Early urinary incontinence is one of the most feared complications of radical prostatectomy (RP) . Bladder neck preservation (BNP), first introduced by Klein , has been proposed as a method to accelerate continence recovery. In fact, several non-randomised controlled trials suggest that BNP is associated with significantly earlier continence recovery when compared with the standard BN dissection (although with similar long-term continence rates) [3-5]. In a multivariate analysis of several factors, Sakai et al.  reported BNP to be the only independent predictor of return to continence at 1 and 3 months.
The rationale for these continence outcomes might be based on the complete preservation of the musculus sphincter vesicae (preprostatic sphincter), that constitutes an integral part of the male urinary continence zone as described by Myers . Additionally, obtaining a BN diameter approximately equal to the diameter of the urethral stump simplifies urethro-vesical anastomosis and reduces the risks of urine leaks. BNP has also been associated with lower risk of BN contracture  and lower rates of ureteric injury .
On the contrary, a large BN requires time-consuming reconstructive tapering and may be more susceptible to anastomotic leak due to the longer suture line. Thus, when there is no positive biopsy of the prostate base or MRI suspicion of basal tumour extension, a BNP technique should be attempted.
Despite its advantages, BNP is probably the least standardised step of both laparoscopic and robotic RP. While in open RP the surgeon maintains the ability to feel the prostate and urethral catheter balloon during BN dissection as an important marker, in the aforementioned minimally invasive techniques, the widely noted absence of palpation renders BNP one of the most challenging steps, especially during the early learning curve [10, 11]. The difficulties in BN identification, expressed through the wide variety of techniques that have been proposed in order to facilitate it (use of a Lowsley retractor or Bèniquè dilator to elevate the BN to define the prostate-vesical junction , use of intraoperative ultrasonography  or simultaneous use of cystoscopy ), further show the difficulty of BNP.
However, the absence of haptics is, at least partially, balanced by the better view offered by laparoscopy/robotics. The enhanced view should aid the surgeon to obtain an accurate dissection in the vesico-prostatic septum, to develop natural planes, to identify and ‘circumscribe’ the urethral tube and finally, to preserve the BN.
The purpose of the present study is to describe our technique and anatomical landmarks for consistent BNP during laparoscopic or robotic RP.
The patient is placed in a steep (30 °) Trendelenburg position. An ‘inverse-U’ peritoneotomy is performed until the level of the vasa on either side to gain access in the Retzius space. The anterior periprostatic fat is ‘swept’ away. The endopelvic fascia is not incised laterally to the prostate. The catheter balloon previously filled to 5 mL is deflated. The robotic atraumatic grasper is advanced in the operating field, and gently retracts the bladder (or the grasper of the assistant in the pure laparoscopic approach). We start the dissection on the right side, in the limit between the detrusor and the base of the prostate, lateral to the pubovesical ligaments. Usually a venous branch underlying the endopelvic fascia and connecting the dorsal venous complex to the neurovascular bundle (NVB) veins is found here; the exact starting point of dissection lies at the level of this vessel.
The subsequent dissection of the medial aspect of the seminal vesicles, reaching the Denonvillier's fascia in depth, defines the lateral and inferior limits of the BN, while the energy-free dissection of the lateral aspect of the seminal vesicle releases the NVB in the area of the prostate-vesicular angle.
Consequently, the lateral and inferior aspect of the BN, the plane of the seminal vesicles and the NVB are exposed, i.e. several key-points of the anatomy of the prostate. Thus, the development of the described plane not only permits BNP but it also prepares and facilitates the subsequent surgical steps.
Once the linear fibres of the BN transitioning to prostatic urethra are identified laterally, a combination of sharp and blunt dissection to tease bladder muscle fibres away from the prostate is used. Monopolar cautery is avoided to reduce the amount of tissue charring and to preserve the visualisation of the native anatomy that allows for identification of bladder muscle fibres, critical for defining the natural tissue plane of the vesico-prostatic junction. The extensive use of bipolar diathermy is also avoided, as it could lead to local necrosis of the BN. Incorporating a necrotic area in the urethrovesical anastomosis could hinder the healing process leading to chronic urine leakage or even anastomotic dehiscence.
A large median lobe, if present, is more easily dissected through this lateral external approach.
When the lateral and the inferior surfaces of the BN are well-defined bilaterally, its anterior transection, performed with cold scissors, becomes easy and safe.
The anterior approach to the BN is often challenging: no standardised anatomical elements are present to safely guide the dissection. The use of the scissors on the ventral surface of the prostate and bladder (once the preprostatic fat has been removed) in order to ‘feel’ the limits of the prostatic base has been proposed; however, the lack of tactile feedback due to the rigid instruments is an evident limitation. Others suggest lifting the adipose tissue of the ventral bladder surface or pulling the bladder catheter (with the balloon filled up to 20 mL) to reveal the limit between the bladder and the prostate. Others suggest that just inspecting the adipose tissue covering the ventral surface of the urinary bladder is sufficient to detect the starting point of dissection of the BN; the fat tissue forms a triangle that covers the ventral prostatic surface at the 12-o'clock position. It is at the tip of this triangle where the BN dissection starts.
The aforementioned techniques exemplify the absence of standardised criteria for BNP.
Furthermore, the anterior approach to the BN presents the risk of suboptimal dissection, as it yields either residual prostate tissue (when performed too distally into prostate) or a gaping BN that may imperil the ureters and require reconstruction (when dissection is too proximal). Furthermore, in the case of the median lobe, the development of a correct plane between the posterior lip of the BN and the third lobe becomes challenging.
With the lateral dissection that we suggest, the lateral and inferior limits of the BN are promptly identified. The identification of the inferior limit of the BN (given by the Denonvillier's fascia underlying the plane of seminal vesicles) aids in avoiding wrong planes of dissection (e.g. between adenoma and capsule, as often occurs with the anterior dissection). Moreover, the lateral approach that we use, could be very helpful in cases of large median lobes. In these cases, an easier external lateral dissection of the median lobe from the posterior lip of the BN is performed without risk of leaving prostatic tissue in place or of compromising the ureteric orifices. Additionally, this approach quickly identifies all the key-points of the prostatectomy (BN, seminal vesicles, NVBs) and consequently it facilitates the subsequent surgical steps. Finally, it offers the possibility of a complete preservation of the pubovesical complex (i.e. detrusor apron with the pubovesical ligaments and associated vascular plexus) recently described by our team : while the anterior approach interrupts the detrusor apron, the lateral one permits its total preservation. The detrusor apron attaches the anterior bladder wall to the pubis through the pubovesical ligaments, offering an additional anterior support to the external urethral sphincter complex and preserving the urethra in its normal position in the pelvic floor. Consequently, its intactness could lead to better continence outcomes, particularly in the early postoperative period.
The lateral dissection that we suggest can be performed both in laparoscopic and robotic RP. The robotic approach offers a stable retraction of the bladder, a better (three-dimensional) view (also due to completely stable image and personal camera control) and abolition of physiological tremors. All these parameters, combined with the increased range of motion of the robotic instruments (7 degrees of freedom), permit, in our opinion, an even more precise (millimetric) and less traumatic dissection compared with pure laparoscopy.
The first author is scholar of Alexander S. Onassis Public Benefit Foundation.