European Urology

It is remarkable that an idea—targeting biopsies to an area of high suspicion within the prostate—previously rejected by the urologic community when ultrasound was used [1], is now deemed worthy of consideration, albeit with a different imaging modality [2]. It is the performance characteristics of the modality (multiparametric magnetic resonance imaging [MRI]), the evaluation of the whole gland, and the preferential detection of lesions of greater volume and Gleason grade, which make it such a promising avenue for further study.

The studies of MRI-targeted biopsy to date lack rigorous methodology and reporting, as outlined in the Standards for the Reporting of Diagnostic Accuracy criteria [3]. We observed a consistently poor characterisation of the populations from which men were drawn [4]. This information is essential in making inferences on appropriateness and sampling efficiency. The choice of denominator in the calculation of the utility of a diagnostic test is also important. The detection rate, on a per-patient basis, of a targeted approach could be reported per man undergoing targeted biopsy (66% in a biopsy-naïve population, 70% in a negative-biopsy population) or for all men in the diagnostic population (with a elevated prostate-specific antigen (PSA) level or abnormal digital rectal examination [DRE]), yielding positive biopsy rates of 41% and 48%, respectively. Other areas of reporting that appeared consistently deficient were the description of the sampling technique used, consideration of the independence of targeted samples from standard cores, the number of cores per target, potting and embedding protocols, and histologic reporting rules. These should all be clearly described. To address these deficiencies, an international working party has been set up to address the specific requirements of reporting in MRI-guided prostate biopsy studies.

An economic evaluation of an MR-guided approach to prostate cancer detection is critical to determining whether this is a feasible approach for men presenting for prostate cancer risk assessment. Economic modelling forms an important part of two recently funded UK trials. The first, Prostate MRI Imaging Study (PROMIS), will assess the utility of MRI as the first investigation of men with an elevated PSA level or abnormal DRE [5]. Multiparametric MRI and ultrasound data are collected and compared to both standard transrectal ultrasound–guided biopsy and template-guided transperineal biopsy with a 5-mm sampling density. In addition, a biobank of imaging, and urine and serum specimens will be set up for future analysis. The second funded trial will investigate the use of computer software (SmartTarget) designed to translate a lesion delineated on any MRI platform to any ultrasound based platform [6]. Use of software such as this may allow the transfer of targeting techniques from centres with considerable expertise to those that wish to develop such an MRI-guided biopsy strategy at minimal cost. The additional cost of imaging earlier in the diagnostic pathway would need to be offset by a reduction in both the cost of repeated investigation of those men who test negative at first standard biopsy and a reduction in the numbers of men who have surveillance or treatment for cancer that is unlikely to become clinically relevant. This will occur only if there is acceptance that some men who have a histologic diagnosis of prostate cancer will not benefit from treatment of that cancer. This idea of selective detection of a condition that is relevant to clinical practice, rather than simply a detectable abnormality, was recently succinctly expressed by Bossuyt and colleagues [7].

We anticipate intensive efforts to apply these principles in the diagnosis of prostate cancer. “Screen smarter” was Ruth Etzioni's challenge at the May 2012 American Urological Association meeting, when addressing the US Preventive Services Task Force recommendations [8]. Replacing random biopsies with a targeted approach might allow us to do that.

Conflicts of interest

Mark Emberton receives research support from the UK National Institute of Health Research University College London Hospitals/University College London Comprehensive Biomedical Research Centre; receives funding from US HIFU, GlaxoSmithKline, and Advanced Medical Diagnostics for clinical trials; is a paid consultant to Steba Biotech; and has received funding from Focused Surgery Inc, Misonix Inc, Oncura (part of GE Healthcare) for medical consultancy and travel to conferences. Caroline M. Moore is a paid consultant to Steba Biotech; has received support for conference attendance from Sanofi, Steba Biotech, Pfizer, and Ipsen; is a proctor for Prostate Mapping; and receives research support from GlaxoSmithKline and Advanced Medical Diagnostics. Samir S. Taneja is a consultant for Eigen Corporation, a consultant and scientific investigator for GTx Biotech, a scientific investigator for Steba Biotech, a speaker for Janssen, an editor for and recipient of book royalties from Elsevier, and a scientific investigator for the US National Institutes of Health (NIH) National Cancer Institute and receives scientific support/funding from the NIH, the US Department of Defence, and the Joseph and Diane Charitable Trust. The other authors have nothing to disclose.