patient characteristics and setting

Inclusion criteria: suspicion of PCa: PSA > 4.0 ng/mL and/or suspicious DRE, and who were Bx‐naïve or had undergone a prior‐negative Bx

Exclusion criteria: none

Setting: Heidelberg, Germany. University hospital

Age: median 65 years (IQR 60‐71)

PSA: median 7.2 ng/mL (IQR 5.3‐10.4)

Prostate volume: median 45 mL (IQR 34‐64)

DRE positive: 291

Index tests

Index tests: MRI only + MRI‐TBx + MRI‐pathway: a 3 Tesla MRI machine (Magnetrom Prisma or Biograph mMR (Siemens Healthcare) was used, with T2, DWI and DCE sequences. The PI‐RADS version 1 Likert 1‐5 score was used, with threshold ≥ 3 for positivity. Transperineal MRI‐TBx were taken from all positive lesions with the Biopsee system (rigid software registration). First MRI‐TBx were taken, subsequently the reference biopsies

Target condition and reference standard(s)

Target condition: GS ≥ 3+4

Reference standard: volume‐based systematic transperineal grid‐directed Bx with a median of 24 cores according to the Ginsburg protocol. Bx operators first performed the MRI‐TBx and had access to MRI data during whole procedure.

Flow and timing

All participants underwent the same reference test. All participants were included in the analysis

Comparative

Notes

Results not reported separately for the two participant groups

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐TBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

No

High

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the performance of MRI‐TBx in diagnosing clinically significant PCa

Type of study: prospective cohort

Selection: consecutive selection

Enrolled/eligible: 1042/1042 (328 Bx‐naïve‐, 324 prior‐negative Bx‐ and 390 active surveillance men)

Inclusion period: September 2009‐February 2015

Patient characteristics and setting

Inclusion criteria: elevated PSA level or abnormal DRE or 2) confirmation of low‐risk PCa for men considering active surveillance

Exclusion criteria: none reported

Setting: Los Angeles, USA. University hospital

Age*: median (IQR) 64.4 years (58.5‐69.4); 65.7 (59.3‐70.2)

PSA*: median (IQR) 5.8 ng/mL (4.4‐8.1); 7,6 (5‐11.5)

Prostate volume*: median (IQR) 45 mL (33‐61.5); 57.7 (39.8‐83.5)

*respectively, for the Bx‐naïve‐ and prior‐negative Bx participant groups

Index tests

Index tests 1: MRI‐pathway: a 3 Tesla MRI machine (Trio Trim/Somatom, Philips) was used, with T2, DWI and DCE sequences. An in‐house Likert 1‐5 score was used, with threshold ≥ 3 for positivity. MRI‐TBx (Artemis fusion device (Eigen, Grass Valley, Calif) were taken first in case of a suspicious lesion, then SBx were taken

Index test 2: transrectal 12‐core SBx were taken in all participants, after MRI‐TBx. No blinding for MRI is reported

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same reference test. All participants were included in the analysis.

Comparative

Notes

Participants on active surveillance (n = 390) were excluded from our analysis. Although in text 328 participants are reported in the biopsy‐naïve group, in the data tables 329 participants are reported

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Unclear

Unclear

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the differences in PCa detection rate and Bx effectiveness between MRI‐TBx and transperineal standard SBx in Bx‐naïve men

Type of study: prospective cohort

Selection: not explicitly reported

Enrolled/eligible: 60/unclear

Inclusion period: October 2014‐April 2016

Patient characteristics and setting

Inclusion criteria: PSA > 4 ng/mL, a PSA density > 0.18 ng/mL/mL, a PSA velocity > 0.75 ng/mL/year or a pathological DRE

Exclusion criteria: previous history of prostate biopsies, prostate surgery or radiotherapy or medical treatment for benign prostate hyperplasia

Setting: Reus, Spain. University hospital

Age: mean 64.1 years (SD 6.7).

PSA: median 7.2 ng/mL (IQR 6‐9.4)

Prostate volume: median 47.8 mL (IQR 34.6‐63.2)

Index tests

Index test 1: MRI‐pathway: a 3 Tesla MRI machine (Signa, GE) was used, with T1, T2 and DWI sequences. The PI‐RADS version 1 Likert 1‐5 score was used, with threshold ≥ 4 for positivity (if no PI‐RADS ≥ 4 lesions were present, also PIRADS 2 and 3 were targeted). (Because study authors provided additional data we were able to use the results for a MRI‐threshold of ≥ 3.) The MRI targets were discussed with radiologist and cognitive fusion transperineal MRI‐TBx on target lesions was performed.

Index test 2: 12‐core transperineal SBx in all men: two cores were directed towards the medial segments of the peripheral zone, two towards the lateral segments of the peripheral and two towards the transition zone for each lobe, with blinding for MRI results. Subsequently, MRI‐TBx were taken.

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same reference test. All participants were included in the analysis.

Comparative

Notes

Study authors provided additional data. In our analysis we were therefore able to use the results for MRI‐threshold ≥ 3 for positivity.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Yes

Low

High

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to determine the sensitivity and specificity of mpMRI for significant PCa with transperineal sector Bx as the reference standard

Type of study: prospective cohort

Selection: consecutive patients

Enrolled/eligible: 201/205 (83 Bx‐naïve‐, 103 prior‐negative Bx‐, 15 active surveillance participants; 4 participants were excluded due to contraindications to MRI)

Inclusion period: July 2012‐November 2013

Patient characteristics and setting

Inclusion criteria: a prior‐negative PBx with ongoing suspicion of PCa because of rising PSA levels (n = 103); those undergoing a primary PBx because of raised PSA level or abnormal DRE (n = 83)

Exclusion criteria: previous history of PBx, prostate surgery or radiotherapy or medical treatment for benign prostate hyperplasia

Setting: London, UK. University hospital

Age*: mean (SD) 65 years (7.6); 64.1 (6.8).

PSA*: mean (SD) 12.6 ng/mL (13.7); 13.3 (12.1)

Prostate volume*: mean (SD) 54 mL (31); 68 (35)

*Although test results are reported only for the mix of the 2 participant groups, these basic characteristics are reported for the 2 groups separately (103 prior‐negative Bx‐; Bx‐naïve patients, respectively)

Index tests

Index test: MRI only. A 1.5 Tesla MRI machine (Signa Excite, GE Healthcare) with T2 and DWI sequences was used. The PI‐RADS version 1 Likert 1‐5 score was used, with threshold ≥ 3 for positivity

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4 and GS ≥ 4+3

Reference standard: transperineal sector Bx, with 24‐40 cores (depending on prostate size) with a brachytherapy grid. MRI‐positive lesions were targeted by cognitive registrated MRI‐TBx, but results were not reported separately

Flow and timing

All participants underwent the same reference test and were included in the analysis.

Comparative

Notes

All active surveillance participants (n = 15) were excluded from our analysis after additional information was received from study authors

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

No

High

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to assess the performance of combining a blood‐based biomarker panel and MRI‐TBx for PCa detection

Type of study: prospective, multicentre, paired diagnostic study

Selection: consecutive selection

Enrolled/eligible: 532/727 (195 participants were excluded due to incomplete data)

Inclusion period: May 2016–May 2017

Patient characteristics and setting

Inclusion criteria: men aged 45‐75 years, no previous PCa, referral for PCa work‐up

Exclusion criteria: previous diagnosis of PCa

Setting: Stockholm, Sweden; Oslo, Norway; and Tonsberg, Norway. University and peripheral hospitals (cancer centre)

Age:

Stockholm (n = 160): mean 63 years (6.2);

Oslo (n = 236): mean 65 years (7.8);

Tonsberg (n = 136): mean 64 years (6.8)

PSA:

Stockholm (n = 160): median 6.2 ng/mL (IQR 4.8‐8,2)

Oslo (n = 236): median 6 ng/mL (IQR 4‐9)

Tonsberg (n = 136): median 7.1 ng/mL (IQR 4.7‐11)

Prostate volume:

Stockholm (n = 160): median 51 mL (IQR 38‐70)

Oslo (n = 236): median 42 mL (IQR 32‐54)

Tonsberg (n = 136): median 44 mL (IQR 33‐55)

DRE positive: not reported

Index tests

Index test 1: MRI‐pathway, a 1.5 Tesla MRI machine (Avanto and Aera, Siemens) was used with T2, DWI sequences, without DCE. PI‐RADS version 2 was used with score 1‐5 and score ≥ 3 for positivity. Transrectal software fused MRI‐TBx were performed from all MRI‐positive lesions, using several machines.

Index test 2: transrectal extended sextant SBx were taken after the MRI‐TBx and therefore not blinded for MRI results.

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same tests. Men with incomplete data were explained and excluded from the analysis, not leading to relevant bias.

Comparative

Notes

Study authors provided additional data.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

No

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to describe the Ginsburg protocol for transperineal MRI‐TBx supported by mpMRI and TRUS image fusion, and report biopsy results

Type of study: prospective cohort study

Selection: consecutive patients

Enrolled/eligible: 571/571 (107 Bx‐naïve‐, 295 prior‐negative Bx‐ and 169 active surveillance men)

Inclusion period: March 2013‐October 2015

Patient characteristics and setting

Inclusion criteria: indication for repeat Bx: either rising PSA or ASAP or multifocal high‐grade prostatic intraepithelial neoplasia on a previous Bx

Exclusion criteria: previous prostate MRI or a transperineal Bx

Setting: Cambridge, UK. University hospital

Age: median 65 years (IQR 59‐69)

PSA: median 7.8 ng/mL (IQR 60‐12)

Prostate volume: median 65 mL (IQR 44‐83)

Index tests

Index test: MRI only, MRI‐TBx and MRI‐pathway. A 1.5 Tesla (MR450) or a 3 Tesla (Discovery MR750 HDx) machine of GE Healthcare was used with T2, DWI and DCE sequences. The PI‐RADS version 1, Likert 1‐5 score was used, with threshold ≥ 3 for positivity.

First transperineal software fusion MRI‐TBx cores were taken (BiopSee platform, Medcom) of every suspicious lesion. Then the reference standard was performed.

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4 and GS ≥ 4+3

Reference standard: an 18‐24 core systematic transperineal Bx according to the Ginsburg protocol, with 1‐2 cores from each of the 12 sectors, using the BiopSee MRI‐TRUS fusion platform with a brachytherapy grid for guidance. Blinding of MRI results not reported

Flow and timing

All participants underwent the same reference test and were included in the analysis.

Comparative

Notes

Study authors provided additional data. In our analysis we excluded the 169 active surveillance participants. Furthermore, we excluded the 106 Bx‐naïve participants because of overlapping data with Hansen 2018.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐TBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Unclear

Unclear

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the detection rates of transperineal MRI‐TBx and SBx for men with previous benign transrectal SBx in 2 high‐volume centres

Type of study: prospective cohort study

Selection: unclear

Enrolled/eligible: 487/487 (200 from centre 1, 287 from centre 2)

Inclusion period: October 2013‐November 2015

Patient characteristics and setting

Inclusion criteria: indication for repeat Bx: rising PSA or a previous SBx specimen showing suspicion of cancer (ASAP) or multifocal high‐grade prostatic intraepithelial neoplasia

Exclusion criteria: none reported

Setting: Heidelberg, Germany. University hospital

Age: median 66 years (IQR 61‐72)

PSA: median 9.7 ng/mL (IQR 7.1‐13.9)

Prostate volume: median 52 mL (IQR 36‐75)

Index tests

Index tests: MRI only,MRI‐TBx and MRI‐pathway. A 3 Tesla MRI machine (Magnetron, Siemens) with T2, DWI and DCE sequences was used. The PI‐RADS version 2, Likert 1‐5 score was used, with threshold ≥ 3 for positivity.

First transperineal software fusion MRI‐TBx cores were taken (BiopSee platform, Medcom) of every suspicious lesion. Then template Ginsburg Bx was performed.

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4 and GS ≥ 4+3

Reference standard: a volume‐based, transperineal template Bx scheme, with a median of 24 cores, according to Ginsburg protocol was performed, using the BiopSee MRI‐TRUS fusion platform with brachytherapy grid for guidance. Blinding of MRI results not reported

Flow and timing

All participants underwent the same reference test.

All participants were included in the analysis.

Comparative

Notes

Only participants from centre 1 (Heidelberg, Germany) were included in our analysis, due to overlap with patients of centre 2 (Cambridge, UK) in Hansen 2016b.

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test MRI‐TBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Unclear

Unclear

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to analyse the detection rates of primary MRI‐fusion transperineal PBx using combined targeted and systematic core distribution in 3 tertiary referral centres

Type of study: prospective cohort

Selection: consecutive patients

Enrolled/eligible: 856/807 (163 participants from centre 1, 402 from centre 2* and 242 from centre 3; 49 participants did not comply with the inclusion criteria)

Inclusion period: October 2012‐May 2016

Patient characteristics and setting

Inclusion criteria: first suspicion of PCa, based on raised PSA levels above age‐related normal range, a suspicious DRE, or other including family history

Exclusion criteria: age > 79 years, PSA level > 30 ng/mL, prior‐negative Bx or previous diagnosis or treatment of PCa

Setting Centre 1: Cambridge UK, tertiary care hospital

Age: median 64 years (IQR 57‐69)

PSA: 6.6 ng/mL (IQR 4.6‐9.0)

Prostate volume: 44 mL (IQR 33‐55)

Positive DRE: 39 participants

Setting Centre 2: Heidelberg, Germany, University Hospital (participants from centre 2 in this study were excluded from analyses in this review to prevent overlapping data with the included study Distler 2017*)

Age: median 65 years (IQR 60‐70)

PSA: 6.9 ng/mL (IQR 5.2‐9.1)

Prosate volume: 47 mL (IQR 32‐62)

Postive DRE: 94 participants

Setting Centre 3: Melbourne, Australia, tertiary care hospital

Age: median 65 years (IQR 60‐70)

PSA: 5.9 ng/mL (IQR 4.6‐8,0)

Prostate volume: 25 mL (IQR 24‐47)

Positive DRE: 54 participants

Index tests

Centre 1: index test: MRI only, a 1.5 Tesla (MR450) or a 3 Tesla (Discovery MR750 HDx) machine of GE Healthcare was used with T2, DWI and DCE sequences. The PI‐RADS version 1 (until 2015) and version 2 (onwards) with a Likert 1‐5 score were used, with threshold ≥ 3 for positivity. Transperineal software fusion MRI‐TBx cores were taken (BiopSee system, Medcom) of every suspicious lesion, followed by template Bx. However, MRI‐TBx results were not reported separately.

Centre 3: index test: MRI only, a 3 Tesla Magnetom (Siemens) was used with T2, DWI and DCE sequences. The PI‐RADS version 1 (until 2015) and version 2 (onwards) with a Likert 1‐5 score were used, with threshold ≥ 3 for positivity. Transperineal cognitive MRI‐TBx cores were taken of every suspicious lesion, followed by template Bx. However, MRI‐TBx results were not reported separately.

Target condition and reference standard(s)

Target condition in both centres: GS ≥ 3+3, GS ≥ 3+4 and GS ≥ 4+3

Reference standard in both centres: volume‐based transperineal template Bx with a median of 24 cores according to the Ginsburg protocol. Bx operators had access to MRI data during whole procedure. MRI‐TBx were taken in addition to the template Ginsburg biopsies and included in the reference standard results.

Centre 1 used the Biopsee system (Medcom) with a 5‐mm spacing brachytherapy grid

Centre 3 used a 5‐mm spacing brachytherapy grid (BK Ultrasound) and a transrectal probe mounted on a stepper

Flow and timing

All participants underwent same reference standard. No participants were excluded for analysis.

Comparative

Notes

*Only the 163 participants from centre 1 and the 242 patients from centre 3 are included in our analysis; we excluded the 402 patients from centre 2 because they are also reported in Distler 2017

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Yes

Did the study avoid inappropriate exclusions?

Yes

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

No

High

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to assess the diagnostic accuracy of MRI and MRI‐TBx using visual registration

Type of study: multicentre study, unclear design

Selection: unclear

Enrolled/eligible: 55/unclear

Inclusion period: April 2011‐March 2013

Patient characteristics and setting

Inclusion criteria: PSA > 4 ng/mL on 2 consecutive measurements in the last 6 months

Exclusion criteria were:

1. abnormal DRE

2. previous PBx

3. diagnosis of PCa

4. previous prostate surgery (e.g. TURP)

5. active or chronic prostatitis

6. contraindication for MRI examination (e.g. pacemaker)

Setting: Turku, Finland/Bratislava, Slovakia. University hospitals

Age: median 66 years (range 47–76)

PSA: median 7.4 ng/mL (range 4–14)

Prostate volume: median 42 mL (range 17–107)

Index tests

Index test 1: MRI‐pathway, a 3 Tesla machine (Magnetom Verio 3T, Siemens) was used with T2, DWI, DCE and spectroscopy sequences. An in‐house MRI Likert 1‐5 scale was used, with threshold ≥ 4 for positivity and MRI‐TBx (but small discrete lesions (maximum diameter of 7–9 mm on mpMRI) were also targeted. Cognitive transrectal MRI‐TBx were taken of all suspicious lesions, after SBx

Index test 2: transrectal extended sextant SBx were taken, blinded for MRI results

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same tests. All participants were included in the analysis; except for 2 participants who did not receive MRI‐TBx due to technical problems, which in our current analysis had to be excluded.

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

No

High

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Yes

Low

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

High

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

Yes

Low

Study characteristics

Patient sampling

Aim of the study: to evaluate the role of a MRI combined with MRI‐TBx for improving risk stratification of men with elevated PSA

Type of study: prospective cohort

Selection: unclear selection

Enrolled/eligible: 161/175 (134 Bx‐naïve, 27 prior‐negative Bx participants and 14 exclusions)

Inclusion period: March 2013‐February 2015

Patient characteristics and setting

Inclusion criteria: 2 repeated measurements of PSA in the range 2.5–20.0 ng/mL and/or abnormal DRE

Exclusion criteria: previous PCa diagnosis, previous Bx within 6 months, prostate surgery, clinical infection or MRI contraindication

Setting: Turku, Finland. University hospital

Age: mean 64.7 years (SD 6.4)

PSA: median 7.5 (IQR 5.7‐9.6).

Prostate volume: median 37 (IQR 27.5‐49)

Index tests

Index tests 1: MRI‐pathway: a 3 Tesla machine (Magnetom Verio 3T, Siemens) was used with T2 and DWI sequences. An in‐house MRI Likert 1‐5 scale was used, with threshold ≥ 3 for positivity and MRI‐TBx. Cognitive transrectal MRI‐TBx were taken of all index lesions, prior to SBx.

Index test 2: 12‐core transrectal SBx, without blinding for MRI results (although strictly following the SBx scheme).

Target condition and reference standard(s)

No reference standard is used in this agreement analyses study (MRI‐pathway vs SBx), therefore the reference standard domain is not applicable and disregarded.

Flow and timing

All participants underwent the same tests. Not all participants were included in analysis. 4 withdrew consent before and 7 after MRI, 1 had a non‐diagnostic MRI, 2 had a PSA < 2.5 or > 20

Comparative

Notes

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test SBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

No

High

Low

DOMAIN 2: Index Test MRI‐pathway

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 3: Reference Standard

Is the reference standards likely to correctly classify the target condition?

Yes

Was the reference standard performed independent from the index test?

Yes

Low

Low

DOMAIN 4: Flow and Timing

Did all patients receive the same reference standard?

Yes

Were all enrolled patients included in the analysis, or were exclusions explained and not leading to a relevant bias?

No

High

Study characteristics

Patient sampling

Aim of the study: to evaluate a volume‐based, computer‐assisted method for TOP‐Bx

Type of study: prospective cohort

Selection: unclear selection

Enrolled/eligible: 172/unclear (mix of 95 Bx‐naïve‐, 51 prior‐negative‐ and 26 active surveillance participants)

Inclusion period: October 2013‐March 2014

Patient characteristics and setting

Inclusion criteria: abnormal PSA or suspicious DRE, persistent suspicion of PCa after prior‐negative Bx

Exclusion criteria: none reported

Setting: Darmstadt, Germany. University hospital

Age*: median 65 years (IQR 58‐71)

PSA*: median 7.2 ng/mL (IQR 5.4‐10.2)

Prostate volume*: median 46 mL (IQR 36‐60)

Positive DRE*: 37 participants

Index tests

Index tests: MRI only + MRI‐TBx + MRI‐pathway. A 3 Tesla machine (Magnetom, Siemens) was used with T1, T2, DWI and DCE sequences. The PI‐RADS version 1, Likert 1‐5 scale was used, with threshold ≥ 3 for positivity and MRI‐TBx. Software fusion transperineal MRI‐TBx were taken of all index lesions independently of the TOP‐Bx, using the BiopSee MRI‐TRUS fusion platform (Medcom).

Target condition and reference standard(s)

Target condition: GS ≥ 3+3, GS ≥ 3+4 and GS ≥ 4+3

Reference standard: Novel, volume‐based, automated core‐placement method for TOP‐Bx placement was performed with a needle distribution sampling each conceivable tumour lesion ≥ 0.5 mL in the complete prostate (100%), with a median of 24 (IQR 23‐27) cores, independent of MRI results.

Flow and timing

All participants underwent the same reference test and were included in the analysis.

Comparative

Notes

Study authors provided additional data. We excluded the 26 active surveillance participants from our analysis.

*However, the basic characteristics are based on all participants (including the active surveillance participants).

Methodological quality

Item

Authors' judgement

Risk of bias

Applicability concerns

DOMAIN 1: Patient Selection

Was a consecutive or random sample of patients enrolled?

Unclear

Did the study avoid inappropriate exclusions?

Yes

Unclear

Low

DOMAIN 2: Index Test MRI‐TBx

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Yes

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low

DOMAIN 2: Index Test MRI

Was the MRI assessed without knowledge of the results of the (reference or other index) biopsies?

Yes

Were the MRI‐TBx performed independent of the (reference or other index) biopsies?

Was the performance of the SBx not influenced by the performance of the (reference or other index) biopsies?

Low

Low