Ultrasound MRI Fusion Biopsy in Prostate Gland

Fig. 55.1

Images from a 59-year-old male with serum PSA 7.92 ng/mL, and one previous biopsy underwent an mp-MRI. The MRI demonstrated a PI-RADS 5 right posterior apex to mid-peripheral zone lesion (white arrow) on axial T2W (a), DWI (b), ADC (c), and DCE (d, e). MRI/US fusion software-based targeted biopsy demonstrated Gleason score 7 (3 + 4) prostate cancer (67 % in three cores)

To describe suspected lesions diagnosed by MRI in a standardized manner, radiologists use standardized suspicion scores and graphical templates to show locations. The most used scores are the 1–5-point Likert scale (based on radiologist’s subjective score) or the prostate imaging reporting and data system (PI-RADS) score (based on determined criteria) [11–13]. In particular, concerning PI-RADS score, the inter-reader agreement performs well, and the inter-reader reproducibility improves with increasing experience.

55.3 MRI-Guided Targeted Biopsy

An MRI targeted biopsy can be performed in three ways: in-bore MRI targeted biopsy, MRI/US fusion visual targeted biopsy, and MRI/US fusion software-based targeted biopsy. These three approaches are all informed by tumor location diagnosed by the MRI. It is just the manner in which the target volume is “represented” to the operator that differentiates them.

Concerning in-bore MRI targeted biopsies, needles are introduced only into the areas of interest by performing a transrectal or transperineal biopsy. Serial MRI scans are performed to confirm biopsy needle placement (Fig. 55.2). Multiple studies demonstrated that in-bore MRI targeted biopsies are feasible with a median detection rate significantly higher than random biopsies. Moreover, this approach reduces the number of sampled cores with a real-time feedback of its placement, allowing a high likelihood of hit target [14, 15]. Nevertheless, in-bore MRI targeted biopsy is time-consuming and costly, not commonly available, and performed in prone position under general anesthesia.

Fig. 55.2

In-bore biopsy. (a) Needle-in control scans are performed in two different planes (axial and coronal); (b) targeted cores are taken from each lesion using an MRI-compatible, 18G, fully automatic biopsy gun

The simplest targeted strategy concerns the use of MRI/US fusion visual targeted biopsies directed to the suspicious areas highlighted on the MRI. The first step, as in the other strategies, is represented by the detection of suspicious lesions on MRI. Then the urologist performs a standard US-guided biopsy, either by a transrectal or a transperineal approach, trying to direct the needles toward the areas suspicious on mp-MRI. Many authors suggest better efficiency and accuracy compared to standard biopsy [16, 17]. The most important disadvantage relates to the learning curve and reproducibility of this strategy. This approach requires an experienced urologist to translate the information of the mp-MRI onto real-time US, which can be challenging according to the deformation and the anatomical characteristics of the prostate.

Finally, MRI/US fusion software-based targeted biopsies represent a novel approach developed to improve the accuracy of prostate biopsy, allow dissemination of the technique, and permit the storage of images for future resampling. MRI/US fusion software-based targeted biopsy devices allow to align the pre-biopsy MR images with intraoperative TRUS in order to enable the urologist to perform targeted biopsy directed toward MR-visible lesions. This approach combines the high diagnostic accuracy of MRI for detecting PCa with TRUS, which represents a procedure well mastered by urologists. The process of coregistration of MRI and US images is automatized by the use of a fusion device, and therefore the results are likely to be more consistent across different centers.

55.3.1 Coregistration of MRI and US Images

MRI to US cognitive fusion is complicated by the significant deformation of the prostate shape that occurs between TRUS and MRI (with or without an endorectal coil). The software-based registration method corrects this effect to achieve better diagnostic accuracy [18].

There are two different methods to register MR images to live TRUS: rigid and elastic registration. Both of them aim to align the MR and US images through the identification of landmarks present on both corresponding images. The outer shape of the prostate is used to match the MRI contour to the live US image.

Elastic registration allows deformation, warping, and dimensional changes between images, based on mathematical algorithms. As every prostate is different in density and elasticity, these calculations are estimations. Rigid registration permits only rotational and translational variations between images, without changing the images themselves. The urologist needs to make some adjustments in case of error due to the rigid registration, using manual correction of the alignment and targeting or using different degrees of pressure/insertion depth of the US probe (Fig. 55.3) [19].

Fig. 55.3

Elastic and rigid methods to register MR images to live TRUS. (a) MRI/US registration with minimal US-probe deformation and use of an endorectal coil (ERC) for MRI; (b) MRI/US registration with increased manual US-probe deformation that can mimic ERC deformation. In middle images in (a) and (b) is shown the simple overlap of US and MR images, resulting in reduced correlation between imaging modalities. Rigid registration permits only rotational and translational variations between images. Elastic registration allows local deformation, e.g., caused by an endorectal coil or TRUS probe. ERC endorectal coil (Reproduced from Logan et al. [19])

While overlapped images obtained from rigid registration usually have discontinuous borders looking less pleasant to the eye than elastic registration, it is difficult to define which method is able to achieve better accuracy. Elastic registration should guarantee better matching, but some experts think the cognitive adjustment might overcome the issues encountered with rigid registration and allow better spatial precision, especially in patients having unusual gland dimensions.

55.3.2 Fusion Platforms

MRI/US fusion software-based targeted biopsy first of all requires a diagnostic mp-MRI with a report scheduling all the suspicious lesions edited by an expert uro-radiologist (Fig. 55.4). mp-MRI images are loaded in the specific software and regions of interest are then outlined. The patient is positioned, and a TRUS is performed, with MR images superimposed on real-time US images (Fig. 55.5). Targeted biopsies directed to mp-MRI-suspicious lesions are then performed (Figs. 55.6 and 55.7).

Fig. 55.4

mp-MRI report scheduling all the suspicious lesions classified by PI-RADS score

Fig. 55.5

Superimposition of MR and US images: the outer shape of the prostate is used to match the MRI contour to the real-time US image. (a) Longitudinal view; (b) transversal view

Fig. 55.6

Targeted biopsies directed to mp-MRI-suspicious lesions: transrectal approach

Fig. 55.7

Targeted biopsies directed to mp-MRI-suspicious lesions: transperineal approach

A standardized report of the biopsy session should be provided, including a detailed notification of MRI/US fusion software-based targeted biopsies, and eventually standard biopsies, that were performed. It can be done in the same manner as for mp-MRI, using a standardized diagram of the prostate, including drawings of the sampled lesions. The report must underline how good the MRI/US fusion software-based targeted biopsy matched the mp-MRI one (e.g., visibility of the lesion). All this information will be useful for analyzing the final histopathology results (Fig. 55.8).

Fig. 55.8

Example of a report of the biopsy session

The different devices currently in use to allow MRI/US fusion software-based targeted biopsy are reported in Table 55.1. To date, there are no available studies directly comparing the different platforms in terms of accuracy, nor detection rate.

Table 55.1

Summary of MR/TRUS fusion software-based targeted biopsy platform specifications

MRI/US fusion software (manufacturer)	US image acquisition	Method of registration	Tracking system	Manipulation	Sampling route	Year of FDA approval
Artemis (Eigen)	Manual	Elastic	Mechanical arm with encoders	Via mechanical arm	Transrectal or transperineal	2008
BioJet (D&K Technologies)	Manual	Rigid (elastic for minor deformations)	Stepper with digital encoders	Via stepper	Transrectal or transperineal	2012
BiopSee (MedCom)	Manual	Rigid (elastic for minor deformations)	Stepper with digital encoders	Via stepper	Transrectal or transperineal	NR
Real-time virtual sonography (Hitachi)	Manual	Rigid	Electromagnetic	Freehand	Transrectal or transperineal	2010
UroNav (Invivo/Philips)	Manual	Rigid	Electromagnetic	Freehand	Transrectal or transperineal	2005
Urostation (Koelis)	Automatic	Elastic	3D ultrasound	Freehand	Transrectal	2010
Virtual Navigator (Esaote)	Manual	Rigid	Electromagnetic	Freehand	Transrectal	2014

MRI/TRUS magnetic resonance imaging/transrectal ultrasound, FDA Food and Drug Administration, NR not reported

55.4 Indications of MRI/US Fusion Software-Based Targeted Biopsy

55.4.1 Main Indications

Re-biopsy in men with persistent suspicion of PCa after first negative prostate biopsy: persistently increased PSA and/or positive digital rectal examination (DRE) [20–25] and/or diagnosis of extensive high-grade prostatic intraepithelial neoplasia (HG-PIN) or atypical small acinar proliferation (ASAP) of the prostate [26]. As expected, a number of studies have shown that in this subgroup of men, MRI/US fusion software-based targeted biopsy allowed the detection of more clinically significant PCa than standard biopsy [27].
Follow-up of patients under active surveillance (AS). Many authors evaluated fusion systems to perform confirmatory targeted biopsy in patients under AS. Hu et al. recently proved in a series of 113 patients that confirmatory MRI/US fusion software-based targeted biopsy resulted in reclassification in 36 % of men, ranging from 24 to 100 % according to the MRI score, from low to high grade, respectively [28]. Sonn et al. demonstrated that in a series of 171 patients, MRI/US fusion software-based targeted biopsy was three times more likely to identify cancer than standard biopsy (21 % versus 7 %, respectively), and of the men with clinically significant PCa initially enrolled for AS, 38 % had disease detected only on targeted biopsies [29]. Moreover, MRI/US fusion software-based targeted biopsy permits to track the location of all biopsy cores, allowing the urologist to perform a re-biopsy in the same suspicious areas, which is mandatory in the correct follow-up of patients under AS.

55.4.2 Other Indications

Other indications, as recommended by many authors but to be confirmed by further studies, could be:

The follow-up of men suspicious for local recurrence after local treatment [30]
The guidance of focal therapy [31]
The characterization of suspicious lesions even at the first biopsy [32, 33]

55.5 Results of MRI/US Fusion Software-Based Targeted Biopsy

55.5.1 Standard Biopsy Versus MRI/US Fusion Software-Based Targeted Biopsy

The two approaches did not differ significantly in overall detection of PCa. When considering a core-by-core analysis, Rastinehad et al. reported an increased detection rate of MRI/US fusion software-based targeted biopsy with respect to standard biopsy (37.9 % vs 12.5 %, respectively, p < 0.001) [34]. The detection rate of clinically significant PCa seems higher performing a MRI/US fusion software-based targeted biopsy than performing a standard biopsy. In the study of Siddiqui et al., MRI/US fusion software-based targeted biopsy diagnosed 30 % more high-risk cancers versus standard biopsy (p<0.001) and 17 % fewer low-risk cancers (p = 0.002) [35]. On the other hand, two recently published randomized controlled trial (RCT) concluded that detection rates for any cancer and clinically significant PCa did not significantly differ between the two approaches, as reported later (see Sect. 55.6) [36, 37].

Only gold members can continue reading. Log In or Register to continue