Study Type – Therapy (case series)
Level of Evidence 4
Study Type – Therapy (case series)
To describe a novel synchronous approach to apical dissection during robotic-assisted radical prostatectomy (RARP) which augments circumferential visual appreciation of the prostatic apex and membranous urethra anatomy, and assess its effect on apical margin positivity.
PATIENTS AND METHODS
Positive surgical margins (PSM) during RP predispose to earlier biochemical recurrence, and occur most frequently at the prostatic apex. Conventional apical transection after early ligation of the dorsal venous complex (DVC) remains suboptimal, as this approach obscures visualization of the intersection between prostatic apex and membranous urethra, leading to inadvertent apical capsulotomy and eventual margin positivity. A synchronous urethral transection commenced via a retro-apical approach was adopted in 209 consecutive patients undergoing RARP by one surgeon (A.T.) between April to September 2009. The apical margin rates for this group were compared with those of 1665 previous patients who received conventional urethral transection via an anterior approach after DVC ligation. Outcomes were adjusted for differences in clinicopathological variables. All RP specimens were processed according to institutional protocols, and examined by dedicated genitourinary pathologists. The location of PSMs was identified as apex, posterior, posterolateral, bladder neck, anterior, base, or multifocal.
Patients receiving synchronous urethral transection had significantly lower apical PSM rates than the control group (1.4% vs 4.4%, P= 0.04). This marked improvement in the retro-apical group occurred despite a significantly higher incidence of aggressive cancer (≥pT3a) documented on final specimen pathology (16% vs 10%, P= 0.027).Technical difficulty was encountered in three of 209 study patients, in whom urethral transection had to be completed using the classic anterior approach.
Improved circumferential visualization of the prostatic apex, membranous urethra and their anatomical intersection facilitates precise dissection of the apex and its surrounding neural scaffold, and optimizes membranous urethral preservation. This has significantly ameliorated apical PSM rates in patients undergoing RARP, despite having to deal with more aggressive cancer on final specimen pathology.
(robot-assisted) radical prostatectomy
(positive) surgical margin
dorsal venous complex
Many patients with prostate cancer choose to undergo radical prostatectomy (RP); it is safe and effective, but in a few cases it is unsuccessful at completely removing the cancer, with patients having positive surgical margins (PSMs) on final specimen pathology. PSMs double a patient’s risk of cancer recurrence [1,2], which requires either radiation treatments or hormones. Apart from being expensive propositions , such treatments are also psychologically distressing for the patient who has already undergone radical surgery.
The probability of PSMs is greatest at the prostatic apex, which remains the ‘Achilles heel’ of prostate cancer surgery. At this location, the prostate subtly merges with the urethral sphincter complex and the urethra, and it is difficult to distinguish these structures. While several modifications and technical refinements have been made to prostate cancer surgery [4–13], including the use of a robotic platform [14–18], apical PSMs continue to be a recurring adverse outcome. Apical PSMs occur primarily because the surgeon cannot differentiate between prostatic tissue, the dorsal venous complex (DVC), and the urethral sphincter. The imperceptible transition from prostate to urethra makes recognition challenging, as does the lack of a formal prostatic capsule at this location. Also, because extraprostatic extension (EPE) at the apex is usually microscopic, surgeons cannot recognize it by feel or by magnification. To locate the junction amidst the ever-changing configurations of apical shapes and beaks, a surgeon must cut through the DVC (the pearly white layer of veins, fat, ligaments and smooth and skeletal muscle fibres). Because it is difficult to judge the thickness of this tissue, small sections of the prostatic apex can remain after surgery. PSMs at the apex can be a result of this scenario. To compensate for poor visualization of the apical junction, surgeons can deliberately cut far from the prostate into the urethra and sphincter, but this approach often results in intrinsic sphincter deficiency, significantly delaying the return of continence .
In our institution, histopathology slides from patients with PSMs are reviewed monthly by a multidisciplinary team comprising the surgical scrub team, and an experienced genitourinary radiologist and pathologist. During these virtual surgical sessions, areas of margin positivity are correlated against preoperative MRI prostate images and relevant intraoperative videotape of the patient’s surgery, to identify technical steps that might be improved to ameliorate recurrence in future cases. It became increasingly evident that the posterior prostatic apex is covered only by Denonvilliers’ fascia (DF) and the rectourethralis muscle, so this afforded a much clearer path to the apex.
We thus developed a retro-apical approach to the apical junction, using a 30° upward-facing lens for optimally visualizing the transition of the prostatic apex into membranous urethra. We found this novel approach afforded ample space to complete the posterior dissection and nerve-sparing, followed by sharp transection of the membranous urethra. This renders the prostate fully mobile, attached anteriorly by only the DVC, anterior fibromuscular stroma and the puboprostatic ligaments. We then switch to a 0° degree lens and transect the DVC anteriorly. Haemostasis is achieved with caudal traction using an inflated Foley balloon (inflated behind the prostate through the open urethra lumen) whilst the anterior anchoring tissue is cut. This synchronous technique of urethral transection was adopted in 209 consecutive cases. We compared its effect on apical PSMs against those of our previous 1665 consecutive cases (during which our approach to urethral transection was primarily an anterior one).
PATIENTS AND METHODS
We postulated that better visualization of the apical-urethral junction facilitates more accurate urethral transection and results in lower apical PSM rates. In brief, our technique of synchronous (posterior followed by anterior) urethral transection diverges from the conventional anterior approach which involves early suture ligation of the DVC followed by sharp dissection through the bunched tissue to identify an often obscured apical-urethral junction. The postulated benefits of this novel approach include: (i) precise release of the retro-apical nerves during posterior apical dissection; (ii) accurate identification of the membranous urethra and its intersection with the prostatic apex, thereby preserving maximum urethral length and thickness for optimizing early return of continence; (iii) reducing accidental capsulotomy into the prostatic apex; and (iv) obviating the risk of rectal injury from an anterior approach.
Our technique of synchronous urethral transection was adopted in 209 consecutive patients undergoing robotic-assisted (RA) RP (RARP) by one surgeon (A.T.) between April and September 2009. The apical PSM rates for this group were compared to those of 1665 previous patients who received conventional urethral transection via an anterior approach after DVC ligation. Outcomes were adjusted for differences in clinicopathological variables.
The specific details of our approach to RARP, with emphasis on technical manoeuvres for optimizing early return of continence [20–22], and balancing aggressiveness of nerve-sparing whilst ensuring complete extirpation of cancer, were described previously [23–27]. The technique highlighted in the present study focuses on synchronous retro-apical urethral transection and management of the DVC.
After transecting the bladder neck, we identify the vas deferens and seminal vesicles behind the retrotrigonal layer . The seminal vesicles and vas deferens are enveloped in a fascial compartment (part of DF) made of loose areolar tissue, veins, arteries, lymphatics and some fat. Each seminal vesicle has its own compartment and most of the blood supply enters near the tip and anterolateral aspect. Lateral to this compartment and near the tip lies the proximal neurovascular plate, which is part of the inferior hypogastric plexus [24–26]. This spatial relationship has profound implications for our nerve-sparing technique because the crucial neural tissue can be damaged (or avulsed) by heat, crushing, and the torque of traction during seminal vesicle dissection. However, most of the medial wall of a seminal vesicle  lies adjacent to a nearly avascular space. This medial avascular plane serves as a safe starting point for lateral dissection of the seminal vesicles. We then mobilize the vasa deferentia and clip the proximal end behind the trigone. The transected distal ends of the vasa deferentia are grasped by the left-side assistant (the fourth robotic arm) and pulled anteriorly, tenting the DF occupying the retroprostatic space. This fascia is incised in the midline and the opening is enlarged laterally toward the medial edge of the pedicles. The under surface of the prostate is identified and a plane is developed within the layer of DF (Fig. 1). A cave-like space is created behind the prostate and dissection is deliberately extended distally to release the apex and the urethra from deeper tissue.
Next, we release the neurovascular tissue using both our trizonal-neural-hammock-release technique (Fig. 2) and our athermal approach [24,26,27] We use sharp athermal dissection of the prostatic pedicles, controlling arteries and veins with Hem-O-Lok® clips (Teleflex Medical Inc, Research Triangle Park, NC, USA) as they enter the prostatic base. The prostate is then completely freed posterolaterally on either side. At this point it is quite mobile and is lifted anteriorly towards the anterior abdominal wall and pubic symphysis (Fig. 3a). This action opens the space behind the urethra and apex, and twists and angles the DVC, thus temporarily occluding it.
We now change to a 30° upward-facing lens for the retro-apical approach to the apical-urethral junction. Care is taken in navigating the lens directly behind the prostate without having it smudged by the bladder or the under surface of the prostate (Fig. 3b). Once positioned at the perfect vantage point, there is a surprising level of clarity with which the glistening white surface of apex can be differentiated from the membranous urethra. The transition from prostatic apex into urethra is further appreciated by manipulating the 20 F Foley catheter in and out of the urethra apical junction. Here, the prostate has few layers of DF covering its posterior surface, and the membranous urethra is posteriorly supported by the rectourethralis muscle (Fig. 3c).
Using sharp, curved scissors we then dissect the superficial layers of the DF and expose the precise prostate-urethral junction. Next, the posterior hemicircumference of the urethra is sharply incised 1 mm distal to the apex (Fig. 4). The Foley catheter is exposed and positioned with its tip at the distal urethral opening (Fig. 5). This permits appreciation of the anterior urethral hemicircumference, with urethral mucosa and muscular wall being seen clearly (Fig. 6). Using blunt and sharp dissections, the urethra is divided circumferentially (Fig. 7). At this point all the muscle fibres are transected and the DVC is left attached to the anterior surface of the prostate (Fig. 8).
Management of the DVC now depends on the width and thickness of the anterior tissue. If it is broad and thick, we ligate the DVC using a CT-1 needle and an O-polyglactin suture using a slip knot (Fig. 9). However, if the remaining tissue is thin, we increase the pneumoperitoneum to 20 cmH2O pressure, and inflate the Foley catheter balloon to 30 mL behind the prostate (Fig. 10). The Foley balloon is placed on caudal traction, resulting in partial occlusion of the DVC and developing the space behind the prostate.
The lens is now changed to 0° and prostatic apical dissection is continued anteriorly. The prostate is retracted cephalad by the left assistant grasping the seminal vesicles and distal ends of the transected vasa deferentia. Using robotic Maryland dissectors on the left arm, the anterior tissue is grasped between its two jaws to ensure that only venous tissue and some parts of the ligaments are gripped (Fig. 11). We sharply cut distally to the grip, and can see the sharp scissors cutting through the venous sinus (Fig. 12). Bleeding is minimal due to the inflated balloon under traction or suture if the DVC is thick and wide. The prostate is freed with a 1-mm hood of ligaments and venous tissue still on the prostatic apex (Fig. 13). Care is taken in preserving the pubo-prostatic ligaments and arcus tendineus for later reconstruction (Fig. 14) [29–31].
Once the prostate is freed, we perform a lymph node dissection and place this specimen in a laparoscopic entrapment bag. The DVC is oversewn robotically with a 0-polyglactin suture if this has not already been done. Care is taken to avoid including the sphincter, urethra and deeper nerves. This is followed by posterior reconstruction of Denonvilliers’ musculofascial plate (Fig. 15), construction of the vesico-urethral anastomosis (Fig. 16), anterior reconstruction of the arcus tendineus and puboprostatic ligaments (Fig. 17), before completing surgery.
All RP specimens were processed according to institutional protocols, and examined by dedicated pathologists with significant experience in genitourinary pathology (J.T., M.S.). Routinely, the right side of the prostate is marked in green ink and the left in black, and specimens serially sectioned from apex to base. After histological examination, the following variables were reported: prostate volume, histological type of cancer, primary Gleason pattern, secondary Gleason pattern, total Gleason score, pathological stage, lymph node positivity, SM positivity, presence of EPE, presence of seminal vesicle invasion, and presence of angiolymphatic invasion. A PSM is reported when cancer is present at the inked margins, and the PSM location identified as apical, posterior, posterolateral, bladder neck, anterior, base, or multifocal. Gross RP specimens were also photographed to improve appreciation of variations in prostate apical anatomy.
To ameliorate interobserver variability in Gleason sum reporting, 182 consecutive RP specimens were previously externally reviewed by an experienced off-site genitourinary pathologist (M.A.). After establishing a concordance rate of 89% between our in-house and off-site pathologists’ reports, external validation was discontinued for financial prudence.
The following variables for each patient undergoing RARP in our programme are routinely entered into a Health Insurance Portability and Accountability Act compliant institutional database under an institutional review board-approved protocol: patient age, body mass index, preoperative PSA level, International Index of Erectile Function, IPSS, comorbidities, medication use, clinical staging and preoperative imaging, systematic biopsy data, intraoperative information, prostate volume, Gleason score on the final histopathology, high-grade prostatic intraepithelial neoplasia, perineural invasion, percentage of cancer in specimen, lymph node positivity, SM status, and EPE of tumour. Results were assessed using univariate and multivariate analysis.
Table 1 summarizes the clinicopathological profiles of patients in both the retro-apical (209) and control groups (1665), the latter receiving conventional anterior urethral transection after DVC ligation. Patients receiving anatomical retro-apical urethral transection had significantly lower apical PSM rates than the control group (1.4% vs 4.4%, P= 0.04). This marked improvement in the retro-apical groups occurred despite a significantly higher incidence of aggressive cancer (≥pT3a) documented on final specimen pathology (16% vs 10%, P= 0.027). In three of the 209 most recent cases in the retro-apical group, retro-apical dissection could not be done safely due to challenging body habitus, local anatomy and a large prostate. In these instances, dissection was completed using the conventional anterior approach. One of these three patients had PSMs. Nonetheless, this patient’s data were included in the retro-apical group results, as we used an intent-to-treat analysis for this study.
|Mean (sd), median (IQR) n/N (%) or % variable||All (1874)||Control (1665)||Retro-apical (209)||P|
|Age, years||59.87 (7.13)||59.91 (7.16)||59.53 (6.89)||0.460|
|Body mass index, kg/m2||27.00 (3.97)||27.00 (3.97)||27.05 (4.03)||0.851|
|PSA level, ng/mL||5.85 (4.92)||5.90 (5.03)||5.44 (3.99)||0.206|
|Max percentage cancer in biopsy core||27.59 (25.80)||27.43 (25.68)||28.86 (26.73)||0.453|
|Core positivity, %||25.74 (29.07)||25.42 (21.56)||28.29 (62.36)||0.511|
|Biopsy Gleason score||0.065|
|Prostate volume, mL||46.0 (38.0–58.0)||46.0 (38.0–58.0)||46.0 (37.0–56.0)||0.292|
|Overall PSM rate, % (pT2–pT4)||9.3||9.8||5.3||0.032|
|Apical PSM (whole group)||4.1||4.4||1.4||0.040|
|Apical PSM (pT2)||53/1571 (3.4)||52/1386 (3.8)||1/185 (0.54)||0.016*|
|Apical PSM (pT3 + pT4)||24/276 (8.7)||22/255 (8.6)||2/21 (9.5)||0.691*|
Table 2 summarizes the data on apical PSMs for the two groups; all clinical and pathological variables were comparable in both these groups, with no significant differences. Figure 18 illustrates the effect that the development of the surgical technique had on apical PSM rates. An actuarial model was created based on data averages from the initial 1665 patients, using both true and projected apical PSM rates from this initial group. Apical PSM rates were significantly reduced in the retro-apical group after incorporating our anatomical retro-apical technique for urethral transection.
|Mean (sd), median (IQR) or % variable||Control||Retro-apical||P|
|Age, years||60.65 (8.11)||54.00 (12.49)||0.454|
|PSA level, ng/mL||8.69 (11.09)||6.70 (5.60)||0.612|
|Biopsy Gleason score||0.424*|
|Core positivity, %||34.42 (22.11)||46.02 (26.48)||0.302|
|Maximum % cancer in biopsy core||32.45 (26.50)||51.00 (37.59)||0.484|
|Prostate volume, mL||48.36 (18.61)||42.00 (10.54)||0.409|
|Body mass index, kg/m2||26.54 (3.69)||28.30 (3.46)||0.418|
|T stage, % in group||0.228*|
|Pathological Gleason score, % in group||1.000*|
|≤ (6, 7 (3 + 4)||75.7||33.3|
|7 (4 + 3)||14.9||66.7|
|% cancer||11.69 (11.50)||17.33 (9.29)||0.407|
Table 3 summarizes the findings of univariate and multivariate analysis for predictors of apical PSM rates. The preoperative serum PSA level, the percentage of positive cores (positive cores/total sampled cores) and the retro-apical technique were all significant factors in predicting whether or not a patient will have apical PSMs.
|Body mass index||−0.037||0.247||0.964||0.906–1.026|
|Pre-op PSA level (log)||1.866||<0.001||6.461||2.669–15.64|
|Core positivity %||0.384||0.076||1.468||0.960–2.243|
|Maximum % biopsy||0.007||0.145||1.007||0.998–1.015|
|Prostate volume (log)||−0.783||0.337||0.457||0.092–2.259|
|Gleason score, biopsy||0.019||0.925||1.019||0.689–1.505|
|Body mass index||NA||NS||NA||NA|
|Pre-op PSA level (log)||1.752||0||5.766||2.364–14.07|
|Core positivity %||0.461||0.043||1.586||1.015–2.478|
|Maximum % biopsy||NA||NS||NA||NA|
|Prostate volume (log)||NA||NS||NA||NA|
|Gleason score biopsy Total||NA||NS||NA||NA|
Complete clearance of cancer at of the prostatic apex remains technically challenging for several reasons: (i) the apex is the narrowest part of the prostate, and so cancer is never too far from the surface; (ii) there is an imperceptible transition between prostate glandular tissue (often containing cancer) and the sphincteric muscle; (iii) unlike other parts of the prostate gland, the apex lacks a true and well-defined capsule; (iv) the anterior part of the apex is covered with large veins intermingled with fibromuscular stroma, making it difficult to identify the precise junction between the apex and the urethra; (v) the apex is the one part of the prostate where a cancer-prone peripheral zone wraps around anteriorly, thus bringing cancer close to the surface [32–34]; (vi) the apex is the most inaccessible part of the prostate, wedged deep in the human bony pelvis beneath the pubic arch (Fig. 19) and DVC, packed tightly between the rectum, nerves, blood vessels and sphincter; (vii) the apex is surrounded by a cavernous mesh-work of veins which is very prone to bleeding (torrential bleeding from poor vascular control can quickly obscure vision of the operative field and impairs precise apical dissection); and (viii) considerable surgeon experience in appreciating the myriad configuration of the prostatic apex (beaks and protrusions) is required to identify the correct plane of dissection.
In addition, balancing cancer clearance with good functional outcomes is very difficult during apical dissection with little latitude either way. The prostatic apex comprises the final common pathway for exit of the erectogenic nerves to the penile tissue (and is where these nerves come closest to the prostate) (Fig. 20). Optimizing membranous urethral length has been shown to correlate with the early retrun of continence . As such, there is little room for wide excision at the apical urethral junction without ultimate compromising sexual and urinary function. Finally, obesity and large prostates also present technical challenges for surgeons in finding an optimal plane of dissection.
Most surgeons approach apical dissection by securing the DVC and cutting through the bunched veins, fatty tissue, puboprostatic ligaments, anterior fibromuscular stroma and occasional arterial bleeders (Fig. 21). This surgical approach remains suboptimal given the paucity of anatomical (visual or tactile) landmarks. With poor appreciation of the apical urethral junction from the distorted anatomy, often compounded by ongoing trickling of venous bleeding from incomplete DVC ligation, accurate transection of the membranous urethra is often a ‘hit and miss’ experience with the risk of cutting a divot into the apical tissue. For these reasons, apical dissection remains the ‘Achilles heel’ of prostate cancer surgery.
To reduce the PSM rate at the apex, several surgeons have described innovative techniques for apical handling. The classical descriptions by Walsh  and Chuang et al.  remain the reference standard for safe and successful management of the apex. These elegant reports not only clarified the anatomy of the apex and DVC, but also highlight technical refinements to help achieve continence and negative SMs. Myers et al. also described the anatomy of this area in great detail and highlighted the benefits of magnification of the apex during surgery, and they addressed the importance of carefully handling the variations in the shape of the apex. Also, Menon et al.  and Laven et al.  discussed the importance of apical dissection in several landmark articles. Joseph Smith, who has made the transition from the open RP to the robotic procedure, also studied apical PSMs and mentioned the important technical refinements to reduce PSM rates . Other innovative approaches towards this goal include blunt dissection, by Namiki et al. , transecting the DVC with no suture, by Guru et al. , and the use of a lateral viewing camera through an assistant’s port, by Sasaki et al. . All of these techniques have their merits and limitations.
We chose to approach this stage of the operation differently in search of better results, and modelled our technique on perineal RP, where a surgeon views the urethra circumferentially, from behind and with no DVC or puboprostatic ligaments obstructing the view (Fig. 22). The potential challenge to this technique was posed by very obese patients with large prostates, who have little space in the pelvis to allow for adequate mobilization (retraction) of prostate. A surgeon must also be willing to approach their dissection from an unfamiliar vantage point when using this technique. However, our team quickly (within 5–10 cases) became so familiar with this new viewpoint that it was difficult for us to go back to the classic anterior approach. (Importantly, our new surgical technique is not solely posterior but is synchronous, involving a combination of posterior and anterior dissections). As previously stated, with very obese patients we might not be able to use the retro-apical approach.
Our approach has the following perceived benefits: (i) neurovascular bundles are released out of harm’s way before we embark on apical dissection; (ii) identification of the prostatic-urethral junction is very easy and precise; (iii) it allows us to gently push the posterior apical beak away from the sphincter, thus gaining a few extra millimetres of length (Fig. 23); (iv) it does not require early suturing of the DVC when the prostate is not yet free (and so avoids unnecessary inclusion of the sphincter or nerves while attempting suture placement with limited mobility and vision); (v) it sometimes allows for balloon tamponade of the DVC, obviating the need for suturing before transection; (vi) it still allows us to preserve maximum urethral length, pubo-prostatic ligaments and arcus tendineus, all of which are critical in preserving continence ; and (vii) there is a greater safety margin with the synchronous approach than with the conventional approach (Fig. 24).
Nonetheless, our approach of synchronous urethral transection has been difficult to execute in obese patients or in those with large prostates. In addition, skilful assistants are needed who retract the prostate just enough to open the retroprostatic fossa and who are able to change the camera to a 30° upward-facing lens without smudging the lens on the bladder or prostate. Also, this approach has only been attempted by one surgeon in 209 patients. These data need further validation by other surgeons and should be analysed from a larger patient pool. However, we think that our transition to this approach accounts for close to a four-fold reduction in the apical PSMs as the factors such as development of technique, focus on apical dissection and attention to detail all contribute to surgical outcomes in both approaches. We are excited by the merits and promising early results of this technique, and recommend it to other surgeons seeking to reduce their apical PSM rates.
In conclusion, we developed a novel approach toward apical dissection during RARP which provides the surgeon with a circumferential view of the prostatic apex, membranous urethra and their anatomical intersection. Improved visualization and transection of the apical urethral junction, with precise dissection of the apical neural scaffold and optimal control of the DVC, has significantly ameliorated apical PSM rates in our patients undergoing RARP, despite having to deal with more aggressive cancer on final specimen pathology.
CONFLICT OF INTEREST
This study was supported in part by a research grant from Intuitive Surgical, Inc., Sunnyvale, CA, USA; Ashutosh Tewari discloses that he has received research grants from Intuitive Surgical, Prostate Cancer Foundation and NIH; he is also the endowed Ronald P. Lynch Professor of Urologic Oncology and Director of the Lefrak Center of Robotic Surgery, Weill Cornell Medical College. Gerald Tan discloses that he receives financial support from the Ferdinand C. Valentine Fellowship in Urologic Research, New York Academy of Medicine; the John Steyn Travelling Fellowship in Urology from the Royal College of Surgeons of Edinburgh; and the Medical Research Fellowship from the National Medical Research Council (Singapore).