- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
Radical prostatectomy is the standard surgical treatment for localized prostate cancer . The oncological outcomes have been widely investigated, with no difference being found between open radical retropubic prostatectomy (RRP), laparoscopic radical prostatectomy (LRP) and robot-assisted laparoscopic prostatectomy (RALP) [2–4]. Efforts are now being made to decrease morbidity and complications after the procedure. Urinary incontinence is a major concern for patients who undergo surgical treatment. The 1-year continence rates are excellent after RALP in larger series [5,6], and almost similar to the results achieved after RRP or LRP . However, the early return of continence remains challenging and has a major impact on the patient's health-related quality of life [7,8].
Several technical modifications aiming to improve postoperative continence have been described, including bladder neck preservation , intussusception of the bladder neck , posterior reconstruction (PR) of the rhabdomyosphincter [11–14] and sparing or reconstruction of the puboprostatic ligaments [15–18].
Restitution of anterior attachments of puboprostatic ligaments using an anterior suspension (AS) stitch anchored to the pubic bone was initially described in RRP [16,17] and found to hasten continence recovery. Subsequently, the technique has been simplified and transferred to RALP in a non-randomized study by Patel et al. .
Similarly, PR, which was first described in both RRP and LPR by Rocco et al. [11,12], has also been shown to decrease continence recovery time. This technical modification has been widely used by RALP surgeons [13,14].
The results reported with these two technical modifications are inconsistent. In the present study, it was therefore decided to combine AS and PR during RALP and to assess outcomes in a randomized, multicentre control study.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
From July 2009 to July 2010, we conducted a prospective study of 72 patients with localized prostate cancer who underwent RALP in three tertiary reference centres. All procedures were carried out by a single surgeon in each institution. Each surgeon had already performed more than 100 RALP before the study began.
Patients were eligible for the present study if they had histologically proven prostate cancer, clinical stage cT2 or less, and had chosen surgery. Exclusion criteria comprised a history of prostatic surgery, local radiotherapy or urethral surgery; a history of a disease that could affect continence, such as insulin-dependent diabetes, neurological disease or urethral stenosis; and the presence of a urinary catheter preventing preoperative continence evaluation.
All eligible patients received oral and written information about the study protocol and were included subsequent to their written consent having been obtained. The study protocol was reviewed and approved by the French Ethical Committee.
Eligible patients were randomly attributed to one of two groups: group A underwent a standard RALP procedure, nd group B underwent RALP with both AS and PR. Randomization was performed using a list consisting of two groups (A and B) and three strata with 72 patients in each. This list, which was generated using dedicated software of Tours University Statistics Laboratory, was known only to the study coordinator. After enrolment screening, the investigator sent an e-mail to the study coordinator requesting the patient's randomization. The coordinator emailed back the result of randomization the day before surgery. Patients were blinded to their allocated group.
The data gathered included age, body mass index (kg/m2), IPSS, prostate volume (evaluated by ultrasonography or MRI), preoperative PSA, American Society of Anesthesiology score, DRE for T stage evaluation, Gleason score on biopsy, neurovascular bundle preservation and duration of indwelling catheter.
The primary outcome of interest was continence, measured at baseline and at 15 days, and at 1, 3 and 6 months after surgery. It was assessed with an internationally recognize French-validated self-administered questionnaire (the University of California Los Angeles Prostate Cancer Index; UCLA-PCI) [19,20] and based on the patient's answers to three items on the questionnaire: (i) ‘Over the past 4 weeks, how often have you leaked urine?’; (ii) ‘Which of the following best describes your urinary control during the last 4 weeks?’; and (3) ‘How many pads or adult diapers per day did you usually use to control leakage during the last 4 weeks?’ (‘4 weeks’ was changed to ‘2 weeks’ for the questionnaire at 15 days after surgery). Patients were considered continent if they answered ‘Not at all’ to (i); ‘total control’ to (ii); and ‘No pads’ to (iii).
All postoperative questionnaires were given to the patients on the day that they were discharged from hospital. The patients fulfilled the questionnaires at home and each appropriate questionnaire was collected during follow-up consultations at 1, 3 and 6 months.
The other data recorded were estimated blood loss, operation duration, postoperative pain (i.e. the presence of perineal or pubic pain), analgesic uptake, complication rate (using the Clavien–Dindo classification) , length of hospital stay and positive surgical margin rate.
The number of patients included in the present study was based on the results of previously published studies [16,17] for a 40–75% improvement in continence 3 months after surgery, with an α-risk of 0.05 and a power of 80%. The number of patients required was 65. A further 10% were added for loss to follow-up, giving a total number of 72 patients.
The only difference in surgical technique between the two groups was the use of AS and PR.
All RALP procedures were performed using a modification of the transperitoneal technique described by Menon et al. . After entering the retropubic space of Retzius, the endopelvic fascia was incised bilaterally and the elevator muscle fibres were moved back from the prostate apex to expose the dorsal venous complex (DVC) and the urethra. The DVC was ligated using an absorbable polyglactine suture.
In group B, AS was then performed using the same suture, passing the needle from right to left between the urethra and the DVC. The needle was then held backwards and passed through the periostium of the pubic bone, from left to right. The stitch was passed through the DVC and the pubic bone again to ensure support, and then tied (Fig. 1). In both groups, the procedure continued with apical dissection of the prostate and division of the urethra. This was followed by dissection of the bladder neck, incision of Denonvilliers' fascia, dissection of seminal vesicles and ligation of the prostate pedicles. Nerve sparing was carried out when required with retrograde or antegrade release of the bundles, which was performed at the discretion of each surgeon. The prostate specimen was then removed in an Endobag (Covidien, Mansfield, MA, USA).
Figure 1. Anterior retropubic suspension. A, After ligation of the dorsal venous complex (DVC), the same suture is used to begin the anterior suspension. The needle is first passed between the DVC and the urethra. B, The suture is passed at the posterior part of the pubic bone in the periostum, holding the needle with the left needle driver at an angle of 90°, passing from left to right. C, The suture is passed again between the DVC and the urethra from right to left and through the pubic bone. D, View of the completed anterior suspension.
Download figure to PowerPoint
PR was performed using a polyglactine absorbable suture passed through the Denonvilliers' fascia, at the posterior wall of the bladder, beginning on the right side. The suture was then passed through the rhabdosphincter, beneath the posterior part of the urethra. PR was completed by passing the suture back through Denonvilliers' fascia and the tissue of the rhabdosphincter twice (Fig. 2).
Figure 2. Posterior reconstruction of the rhabdomyosphincter. A, After removing the prostate specimen, the stitch is first passed through the Denonvilliers' fascia at the posterior face of the bladder. B, The suture is passed through the musculo-fascial tissue beneath the urethra. Then it is passed two more times through the Denonvilliers' fascia and rhabdosphincter. C, View of the completed posterior reconstruction, ready to begin vesico-ureteral anastomosis.
Download figure to PowerPoint
Vesico-ureteral anastomosis was performed with a continuous running suture using two monocryl sutures tied together, using the technique of Van Velthoven et al. .
Data were analyzed in accordance with the guidelines of the CONSORT group , using a t-test or Mann–Whitney U-test to compare numeric variables, and a chi-squared test or Fisher's exact test to compare categorical variables. For multivariable analysis, a logistic regression was used including all the variables with P < 0.15 in univariable analysis. P < 0.05 was considered statistically significant. All statistical calculations were overseen by one of our institution's statistician. Data analysis was performed using IBM SPSS Statistics, version 16 (IBM Corporation, Somers, NY, USA).
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
The 72 patients enrolled in the present study were randomly allocated to two groups: 33 to group A and 39 to group B. In total, 41 patients were included in centre 1, nine patients in centre 2 and 22 in centre 3. Preoperative data are shown in Table 1. Data were missing for 10 (13.9%) patients at 1 month, five patients at 3 months and 12 patients at 6 months The study flow chart is shown in Fig. 3.
Table 1. Preoperative data
|Variable||Group A (n = 33)||Group B (n = 39)||Overall (n = 72)|
|Age (years), mean (SD)||62.4 (5)||62.5 (6.8)||62.4 (6)|
|Body mass index (kg/m2), mean (SD)||25.7 (3.6)||26.3 (3.5)||26 (3.5)|
|IPSS score, median (range)||7 (0–17)||5 (0–16)||5 (0–17)|
|IIEF-5 score, median (range)||23 (1–25)||20 (6–25)||22 (1–25)|
|Prostate volume (mL), median (range)||40 (25–80)||38.5 (20–90)||40 (20–90)|
|PSA (ng/mL), median (range)||7.9 (3.6–23.6)||6.4 (2.7–16.6)||7 (2.7–23.6)|
|ASA score, n (%)|| || || |
| 1||20 (64.5)||14 (37.8)||34 (50)|
| 2||11 (35.5)||22 (59.5)||33 (48.5)|
| 3||0||1 (2.7)||1 (1.5)|
|Gleason score on biopsy, n (%)|| || || |
| <7||18 (54.6)||24 (61.5)||42 (58.3)|
| 7||11 (33.3)||11 (28.2)||22 (30.6)|
| >7||4 (12.1)||4 (10.3)||8 (11.1)|
|TNM classification, n (%)|| || || |
| T1||25 (75.8)||32 (86.5)||57 (81.4)|
| T2||7 (21.2)||5 (13.5)||12 (17.2)|
| T3||1 (3)||0||1 (1.4)|
|Neurovascular bundle preservation, n (%)||20 (60.6)||29 (76.3)||49 (71)|
|Pelvic lymph node dissection, n (%)||9 (28.1)||8 (21.6)||17 (23.9)|
There was no significant difference between groups regarding operation duration, estimated blood loss, length of stay, catheter duration, postoperative pain or the consumption of analgesics.
No difference was found regarding complication rates between groups: six patients in group A and eight in group B had complications (18.2% vs 20.5%, P = 0.8). There was no significant difference in the severity of complications, based on the Clavien–Dindo classification, in the 14 patients with complications (P = 0.156). There were no blood transfusions. The postoperative pain reported by patients was similar in the two groups at 1 and 3 months (Table 2).
Table 2. Comparison of peri-operative outcome measures, complications and postoperative pain between groups
|Variable||Group A||Group B||Overall|| P |
|Operation duration (min), median (range)||220 (105–350)||200 (100–400)||200 (100–400)||0.282|
|Estimated blood loss (mL), median (range)||300 (20–1500)||300 (100–1100)||300 (20–1500)||0.540|
|Bladder catheterization (days), median (range)||7 (5–20)||6 (5–40)||7 (5–40)||0.203|
|Hospital stay (days), median (range)||6 (2–18)||6 (2–11)||6 (2–18)||0.999|
|Complications, n (%)||6 (18.2)||8 (20.5)||14 (19.4)||0.803|
|Complications grade I, n (%)||2 (33.3)||6 (75)||8 (57.1)||0.156|
|Complications grade ≥ II, n (%)||4 (66.6)||2 (25)||6 (42.9)|
|Analgesic uptake (1 month), n (%)||2 (6.1)||1 (2.9)||3 (4.2)||0.437|
|Postoperative pain (1 month), n (%)||10 (30.3)||6 (15.4)||16 (22.2)||0.129|
|Postoperative pain (3 months), n (%)||2 (6.1)||3 (7.7)||5 (6.9)||0.58|
Using the UCLA-PCI definition, the continence rate at 15 days, and at 1, 3 and 6 months, after surgery was 3.6%, 7.1%, 15.4% and 57.9%, respectively, in group A, and 5.9%, 26.5%, 45.2% and 65.4%, respectively, in group B (Fig. 4). AS and PR resulted in significantly higher continence rates at 1 and 3 months (P = 0.047 and 0.016, respectively). Using 0–1 pads/day as a definition of continence, AS and PR also resulted in significantly higher continence rates at 1 and 3 months (17.9% vs 50%, P = 0.008; 42.3% vs 80.6%, P = 0.003) but not at 15 days and 6 months after surgery.
Overall continence rates at 1, 3 and 6 months were 17.7%, 31.6% and 62.2%, respectively, using the UCLA-PCI definition, and 35.5%, 63.2% and 93.2% with the 0–1 pads/day definition.
No significant urinary stricture was found in either group.
With regard to oncological outcome, the overall PSM rate in the present study was 16.9%. There was no significant difference between the two groups regarding PSM rate and pathological stage (Table 3).
Table 3. Comparison of oncological outcome between groups: pathological stages, Gleason score and positive surgical margins
|Variable||Group A||Group B|| P |
|pTNM, n (%)|| || || |
| pT0||0||2 (2.8)||0.405|
| pT2a||3 (4.2)||2 (2.8)|
| pT2b||3 (4.2)||3 (4.2)|
| pT2c||18 (25)||24 (33.6)|
| pT3a||6 (8.4)||5 (6.9)|
| pT3b||3 (4.2)||3 (4.2)|
|Gleason score on surgical specimen, n (%)|| || || |
| <7||12 (16.8)||11 (15.3)||0.547|
| 7||18 (25)||26 (36.1)|
| >7||3 (4.2)||2 (2.8)|
|Positive surgical margin, n (%)||4 (12.1)||8 (21.1)||0.317|
At 1 month, group was the only predictive factor of continence in the univariable analysis (P = 0.047) (Table 4). Multivariable analysis included preoperative IPSS, ASA score and estimated blood loss (P = 0.072, 0.089 and 0.092, respectively); no independent predictive factors were found when comparing continent and non-continent groups.
Table 4. Predictive factors of continence (univariable)
|Variable||At 1 month||At 3 months|
|Continent (n = 11)||Incontinent (n = 51)|| P ||Continent (n = 18)||Incontinent (n = 39)|| P |
|Centre, n (%)|| || || || || || |
| 1||8 (73)||33 (64.7)||0.421||11 (61.1)||28 (71.8)||0.479|
| 2||0||7 (13.7)||4 (22.2)||4 (10.3)|
| 3||3 (27)||11 (21.6)||3 (16.7)||7 (17.9)|
|Age (years), mean (SD)||60.7 (6.3)||63.2 (5.2)||0.180||62.4 (6.6)||62.4 (5)||0.989|
|BMI (kg/m2), mean (SD)||25.9 (3.1)||25.5 (3.1)||0.752||26.1 (2.8)||25.2 (3.2)||0.316|
|Preoperative IPSS, median (range)||4 (0–10)||6.5 (0–17)||0.072||4.5 (0–17)||7 (0–17)||0.072|
|Preoperative IIEF-5, median (range)||23 (6–25)||21 (1–25)||0.946||21 (6–25)||22 (1–25)||0.957|
|ASA score, n (%)|| || || || || || |
| 1||2 (22.2)||26 (53)||0.089||8 (50)||20 (54)||0.786|
| >1||7 (77.8)||23 (47)||8 (50)||17 (46)|
|Prostate volume (mL), median (range)||40 (24–60)||39 (20–80)||0.607||40 (20–60)||37 (23–80)||0.900|
|TNM|| || || || || || |
| T1||9 (90)||43 (84.3)||0.854||14 (82.4)||33 (84.6)||0.728|
| T2||1 (10)||7 (13.7)||3 (17.6)||5 (12.8)|
| T3||0||1 (2)||0||1 (2.6)|
|EBL (mL), median (range)||200 (100–500)||300 (20–1400)||0.092||225 (100–600)||300 (20–1400)||0.236|
|OT (min), median (range)||190 (100–290)||200 (110–400)||0.458||170 (100–290)||200 (110–340)||0.061|
|Complications|| || || || || || |
| Yes||2 (18.2)||11 (21.6)||0.583||3 (16.7)||9 (23.1)||0.43|
| No||9 (81.8)||40 (78.4)||15 (83.3)||30 (76.9)|
|Specimen weight (g), median (range)||53 (37–85)||50 (34–100)||0.638||48 (37–85)||50 (34–100)||0.891|
|NBP|| || || || || || |
| Yes||7 (70)||34 (66.7)||0.576||13 (76.5)||25 (64.1)||0.362|
| No||3 (30)||17 (33.3)||4 (23.5)||14 (35.9)|
At 3 months, again, group was the only predictive factor of continence in the univariable analysis (P = 0.016), whereas multivariable analysis showed no independent predictive factors.
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
In the present randomized study, the performance of anterior suspension associated with posterior reconstruction during RALP led to improved early continence at 1 and 3 months after surgery compared to the control group, although there was no significant improvement in very early (15 days) and late (6 months) continence.
Eastham et al.  first described the effects of AS on continence in RRP. Subsequently, Campenni  used a two-suspension stitch technique during RRP, with improved continence in the suspension group (P = 0.048). In a prospective randomized study of RRP, Noguchi et al.  found a difference in the continence rate at 1 month (P = 0.029) and 3 months (P = 0.034) between a suspension group and a control group. Their suspension technique differed from that employed in the present study in that they used the DVC and the ligaments rather than the pubic bone to anchor their ligation.
The AS performed in the present study was recently described in RALP by Patel et al.  in a non-randomized prospective trial comparing the continence of 237 patients with and 97 patients without AS. There was a significant difference between groups at 3 months (P = 0.013) but not at 1, 6 or 12 months.
The present study found a significant difference in early continence, at 1 month, which we consider to be related to the concomitant use of posterior reconstruction.
Rocco et al.  first described PR in a study involving 300 RRP. Continence improved in the reconstruction group at 1 and 3 months (P < 0.05). When this technique was transferred to LRP , a significant difference was found at 1 month (P = 0.001) but not at 3 months (P = 0.25).
During RALP, in a well-designed, although not randomized trial including 803 patients, Coelho et al.  showed that PR shortens continence recovery time (P = 0.037) and improves early continence (P = 0.016). Other non-randomized studies during RALP showed inconsistent results. Nguyen et al.  found a difference in the 1-month continence rate, whereas Joshi et al.  found no difference in the 3-month continence rates of two groups with or without PR.
There have been three randomized studies examining the effects of reconstruction on continence during RALP. Using only posterior reconstruction, Sutherland et al.  found no difference in 3-month continence in 94 patients with or without PR (P = 0.1). The posterior reconstruction technique employed in the present study differed from the PRs of Sutherland et al.  and Coelho et al.  because we performed a single-layer reconstruction, aproximating the Denonvillier's fascia and the posterior tissues of the rhabdomyosphincter. The present study used a separate running suture for the whole anastomosis. We consider that this easy-to-perform technique sufficiently tightens up the posterior support of the vesico-uretral anastomosis.
In a randomized study of AS and PR during RALP, Menon et al.  found no difference in very early continence at 1, 2, 7 and 30 days (P > 0.1). Koliakos et al.  found a difference in the continence rate at 7 weeks after surgery (P = 0.029). Unlike the present study, Koliakos et al.  performed AS after completing the vesico-ureteral anastomosis, using stitches anchored to the DVC rather than to the pubic bone.
Discrepancies in the continence rates reported in the studies (Table 5) may be caused partly by the lack of a standardized definition of continence. Some studies define continence as no pads used and no leak of urine [16,18,25], some as the use of no pads , and others as the use of zero or one pad per day [11–13]. To avoid biases, the present study used a validated self-administered questionnaire with a strict definition of continence. We consider that this explains why our overall continence rate may appear to be lower than in other studies. Indeed, using definition of ‘0–1 pads/day’, the continence rates in the present study are much higher (63% at 3 months and 93% at 6 months).
Table 5. Comparison of continence with previous studies of anterior suspension and/or posterior reconstruction
|Study|| n ||Procedure||Study design||Evidence*||Definition of continence||Continence (%) (treatment vs control)|| P |
|Present study||72||AS + PR (RALP)||R||1b||0 pad + 0 leak (UCLA-PCI)||1 month: 26.5 vs 7.1||0.047|
|3 months: 45.2 vs 15.4||0.016|
|Noguchi et al. ||60||AS (RRP)||R||1b||0 pad + 0 leaks||1 month: 53 vs 20||0.029|
|3 months: 73 vs 47||0.034|
|Campeni et al. ||50||AS (RRP)||NR||2b||0 pad + 0 leaks||3 months: 32 vs 12||0.048|
|Patel et al. ||331||AS (RALP)||NR||2a||0 pad + 0 leaks||1 month: 40 vs 33||0.282|
|3 months: 92.8 vs 83||0.013|
|Sutherland et al. ||94||PR (RALP)||R||1b||0 or 1 pad||3 months: 63 vs 81||0.07|
|Rocco et al. ||300||PR (RRP)||NR||2b||0 or 1 pad||1 month: 74 vs 30||<0.001|
|3 months: 85.2 vs 46||<0.001|
|Rocco et al. ||62||PR (LRP)||NR||2a||0 or 1 pad||1 month: 83.2 vs 32.3||<0.001|
|3 month: 92.3 vs 76.9||0.25|
|Nguyen et al. ||62||PR (RALP)||NR||2b||0 or 1 pad||6 weeks: 56 vs 17||0.006|
|Joshi et al. ||107||PR (RALP)||NR||2a||0 leaks||3 months: 25 vs 31||0.393|
|Coelho et al. ||803||PR (RALP)||NR||2a||0 pad||1 week: 28.7 vs 22.7||0.048|
|4 weeks: 51.6 vs 42.7||0.016|
|Koliakos et al. ||47||AS + PR (RALP)||R||1b||0 or 1 pad||7 weeks: 65 vs 33||0.029|
|Menon et al. ||116||AS + PR (RALP)||R||1b||0 pad||1 month: 42 vs 47||>0.1|
Unlike Menon et al. , the present study showed a difference in the 1-month rate continence. The continence rate in the control group at 1 month is less important than that reported by Menon et al. . The most important difference the study by Menon et al.  and the present study is the number of RALP cases previously experienced. The present study confirms that adding AS and PR during RALP may provide more benefits to the patient when the surgeon is moderately experienced.
The mechanism of continence after surgery is complex and not totally understood. In line with Patel et al. , we consider that AS provides an anatomical support for the urethra, allowing improvement of the urethral length during the apex dissection, and stabilization of the urethra and the striated sphincter in an anatomical position. Moreover, adding PR enables the surgeon to perform a tension-free anastomosis and recreate a posterior support for the sphincter. Some studies have reported a decrease in anastomotic leakage using PR . We did not systematically evaluate anastomotic leakage in the present study (a cystogram was made only if there was clinical evidence of leakage).
In the present study, morbidity did not increase in the AS and PR group. Complication rates and perineal pain were similar between the groups and no pubic osteitis was observed in the suspension group. Moreover, the PSM rate in the present study was not significantly affected by AS and PR, with oncological control remaining the primary endpoint of the prostatectomy. Indeed, it may be assumed that the anterior suspension, performed before apical dissection and urethral section, could influence the apical margin rate. The results obtained in the present study invalidate this assumption.
The present study has a number of limitations. First, there was no protocol for peri-operative care, particularly with regard to rehabilitation, which could have led to differences between centres. However, the centre variable was not significantly associated with continence in the statistical analysis, and AS and PR appear to increase the continence independently of peri-operative care. Another limitation is the number of patients lost to follow-up, which exceeded the number predicted. There were many out-of-state patients who were referred to our centres. Because these patients did not live near the study centres, they required several hours to complete their visits. To some extent, this may explain our attrition rate. Nevertheless, the results of the present study are statistically significant because the effects of the techniques employed were greater than anticipated. Indeed, we have conducted analysis that consider patients lost to follow-up as continent and incontinent aiming to determine how stable the estimated continence rates were. At 3 months, in the present study, we report a significant difference in continence between groups A and B (P = 0.016). Considering all patients lost to follow-up at 3 months as continent, 11 out of 33 patients in group A (33.3%) and 22 out of 39 patients in group B (56.4%) were continent. The difference was statistically significant (P = 0.05). Considering all patients lost to follow-up as incontinent, there were four out of 33 patients in group A (12.1%) and 14 out of 39 patients in group B (35.9%) who were continent. This difference was still significant (P = 0.02) and we consider that it shows the stability of the continence rates that were analyzed.
The present study comprises one of the few randomized-control trials on the topic and the only multi-institutional study.
In conclusion, anterior suspension combined with posterior reconstruction during RALP significantly improves the continence recovery at 1 and 3 months after surgery. These techniques are easy to perform, are safe and do not compromise oncological control.