Prebiopsy multiparametric prostate MRI (mp-MRI), followed by transrectal ultrasound-guided (TRUS-G) target biopsies (TB) of the prostate is a key combination for the diagnosis of clinically significant prostate cancers (CSPCa), to avoid prostate cancer (PCa) overtreatment. Several techniques are available for guiding TB to the suspicious mp-MRI targets, but the simplest, cheapest, and easiest to learn is “cognitive,” with visual registration of MRI and TRUS data. This review details the successive steps of the method (target detection, mp-MRI reporting, intermodality fusion, TRUS guidance to target, sampling simulation, sampling, TRUS session reporting, and quality insurance), how to optimize each, and the global indications of mp-MRI-targeted biopsies. We discuss the diagnostic yield of visually-registered TB in comparison with conventional biopsy, and TB performed using other registration methods.
The positive diagnosis of prostate cancer (PCa) requires a direct sampling of the gland. This is an invasive procedure, with rare, but not negligible, potential complications [
Prostate MRI has long time been used to assess PCa stage only, after positive biopsies, in selected patients. The advents of new imaging techniques (dynamic contrast-enhanced (DCE) imaging; diffusion-weighted imaging (DWI)), and new imaging protocols (high-resolution external phased array coils) have radically modified the way this examination is used [
Various techniques have been described to guide biopsies on mp-MRI targets. The simplest is to mentally perform a “
In this paper, we will describe how and when
Comparatively to MRI “in-bore” biopsies, MRI-targeted prostate biopsies performed under TRUS guidance have three constraints: (a) they are based on mp-MRI data acquired under different circumstances (without or with an endorectal balloon), implying a different geometry and a different environment (bladder or rectal repletion); (b) they are usually performed by different physicians; (c) they are manually guided. Thus, TRUS-G mp-MRI targeted biopsies (TB) will require a good TRUS operator experience, a clear communication between physicians, and a reliable registration of mp-MRI and TRUS data.
TRUS-G mp-MRI targeted biopsies can be performed using different techniques which slightly vary, depending on the biopsy route (transrectal or transperineal), on the registration method (visual or software-assisted), and on the guidance method (manual, robot-assisted), but can be summarized into 7 common successive steps (Table
Seven-step protocol for prostate biopsies performed under TRUS guidance with mp-MRI targeting.
Actors | Steps | Summary | Description | |
---|---|---|---|---|
MRI room | |
1 | Target detection | Detection of mp-MRI lesions having a cancer suspicion score ≥3 and a significant size. A mean of two targets is a good compromise, with a primary target clearly identified. |
2 | Target reporting | Transmission of intelligible and accurate information to the physician that will perform the biopsy procedure. | ||
|
||||
TRUS biopsy room | Urologist/radiologist/both | 3 | Intermodality fusion | Registration (or “fusion”) of static, asynchronous, and multiparametric MRI data with that of a real-time and dynamic TRUS data. |
4 | TRUS guidance to target(s) | Guidance of the biopsy needle gun to the correct location of the mp-MRI target within the TRUS image volume. | ||
5 | Sampling simulation | Simulation the sampling helps the physician assess the quality of the sampled core, as well as the safety of the sampling. | ||
6 | Tissue sampling | Tissue sampling is usually performed using a semiautomatic needle biopsy gun triggered by the physician. | ||
7 | TRUS biopsy report and quality insurance | Reporting of the biopsy procedure, including the location of this additional cores and the correspondence with suspicious mp-MRI image. |
This first step is common to all MRI-targeted biopsies. Radiologists usually describe MRI targets with minimal threshold sizes (6-7 mm in the peripheral zone (PZ) and 10 mm in the transition zone (TZ)), in a standardized manner described in the 2012 ESUR guidelines, to avoid describing a myriad of suspicious images. Because of this threshold size, suspicious mp-MRI lesions (having a score ranging from 3 to 5 out of 5 on the Pi-RADS scale) are de-facto considered as suspicious of clinically significant prostate cancer (CSPCa), and several studies have emphasized the fact that prebiopsy mp-MRI targeting increases diagnosis yield of PCa and CSPCa on biopsies [
As mp-MRI and TRUS-G biopsies will be performed in different rooms, at different times, and usually by two different operators having two different specialties, the quality and accuracy of mp-MRI reporting is fundamental. The decision of performing (or not) a TB on an image, the selection of the best targets, and the matching of MRI and ultrasound data will highly depend on it.
The simpler reporting is a plain-text prose description of the lesion(s), detailing its (their) location (based on the recommended 27 regions standard diagram), size (in mm) in at least 2 dimensions, MRI appearance on T2, DWI and DCE sequences, Pi-RADS suspicion score (5-point scale), and likelihood of extraprostatic spread.
It is recommended to enhance the report with a graphic diagram of the prostate, on which mp-MRI targets are drawn (manually or electronically), optionally using a color scale for the Pi-RADS score (Figure
Standardized mp-MRI target reporting. 69-year-old man enrolled into active surveillance without MRI, one year ago. First AS control with prebiopsy MRI shows a suspicious (4/5) image in the anterior horn of the right PZ. Transmission of this information to the urologist performing the TB is made using the recommended standardized 27-sector diagram, common to radiologists, urologists, and pathologists in the institution. It is simply pasted at the end of the traditional text report. A screenshot of the workstation, centred on the image, is also copy/pasted in the report and saved into the PACS system. This information is available at time of TRUS biopsies. TB confirmed the diagnosis with 1 out of 2 cores positive (5 mm; Gleason
Another enhancement of the reporting is the addition of key images on the diagram for each lesion. These key images may include orthogonal projection for accurate craniocaudal localization (Figure
In any event, we recommend physicians performing the TRUS-G biopsies to review the MRI examination a few minutes before the TRUS sessions on the PACS (if possible in presence of the radiologist who interpreted the case) to select targets in the best conditions.
Performing TB under TRUS guidance, with the visual help of the MRI images alone is called “
Mp-MRI targeted TRUS-G biopsy of clinically significant prostate cancer. 66-year-old patient with a PSA of 8 ng/mL. Prebiopsy mp-MRI (a) shows a 9 mm low T2 nodule (image (a); upper row), with high retriction of water diffusion and hypervascularization (image (a); lower row) in the anterior horn of the right apical PZ, ahead of the posterior 18 mm of gland sampled by SB cores. A TB was performed with knowledge of this information. It diagnosed a CSPCa (4 out of 4 positive TB; no SB was positive). Image (b) shows the trace of the needle biopsy gun inside the nodule. This lesion was aimed with visual registration, thanks to its zonal anatomy (ahead of the anterior TZ, at the edge of the anterior prostate surface), its size, and the presence of a small cyst in the right TZ (not visible on image (b), but used to locate the craniocaudal location of the lesion at time of biopsy).
Use of anatomical and stereotactic landmarks for visual registration of 3 sample mp-MRI targets located in the “gray zone.” This figure shows three schematic representations of prostate base, midgland and apex ((a); top to bottom) sections, represented by gray dashed lines on the sagittal and coronal views ((b); top and bottom, resp.). The standard 12 systematic biopsy (SB) cores are represented by dotted arrows. The red lesion is a typical anterior apex cancer; the blue lesion is right TZ cancer, and the purple lesion a PZ cancer located on the midline. All have a clinically significant volume. Visual registration of a mp-MRI target on the TRUS image can be helped by multiple anatomical landmarks or simple three-dimensional distances, both on the craniocaudal plane (ejaculatory ducts (
When the physician has located the visually registered TRUS target, he has to guide the biopsy needle gun to the target. This guidance is usually helped by the overlay of a dotted line symbolizing the needle direction on the real-time image.
Targets located in the PZ will be easy to aim at, because they are immediately located at the tip of the biopsy needle guide. Inversely, those located in the TZ will require a small learning curve: in a couple of seconds, the operator has to detect the TRUS target (step 3) with freehand motion, manually lock the probe position on the target, insert the needle into the gland, advance beyond the peripheral zone, stop the needle at the contact of the target, check that the needle will not transfix the gland, then trigger the semiautomatic needle gun. Depending on the target’s depth and on the gland elasticity, the prostate or the probe may slightly move during this process, and this explains why TB may not sample the target as accurately as desired and have to be repeated at least twice in order to maximize the chance of sampling a target correctly.
A rapid simulation of the biopsy path should be done, to avoid transfixion through the bladder or urethra. The operator has to check that there is no risk of bladder wall or urethra wound in the 20 mm beyond the tip of the needle. This may happen for TB located at base or for lesion located at the anterior apex (e.g., in the anterior fibromuscular stroma (AFMS)). In this case, freehand guidance of the needle to the target can be performed by brushing the urethral sphincter (thick hypeoechoic rim), then angulating the probe medially to aim the lesion. Some targets located on the midline may also be difficult to aim due to the lateralization of the needle guide on the ultrasound probe. Applying a high angulation on the probe and performing a TB almost in parallel to the posterior prostate surface can solve this issue (Figure
TRUS-G biopsy of two lesions requiring special needle orientation and experience for aiming TB. This 62-year-old man with a PSA of 8.25 ng/mL referred to our centre because persisting elevated PSA after 2 biopsy sessions. Last biopsy showed 2 positive cores at the right apex (2 mm each,
Guidance is the key step to the quality of TRUS-G TB (with visual or software registration). It can be improved by dedicating a physician to the probe manipulation and another one to cores sampling. An alternative option is to use a mechanical arm to lock the probe on the TRUS target, in order to handle the biopsy gun comfortably [
A standardized report of the biopsy session should be provided, including detailed notification of SBs and TBs that were performed. It can be done in the same manner as for MRI, using a standardized 27-sectors diagram of the gland, including drawings (freehand or computerized) of the sampled lesions, which may be different than those described at mp-MRI. It is also important to report how good the TRUS TB matched the mp-MRI one (e.g., visibility of the lesion). All this information will be useful for analysing final histopathology results. In case of positive TB, physicians will be able to compare core cancer length with the TRUS ballistic and determine the likelihood of having sampled a significant cancer. In case of negative sampling, biopsy ballistic report will help decide whether the lesion was a false mp-MRI positive or a significant image that may have not been correctly sampled.
Prebiopsy MRI was first proposed in patients after a negative first round of biopsies. This attitude was quickly suggested [
New treatment options (including focal therapy) and active surveillance (AS) highly depend on imaging results. In a series of 388 consecutive men eligible for AS, Vargas et al. have shown that a negative MRI (Pi-RADS scores 1-2) had an excellent (0.96–1) negative predictive value of CSPCa at confirmation biopsies, whereas a positive MRI (scores 3–5) was highly sensitive for upgrading on confirmation biopsies [
Consequently, main clinical indications of prebiopsy mp-MRI combined with TRUS-G TBs (visually or electronically registered) are (a) diagnosis of CSPCa in patients with a clinical or biological suspicion of PCa (first or
The evolution of prostate MR imaging has enabled its use before biopsies to detect cancers located in areas usually undersampled (anterior TZ, AFMS, anterior horns of the PZ, extreme apex, and base) by SB protocols and poorly detected at TRUS imaging. Consequently, in many centres, the practice of prostate biopsy has slightly evolved to prebiopsy mp-MRI, followed by SB and optional TB in case of suspicious mp-MRI target.
There are three broad categories of mp-MRI TB guidance under TRUS imaging: (a) cognitive fusion using “visual” registration (TB-VI); (b) software-assisted fusion using rigid registration (TB-FUr) (not changing MRI or TRUS data geometry) [
Cognitive fusion with visual registration (TB-VI) is the oldest, simplest, fastest, and cheapest technique. It was described in the mid 2000 [
In an extensive review written in 2011, Moore et al. [
It is not possible to compare geometric accuracies of cognitive and software fusion techniques together, but some studies have compared the diagnostic yield of PCa from TBs performed using both visual (TB-VI) and software-assisted (TB-FU) fusion techniques, in terms of positivity for cancer, cancer length per core, or Gleason score: Mouraviev et al. published a series of 32 patients and found significantly different detection rates of 33.3% versus 46.2%, and sensitivities of 45.5% versus 61.9% for TB-VI and TB-FUr, respectively [
Some disadvantages of TB-VI should be highlighted: (a) they are operator dependent. TB-VI require more experience and precision than SB, with good knowledge of both MRI and ultrasound semiology to be able to match MRI and TRUS images; (b) diagnostic accuracy might vary depending on lesion visibility on TRUS imaging, and lesion location, as TRUS and MRI do not have the same exploration planes: lesions located at the inferior part of the gland, MRI and TRUS axial sections will be visible on slices acquired on nearly identical planes, whereas images located at the upper part of the gland, or anteriorly, will not (Figure
Comparison of MR and TRUS imaging acquisition planes of the prostate. MRI (a) and TRUS (b) imaging are performed with different acquisition planes. MR imaging is usually acquired perpendicularly to the prostate wall (or strictly transverse), with a slice thickness varying from 2.5 to 5 mm, homogeneously distributed (red lines overlaid on image (a)). TRUS imaging is acquired using an endorectal probe inserted in rectum, through the anal canal which is a fixed point that limitates probe translation and angulation. Hence, TRUS images acquired at the apex will often match MRI, whereas those acquired at the base will require a registration step (visual or computer-assisted) to match MRI.
Prebiopsy mpMRI combined with TRUS-guided target biopsies of the prostate is a major step forward over systematic biopsies alone. It is becoming a key for the diagnosis of prostate cancer, because it detects significant cancers in areas usually undersampled by systematic biopsies. Those biopsies can be performed using visual or software-assisted registration of MRI and US data. In comparison with other MRI-US fusion techniques, visual registration is easier to learn, cheapest and simpler, making it compatible with daily office practice and a potential inclusion in the standard diagnostic pathway of PCa.
Anterior fibromuscular stroma of the prostate
Benign prostatic hyperplasia
Clinically significant prostate cancer
Dynamic contrast-enhanced MR imaging
Diffusion-weighted imaging
European Society of Uroradiology
Multiparametric prostate MRI
Picture archiving and communication system
Prostate cancer
Peripheral Zone of the prostate
Extended TRUS systematic posterior biopsies
Target biopsy
TBs performed manually under TRUS-G with software-assisted registration (rigid or elastic)
TBs performed manually under TRUS-G with software-assisted registration (elastic)
TBs performed manually under TRUS-G with software-assisted registration (rigid/isometric)
TBs performed manually under TRUS-G with visual registration
Transrectal ultrasound-guided
Transition Zone of the prostate.
The authors declare that there is no conflict of interests regarding the publication of this paper.