Experience with radical perineal prostatectomy in the treatment of localized prostate cancer

Abstract

Radical perineal prostatectomy is an alternative to the retropubic, robotic-assisted or the laparoscopic approach and it is one of the oldest surgical procedures for prostate cancer, but its use has declined over the past 30 years. The desire for lower morbidity, short hospital stay, and decreased costs has resulted in a renaissance of radical perineal prostatectomy. The combination of low morbidity, short hospital stay, and low economic costs brought to a renaissance of radical perineal prostatectomy.

Introduction

Prostate cancer (PC) is the most frequently diagnosed noncutaneous cancer in European (Figure 1, [Globocan 2008 (IARC)-2011]) and American men [Hartke and Resnick, 2007] and although several options are available to the patient, surgical cure continues to be a mainstay of therapy.

Figure 1.

Estimated number of cancer cases in Europe (all ages).

History

As early as 1867 Billroth first used the perineal access [Stackl, 1999] to the prostate but this access was not to become popular. In spite of this, Hugh Hampton Young in Chicago [Young, 1905] used a perineal approach (the dissection is carried anteriorly along the external anal sphincter until arriving at the rectourethralis muscle) in 1905 for treatment of cancer. Belt in 1942 [Belt, 1942] modified this approach (the fibres of the external anal sphincter are dissected and retracted anteriorly with an appendiceal retractor; these fibres are not incised) and nowadays this access, with the modifications of Dees in 1970 [Dees, 1970] and Vest in 1994 [Vest, 1994], is the most frequently used approach in the radical perineal prostatectomy (RPP).

Before the 1970s many urologists practised RPP which was indicated for patients with well-differentiated small localized cancers [Sullivan et al. 2000].

On the basis of the anatomical studies by Walsh and Donker [Walsh and Donker, 1982] the retropubic approach with concomitant pelvic lymphadenectomy broadened. The last modification was carried out 1985 by Weiss and colleagues [Weiss et al. 1985] who performed the first nerve-sparing RPP. At the moment, the technique of Weiss and colleagues is the basis of all nerve-sparing approaches used so far.

The surgical technique

The patient is placed in an exaggerated lithotomy position with the perineum almost parallel to the floor or in a modified exaggerated lithotomy position with the perineum only slightly elevated.

After positioning of the curved Lowsley retractor into the bladder and opening its wings, the perineal incision (curvilinear) is placed with the apex in the midperineum and the ends medial to the ischial tuberosities. The incision should not extend posteriorly beyond the 3 and 9 o’clock positions relative to the anus. The central tendon is incised by electrocautery as well as the rectourethralis muscle. Now the rectum is brought dorsally revealing the Denonvillier’s fascia. A vaginal retractor valve or a dampened sponge is placed.

Without any traction on the Lowsley tractor, the rectourethralis muscle is divided close to the apex of the prostate with vertically oriented scissors, allowing the rectum to fall dorsally; caution must be exercised to prevent rectal injury at this point of the procedure.

The Denonvillier’s fascia is then incised from the midpoint of the seminal vesicles to the mid-apex. The vas deferens is identified and cut on both sides followed by the liberation of the two seminal vesicles. If preservation of the neurovascular bundle is intended, the exposed anterior layer of the Denonvillier’s fascia is incised vertically in the midline from the vescicoprostatic junction to the apex of the prostate. Careful lateral dissection (without use of electrocautery) with gentle lateral traction preserves the neurovascular bundles as they course between the layers of Denonvillier’s fascia at the posterolateral edge of the prostate.

As preferred by the present authors, the two layers of the fascia are now injected with saline solution centrally over the prostate gland to provoke a hydrodistention of two parts and coloured with methylene blue. After 2 minutes the colour is washed away, once again with saline solution and the neurovascular bundles remain a deep blue colour.

With careful sharp dissection the cavernosal nerve bundles and attached fascia are separated from the prostate from the apex to the base. This procedure is performed bilaterally.

Following the preparation of the apex the urethra is dissected with a scalpel and divided just out of the apex. The curved Lowsley tractor is then replaced by a straight Lowsley tractor and the prostate liberated, step by step from the attachment to the bladder. At the end the bladder neck is incised and the complete surgical specimen is removed.

After accurate haemostasis and identification of the ureteral orifices, the bladder neck is reconfigured according to the tennis racket fashion followed by the eversion of the bladder mucosa. The first three stitches of the anastomosis at the anterior face are placed followed by a running suture for the remaining circumference. Once anastomosis has been completed a 14–18 French catheter is placed transurethrally under digital protection of the anastomosis. The bladder is filled with 300 ml of saline solution to check the water tightness.

After visual and digital control of the anterior rectal wall a Jackson–Pratt or Penrose drain is placed near the vesicoureteral anastomosis and brought lateral of the incision, the central tendon and Colles’ fascia are reapproximated with two single absorbable sutures and the skin closed with metallic clips or absorbable sutures.

Ambulation and normal diet are allowed the morning after surgery and stool softeners are provided to the patient; the drain is typically removed when secretion has stopped or on postoperative day 1. As preferred by the authors on postoperative day 5 a cystogram is performed and if there is no leakage the catheter is removed.

The postvoidal residual volume is controlled three times by means of sonography and has to be below 50 ml. After 3 weeks all activities are unrestricted.

Background

Radical prostatectomy is considered the most reliable approach in the eradication of localized PC [Bartsch et al. 2001]. The goal of modern radical prostatectomy is to excise all cancer with the least morbidity and to increase the chances of a full recovery of continence and potency.

Nowadays there are four options for radical prostatectomy: routinely the procedure is performed either retropubically or with a perineal approach (open surgery), although mini-invasive prostatectomy also exists via the laparoscopic and more recently the robotic-assisted laparoscopic technique (Da Vinci®).

While RPP has been largely supplanted by retropubic and minimally invasive radical prostatectomy, it was the predominant surgical approach for PC for many years.

Risk estimation

The use of prostate-specific antigen (PSA) level testing and transrectal ultrasonography (TRUS)-guided needle biopsy has made it possible to more frequently detect early stage PC [Harris, 2007].

Radical curative surgery is offered to those patients with a life expectancy of more than 10 years who demonstrate a high likelihood of organ-confined disease. There is no rigid age limit for radical prostatectomy and a patient should not be denied this procedure on the grounds of age alone [Corral and Bahnson, 1994; Heidenreich et al. 2011a].

There exist nomograms that can identify patients at low risk for lymphatic metastases, on the basis of preoperative information. The preoperative ability to accurately predict pathologic stage of PC permits improved patient counselling, more appropriate selection of treatment plan or risk stratification for novel clinical trials for those with more advanced disease. The work of Partin and colleagues has allowed a relatively accurate prediction of disease stage and pelvic lymph node involvement through the analysis of PSA level, clinical stage, and pathologic Gleason score [Partin et al. 1997]. This work has allowed a more selective use of staging lymphadenectomy [Parra et al. 1996]. Another prognostic tool to estimate the risk is the nomogram of D’Amico and colleagues [D’Amico et al. 1999]; there are three risk groups:

• - Low risk: local confined tumour, PSA < 10 ng/ml, Gleason score ≤ 6.

• - Intermediate risk: PSA 10–20 ng/ml, Gleason score > 7.

• - High risk: local advanced cancer, PSA > 20 ng/ml, Gleason score > 8.

For the prediction of the postoperative relapse there exists also a nomogram of Kattan and colleagues [Kattan et al. 1998] and Graefen and colleagues [Graefen et al. 1999].

A (laparoscopic) pelvic lymphadenectomy is deemed unnecessary if the biopsy Gleason score is below 7 and the serum PSA level is less than 10 ng/ml.

Nowadays the pelvic lymph node dissection in RPP is performed with a second laparoscopic access, and has allowed a combined laparoscopic pelvic lymphadenectomy (and so minimally invasive) and standard RPP with little morbidity [Lerner et al. 1994; Levy and Resnick, 1994; Parra et al. 1994]. This procedure requires a change in the patient’s position and increases the operative time. Saito and Murakami [Saito and Murakami, 2003] first described a limited lymphadenectomy during RPP via the same incision and in 2006 Keller and colleagues [Keller et al. 2006] proved that also the concomitant extended lymph node dissection in the RPP under direct vision via the same incision is possible and the major disadvantage of RPP is resolved.

Patient’s factors

While body habitus and prior abdominopelvic surgery have a negative impact on a given patient’s candidacy for laparoscopic radical prostatectomy (LRP) and to some extent RRP, almost all patients who are candidates for prostatectomy can undergo RPP. A very high body mass index (BMI) does not increase the difficulty of the operation, in contrast to other techniques. Also Imperatore and colleagues showed in 2011 that after previous prostate surgery the transperineal approach should be performed rather than the retropubic approach [Imperatore et al. 2011].

Patients with severe ankylosis of the hips or spine and those with unstable artificial hip replacements may not tolerate the exaggerated lithotomy position. However, common degenerative disk diseases are not a contraindication to positioning.

Some patients, such as those having renal transplantation or those with severe inflammation secondary to placement of synthetic mesh for repair of a hernia or the morbidly obese, who may not be amenable to RRP, typically may undergo prostatectomy by the perineal approach [Yiou et al. 1999; Boczko and Melman, 2003].

A limitation for RPP is a big prostate weight, even if studies have shown that a prostate with a weight over 90 g [Comploj et al. 2011] can be removed.

Oncological outcomes

RPP demonstrates excellent pathologic outcomes consistent with those of RRP.

It has been suggested that RPP might result in positive surgical margins more often than the retropubic approach, but this has not been confirmed [Weldon et al. 1995].

A review of the literature demonstrates an incidence of positive margins in 15% to 44% of radical perineal specimens [Weldon et al. 1995; Iselin et al. 1999; Lance et al. 2001; Harris, 2003; Gillitzer et al. 2004]. Comploj and colleagues showed that in their series with 212 RPPs a positive margin was present in 22% [Comploj et al. 2011]: 62% of patients with positive margins did not develop a biochemical recurrence after a mean follow-up time of 48 (range 6–117) months in contrast to the study of Paulson, who demonstrated that only 20% of patients with positive margins did not show biochemical failure [Paulson, 1995]. Harris was able to demonstrate better results with freedom from PSA detectability in 96.3% for organ-confined cancer [Harris, 2003].

In an actual overview, a positive margin rate for LRP from 16% up to 30% [Harris, 2003; Salomon et al. 2002; Touijer and Guillonneau, 2006] has been reported. No significant advantage of one technique could be identified.

Gibbons, Iselin and colleagues demonstrated excellent cancer control in patients who were followed for 20 years after RPP and showed that locally confined cancers can be safely removed with good long-term tumour-free survival using the perineal approach [Iselin et al. 1999; Gibbons et al. 1989]. Comploj and colleagues noted an overall biochemical progression-free survival of 86% [Comploj et al. 2011]. These data are comparable with those of the series by Guillonneau and colleagues [Guillonneau et al. 2003] and the series by Salomon and colleagues [Salomon et al. 2002], which compared all three therapeutic options.

Unfortunately we have to state that the data of the studies are very heterogeneous regarding stage, grade, risk scores, age, PSA value and BMI. A standardization of a general accepted reporting system is completely lacking and most of the studies are retrospective. Therefore, the comparability is compromised in terms of evidence.

Functional outcome

The perineal approach permits excellent exposure of the apex and facilitates an exact anastomosis of the urethra, reducing the risk of postoperative urinary incontinence when compared with RRP [Harris, 2007].

In a review article by Rassweiler and colleagues [Rassweiler et al. 2006], the range of continence rate after LRP, RRP, and RPP was between 50% and 91.7%. Comploj and colleagues showed that 81% of the patients were completely dry (no drop) after a follow up of 48 months [Comploj et al. 2011]. These are slightly better results as shown by Schmeller and colleagues (for the perineal group), and much better results than for the laparoscopic approach after 2 years of follow up [Schmeller et al. 2007]. The perineal access has the advantage of an easy dissection and water tight anastomosis of the urethra, and prior operations in the pelvis have no negative impact on the surgical procedure as well as a suprapubic femoro-femoral bypass [Harris, 2003].

Anatomic techniques of the retropubic prostatectomy were applied to the perineal approach, resulting in potency comparable to that with RPP [Weiss et al. 1985; Weldon and Tavel, 1988].

A point of concern in nerve-sparing RPP is the need to move the rectum backwards. During this manoeuvre the neurovascular bundles can be lacerated even if they are running laterally. Afterwards they have to be dissected from the prostate when the prostate is removed. This can result in a significant higher trauma to the neurovascular bundles and decreases the chances of recovering erectile function.

However, all comparisons are limited by variations in the study population size, patient characteristics and multiple other factors that are known to affect potency.

Complications

Bishoff and colleagues [Bishoff et al. 1998] reported a higher rate of faecal soilage following RPP versus RRP; however, their perineal incision extended 2 cm posterior to the anal orifice.

Comploj and colleagues observed neither stool incontinence nor faecal urgency or flatus during valsalva manoeuvres [Comploj et al. 2011], in contrast to the series by Harris [Harris, 2003]. Therefore, this is probably the reason that this approach, according to Young, avoids any sphincteric structures of the rectum.

Wound infections/dehiscence are a bigger problem in RPP than in the other techniques [Schmeller et al. 2007; Comploj et al. 2011].

Reinsertion of a catheter due to urinary retention after RPP is needed a little more often than in the other approaches [Comploj et al. 2011; Schmeller et al. 2007], and also the median time for urinary catheter drainage is longer in the RPP (median 10 days) than in RRP or RLP (7 days). This is in contrast to postoperative pain which is significantly lower in RPP than in the other procedures [Schmeller et al. 2007; Paiva et al. 2009].

One of the greatest advantages of RPP is the low blood loss and transfusion rate (because of the lower wound surface), the lower morbidity than in the abdominal laparotomy for the RRP and early time to ambulation and discharge after RPP [Lance et al. 2001; Paiva et al. 2009].

Anastomotic stricture is a well-recognized complication after open radical prostatectomy, with a reported incidence of 0.5–14.6% [Harris, 2003; Gillitzer et al. 2004; Heidenreich et al. 2011b]. Gillitzer and colleagues showed that an anastomotic stricture is more common after RRP than after RPP [Gillitzer et al. 2010].

A rare complication is a neuropraxia of lower extremity after the perineal approach with an incidence up to 21% [Price et al. 1998].

Renaissance

RPP demonstrates proven long-term cancer control of RRP with low morbidity, rapid convalescence (therefore, a high patient acceptance), and lower operative times and lower hospital costs than with RRP or LRP [Sullivan et al. 2000; Harris, 2007; Comploj et al. 2011; Harris, 2003; Guillonneau et al. 2003; Lotan et al. 2004; Silverstein et al. 2004].

This technique is a low overall cost therapy because of simplicity (in experienced hands) and less use of resources with immediate morbidity, and could be used in many countries where robotic or even laparoscopic surgery is not available [Nargund and Aman, 2011; Harris, 2003; Prasad et al. 2011].

Also the statement that the operative time and the learning curve in RPP is longer than in the other techniques could be disproved by various recent studies [Keller et al. 2006; Comploj et al. 2011; Schmeller et al. 2007; Eliya et al. 2011] and excellent, reproducible results can be obtained using basic surgical principles [Eliya et al. 2011].

The anatomic RPP is safe and versatile [Harris, 2007].

Conclusion

There has recently been a resurgence of interest in RPP for the following reasons. Treating patients with localized PC with RPP has similar characteristics as minimally invasive surgery. Furthermore, this approach provides the most favourable exposure for urethral dissection and anastomosis, thus reducing the risk of postoperative urinary incontinence when compared with RRP.

It could be shown that RPP is not only minimally invasive but beneficial from an economic perspective and should not yet be abandoned in the treatment of early PC.