Efficacy of Using Three-Tesla Magnetic Resonance Imaging Diagnosis of Capsule Invasion for Decision-Making About Neurovascular Bundle Preservation in Robotic-Assisted Radical Prostatectomy

Prostate cancer (PC) increasingly presents as early-stage disease clinically owing to increased screening, including prostate-specific antigen (PSA) screening. The standard of care for organ-confined cancers has been retropubic radical prostatectomy, which carries a substantial risk of morbidity, including incontinence and impotence [1]. Robot-assisted radical prostatectomy (RARP) provides improved visualization of the surgical field and improved instrument control compared with open and laparoscopic prostatectomy [2]. However, surgeons performing RARP lack the tactile feedback upon which they have traditionally relied to determine the extent of resection [3].

In this situation, preoperative detection of extracapsular extension (ECE) may be necessary to guide the surgical strategy in radical prostatectomy, to achieve PC-negative margins, and to spare the neurovascular bundles (NVBs) as much as possible to preserve erectile function and good postoperative continence [4-6].

In most cases, current PC staging is based on clinical assessment, notably, digital rectal examination (DRE), to sense a nodule or an extraprostatic rigid mass during prostate palpation. This clinical approach seems outdated, however, because DRE has low specificity [7]. Prostatic magnetic resonance imaging (MRI) appears to be a promising method for detecting PC and even for evaluating ECE during the pretreatment workup [8-10]. However, data about the specific role of prostatic MRI in PC staging are still lacking [11]. Moreover, racial differences have been reported in PC tumor aggressiveness and invasion characteristics [12]. Specific racial guidelines for decision making about nerve sparing may need to be established.

In this study, we evaluated the utility of 3-T MRI for assessing ECE and indicating the appropriateness of NVB sparing during RARP in a Japanese patient population.

Approximately 40% of patients with localized PC choose some form of surgical resection for treatment [15]. In any surgical approach, surgeons must balance the desire to achieve a cancer-free resected margin with the need to minimize postoperative morbidity, which may involve incontinence and erectile dysfunction [16]. The NVB, which mediates erectile function, lies posterolateral or lateral on the prostatic capsule and adjacent to the PZ of the prostate, where 70% of PCs arise [17]. In RP, surgeons typically identify and spare the NVB if possible; however, cases with suspicious ECE need to be widely resected to include the NVB and surrounding tissues to achieve negative surgical margins. This procedure may not be easy in traditional open RP owing to individual patient anatomy or severe blood loss [18].

The goal in nerve-sparing RP is to preserve the greatest amount of nerve tissue possible without compromising surgical margins. Robotic-assisted surgery has been spreading in Japan, and RARP was included under government medical insurance in 2012. Robotic technology is a step forward, as it provides increased magnification, high-definition imaging, and wristed instrumentation and is associated with significantly less blood loss than open surgery [19]. For optimal nerve-sparing outcome, the preoperative search for ECE in the prostate and the accurate staging of PC appear to be key points in the pretherapeutic workup for indicating whether the nerve-sparing approach is feasible. There are currently no guidelines for this in Japan.

Unlike clinical variables (PSA values and findings from DRE), results from MRI are spatially localized and allow surgeons to individually sculpt the extent of surgical resection as mentioned above [3]. In recent years, MRI with field intensities of 3 T, a significant increase in the signal compared with 1.5-T MRI, has become commonplace [20]. Three-T MRI maintains imaging quality while significantly reducing imaging time and increasing the signal-to-noise ratio up to twofold. Because of the increased signal-to-noise ratio and the improved spatial resolution at 3 T, improvements in the localization and detection of PC can be expected [21]. Regarding the efficacy of 3-T MRI for ECE determination, a previous study showed that sensitivity was 66.7% and specificity 100% for the detection of ECE in 27 PC cases [22]. In another study, the accuracy of 3-T endorectal MRI prediction of ECE was 75% [23]. Our data showed an overall sensitivity and specificity for predicting ECE of 60.0% and 86.0%, respectively, which is comparable with previous studies.

According to the literature, 1.5-T MRI performed with an endorectal coil is currently the standard imaging method for staging PC [9]. However, this method has several problems related to examination tolerance, movement, near-field effect, capsular profile deformation artifacts connected to coil use, and cost [24]. Staging by 1.5-T MRI with an endorectal coil shows extremely variable results (a detection range of 13% to 95% for ECE and 25% to 72% for extension to seminal vesicles) [9, 11]. Our study used 3-T MRI with a phased-array pelvic coil that allows fewer artifacts and thus provides comparatively acceptable quality images for decision-making about nerve-sparing surgery.

DWI is a complementary functional technique that may have utility in the detection, quantification, and grading of PC [25]. DWI data can be postprocessed to give apparent diffusion coefficient maps, which assist in detection and localization. Dynamic contrasted-enhanced (DCE) MRI is another complementary functional MR technique that assesses the relative tissue perfusion within the prostate. Detection and characterization are improved by the addition of DCE-MRI to T2-weighted images [26]. For overall PC detection, multiparametric MRI showed better quality than any individual MRI sequence [27]. In this study, we used T2-weighted imaging, DWI, and DCE-MRI for PC staging, which may have contributed to our results showing a statistically significant trend for the surgeon to perform fewer NVB-sparing procedures if the MRI reported ECE than if no ECE was reported. The same trend was also mentioned by Roethke et al as significant (p<0.01) in their study [28]. An important question is the influence of preoperative MRI on the positive surgical margin rate. In our study, patients with ECE on MRI had a higher positive surgical margin rate than did patients who were not suspicious for ECE. Additionally, there were no positive surgical margins with nerve-sparing procedures in the group shown to be ECE negative on MRI.

This study have some limitations. First, the number of cases may not have been enough for definitive conclusions. Second, we did not use an endorectal coil. Even though an endorectal coil could have provided better spatial resolution, this approach has several limitations, including increased cost and examination time, a nonuniform signal across the prostate, and an increase in motion artifacts owing to rectal peristalsis. Third, this was a single-arm study and did not include a comparative group, for instance, a 1.5-T MRI group. These limitations will be overcome in our future studies.

We found that 3-T MRI showed comparatively acceptable results for staging PC and accurately detecting ECE to guide decision-making for nerve-sparing surgery in RARP. Our data offer evidence that 3-T MRI might improve decision-making about nerve-sparing surgery, although a prospective study with a comparison group and larger number of cases is still needed.