URO ONCO

Welcome, this website is intended for all international healthcare professionals in uro-oncology. By clicking the link below you are declaring and confirming that you are a healthcare professional.

You are here

Limited evidence for the use of imaging to detect prostate cancer: A systematic review

European Journal of Radiology, 9, 83, pages 1601 - 1606

Comment from Henk van der Poel:
Imaging options for prostate cancer detection are limited. In 4852 identified studies, only 6 studies were included in the systematic review. Imaging currently does not aid in prostate cancer detection.

Highlights

 

  • In men with clinical suspicion of prostate cancer, ultrasound guided systematic biopsies is the golden standard for diagnosis.
  • Diagnostic imaging techniques, especially magnetic resonance imaging, is being used in trials to aid detection of prostate cancer.
  • To date, there is insufficient scientific evidence for the use of imaging techniques to detect prostate cancer.

Abstract

Objective

To assess the diagnostic accuracy of imaging technologies for detecting prostate cancer in patients with elevated PSA-values or suspected findings on clinical examination.

Methods

The databases Medline, EMBASE, Cochrane, CRD HTA/DARE/NHS EED and EconLit were searched until June 2013. Pre-determined inclusion criteria were used to select full text articles. Risk of bias in individual studies was rated according to QUADAS or AMSTAR. Abstracts and full text articles were assessed independently by two reviewers. The performance of diagnostic imaging was compared with systematic biopsies (reference standard) and sensitivity and specificity were calculated.

Results

The literature search yielded 5141 abstracts, which were reviewed by two independent reviewers. Of these 4852 were excluded since they did not meet the inclusion criteria. 288 articles were reviewed in full text for quality assessment. Six studies, three using MRI and three using transrectal ultrasound were included. All were rated as high risk of bias. Relevant studies on PET/CT were not identified.

Conclusion

Despite clinical use, there is insufficient evidence regarding the accuracy of imaging technologies for detecting cancer in patients with suspected prostate cancer using TRUS guided systematic biopsies as reference standard.

Abbreviations: TRUS - transrectal ultrasonography, DWI - diffusion-weighted imaging, DCE-MRI - dynamic contrast enhanced magnetic resonance imaging, MRSI - magnetic resonance spectroscopic imaging, PSA - prostate specific antigen, Se - Sensitivity, Sp - Specificity, NHS - National Health Service United Kingdom, SBU - Swedish Council on Health Technology Assessment.

Keywords: Prostate cancer, Diagnosis, Biopsy, Imaging, Meta-analysis, Systematic review.

1. Introduction

Prostate cancer (PC) diagnostics is one of the largest and most important challenges in today's health care. New techniques are introduced but few have been thoroughly evaluated before they have been adopted in clinical practice.

In the pre-PSA era PC was in most cases diagnosed at a clinical stage when the disease was symptomatic. With the introduction of PSA, the landscape of prostate cancer has totally changed. Today the majority of tumors are detected as a result of PSA testing and at early stage 1. Although screening with PSA testing, as in the European screening study (ERSPC), has been shown to reduce disease specific mortality in prostate cancer [1] , there still remain issues that have to be resolved. The PSA test has good sensitivity for PC but poor specificity [2] , which means that a large number of men without PC will undergo unnecessary ultrasound guided prostate biopsies, which today is the golden standard technique for diagnosis of prostate cancer. Ultrasound guided prostate biopsy has a very high specificity but a lower sensitivity [3] which results in repeated biopsies in men with a first set of negative biopsies but a continuously high PSA level. In addition, many men experience prostate biopsy as an unpleasant and painful procedure that is accompanied by a substantial risk of complications such as infection and bleeding [4] . All together prostate cancer diagnostic work up constitutes today a substantial clinical challenge, a fact that becomes even more obvious acknowledging PC as the most common non-cutaneous malignancy in the western world [5] . Hence, the economic costs derived from prostate cancer diagnostics imply a burden for the health care system and the society in general. In a recent population-based cost study, the economic burden of prostate cancer was estimated to be €8.43 billion in the European Union [6] .

The problems with PC diagnosis so far have led to the search for new and better imaging techniques of the prostate that could lead to a more simple and valid diagnosis.

There are three main areas of imaging that have been put forward. First, transrectal ultrasound that today is used for guiding the biopsies in a systematic manner. Here, different techniques have been developed and marketed such as Doppler, Histoscanning® and elastography. Second, positron emission tomography/computed tomography (PET/CT) with new tracers has also been introduced as a possible imaging modality of PC.

Third, there is a growing interest in the use of magnetic resonance imaging (MRI) in PC diagnosis. The fast development of MRI using multi parametric imaging including diffusion-weighted, dynamic contrast enhanced, diffusion weighted and spectroscopic imaging has put MRI as one of the most promising new imaging techniques for PC.

The urgent need to improve PC diagnosis has led to a wide introduction of these techniques into the clinic, often without prior evaluation of their performance when it comes to sensitivity and specificity. There is also a risk of increasing costs of PC diagnosis if imaging techniques do not lead to improvement of the accuracy of the diagnosis. Performing a diagnostic investigation with MRI and PET costs about €720 (SEK 6500) and €1440 (SEK 12 900) more respectively than performing a prostate biopsy ( http://www.sbu.se/sv/Publicerat/Alert/Bilddiagnostik-vid-misstankt-prostatacancer ). Hence, it is of great importance to investigate how well these new imaging techniques perform. Thus, the aim of this systematic review was to assess the diagnostic accuracy of these different imaging techniques.

2. Methods

2.1. Protocol and registration

This systematic review was conducted at The Swedish Council on Health Technology Assessment (SBU) [ www.sbu.se ] and published as a report in Swedish in 2014. SBU uses a peer-reviewed protocol including pre-specified objectives and questions according to PICO (Handbook SBU; www.sbu.se ). The protocol is available from SBU upon request.

2.2. Eligibility criteria

Only published studies in English, German, Danish, Norwegian and Swedish were accepted. The criteria for eligibility included the following characteristics.

Population: men with suspected prostate cancer, i.e. an elevated PSA-value or a suspicious finding on clinical examination, and with no previous diagnosis or treatment of prostate cancer.

Interventions (index test): magnetic resonance, diffusion weighted magnetic resonance, dynamic contrast enhanced magnetic resonance, spectroscopy, positron emission tomography, positron emission tomography/computed tomography, transrectal ultrasound with Doppler, contrast-enhanced ultrasound, elastography or Histoscanning®. Regarding MRI, only articles fulfilling European Consensus statements regarding minimal requirements for multi parametric imaging were included [7] and [8].

Control (reference test): ultrasound guided systematic prostate biopsies (≥10) or histopathologic examination after prostatectomy.

Outcome measures: sensitivity and specificity for diagnostic accuracy. Sensitivity relates to the ability of a method to correctly identify true sick (here prostate cancer according to biopsy) individuals. Specificity, on the other hand, is the ability of a method to correctly identify true healthy individuals (here no prostate cancer according to biopsy).

2.3. Information sources

The electronic literature search was performed by an information specialist and included the databases PubMed, EMBASE and The Cochrane Library. The last search was performed in June 2013. In addition to the electronic search, a hand search was made and references from narrative reviews and articles in international journals not identified in the main search were included. Grey literature was not included.

2.4. Study selection

Two reviewers independently screened the titles and abstracts identified by the search strategy. All studies of potential relevance according to the inclusion criteria were obtained in full text and two reviewers independently assessed them for inclusion. Any disagreement was resolved by discussions. Reference lists were screened for additional studies of relevance.

2.5. Data collection process

From each included study, data was extracted and inserted in a table by one reviewer. The other reviewers audited the data extraction. Any disagreements were resolved by discussion.

2.6. Data items

Information was extracted from each included article concerning recruitment (prospective, retrospective, consecutive and time period for collecting the data) and characteristics of the population (age, PSA-level, previous negative biopsy or not, disease prevalence). Concerning the index test, the type and details of the test was recorded, the number of observers and reporting on blinding to the reference test was also noted. For the reference test, the number of systematic biopsies was recorded as well as the number of observers and blinding to the index test. As to the outcome, data on sensitivity and specificity was extracted. When confidence intervals were lacking, these were calculated from data presented in the articles.

2.7. Risk of bias in individual studies

Two reviewers independently assessed the risk of bias in individual studies using the QUADAS tool [9] . Each study was rated as low, moderate or high risk of bias according to predetermined criteria given in Table 1 . For health economic studies, the quality was assessed based on a SBU tool for critical assessment of health economic studies [10] .

Table 1 Pre-specified inclusion criteria according to the PICO approach.

P Men with suspected prostate cancer, i.e. increased PSA level (≥4 ng/ml) or suspected findings on clinical examination
 
I (index test) Imaging methods (MRI, transrectal ultrasound with Doppler, or PET/DT for directing biopsies)
 
C (reference test) Histopathologic analysis of tissue taken from prostate biopsies (n ≥ 10) or corresponding analysis of tissue after prostatectomy
 
O (outcome measures) Data on sensitivity and specificity on patient level available or can be calculated.
 
Pre-specified criteria for low moderate and high risk of bias
Low risk of bias Prospective study. Adequately described population, consecutive inclusion of patients. Number and experience of observers of index test reported, number of observers of biopsy reported. Blinding to both index test and reference test. Imaging performed before biopsy.
 
Moderate risk of bias Retrospective study. Adequately described population, consecutive inclusion of patients. Blinding to both index test and reference test. Number and experience of observers of index test reported. Imaging performed before biopsy or after one or more negative biopsies.
 
High risk of bias Requirements for moderate study quality not met.

2.8. Planned method of analysis

The intention was to synthesize sensitivity and specificity, respectively, of reasonably homogeneous studies with low or moderate risk of bias and to present the results as forest plots and ROC-curves.

2.9. Quality of evidence (GRADE)

The international grading system GRADE [11] was used to assess the quality of evidence for effects and harms according to the following four levels:

  • High quality – We are very confident that the true effect lies close to that of the estimate of the effect.
  • Moderate quality – We are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
  • Low quality – Our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect.
  • Very low quality – We have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of the effect.

Factors that can weaken the quality of evidence are study limitations, indirectness, inconsistency, imprecision and publication bias. The quality of evidence was decided upon through discussions and any disagreements were solved by consensus.

3. Results

3.1. Study selection

The literature search yielded 5141 abstracts, which were reviewed by two independent reviewers. Of these 4853 were excluded since they did not meet the inclusion criteria. 288 articles were reviewed in full text for quality assessment. Six studies were finally included, three regarding MRI and three regarding transrectal ultrasound with Doppler ( Fig. 1 ). A list of excluded full-text articles, with the main reason for exclusion, is available at http://www.sbu.se/sv/Publicerat/Alert/Bilddiagnostik-vid-misstankt-prostatacancer .

gr1

Fig. 1 Flowchart showing search strategy, number of included and excluded articles and study quality of included studies in the systematic review. SR, systematic review.

3.2. Study characteristics

Study characteristics for the six included studies are shown in Table 2 .

Table 2 Accuracy of magnetic resonance imaging (MRI) or ultrasound compared with biopsy for detecting and localizing prostate cancer in patients with suspected cancer without previous biopsy or with a previous negative biopsy.

Author

Year

Reference

Country
Study design

Participants
Index test Reference test Results (95% CI, calculated by SBU) a Study quality

Comments
MRI
Girometti

2012

[16]

Italy
Prospective

Recruitment: consecutive with ≥1 previous negative biopsy, 2010–2011

Participants: n = 26

Age: median 64 (51–74)

PSA: median 6.0 (2.5–9.7) ng/ml

Disease prevalence: 5/26 = 19%
3T, T2W, DWI, DCE-MRI

Observers: n = 2

Blinding: yes
≥8 systematic biopsies

Observers: not reported

Blinding: not reported
T2W, DWI, DCE-MRI, MRSI

Se: 5/5 = 100%

(48; 100)

Sp: 8/21 = 38%

(21; 59)
Low

Poorly described methodology

Small sample
 
Tamada

2011

[17]

Japan
Retrospective

Recruitment: 2006–2009

Participants: n = 50

Age: mean 70 (40–84)

PSA: median 6.7 (4.1–9.9) ng/ml

Disease prevalence: 35/50 = 70%
1.5T, T2W, DWI, DCE-MRI

Observers: n = 2

Blinding: yes
12 systematic biopsies

Observers: not reported

Blinding: not reported
T2W, DWI, DWI, DCE-MRI

Se: 29/35 = 83%; (66; 93)

Sp: 12/15 = 80%, (52; 96)
Low

Retrospective

Poorly described population

Probable selection bias (high cancer prevalence)
 
Vilanova

2011

[18]

Spain
Retrospective

Recruitment: no previous biopsy, 2008–2009

Participants: n = 70

Age: mean 64 (43–87)

PSA: median 7.4 (4.0–17.2) ng/ml

Disease prevalence: 38/70 = 54%
Endorectal

1.5T, T2W, DWI, DCE-MRI, MRSI

Observers: n = 3

Blinding: yes
10 systematic biopsies

(n = 57) or histopathology after prostatectomy (n = 13)

Observers: not reported

Blinding: not reported
T2W, DWI, DCE-MRI

Se: 30/38 = 79% (63; 90)

Sp: 26/32 = 81% (64; 93)

T2W, DWI, DCE-MRI, MRSI:

Se: 36/38 = 95% (83; 99)

Sp: 26/32 = 81% (65; 91)
Low

Retrospective

Poorly described population
 
Ultrasound
Ho

2012

[19]

Malaysia
Prospective

Recruitment: no previous biopsy, time period not stated

Participants: n = 140

Age: not reported

PSA: above 4.0 ng/ml

Disease prevalence: 30/140 = 21%
Transrectal ultrasound with Doppler

Observers: n = 1

Blinding: yes
Biopsies according to Vienna nomogram

Observers: not reported

Blinding: not reported

Index- and reference test performed at the same time
Se: 29/30 = 96.7% (83; 99)

Sp: 27/110 = 24.5% (18; 33)
Low

Poorly described methodology
 
Kuligowska

2001

[20]

USA
Retrospective

Recruitment: 1995–1999

Participants: n = 544

Age: mean 63 (36–90)

PSA: 85% of participants above 4.0 ng/ml

Disease prevalence: 190/544 = 35%
Transrectal ultrasound with Doppler

Observers: n = 1

Blinding: yes
10–14 systematic biopsies

Observers: not reported

Blinding: not reported
Se: 82/190 = 43% (36; 51)

Sp: 235/354 = 66% (62; 71)
Low

Retrospective

Poorly described population
 
Pepe

2010

[21]

Italy
Prospective

Recruitment: consecutive, 2007–2008

Participants: n = 60

Age: median 64 (53–72)

PSA: not reported

Disease prevalence: 23/60 = 38%
Contrast enhanced transrectal ultrasound with Doppler

Observers: not reported

Blinding: yes
18–34 systematic biopsies. (different for re-biopsy and first time biopsy)

Observers: not reported

Blinding: not reported

Index- and reference test performed at the same time
Se: 7/23 = 30.4% (13; 53)

Sp: 22/37 = 59.4% (42; 75)
Low

Poorly described methodology and population

a DWI, diffusion-weighted imaging. DCE-MRI, dynamic contrast enhanced magnetic resonance imaging; MRSI, magnetic resonance spectroscopic imaging; PSA, prostate specific antigen; Se, Sensitivity; Sp, Specificity. # Confidence intervals calculated by us.

3.3. Risk of bias within studies

All six studies were graded as high risk of bias according to QUADAS combined with more specific predetermined criteria ( Table 1 ). The main reason for high risk of bias was inadequately described study population or methodology part ( Table 2 ). The intention was to combine sensitivity and specificity of separate reasonably homogeneous studies of low or moderate risk of bias only. However, only studies with high risk of bias were identified. In order to not lose information it was decided to present and combine data from these studies.

3.4. Results of individual studies

Characteristics and results of included studies are presented in Table 2 .

3.4.1. MRI

Three studies, published between 2011 and 2012 examined the diagnostic accuracy of magnetic resonance imaging. The studies included patients between 2006 and 2011. The study by Girometti et al. [12] was a prospective study performed at 3 T while the other two studies were retrospective and performed at 1.5  T [13] and [14]. The cancer prevalence in the studies ranged between 19 and 70%. In Vilanova et al. [14] an endorectal surface coil was used, while the other two studies used external coils. In Tamada [13] and Vilanova [14] MRI was performed before biopsy, and in Girometti MRI was performed after at least one negative biopsy, which was performed within 2–35 months prior to the MRI. Tamada and Vilanova showed higher specificity (72–80%) per patient compared to the study by Girometti in which MRI was performed after at least one negative biopsy round (38%). Girometti et al. showed a negative predictive value of 100%.

3.4.2. Transrectal ultrasound with Doppler

Three studies, published between 2001 and 2012, examined the diagnostic reliability of transrectal ultrasound with Doppler. The studies by Ho et al. [15] and Pepe et al. [16] were prospective, while the study by Kuligowska et al. [17] was retrospective. All three studies used Doppler technique to enhance the diagnostic information. In the study by Pepe ultrasound contrast was used in combination with Doppler. Kuligowska included both patients with prior negative biopsy and patients with no prior biopsy. Ho and Pepe included only patients without prior biopsy. Ho showed the highest sensitivity of 67%, but at the expense of a low specificity of 25%. Pepe showed the lowest sensitivity of 30%, but with a higher specificity of 59%.

3.5. Synthesis of results

3.5.1. MRI

A pooled analysis of the results from the three MRI studies showed a sensitivity of 82% (95% CI = 72, 90) and a specificity of 68% (95% CI = 55, 79).

3.5.2. Transrectal ultrasound with Doppler

A pooled analysis of the results from these three studies performed with transrectal ultrasound and Doppler showed a sensitivity of 49% (95% CI = 42, 55) and a specificity of 57% (95% CI = 52, 61).

3.6. Quality of evidence (GRADE)

The quality of evidence was insufficient to assess the diagnostic value of imaging since all selected studies had high risk of bias. The most common reasons were methodology for assessment of images including blinding between index and reference tests.

4. Discussion

This systematic review shows that the evidence to support the use of imaging for diagnosis of prostate cancer is limited and emphasizes the need for more high-quality prospective trials in the field of PC detection. In total, only six publications were considered relevant. These publications all were regarded as having a high risk of bias. Methodological limitations regarding evaluation of images and lack of blinding to reference methods were common reasons. There was also a wide range of prevalence of cancer in the published studies (19–70%) indicating different selection of patients. Our findings are in line with recently reported clinical guidelines by the National Institute for Health and Care Excellence (NICE) where the reviewed studies specifically evaluating the accuracy of MRI before biopsy were assessed as being of low quality ( http://publications.nice.org.uk/prostate-cancer-diagnosis-and-treatment-cg175 ).

A challenge in this group of patients is that the current reference test, transrectal ultrasound guided biopsies, is only a minor sample of the prostate and not a true reference regarding tumor extent within the prostate. Cancer detection rate or number of biopsies needed to achieve diagnosis are other endpoints that can be used, especially when evaluating the usability of new imaging directed biopsy techniques, such as ultrasound-MRI image fusion or direct MR-guided robotic biopsies.

The rapid development of imaging technologies is a challenge when it comes to establishing scientific evidence. This is obvious in this review regarding MRI, where no studies published before 2011 remained. One may argue that this review restricted the number of possible studies using MRI by considering ESUR expert consensus minimal standard recommendations [8] as part of the inclusion criteria. The question of whether or not more restricted MR imaging protocols can be applied for detection of PC was not the focus of this review.

One of the most important strengths of this study is the profound methodology being performed in line with international standards for systematic reviews (PRISMA). The included studies have been assessed with validated instruments (QUADAS and AMSTAR) and the report has been reviewed in various instances, both internally by the scientific councils of the agency and externally by four independent experts in the field (acknowledgements).

Based on the current evidence using TRUS guided systematic biopsies, the costs of the studied imaging techniques for detection of prostate cancer do not motivate use outside clinical trials. In a complementary search for cost-effectiveness studies, 403 studies were identified ( http://www.sbu.se/sv/Publicerat/Alert/Bilddiagnostik-vid-misstankt-prostatacancer ). Of these, only one (Stadlbauer et al., 2011) was included [18] . The study was considered to have high risk of bias since it had not included all relevant aspects on cost and effects (e.g. only the costs of the diagnostic technologies were included). Stadlbauer et al. compared MRI followed by prostate biopsies in patients with positive findings in the MRI with taking prostate biopsies only on all patients and the conclusion was that no clear recommendation could be made based on the results. The lack of studies dealing with relevant aspects of costs indicates a need for more knowledge in this field. However, to determine the cost-effectiveness of the new imaging techniques, the increased costs must be weighed not only against the diagnostic efficacy in terms of sensitivity and specificity but also in relation to the value of achieving a non-invasive diagnosis and the potential possibility of avoiding costs related to complications of interventions such as multiple biopsies. To assess the value of the technologies for patients, in terms of quality of life and survival, it is also necessary to establish the clinical value of improving the diagnostic accuracy.

In summary, the results of this systematic review indicate that despite clinical use, the quality of evidence is insufficient to assess the diagnostic value of imaging for detection of PC using TRUS guided systematic biopsies as reference standard. Thus, considering the costs and requirements to appropriately use imaging technology in this large patient group, it should currently be used under careful consideration and preferably restricted to clinical trials.

In conclusion, the number of imaging studies with sufficient scientific quality is presently too limited to be able to recommend use outside clinical trials for this patient group.

Conflict of interest

None declared.

Acknowledgements

For external review of manuscript: Jonas Hugosson, Professor, Sahlgrenska University Hospital, Gothenburg; Anders Magnusson, Professor, Uppsala University Hospital, Uppsala; Sten Nilsson, Professor, Karolinska University Hospital, Stockholm; Camilla Thellenberg Karlsson, M.D., Ph.D., Norrland University Hospital, Umeå.

For literature search: Agneta Brolund, Information Specialist, Swedish Council on Health Technology Assessment, Stockholm.

For input regarding evaluation of diagnostic tests: Sofia Traneus, Project Leander, Swedish Council on Health Technology Assessment, Stockholm.

For administrative support: Anna Attergren Granath, Project Assistant, Swedish Council on Health Technology Assessment, Stockholm.

References

  • [1] F.H. Schröder, J. Hugosson, M.J. Roobol, T.L. Tammela, S. Ciatto, V. Nelen, et al. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med. 2009;360:1320-1328
  • [2] T.A. Stamey, N. Yang, A.R. Hay, J.E. McNeal, F.S. Freiha, E. Redwine. Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med. 1987;317:909-916 Crossref
  • [3] B. Djavan, V. Ravery, A. Zlotta, P. Dobronski, M. Dobrovits, M. Fakhari, et al. Prospective evaluation of prostate cancer detected on biopsies 1, 2, 3 and 4: when should we stop?. J Urol. 2001;166:1679-1683
  • [4] R. Raaijmakers, W.J. Kirkels, M.J. Roobol, M.F. Wildhagen, F.H. Schrder. Complication rates and risk factors of 5802 transrectal ultrasound-guided sextant biopsies of the prostate within a population-based screening program. Urology. 2002;60:826-830 Crossref
  • [5] M.J.A. Garcia, E.M. Ward, M.M. Center, Y. Hao, R.L. Siegel, M.J. Thun. Global cancer facts & figures. (American Cancer Society, Atlanta, GA, 2007)
  • [6] R. Luengo-Fernandez, J. Leal, A. Gray, R. Sullivan. Economic burden of cancer across the European Union: a population-based cost analysis. Lancet Oncol. 2013;14:1165-1174 Crossref
  • [7] L. Dickinson, H.U. Ahmed, C. Allen, J.O. Barentsz, B. Carey, J.J. Futterer, et al. Magnetic resonance imaging for the detection, localisation, and characterisation of prostate cancer: recommendations from a European consensus meeting. Eur Urol. 2011;59:477-494 Crossref
  • [8] J.O. Barentsz, J. Richenberg, R. Clements, P. Choyke, S. Verma, G. Villeirs, et al. ESUR prostate MR guidelines 2012. Eur Radiol. 2012;22:746-757 Crossref
  • [9] P. Whiting, A.W. Rutjes, J.B. Reitsma, P.M. Bossuyt, J. Kleijnen. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol. 2003;3:25 Crossref
  • [10] Assessment SCoHT. Evaluation and synthesis of methods in healthcare: a handbook. (SBU, Stockholm, Sweden, 2013)
  • [11] H. Balshem, M. Helfand, H.J. Schunemann, A.D. Oxman, R. Kunz, J. Brozek, et al. GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64:401-406 Crossref
  • [12] R. Girometti, M. Bazzocchi, G. Como, G. Brondani, M. Del Pin, B. Frea, et al. Negative predictive value for cancer in patients with “Gray-Zone” PSA level and prior negative biopsy: preliminary results with multiparametric 3.0 tesla MR. J Magn Reson Imaging. 2012;36:943-950 Crossref
  • [13] T. Tamada, T. Sone, H. Higashi, Y. Jo, A. Yamamoto, A. Kanki, et al. Prostate cancer detection in patients with total serum prostate-specific antigen levels of 4–10 ng/mL: diagnostic efficacy of diffusion-weighted imaging, dynamic contrast-enhanced MRI, and T2-weighted imaging. AJR Am J Roentgenol. 2011;197:664-670 Crossref
  • [14] J.C. Vilanova, C. Barcelo-Vidal, J. Comet, M. Boada, J. Barceló, J. Ferrer, et al. Usefulness of prebiopsy multifunctional and morphologic MRI combined with free-to-total prostate-specific antigen ratio in the detection of prostate cancer. AJR Am J Roentgenol. 2011;196:W715-W722 Crossref
  • [15] C.C. Ho, T.W. Khor, P. Singam, E.H. Goh, G.H. Tan, B. Bahadzor, et al. Evaluation of power Doppler ultrasonography for prostate biopsy in men with elevated serum prostate specific antigen levels. Clin Ter. 2012;163:211-214
  • [16] P. Pepe, G. Candiano, M. Pennisi, F. Aragona. Can Sonovue targeted biopsy replace extended or saturation biopsy in prostate cancer diagnosis? Our experience at primary and repeat biopsy. Arch Ital Urol Androl. 2010;82:155-159
  • [17] E. Kuligowska, M.A. Barish, H.M. Fenlon, M. Blake. Predictors of prostate carcinoma: accuracy of gray-scale and color Doppler US and serum markers. Radiology. 2001;220:757-764 Crossref
  • [18] A. Stadlbauer, R. Bernt, E. Salomonowitz, E. Plas, G. Strunk, K. Eberhardt. [Health-economic evaluation of magnetic resonance imaging before biopsy for diagnosis of prostate cancer]. Rofo. 2011;183:925-932 Crossref

Footnotes

a Department of Diagnostic Radiology, Karolinska University Hospital, Solna, Sweden

b Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden

c Department of Urology, Karolinska University Hospital, Solna, Sweden

d Department of Clinical Physiology, Sahlgrenska University Hospital, Gothenburg, Sweden

e School of Health and Medical Sciences, Örebro University, Örebro, Sweden

f Department of Urology, Örebro University Hospital, Örebro, Sweden

g The Swedish Council on Health Technology Assessment, Stockholm, Sweden

lowast Corresponding author at: Department of Diagnostic Radiology, Karolinska University Hospital, Solna, S-171 76 Stockholm, Sweden. Tel.: +46 851776117; fax: +46 851774583.