A prospective, non-randomized trial comparing robot-assisted laparoscopic and retropubic radical prostatectomy in one European institution


Vincenzo Ficarra, Associate Professor of Urology, Department of Oncologic and Surgical Sciences, Urology Clinic, University of Padua, Monoblocco Ospedaliero – IV Floor, Via Giustiniani 2-35128, Padua, Italy.
e-mail: vincenzo.ficarra@unipd.it



To compare the functional results of two contemporary series of patients with clinically localized prostate cancer treated by robot-assisted laparoscopic prostatectomy (RALP) or retropubic radical prostatectomy (RRP).


This was a non-randomized prospective comparative study of all patients undergoing RALP or RRP for clinically localized prostate cancer at our institution from February 2006 to April 2007.


We enrolled 105 patients in the RRP and 103 in the RALP group; the two groups were comparable for all clinical and pathological variables, except median age. For RRP and RALP the respective median operative duration was 135 and 185 min (P < 0.001), the intraoperative blood loss 500 and 300 mL (P < 0.001) and postoperative transfusion rates 14% and 1.9% (P < 0.01). There were complications in 9.7% and 10.4% of the patients (P = 0.854) after RRP and RALP, respectively; the positive surgical margin rates in pT2 cancers were 12.2% and 11.7% (P = 0.70). For urinary continence, 41% of patients having RRP and 68.9% of those having RALP were continent at catheter removal (P < 0.001). The 12-month continence rates were 88% after RRP and 97% after RALP (P = 0.01), with the mean time to continence being 75 and 25 days (P < 0.001), respectively. At the 12-month follow-up, 20 of 41 patients having bilateral nerve-sparing RRP (49%) and 52 of 64 having bilateral nerve-sparing RALP (81%) (P < 0.001) had recovery of erectile function.


RALP offers better results than RRP in terms of urinary continence and erectile function recovery, with similar positive surgical margin rates.


retropubic radical prostatectomy


laparoscopic RP


robot-assisted LRP


neurovascular bundle


International Index of Erectile Function


body mass index


Eastern Cooperative Oncology Group


positive surgical margin


International Consultation of Incontinence Questionnaire – Urinary Incontinence


interquartile range.


Anatomical retropubic radical prostatectomy (RRP), as described by Walsh et al.[1] in 1982, has been the most commonly used surgical treatment for clinically localized prostate cancer in the last few decades. Since 1999, some authors suggested the possibility of achieving better functional results through the magnified vision made possible by the laparoscopic approach [2,3]. However, laparoscopic radical prostatectomy (LRP) has been shown to be a difficult technique to learn and most of the available functional results have shown no significant advantage, both in terms of urinary continence and recovery of erectile function when compared with RRP [4].

In recent years the development of traditional LRP has lead to robot-assisted LRP (RALP). Although RALP has been used widely in the last 5 years, especially in the USA, only a few non-randomized studies comparing RRP or pure LRP with RALP are currently available. These studies show a lower blood loss, lower transfusion rates during and after surgery, shorter catheterization time, and shorter hospital stay for patients undergoing RALP [5]. The lower invasiveness of RALP than RRP was recently reported by Fracalanza et al.[6], who showed that the levels of interleukin-6, C-reactive protein and lactate were significantly lower in patients who had undergone RALP than those treated with RRP.

Although important, the advantages in terms of perioperative outcomes might be insufficient to justify the higher costs of RALP. The success of robotic surgery should be based on confirming equal oncological outcomes and superior functional results. Although the published series of RALP reported excellent continence and potency rates at an appropriate follow-up [7–9], few comparative studies evaluated the functional outcome in RALP and RRP. Only Tewari et al.[10] reported advantages in terms of earlier urinary continence and recovery of erectile function in those patients undergoing RALP compared with a contemporary series of RRP.

The purpose of the present study was to compare the functional results achieved in two contemporary series of patients with clinically localized prostate cancer treated by either RALP or RRP at the same institution.


In a non-randomized, prospective comparative study, all patients undergoing RALP or RRP for clinically localized prostate cancer at our institution were enrolled from February 2006 to April 2007. The choice between RRP or RALP was based on a joint decision by the patients and physicians, and all the patients were theoretically suitable for the other surgical approach. There was no attempt to randomize patients.

All RALP procedures were done by two surgeons who had completed at least 50 RALPs each before the beginning of the study. The ports for the three-arm robot (da Vinci Intuitive Surgical, Inc., Sunnyvale, CA, USA) and the traditional laparoscopic tools were placed as previously reported [11]. Once the peritoneal cavity has been inspected, a transverse peritoneal incision is made extending from the right to the left medial umbilical ligament. The extraperitoneal space is developed after transecting the medial and median umbilical ligaments. This dissection permits the bladder and prostate to fall back. Subsequently, the prostatovesical junction is identified at the point at which loose fat can no longer be swept off the prostate, and is incised. Once the anterior ureteric wall has been incised and the Foley catheter identified, the posterior urethral wall is incised, preserving the bladder neck. After this, the posterior layer of Denonvilliers’ fascia is exposed and incised, exposing the vasa and seminal vesicles, which are grasped and pulled upward individually. The vasa are transected and the seminal vesicles are skeletonized. Subsequently, the plane between the rectum and the prostate is dissected towards the prostate apex. Then, the anterolateral surface is dissected, following intrafascial or interfascial planes. The lateral pedicles are controlled using monopolar forceps, using electrocautery as little as possible and as far as possible from the neurovascular bundles (NVBs). Anterior prostate dissection is completed and the dorsal vein complex is transected. Haemostatic sutures are usually placed after the section. Once the dorsal vein complex has been sectioned, the dissection of the prostatic apex is completed. The urethra is transected at the level of the prostatic apex and the prostate specimen is placed in an Endocatch bag. After further haemostatic sutures on the prostatic bed, the vesico-urethral anastomosis is made using the Van Velthoven technique. Patients with an intermediate or high risk of lymph node involvement according to the Partin table for progression undergo bilateral iliac and obturator lymphadenectomy. A drain is positioned through the assistant’s lateral 5-mm working trocar. The prostate is removed via the peri-umbilical port and the wounds are closed in two layers with polyglactin fascial sutures.

The RRP was done by four surgeons who had previously completed at least 400 RRP each. The procedure was extraperitoneal through an infraumbilical-pubic incision, with a standard iliac and obturator lymph node dissection in every patient. The RRP followed these steps: incision of the endopelvic fascia; ligature and section of the deep venous complex; controlled section of the puboprostatic ligaments, with preservation of the pubo-urethral components; dissection of the prostatic apex and section of the urethra; section of the pubo-urethralis muscle and retrograde dissection of the posterior surface of the prostate and seminal vesicle; preservation of the NVBs, as described by Walsh et al.[1]; bladder neck dissection and racket handle reconstruction after eversion of the bladder mucosa. A urethro-vesical anastomosis was made with six to eight interrupted sutures on a 18 F catheter.

In both groups, every attempt was made to preserve the NVBs in those patients with cT1c-T2a prostate cancer, a biopsy Gleason score of ≤7 and a preoperative International Index of Erectile Function (IIEF) score of >26.

All the patients had a cystogram at 4 or 5 days after surgery to assess the status of the vesico-urethral anastomosis. The catheter was removed if there was a water-tight anastomosis or if there was little extravasation of contrast medium (<5%). All the patients undergoing unilateral or bilateral nerve-sparing procedures had early rehabilitation by intracavernosal injections with prostaglandin E1 and/or phosphodiesterase-5 inhibitors (sildenafil 100 mg daily or tadalafil 20 mg every third day).

The RP specimen was formalin-fixed in the standard manner. The whole-mount sections were identified in a consecutive manner with capital letters, always starting from the section closest to the apex, so that the whole specimen was available for histological examination. The en-face section was then processed as a single section. In particular, the paraffin-embedded specimen was examined histologically as 4 mm-thick, whole-mount, haematoxylin and eosin-stained sections.

In each patient we collected prospectively the following clinical and pathological variables: age, body mass index (BMI), Eastern Cooperative Oncology Group (ECOG) performance status, Charlson score, preoperative total PSA level, Gleason score at biopsy, clinical stage (cT, TNM 2002), Gleason score at RP, pathological stage (pT) and positive surgical margin (PSM) rate.

We also evaluated the following perioperative variables: surgical time for RP, blood loss, perioperative transfusion rate, complications (during surgery, and early, within the first 30 days after surgery), status of the vesico-urethral anastomosis at cystography, time to catheter removal and hospital stay.

Perioperative complications were defined as any deviation from the standardized RALP pathway and recorded according to the classification described by Clavien et al.[12]. Briefly, the Clavien system classifies complications into five grades: grade I, deviation from the normal course but with no need for therapy; grade II, complication requiring pharmacological or bedside treatment; grade III, complication with the need for surgical, endoscopic or radiological intervention (further IIIa/b: without/with the need for general anaesthetic); grade IV, life-threatening complication requiring intensive care with residual disability; and grade V, death [12].

The following functional outcomes were addressed: urinary continence at the time of catheter removal, urinary continence at the 12-month follow-up, time to recovery of urinary continence, presence of bladder neck stenosis; erectile function at the 12-month follow-up, time to recovery of erectile function.

At the time of catheter removal all patients who had a dry safety pad within the first 24 h were defined as continent. Urinary continence and erectile function at the follow-up were evaluated using the International Consultation of Incontinence Questionnaire – Urinary Incontinence (ICIQ-UI) short-form instrument and the IIEF-5, respectively. Questionnaires were self-completed before surgery and at the 12-month follow-up. All the patients reporting no leak or leaks ‘about once a week or less often’ for the question ‘How often do you leak urine?’ were defined as continent. All the patients with a IIEF-5 of >17 were defined as potent.

Parametric continuous variables are given as the mean (sd) and nonparametric continuous variables as the median (interquartile range, IQR). Student’s t-test, the Mann–Whitney U-test and Pearson’s chi-square were used to compare parametric continuous, nonparametric continuous and categorical variables, respectively, with a two-sided P < 0.05 considered to indicate statistical significance.


In all, 105 patients were enrolled in the RRP and 103 in the RALP group; Table 1 summarizes the clinical and pathological characteristics of the evaluated patients. The two groups of patients were comparable for comorbidity, BMI, ECOG performance status, total PSA level, Gleason score at biopsy, clinical stage, pathological stage, and overall positive surgical margin rates, and positive surgical margin rates in pT2 cancers. The median age was significantly different between the groups of patients, at 65  (61–69) years for those undergoing RRP and 61 (57–67) years for those undergoing RALP (P < 0.001).

Table 1.  The clinical and pathological characteristics of the patients enrolled in the study, and the complications
Median (IQR) or n (%) variableRRPRALPP
  1. * data available in 104 RRP and 97 RALP only.

No. of patients105103 
Age, years 65 (61–69) 61 (57–67)<0.001
Charlson score >2 27 (26) 14 (14)0.06
BMI, kg/m2 26 (24–28) 26 (24–28)0.22
ECOG performance status >0  2 (1.9)  2 (1.9)0.98
Total PSA level, ng/mL  6 (5–10)  6.4 (4.6–9)0.32
Gleason score at biopsy*  0.63
 6 67 (64) 71 (73) 
 7 29 (28) 18 (19) 
 8–10  8 (8)  8 (8) 
Clinical T stage  0.12
 T1c 66 (63) 77 (75) 
 T2a-b 32 (30) 22 (21) 
 T2c  7 (7)  4 (4) 
Prostate volume (by TRUS) 40 (30–47) 37.5 (30–48)0.40
Pathological T stage  0.19
 pT2 49 (47) 60 (58) 
 pT3a 42 (40) 39 (38) 
 pT3b 14 (13)  4 (4) 
PSM rate   
 Overall 21 (30) 35 (34)0.97
 In pT2 cancers  6 (12)  7 (12)0.70
Perioperative complications
Colon lesion  0  1 (1) 
Rectal lesion  0  1 (1) 
Postoperative bleeding  7 (7)  7 (6) 
Paralytic ileus  1 (1)  1 (1) 
Cardiovascular complications  2 (2)  0 
Wound dehiscence  1 (1)  0 
Overall  11 (13) 10 (10)0.854

Forty-one patients (39%) in the RRP group and 64 (62%) in the RALP group had a bilateral nerve-sparing prostatectomy. The median operative duration was significantly longer in the RALP group (P < 0.001), and the median intraoperative blood loss was significantly greater in the RRP group (P < 0.001; Table 1).

During the early course after RP, 15 (14%) patients undergoing RRP and two (1.9%) undergoing RALP had homologous blood transfusions (P < 0.01).

At cystography, the urethrovesical anastomosis was watertight in 68 patients (65%) after RRP and in 76 (74%) after RALP. A small extravasation of contrast medium was present in 16 patients (15%) in the RRP group and in 19 (18%) after RALP, while there was more extravasation in 21 (20%) and eight (8%) (P = 0.03), respectively. The catheter was removed after a median of 6 (5–12) days in the RRP group and 5 (4–7) days after RALP (P < 0.001). The median hospital stay was 7 (6–9) days after RRP and 6 (5–8) days after RALP (P = 0.01). There were perioperative complications in 11 (11%) patients undergoing RRP and in 10 (10%) undergoing RALP (P = 0.854). Table 1 summarizes the complications recorded in the two groups of patients. Specifically, surgical re-exploration was needed in four patients (4%) undergoing RALP, due to bleeding, in one case from a port site and in three from the small pelvis.

According to the Clavien classification, in the RRP group one patient had a complication determined to be Clavien grade I, and 10 that were grade II. In the RALP group one patient had grade I complications, four had grade II, and five (5%) had grade III.

Generally PSMs were detected in 21 patients after RRP (21%) and in 35 (34%) after RALP (P = 0.972). However, considering only those patients with pathologically localized prostate cancers, PSMs were identified in six of 49 patients (12%) after RRP and in seven of 60 (12%) after RALP, respectively (P = 0.708).

After catheter removal, 43 patients treated by RRP (41%) and 71 (69%) by RALP were continent (P < 0.001). At a minimum follow-up of 12 months, 92 (88%) of the patients in the RRP group and 100 (97%) of those in the RALP group were continent (P = 0.01). The mean time to continence was 75 (116) days in the RRP and 25 (39) days in the RALP group, respectively (P < 0.001). Stenosis of the urethrovesical anastomosis was found in six patients (6%) after RRP and in three (3%) after RALP (P = 0.32).

Considering only patients undergoing bilateral nerve-sparing surgery, at ≥12 months of follow-up, 20 of 41 (49%) in the RRP and 52 of 64 (81%) in the RALP group were potent (P < 0.001). Similarly, evaluating only the patients aged <65 years and with a Charlson score of ≤2, 15 of 26 (58%) in the RRP and 41 of 49 (84%) in the RALP group were potent (P = 0.01). In this subgroup of patients, the mean time to recovery of erectile function was 6.7 (3.6) months in the RRP group and 3.9 (3) months in the RALP group (P < 0.01).


To our knowledge, the present study is the first prospective comparative assessment comparing RRP and RALP with the use of validated questionnaires to evaluate the functional results. Our data showed that RALP can offer better functional results than RRP in terms of both urinary continence and recovery of erectile function at ≥12 months of follow-up.

In the last few years RALP has emerged as the main alternative to RRP. The first reported data on RALP showed a significantly shorter ‘learning curve’ than for pure LRP [5,13]. Moreover, other data suggested that the ‘learning curve’ can be completed within ≈ 20 cases also by those surgeons with no previous experience with pure LRP [11,14].

Some of the available comparative studies showed that, compared with RRP, RALP is associated with a significant reduction in intraoperative blood loss [10,13,15,16], blood transfusions and postoperative pain [10,12], as well as with a shorter hospital stay [10,12,15]. However, most of these comparative studies were retrospective and enrolled only a few patients. Our analysis reconfirmed the advantages of RALP for some outcomes in the context of a prospective study, evaluating two contemporary groups of patients managed using the same postoperative path. In our series the complication rates were similar between RRP and RALP. However, five patients in the RALP group had Clavien grade III complications. Specifically, there was postoperative bleeding needing surgical exploration in four patients, while two potentially severe complications, e.g. colon and rectal injuries, were recognized during RALP and promptly repaired, with no significant effect on the postoperative course. On the whole, these data indicate only an insignificant trend in favour of RRP, which probably reflects the difference in experience with the two techniques in our centre.

For the hospital stay our values were far longer in both study arms than those reported in all the series from the USA, which probably reflected different pathways of management. Specifically, due to the differences in economic health systems and cultural background, in Italy patients often stay in hospital until the urinary catheter is removed, while patients in the USA are usually discharged very quickly if there is an uneventful early postoperative course.

However, some of the statistically significant differences in the perioperative outcomes might not translate into a clinically significant advantage for patients, and thus could not justify the costs related to the use of the robotic system. For these reasons we believe that oncological and functional outcomes should be the basis of any meaningful comparison between different surgical approaches. Most of the published studies on RALP provided only preliminary oncological data, mainly for organ-confined disease. The PSM rates in patients with pT2 cancers in the present RALP series can be considered similar to those reported in the different RRP or pure LRP series [3,5], and to those in other available comparative studies (Table 2) [10,15,17–19]. The data from our comparative study reconfirmed that the PSM rates stratified by pathological stage were comparable in the two different techniques, using the same pathological protocols to sample the RP specimen. The PSM rates in the RRP series were similar to those reported in other published series of RRP. By contrast, our rate of PSMs in the RALP series seemed higher than those reported in other series, where values were 1.5–2.5%[7–9]. This difference might be due to the different pathological protocols used to process the RP specimens in different centres, and to the different expertise of the surgeons. Our PSM rates were similar to those reported by the same authors evaluating earlier phases of their experience [20–24].

Table 2.  PSM rates, urinary continence and erectile function in the available comparative studies evaluating RRP and RALP
RefStudy designNo. of patientsPSM, %Urinary continence, %Erectile function, %
OverallpT2MethodCriteria6 m12 mMethodCriteria3 m12 m
  1. Median time to *urinary continence, †presence of erections, or ‡intercourse. Unv Q, unvalidated questionnaire; d, days; m, months.

[10]ProspectiveRRP 10023InterviewNo pad160 d*InterviewErection440 d
RALP 200 6   44 d*  180 d
RRP      Intercourse700 d
RALP       340 d
[15]RetrospectiveRRP 6020 9Unv QNo pad 75    
RALP 6016.7 4.5   76    
[17]ProspectiveRRP 3029        
RALP 3026        
[18]RetrospectiveRRP 20035.724        
RALP 20015 9.4        
[19]RetrospectiveRRP 56417Unv QNo leak93.7Unv QIntercourse63
RALP 28615.6   91.8  70
PresentProspectiveRRP 1052112.2ICIQ-UINo leak 88IIEF-5IIEF-5 >1749
RALP 10334 11.7   97  81

Reported 12-month urinary continence rates after RALP are 84–98%[5]; specifically, in the most important series, Menon et al.[7] found that 92% of patients were continent, while others [8–9] reported 12-month urinary continence rates as high as 97% and 95%, respectively . For the available comparative studies, the data are very limited and none of the previous reports used validated questionnaires to evaluate urinary continence (Table 2). Only the comparative study by Tewari et al.[10] showed an earlier recovery of urinary continence in patients having RALP than after RRP. The results of the present study reconfirmed, using a validated questionnaire such as the ICIQ-UI short-form, the earlier recovery of continence in patients undergoing RALP, and showed a significant advantage in the 12-month continence rate in the RALP group. The most original data was the high percentage of patients who were continent at catheter removal, which might be due to the more accurate apical dissection and urethrovesical anastomosis.

The use of the robot-assisted approach could probably facilitate the easier preservation of the so-called ‘Aphrodite’s veil’ during nerve-sparing procedures [25], compared to what can be conserved with retropubic surgery [26]. The reported 12-month potency rates are 70–90%[7–9,27] and Menon et al.[7] suggested even higher values at a longer follow-up. Similar to urinary continence, the available data on recovery of erectile function from comparative studies were few and none of the previously reports used a validated questionnaire (Table 2). Only Tewari et al.[10] showed a significant advantage in terms of recovery of erectile function for patients treated by robotic surgery compared to RRP. Our results were similar but in the context of a prospective study comparing two contemporary series of patients using a validated questionnaire to assess erectile function, and the same protocol of corpora cavernosal rehabilitation after surgery.

The main strengths of the present study are the prospective collection of the data in the two arms, and the use of validated questionnaires to evaluate both urinary continence and erectile function, specifically the ICIQ-UI short form and IIEF-5, while all the other available studies used unvalidated questionnaires or open interview [10,15,19]. Moreover, the same protocol for preoperative diagnosis and staging of prostate cancer, perioperative treatment, and pathological evaluation were adopted in the both study groups, and the follow-up was long enough to evaluate complications and functional outcomes. However, the main limitation of the present study was the lack of randomization. The selection between the treatments was based mainly on the preferences of the surgeons evaluating the patients. We believe that, to date, a randomized trial might be difficult to conduct because patients are usually unwilling to be randomly assigned to different treatments and more are willing to accept only robotic surgery. Due to the lack of randomization, the two study groups were different in median age, with patients undergoing RALP being younger than those in the RRP group. However, the BMI, ECOG performance status and Charlson comorbidity scores were not significantly different. Other possible limitations of the study were the single-centre, multi-surgeon basis and the relatively few patients enrolled. Two surgeons performed all the robotic procedures, while four different surgeons did the RRP. Although all of the surgeons were experienced it is likely that because four surgeons performed RRPs this might have accounted for some variability in the results. However, it reflects the daily clinical practice in many centres proposing different surgical approaches for the same procedures.

In conclusion, the present prospective, non-randomized, comparative study showed that RALP can offer better results than RRP in terms of the recovery of urinary continence and erectile function, without, to date, jeopardizing the oncological outcome, evaluated according to the PSM rates. Our data justify the use of robotic surgery for treating clinically localized prostate cancer as an alternative to the traditional retropubic approach, although at present the costs of RALP are higher.


This study was partially funded by the Italian Ministry for University and Research (MIUR). MIUR had no role in study design and realization, data collection, analysis and interpretation.


None declared.