- Top of page
- Patients and Methods
- Conflict of Interest
For many years now radical prostatectomy (RP) has been the preferred therapeutic option for patients with localized or locally advanced prostate cancer . Open radical prostatectomy (ORP) has been the most commonly used surgical technique for decades, but recently, robot-assisted radical prostatectomy (RARP) has become an alternative option . Despite the large number of RPs performed, urinary incontinence (UI) remains a common postoperative outcome [3-6].
Several studies have compared UI after ORP and RARP. Some studies found that patients achieved continence much earlier after RARP than after ORP [7-9], but other studies could not confirm this [10-12]. The IPSS for the evaluation of voiding symptoms was used in several studies [13-15] and in all these studies the IPSS results were found to improve after surgery.
Miller et al.  compared health-related quality of life between ORP and RARP in a small cohort of 162 patients. They found a faster return to baseline quality of life in patients after RARP but these results were not confirmed in the study by Malcolm et al. . Additionally, Hara et al.  found that quality of life with regard to difficulty of UI was significantly disturbed by surgery.
The objective of the present study was to compare functional outcomes, i.e. UI, voiding symptoms and quality of life in patients who underwent ORP with functional outcomes in those who underwent RARP.
Patients and Methods
- Top of page
- Patients and Methods
- Conflict of Interest
A total of 116 and 64 patients with localized or locally advanced prostate cancer who were planned to undergo ORP or RARP, respectively, in the University Hospitals of Leuven between September 2009 and July 2011 agreed to participate in this trial. All patients gave written informed consent. Patients with cognitive problems, non-Dutch speaking patients and those simultaneously planned for a salvage procedure or other surgery in the pelvic region were excluded.
Before surgery, patients were recruited at the outpatient clinic of urology. Patients could not be randomized on surgical approach because most patients chose this based on personal preferences for open or robot-assisted surgery or for a specific surgeon. A patient had no choice of surgical approach only when their medical condition did not permit one or the other technique. Patients with low-risk disease more often underwent RARP while almost all patients with high-risk disease (≥ cT3, PSA ≥ 20 and Gleason Score ≥ 8) underwent open surgery because they required an extended lymph node dissection. Three surgeons, each specialized in ORP and/or RARP, completed all operations. One surgeon had performed >3000 open surgeries (H.V.P.), the second surgeon had performed >700 ORPs and >50 RARPs and the third surgeon had performed >150 RARPs. Open surgery for patients in the low-/intermediate- and high-risk groups was performed according to the technique previously described [19, 20]. Robot-assisted surgery was performed as described by Menon et al. .
All patients followed an individual pelvic floor muscle training programme, on an outpatient basis, once a week until total continence was achieved. Continence was defined as 3 consecutive days of 0 g urine leak using the 24-h pad test. After surgery, the pelvic floor muscle training programme was started on the day of catheter removal and consisted of exercises of the pelvic floor, manually controlled by the therapist and supplied with elektromyography-biofeedback. All patients were treated by a dedicated and specialized therapist. Additionally, patients performed 60 contractions per day at home.
All patients performed a 24-h pad test each day for 3 days before surgery. After catheter withdrawal urine leak per 24 h was recorded daily until continence was achieved. Auto-measurements of the patients were double-checked on a regular basis by weighing the pad the patients wore, when they came to therapy. Furthermore, from time to time, patients were asked to collect all pads used in a 24-h period in a plastic bag and to take this to the hospital for an additional measurement. Types of incontinence (stress, urge) were not differentiated in the analysis.
All patients were prospectively assessed before and 1, 3, 6 and 12 months after surgery at the Department of Physiotherapy. Patients had to perform a 1-h pad test, fill in a score on a visual analogue scale (VAS) concerning their subjective feeling about UI and complete the IPSS, a questionnaire to evaluate voiding symptoms (score 0–35). Additionally the King's health questionnaire (KHQ), a self-administered questionnaire designed to assess the impact of UI on quality of life, was completed. The nine domains of the questionnaire are general health perception, incontinence impact, role limitations, physical limitations, social limitations, personal relationships, emotions, sleep or energy and severity measures. Patients' weight and height were also assessed. A trained assessor performed the measurements. Urodynamic measurements were not performed in the first year after surgery. In the initial postoperative period no anticholinergics were prescribed.
The primary outcome measures were time to continence and cumulative incidence of continence (24-h pad test). The secondary outcomes were the point prevalence of continence, measured using the 1-h pad test and the VAS score at 1, 3, 6 and 12 months after surgery. At the same time points, IPSS and KHQ were assessed.
Patient characteristics were compared between the ORP and RARP groups. An independent t-test was used for continuous, normally distributed data and the Fisher's exact/ chi-squared test was used for categorical data.
Data were analysed according the intention-to-treat principle. Kaplan–Meier analyses with a log-rank test were used to compare the time to continence between both types of surgery. Dropouts were censored at the point of last follow-up. Afterwards, Cox regression was applied to compare the different groups with regard to time to continence, with correction for the different covariates. Fisher's exact test was used to compare objective and subjective point prevalences of urinary continence, defined as 0 g on the 1-h pad test, and the VAS score measured at 1, 3, 6 and 12 months after surgery. For comparison of the voiding symptom severity and quality of life at 1, 3, 6 and 12 months after surgery, the Mann–Whitney U-test was used, because the data were not normally distributed. All data were analysed using SPSS 19.0.
The study received ethical approval from the medical ethics commission of the University Hospitals of Leuven responsible for human/animal experimentation (ML5470).
- Top of page
- Patients and Methods
- Conflict of Interest
In all, 116 patients underwent ORP and 64 patients underwent RARP. Seven patients were lost to follow-up immediately after surgery because of sudden death (n = 1), cerebrovascular accident (n = 1) or transport problems (n = 5). Three other patients dropped out of the study at 106, 177 and 204 days after surgery with 4, 17 and 28 g urine leak, respectively (24-h pad test). The indwelling catheter was removed after a median of 11 and 12 days after ORP and RARP, respectively. Nine patients were classified as outliers (6, 32, 34, 40, 41, 43, 49, 75 and 101 days of catheter-wearing), but were not omitted from the analysis. Three of these patients underwent ORP and six RARP. Because randomization could not be performed at surgery level, D'Amico risk groups and nerve-sparing status differed between the ORP and RARP groups. All other baseline characteristics were similar between the two groups (Table 1).
Table 1. Characteristics of patients according to type of surgery
|N = 116||N = 64|
|Mean (sd) age, years||62.22 (6.12)||61.48 (6.08)||0.436*|
|Median time urinary catheter, days (SD)||11||12||0.252*|
|D'Amico risk group, n (%) || || ||0.000†|
|I||8 (7)||14 (22)|| |
|II||50 (43)||43 (67)|| |
|III||57 (49)||7 (11)|| |
|Missing||1 (1)||0 (0)|| |
|Nerve-sparing, n (%) || || ||0.000†|
|Non-nerve-sparing||24 (21)||0 (0)|| |
|Unilateral nerve-sparing||40 (34)||4 (6)|| |
|Bilateral nerve-sparing||52 (45)||60 (94)|| |
|Missing||0 (0)||0 (0)|| |
|Surgical margin status, n (%)|| || ||0.204†|
|Negative||91 (78)||45 (70)|| |
|Positive/doubtful||24 (21)||19 (30)|| |
|Missing||1 (1)||0 (0)|| |
|Preoperative continence status, n (%)|| || ||0.500†|
|Continent||76 (66)||40 (63)|| |
|Incontinent||33 (28)||22 (34)|| |
|Missing||7 (6)||2 (3)|| |
|Body mass index, n (%)|| || || |
|≤25.0 kg/m2||33 (29)||21 (33)||0.501†|
|25.1—30.0 kg/m2||63 (54)||36 (56)|| |
|>30.0 kg/m2||20 (17)||7 (11)|| |
|Mean (sd) preoperative IPSS score||6.80 (5.46)||8.14 (6.03)||0.087‡|
|Mean (sd) preoperative KHQ score|| || || |
|General health perceptions||78.66 (17.83)||77.34 (16.50)||0.468‡|
|Incontinence Impact||86.49 (21.52)||80.21 (25.69)||0.072‡|
|Role limitations||94.11 (14.62)||94.27 (14.91)||0.949‡|
|Physical limitations||94.40 (14.40)||96.87 (8.33)||0.431‡|
|Social limitations||97.89 (7.01)||98.61 (5.42)||0.562‡|
|Personal relationships||95.31 (16.41)||92.90 (19.32)||0.536‡|
|Emotions||90.72 (19.69)||86.98 (22.79)||0.130‡|
|Sleep/Energy||80.03 (20.76)||78.64 (19.35)||0.442‡|
|Severity measures||92.38 (15.72)||92.71 (15.10)||0.653‡|
The median time to continence was 46 days for patients after ORP and 16 days after RARP (P = 0.026). The median amount of first day incontinence was 186 g and 44 g for the ORP and RARP group, respectively (P < 0.05).
Figure 1 shows the Kaplan–Meier survival analysis for the time to urinary continence according to type of surgery. Patients achieved continence significantly faster after RARP than after ORP (P = 0.007). After correction for the different patient characteristics (age, D'Amico risk group, nerve-sparing status, surgical margin status, preoperative urine leak, body mass index) the difference remained significant (P = 0.036, hazard ratio [HR] 1.522 [1.027–2.255]) (Table 2). Additionally, age (P = 0.029, HR 0.666 (0.454–0.958) and preoperative UI (P = 0.004, HR 1.694 (1.179–2.433)) were significant contributing factors, indicating that younger men and men without preoperative incontinence achieved continence sooner; however, nerve-sparing status and D' Amico risk group significantly differed between ORP and RARP (Table 1). Clearly, a comparison of the time to continence between ORP and RARP was only meaningful in the intermediate-risk group and/or bilateral nerve-sparing group, because only these groups had a sufficient number of patients. These subgroup analyses still yielded a faster return of continence after RARP (HR ≥ 1.2), but the effect was decreased in size and was nonsignificant (Table 3).
Table 2. Cox regression for the time to urinary continence according to type of surgery
| ||df||Univariate analysis||Multivariate analysis|
|P||HR||95% CI||P||HR||95% CI|
|RARP vs ORP||1||0.007||1.548||1.123||2.132||0.036||1.522||1.027||2.255|
|Age: ≥65 vs <65 years||1||0.002||0.588||0.420||0.825||0.029||0.666||0.454||0.958|
|Risk group||2||0.300|| || || ||0.755|| || || |
|Intermediate vs low||1||0.380||0.810||0.506||1.297||0.966||0.989||0.602||1.627|
|High vs low||1||0.134||0.686||0.420||1.122||0.615||1.159||0.653||2.056|
|Nerve-sparing status||2||0.055|| || || ||0.572|| || || |
|Unilateral vs non-nerve- sparing||1||0.372||1.289||0.738||2.251||0.732||1.110||0.610||2.021|
|Bilateral vs non-nerve-sparing||1||0.034||1.719||1.041||2.837||0.343||1.364||0.718||2.589|
|Preoperative urine leak: incontinent vs continent||1||0.012||1.554||1.103||2.191||0.004||1.694||1.179||2.433|
|BMI||2||0.705|| || || ||0.574|| || || |
|25.1–30.0 vs ≤25.0 kg/m2||1||0.956||1.010||0.712||1.431||0.796||0.953||0.663||1.371|
|>30.0 vs ≤25.0 kg/m2||1||0.475||0.837||0.513||1.365||0.304||0.763||0.456||1.278|
Table 3. Cox regression for the time to urinary continence according to type of surgery (subanalyses)
|ORP vs RARP||N||P||HR||CI (95%)|
|No correction for confounders||173||0.007||1.548||1.123||2.132|
|Correction for all confounders (age, D'Amico risk group, nerve-sparing status, preoperative urine leak)||165||0.036||1.522||1.027||2.255|
|Subgroup bilateral nerve-sparing||110||0.191||1.292||0.880||1.899|
|Subgroup intermediate-risk group||89||0.287||1.264||0.821||1.944|
|Subgroup bilateral nerve-sparing and intermediate-risk group||71||0.516||1.173||0.724||1.900|
Four patients after ORP and two after RARP were not continent 12 months after surgery (Table 4). Compared with patients in the ORP group, significantly more patients in the RARP group were continent at 1 month after surgery. Additionally, patients who underwent RARP had significantly less urine leak at 1 month after surgery
Table 4. Cumulative incidence of continence and mean urine leak at 1, 3, 6 and 12 months after ORP and RARP (24-h pad test)
|Time since catheter removal||Continent patients, n (%)||Mean urine leak, g|
|1 month||46 (42)||42 (69)||0.010||108||50||0.036|
|3 months||85 (78)||53 (87)||0.162||16||14||0.816|
|6 months||102 (94)||58 (95)||0.540||10||4||0.492|
|12 months||105 (96)||59 (97)||0.896||2||4||0.454|
The point prevalence of continence, defined as 0 g on the 1-h pad test, and the VAS scores only differed significantly at 1 month after surgery (Table 5). Furthermore, the RARP group had significantly fewer voiding symptoms than the ORP group at 1 (P = 0.013) and 3 (P = 0.038) months after surgery. At 1 month after surgery, the RARP group scored better in all aspects of the KHQ, than the ORP group. Furthermore, the scores for ‘sleep/energy’ and ‘severity measures’ were significantly better after RARP at 3 months postoperatively. At 12 months, patients in the ORP group were more physically limited and took more precautions to avoid urine leak than those in the RARP group (P = 0.014 and P = 0.011, respectively).
Table 5. Comparison of the point prevalence of continence at 1, 3, 6 and 12 months after RP, according to type of surgery (RARP vs. ORP)
|Secondary outcome variables||ORP||RARP||Odds ratio (95% CI)||P*|
|n (%)||n (%)|
|Point prevalence of continence, defined as VAS score ≤1|| || || || |
|At 1 month (115/62)||36 (31.3)||37 (59.7)||3.248 (1.708–6.175)||0.000|
|At 3 months (113/62)||70 (61.9)||46 (74.2)||1.766 (0.891–3.500)||0.132|
|At 6 months (113/62)||87 (77.0)||51 (83.6)||1.524 (0.680–3.416)||0.334|
|At 12 months (109/59)||84 (77.1)||50 (84.7)||1.653 (0.715–3.824)||0.315|
|Point prevalence of continence, defined as 0 g on 1-h pad test|| || || || |
|At 1 month (110/61)||47 (42.7)||36 (59.0)||1.930 (1.023–3.642)||0.055|
|At 3 months (110/62)||75 (68.2)||49 (79.0)||1.759 (0.847–3.655)||0.157|
|At 6 months (109/62)||93 (85.3)||56 (90.3)||1.606 (0.594–4.344)||0.477|
|At 12 months (105/59)||88 (83.8)||53 (89.8)||1.706 (0.633–4.598)||0.353|
Fifteen patients (12.9%) in the ORP group and eight patients (12.5%) in the RARP group received additional radiotherapy. Radiotherapy was always started after continence was achieved, except in three patients. In these patients, mean urine leak was 7, 8 and 64 g per 24 h, respectively, at the start of additional radiotherapy. A total of 30% of patients had some preoperative urine leak (range 1–10 g/day). These patients achieved continence significantly slower than those who were preoperatively continent (P = 0.01). Of the six patients who remained incontinent 1 year after surgery, three had minimal urine leak preoperatively (range 1–3 g).
Five patients underwent a conversion from RARP to ORP. The mean (sd; median) duration of incontinence was 17.6 (20.08; 9.0) days, the mean (sd; median) incontinence on the first day after catheter withdrawal was 270.6 ( 520.46; 51.0) g, the catheter remained for a mean (sd) of 11.6 (0.89) days in situ, and the mean (sd) patient age was 59.85 (2.83) years.
Data were analysed according to the intention-to-treat principle. The positive surgical margin rate did not significantly differ between the ORP and RARP groups, although raw data differ greatly: 21 vs 30% positive/doubtful surgical margins for the ORP vs RARP group, respectively (Table 2). There was no difference between the groups in apical surgical margin rate.
- Top of page
- Patients and Methods
- Conflict of Interest
Patients in the RARP group regained urinary continence sooner than those in the ORP group (24-h pad test). The median time to continence and the median amount of UI on the first day after catheter removal were significantly less after RARP. This was further confirmed by the significant difference in the 1-h pad test and VAS score at 1 month postoperatively in favour of RARP. After correction for the different patient characteristics (Table 1) the statistical evidence in favour of RARP remained (P = 0.036, HR 1.522 [1.027–2.255]). Comparison of time to continence between ORP and RARP with a sufficient number of patients (intermediate risk and/or bilateral nerve-sparing) still yielded a faster return of continence after RARP, but the effect decreased in size and lacked statistical significance.
Additionally, voiding symptoms severity and quality of life were significantly better after RARP.
Only six patients, four in the ORP and two in the RARP group, still had involuntary urine leak after 1 year (range 6–167 g/day).
Our study had several strengths. This was a prospective study in which we studied the evolution of UI, voiding symptoms and quality of life. All of our patients were followed up for 12 months and evaluated at regular time intervals. All patients measured their urine leak daily during 24 h by weighing their pads accurately to 1 g. Patients performed a 1-h pad test and a VAS score concerning the subjective feeling of incontinence at regular time intervals (1, 3, 6 and 12 months postoperatively). Three experienced surgeons performed the ORP or RARP, using highly standardized surgical procedures and voiding symptoms and quality of life were assessed at 1, 3, 6 and 12 months after surgery.
A limitation of the study was that the number of patients in each surgical group was not equal.
In the literature, the continence rate immediately after catheter removal is reported to be 10–41% after ORP [3, 8] and 13.1–68.9% after RARP [7, 8, 13, 21]. Continence rates increase to 63–83% after ORP [9, 10, 22, 23] and 70–95% after RARP at 3 months [9, 10, 22], and 6 months after surgery the continence rates are 83–89.6% (ORP) [9, 23, 24] and 75–95% (RARP) [9, 24]. At 12 months after surgery 80–94% of patients who have undergone ORP and 89–97% who have undergone RARP have regained continence [8-11, 13, 23]. The present study achieved even better results, with 96% continence after ORP and 97% after RARP. Only six patients remained incontinent 1 year after surgery. Four patients underwent a male sling procedure and were continent afterwards. Two patients refused incontinence surgery: one patient (with 130 g urine leak) because of several other comorbidities and the other patient because of minimal urine leak (6 g). The catheterization time was fairly long for both groups; all patients at our institution go home after RP with the catheter in situ on around day 6 and return to the hospital around day 12, when the catheter is then removed without cystogram.
By contrast to several other studies [7-9], we used objective and subjective measures to evaluate urine leak. In our study, patients performed a 24-h pad test preoperatively (for 3 days) and postoperatively daily until continence was achieved. Furthermore all patients performed a 1-h pad test and filled in a VAS score at fixed timepoints. For the VAS, a score of 0 or 1 was interpreted as continent. Many patients scored themselves as 1 on the VAS, although they were completely dry on the 1h- and the 24-h pad tests. Usually, this was for safety reasons or because of post-micturition dribble of <1 g/ day.
A comparison of ORP with RARP showed that the median time to continence decreased from 160 days to 44 days with RARP , but at 6 months no difference in continence rates could be found in two other matched comparison series [11, 22] In the present study, the median time to continence was 46 days after ORP and 16 days after RARP. Similarly, Krambeck et al.  and Ahlering et al. , reported no differences in continence rates at 6 months after surgery, with rates of 94% (ORP) and 95% (RARP).
According to Namiki et al. , voiding symptoms, measured using the IPSS, significantly improved after ORP. Similarly Sammon et al.  and Menon et al.  indicated that voiding symptoms also improved after RARP, but no level of significance was indicated. In the present study, the IPSS improved in both groups, but the RARP group performed significantly better in the short term (1 and 3 months after surgery) than the ORP group.
Comparing different studies with regard to quality of life is difficult, because different questionnaires were used among studies. Miller et al.  compared quality of life during the first 6 weeks after ORP and RARP in a small cohort of 162 patients using the 12-item Short-Form questionnaire. They found better physical scores and a faster return to baseline quality of life in patients after RARP. Similarly to Miller et al, we found a faster improvement in quality of life in patients who underwent RARP. By contrast, Malcolm et al.  could not find any pronounced advantages to RARP vs ORP from the standpoint of quality of life outcomes, using the UCLA-Prostate Cancer Index questionnaire. In our study, the KHQ was used, a questionnaire designed only to assess the impact of UI on quality of life. Furthermore, Hara et al.  used the European Organisation for the Research and Treatment of Cancer prostate cancer questionnaire for a subanalysis comparing patients before and after ORP and found that quality of life associated with voiding dysfunction was impaired before but significantly improved after ORP. By contrast, the quality of life associated with difficulty of UI was significantly worsened by surgery. In our trial, patients in the RARP group had significantly fewer voiding symptoms than those in the ORP group at 1 and 3 months after surgery. Additionally, patients in the ORP group took more precautions to avoid urine leak at 12 months postoperatively. The differences between the surgical approaches were greatest in the short term. This was to be expected because patients progress the most in the early postoperative period.
The literature is not unanimous concerning the predictive value of nerve-sparing status in time to continence [25-28]. In the present study, the difference in continence outcomes between ORP and RARP was partially attributable to the different degrees of nerve-sparing performed, but only in the univariate analysis (Table 2). According to Ferronha et al. , no substantial differences could be found between the two surgical techniques regarding positive margins. In our study the positive surgical margin rate did not differ significantly between the ORP and RARP groups, although raw data would suggest otherwise. Furthermore we found no difference in apical surgical margin rate, indicating that patients in the RARP group were not more continent because part of the apex was not removed and the urethral sphincter was less damaged.
In conclusion, in this prospective trial patients who underwent RARP tended to regain urinary continence sooner than those who underwent ORP (24-h pad test) but, in subgroup analyses, statistical significance disappeared and effect size decreased dramatically, indicating that results must be interpreted with caution. Nevertheless, the analysis also indicated significant better scores for voiding symptom severity and quality of life after RARP.