Posterior musculofascial reconstruction in robotic‐assisted laparoscopic prostatectomy for the treatment of clinically localized prostate cancer

Abstract

Background

Delayed recovery of urinary continence is a major adverse effect of robotic‐assisted laparoscopic prostatectomy (RALP) in men undergoing prostate cancer treatment. To address this issue, a number of surgical techniques have been designed to reconstruct the posterior aspect of the rhabdosphincter, which is responsible for urinary continence after removal of the prostate; however, it is unclear how well they work. 

Objectives

To assess the effects of posterior musculofascial reconstruction RALP compared to no posterior reconstruction during RALP for the treatment of clinically localized prostate cancer.

Search methods

We performed a comprehensive search of the Cochrane Library, MEDLINE, Embase, three other databases, trials registries, other sources of the grey literature, and conference proceedings, up to 12 March 2021. We applied no restrictions on publication language or status.

Selection criteria

We included randomized controlled trials (RCTs) in which participants were randomized to undergo variations of posterior musculofascial reconstruction RALP versus no posterior reconstruction during RALP for clinically localized prostate cancer.

Data collection and analysis

Two review authors independently classified studies and abstracted data from the included studies. Primary outcomes were: urinary continence recovery within one week after catheter removal, at three months after surgery, and serious adverse events. Secondary outcomes were: urinary continence recovery at six and twelve months after surgery, potency recovery twelve months after surgery, positive surgical margins (PSM), and biochemical recurrence‐free survival (BCRFS). We performed statistical analyses using a random‐effects model. We rated the certainty of evidence (CoE) according to the GRADE approach.

Main results

Our search identified 13 records of eight unique RCTs, of which six were published studies and two were abstract proceedings. We included 1085 randomized participants, of whom 963 completed the trials (88.8%). All participants had either cT1c or cT2 or cT3a disease, with a mean prostate‐specific antigen level of 8.15 ng/mL.

Primary outcomes

Posterior reconstruction RALP (PR‐RALP) may improve urinary continence one week after catheter removal compared to no posterior reconstruction during RALP (risk ratio (RR) 1.25, 95% confidence interval (CI) 0.90 to 1.73; I² = 42%; studies = 5, participants = 498; low CoE) although the CI also includes the possibility of no effect. Assuming 335 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 84 more men per 1000 (33 fewer to 244 more) reporting urinary continence recovery. 

Posterior reconstruction may have little to no effect on urinary continence three months after surgery compared to no posterior reconstruction during RALP (RR 0.98, 95% CI 0.84 to 1.14; I² = 67%; studies = 6, participants = 842; low CoE). Assuming 701 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 14 fewer men per 1000 (112 fewer to 98 more) reporting urinary continence after three months.

PR‐RALP probably results in little to no difference in serious adverse events compared to no posterior reconstruction during RALP (RR 0.75, 95% CI 0.29 to 1.92; I² = 0%; studies = 6, participants = 835; moderate CoE). Assuming 25 per 1000 men undergoing standard RALP experience a serious adverse event at this time point, this corresponds to six fewer men per 1000 (17 fewer to 23 more) reporting serious adverse events. 

Secondary outcomes

PR‐RALP may result in little to no difference in recovery of continence 12 months after surgery compared to no posterior reconstruction during RALP (RR 1.02, 95% CI 0.98 to 1.07; I² = 25%; studies = 3, participants = 602; low CoE). Assuming 918 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 18 more men per 1000 (18 fewer to 64 more) reporting urinary continence recovery. 

We are very uncertain about the effects of PR‐RALP on recovery of potency 12 months after surgery compared to no posterior reconstruction during RALP (RR 1.02, 95% CI 0.82 to 1.26; I² = 3%; studies = 2, participants = 308; very low CoE). Assuming 433 per 1000 men undergoing standard RALP are potent at this time point, this corresponds to nine more men per 1000 (78 fewer to 113 more) reporting potency recovery. 

PR‐RALP may result in little to no difference in positive surgical margins compared to no posterior reconstruction during RALP (RR 1.24, 95% CI 0.65 to 2.33; I² = 50%; studies = 3, participants = 517; low CoE). Assuming 130 per 1000 men undergoing standard RALP have a positive surgical margin, this corresponds to 31 more men per 1000 (46 fewer to 173 more) reporting positive surgical margins. 

PR‐RALP may result in little to no difference in biochemical recurrence compared to no posterior reconstruction during RALP (RR 1.36, 95% CI 0.74 to 2.52; I² = 0%; studies = 2, participants = 468; low CoE). Assuming 70 per 1000 men undergoing standard RALP have experienced biochemical recurrence at this time point, this corresponds to 25 more men per 1000 (18 fewer to 107 more) reporting biochemical recurrence. 

Authors' conclusions

This review found evidence that PR‐RALP may improve early continence one week after catheter removal but not thereafter. Meanwhile, adverse event rates are probably not impacted and surgical margins rates are likely similar. This review was unable to determine if or how these findings may be impacted by the person's age, nerve‐sparing status, or clinical stage. Study limitations, imprecision, and inconsistency lowered the certainty of evidence for the outcomes assessed. 

Plain language summary

Should we perform posterior reconstruction RALP or standard RALP for clinically localized prostate cancer?

Review question

In men with prostate cancer who are having their prostate removed using surgery assisted by a robotic device (called robotic‐assisted laparoscopic prostatectomy, or RALP), how does connecting the tissue behind the urethra (so‐called posterior reconstruction) compare to surgery where these connections are not made (standard RALP)?

Background

Urologists often use a robot to remove the prostate in men with prostate cancer. After surgery, most men leak urine for some time. This problem is called incontinence and usually improves six to 12 months after surgery in most men. However, it can be very bothersome during this time.

Study characteristics

We included eight studies in which chance determined whether men had posterior reconstruction RALP or standard RALP. These studies included 1085 men with an average age ranging from 60 to 67 years. The average prostate‐specific antigen (PSA) level in the men was 8.15 ng/mL. Higher levels of PSA may indicate worse prostate cancer.

Key results

We found that posterior reconstruction RALP may result in better continence one week after the catheter comes out compared to standard RALP (although it is also possible that it is no better), but it may make little to no difference at either three or 12 months after surgery. Posterior reconstruction RALP probably results in little to no difference in serious unwanted effects compared with the standard way of doing the surgery. There may also be little to no difference in positive surgical margins, meaning the risk of there being cancer cells right at the cut edge of the prostate when viewed under the microscope. There may also be no difference between the two techniques in terms of the risk of a PSA level that goes up within 12 months of surgery, which often signals that there is cancer left behind. We are very uncertain how posterior reconstruction RALP effects the ability to achieve an erection, compared to standard RALP.

Certainty of the evidence

The certainty of the evidence ranged from moderate to very low depending on the outcome, meaning that we have moderate to very little confidence in the results.

Summary of findings

Summary of findings 1. Posterior reconstruction compared to no posterior reconstruction for the treatment of prostate cancer.

Posterior reconstruction compared to no posterior reconstruction for the treatment of prostate cancer
Patient or population: men (ages > 18 years) with clinically localized prostate cancer Setting: inpatient setting Intervention: posterior reconstruction RALP Comparison: no posterior reconstruction RALP
Outcomes	№ of participants (studies)	Certainty of the evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects* (95% CI)		What happens
				Risk with no posterior reconstruction	Risk difference with posterior reconstruction
Urinary continence within 1 week after catheter removal   MCID: 5% absolute difference	498 (5 RCTs)	⊕⊕⊝⊝ LOWa, b, c	RR 1.25 (0.90 to 1.73)	Study population		Posterior reconstruction may improve urinary continence one week after catheter removal (although the CI also includes the possibility of no effect).
				335 per 1000	84 more per 1000 (33 fewer to 244 more)
Urinary continence 3 months after surgery   MCID: 5% absolute difference	842 (6 RCTs)	⊕⊕⊝⊝ LOWa, b, c	RR 0.98 (0.84 to 1.14)	Study population		Posterior reconstruction may result in little to no difference in urinary continence 3 months after surgery.
				701 per 1000	14 fewer per 1000 (112 fewer to 98 more)
Serious adverse events   Follow‐up: 12 months MCID: 2% absolute difference	835 (5 RCTs)	⊕⊕⊕⊝ MODERATE a	RR 0.75 (0.29 to 1.92)	Study population		Posterior reconstruction probably results in little to no effect on serious adverse events.
				25 per 1000	6 fewer per 1000 (17 fewer to 23 more)
Urinary continence 6 months after surgery   MCID: 5% absolute difference	741 (5 RCTs)	⊕⊕⊝⊝ MODERATEa	RR 1.01 (0.97 to 1.05)	Study population		Posterior reconstruction probably results in little to no difference in urinary continence 6 months after surgery.
				917 per 1000	16 more per 1000 (26 fewer to 44 more)
Urinary continence 12 months after surgery   MCID: 5% absolute difference	602 (3 RCTs)	⊕⊕⊝⊝ LOWa, d	RR 1.02 (0.98 to 1.07)	918 per 1000	18 more per 1000 (18 fewer to 64 more)	Posterior reconstruction may result in little to no difference in urinary continence 12 months after surgery.
Potency recovery 12 months after surgery   MCID: 5% absolute difference	308 (2 RCTs)	⊕⊝⊝⊝ VERY LOWa, e	RR 1.02 (0.82 to 1.26)	Study population		We are very uncertain how posterior reconstruction affects potency recovery 12 months after surgery.
				433 per 1000	9 more per 1000 (78 fewer to 113 more)
Positive surgical margins   MCID: 5% absolute difference	517 (3 RCTs)	⊕⊕⊝⊝ LOWa, b, c	RR 1.24 (0.65 to 2.33)	Study population		Posterior reconstruction may result in little to no difference in positive surgical margins.
				130 per 1000	31 more per 1000 (46 fewer to 173 more)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; MCID: minimal clinically important difference; RALP: robotic‐assisted laparoscopic prostatectomy; RCT: randomized controlled trial; RR: Risk ratio.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate. The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited. The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate. The true effect is likely to be substantially different from the estimate of effect.

^a Downgraded by one level for study limitations: high or unclear risk of selection bias (allocation concealment), performance, and reporting bias.
^b Downgraded by one level for inconsistency due to substantial heterogeneity.
^c Not rated down for imprecision (further), since wide CI appears mainly secondary to inconsistency.
^d Downgraded by one level for imprecision: confidence interval crosses assumed threshold of clinically important difference.
^e Downgraded by two levels for imprecision: very wide confidence interval crosses assumed threshold of clinically important difference.

Background

Description of the condition

The most common solid organ malignancy in men is prostate cancer (Siegel 2019). In 2018, the incidence of prostate cancer was 13.5% and the mortality rate 6.7%, with varying prevalence among different racial groups (Bray 2018; Rawla 2019). Risk factors for prostate cancer include having a family history of prostate cancer, being advanced in age, and being of African descent (Rawla 2019). Radical prostatectomy is a well‐established treatment for clinically localized prostate cancer, with trial evidence showing improved long‐term oncological outcomes compared to watchful waiting (Bill‐Axelson 2018; Wilt 2012; Wilt 2017; Vernooij 2020; Wilt 2020; Wilt 2021; Wilt 2021a). In recent years, robotic‐assisted laparoscopic prostatectomy (RALP) has become the main approach, especially in the USA (Menon 2018; Sayyid 2017) although there is little high quality evidence to support its superiority when it comes to patient‐important outcomes (Ilic 2017; Ilic 2018).

Description of the intervention

Prostatectomy outcomes are influenced by the surgical skill, technique, and experience of the surgeon (Vickers 2007). The two major adverse effects following a radical prostatectomy are erectile dysfunction and temporary or persistent urinary incontinence (Grasso 2016). Urinary continence is influenced by the muscular component, sphincter support structures, and supportive ligaments of the bladder and urethra towards the pelvis and anterior abdominal wall (Arroyo 2019). Numerous technical modifications to restore the normal anatomy have been implemented in efforts to retain continence after standard RALP (Asimakopoulos 2019; Checcucci 2019b; Lim 2014). One method involves reconstructing the posterior musculofascial plate (Rocco 2001).

The musculofascial plate is a suspension system which comprises Denonvilliers' fascia, the rhabdosphincter, median fibrous raphe, the posterior fascia of the prostate, and the central tendon of the perineum (Rocco 2012). The rhabdosphincter comprises circular striated muscle fibers surrounding the urethra from the membranous aspect to the prostatic apex. The anterolateral walls of the rhabdosphincter contract along the more rigid posterior wall, acting as a fulcrum for the muscular action (Rocco 2012). The purpose of reconstructing the posterior musculofascial plate is to minimize urethrosphincteric sliding after prostatectomy, support the vesicourethral anastomosis, approximate the bladder neck to the urethral stump, and provide a fulcrum for contraction of the rhabdosphincter (Arroyo 2019).

Reconstruction of the posterior aspect of the rhabdosphincter was first described by Rocco and colleagues in 2001 for open surgery, and in 2006 a study suggested that it might shorten time to continence after radical prostatectomy (Rocco 2001; Rocco 2006). In 2007, this type of reconstruction was described in transperitoneal laparoscopic radical prostatectomy (Rocco 2007a; Rocco 2007b). Subsequently, Coelho and colleagues modified the reconstruction for RALP, and reported no increase in operative time or increase in potential harms to the participant (Coelho 2011; Coelho 2018). Since its worldwide spread, multiple different modifications of posterior musculofascial reconstruction (PMR) have evolved, but there is disagreement in the existing literature as to whether this improves continence when compared to no posterior reconstruction during RALP (Gautam 2010). Many different reconstructive techniques are well described in the existing literature and illustrated surgical videos (Vis 2019). Differing continence rates have been reported in trials; this may be due to variations of the original technique, varying degrees of surgical experience, and varying definitions of assessments of urinary continence (Coelho 2011; Coelho 2018; Rocco 2012). Additionally, there is a distinction between surgical techniques to preserve natural urinary continence mechanisms and techniques to reconstruct pelvic anatomy. Some argue that preserving natural continence mechanisms should remain the mainstay to improve urinary continence after RALP (Vora 2013). A prior, closely related Cochrane review has assessed the Retzius‐sparing approach as a way to improve continence outcomes (Rosenberg 2020b) after RALP.

Adverse effects of the intervention

Theoretically, increasing the number of sutures in the area with PMR may increase urethral stenosis and acute urinary retention, requiring secondary interventions (Grasso 2016). Other adverse effects of PMR are similar to those associated with standard RALP. PMR provides support for the new vesicourethral anastomosis, but persistent urinary incontinence can still occur. Additional adverse effects of the intervention include blood loss, need for transfusions, erectile dysfunction, and intraoperative injury to adjacent structures. 

How the intervention might work

The original technique described by Rocco and colleagues in 2001 for open prostatectomy used interrupted sutures on both sides of the midline to approximate the Denonvilliers' fascia to the posterior aspect of the rhabdosphincter and posterior median raphe (Gautam 2010; Rocco 2001; Rocco 2007a). The technique has since been adapted for RALP by placing continuous sutures, which is faster and technically easier. Multiple different modifications of this technique have since evolved (Gautam 2010). In general, the free edge of the Denonvilliers' fascia after the prostatectomy is identified and approximated to the posterior aspect of the rhabdosphincter and posterior median raphe using a continuous suture. A second layer of reconstruction can be performed, approximating the posterior lip of the bladder neck and the rectourethralis muscle to the posterior urethral edge and to the previously reconstructed median raphe (Coelho 2018). We recognize that there are many different nuances in posterior musculofascial reconstruction techniques, along with additional reconstruction techniques such as anterior reconstruction (Checcucci 2019b). We have attempted to account for this in our review by several preplanned subgroup analyses.

Why it is important to do this review

Many surgeons around the world are performing posterior musculofascial reconstruction after RALP. It is thought to be easily reproducible and quickly executed, with no increased risk of perioperative complication. However, varying reports of urinary continence and peri‐anastomotic urinary leakages have been reported, which brings into question the value of this technique (Grasso 2016; Coelho 2018).

A number of randomized and non‐randomized studies have compared posterior reconstruction RALP to standard RALP, and their results have been summarized in two systematic reviews (Grasso 2016; Rocco 2012). However, neither applied the same methodological rigor as in a Cochrane Review, which includes an a priori published protocol, an exhaustive search of the published and unpublished literature (irrespective of language of publication), a focus on patient‐important outcomes and clinically meaningful differences, and the use of the GRADE approach to assess the certainty of evidence for each outcome (Guyatt 2011 [Guyatt 2011]). Our Cochrane Review will therefore address an important gap in the knowledge on the effectiveness of PMR to guide clinical practice and future research. This brings the potential to change the practice of urologic surgery to minimize the common adverse events experienced after undergoing RALP. The findings of this review will help promote the further refinement and dissemination of a standardized approach to RALP. 

Objectives

To assess the effects of posterior musculofascial reconstruction versus no posterior reconstruction during robotic‐assisted laparoscopic prostatectomy for the treatment of clinically localized prostate cancer.

Methods

Criteria for considering studies for this review

Types of studies

We included only parallel group randomized trials. We excluded cross‐over and cluster‐randomized trials as they are not relevant to this comparison. We did not consider pseudo‐randomized controlled trials or observational studies, given their increased risk of selection bias. We included studies regardless of their publication status or language of publication.

Types of participants

We included studies of men (aged 18 years or older) with clinically localized prostate cancer (clinical stage T1 to T2, N0, M0), who planned to undergo RALP.

We planned to exclude studies of men with pre‐existing urinary incontinence. Should we have identified studies in which only a subset of participants was relevant to this review, we would only have included them if they reported data separately for the relevant subset.

Types of interventions

We investigated the following comparisons of experimental intervention versus comparator intervention. We included concomitant interventions providing they were the same in the experimental and comparator groups.

Experimental intervention

• Posterior musculofascial reconstruction during RALP

Comparator intervention

• No posterior reconstruction during RALP

Comparison

• Posterior musculofascial reconstruction during RALP versus no posterior reconstruction during RALP

Types of outcome measures

We included studies regardless of whether they measured the outcomes to be assessed in this review.

Primary outcomes

• Urinary continence within one week after catheter removal (dichotomous outcome)

• Urinary continence three months after surgery (dichotomous outcome)

• Serious adverse events (dichotomous outcome)

Secondary outcomes

• Urinary continence six months after surgery (dichotomous outcome)

• Urinary continence 12 months after surgery (dichotomous outcome)

• Potency 12 months after surgery (dichotomous outcome)

• Positive surgical margins (dichotomous outcome)

• Biochemical recurrence 12 months after surgery (dichotomous outcome)

• Urinary function scores on quality of life scale (continuous outcome)

• Sexual function scores on quality of life scale (continuous outcome)

Method and timing of outcome measurement

• Urinary continence

  • Self‐reported absence of leakage or use of zero or one pads/day.

  • We assessed this outcome up to 12 months after surgery.

  • We considered a 5% absolute difference in continence rates as clinically important.

• Serious adverse events

  • Measured as Dindo‐Clavien system grade III, IV or V (Dindo 2004).

  • We assessed this outcome up to 12 months after surgery.

  • We considered a 2% absolute difference in serious adverse event rates as clinically important.

• Potency

  • Number or percentage of participants achieving potency according to validated potency scales, such as the International Index of Erectile Function (IIEF) and IIEF‐5 scores (Rosen 1997; Rosen 2011). We define achieving potency as an IIEF‐EF score of 19 or greater (mild erectile dysfunction) and IIEF‐5 score of 17 or greater (no erectile dysfunction).

  • We assessed this outcome at 12 months after surgery.

  • We considered a 5% absolute difference in potency rates as clinically important.

• Positive surgical margins

  • Positive when cancer cells are found at the ink‐marked resection margin.

  • We assessed this outcome following surgery, on the basis of the prostatectomy specimen.

  • We considered a 5% absolute difference in positive surgical margins as clinically important.

• Biochemical recurrence‐free survival

  • Defined as any prostate‐specific antigen values greater or equal to 0.2 ng/mL.

  • In the published protocol we had defined this as a time‐to‐event outcome (Rosenberg 2020). However, since no study reported this outcome in this form, we analyzed the best available evidence in the form of risk ratios (see also Differences between protocol and review).

  • We assessed this outcome up to 60 months (five years) after surgery.

  • We considered a 2% absolute difference in biochemical recurrence rates at 60 months as clinically important.

• Urinary function scores on quality of life scale

  • Final value or change assessed with validated questionnaires such as the urinary domain of the Expanded Prostate Cancer Index Composite questionnaire (Wei 2000).

  • We assessed this outcome up to 12 months after surgery.

  • We considered a 5% absolute difference in quality of life as clinically important.

• Sexual function scores on quality of life scale

  • Final value or change assessed with validated questionnaires, such as the sexual domain of the Expanded Prostate Cancer Index Composite questionnaire (Wei 2000).

  • We assessed this outcome up to 12 months after surgery.

  • We considered a 5% absolute difference in quality of life as clinically important.

For all outcomes except potency, for which we found an established minimally clinically important difference (MCID) reported in the literature, all proposed thresholds were based on the clinical experience of the review authors.

Search methods for identification of studies

We performed a comprehensive search with no restrictions on the language of publication or publication status. We repeated searches within three months prior to anticipated publication of the review.

Electronic searches

We searched the following sources, from the inception of each database to 12 March 2021. Our search strategies are detailed in Appendix 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 9.

• Cochrane Library via Wiley

  • NHS Economic Evaluation Database (NHSEED)

  • Database of Abstracts of Reviews of Effects (DARE)

  • HealthTechnology Assessment database (HTA)

• MEDLINE via PubMed (from 1946)

• MEDLINE via Ovid (from 1946)

• Embase via Ovid (from 1947)

• Web of Science Core Collection

• Scopus

• Global Index Medicus

We also searched the following resources.

• ClinicalTrials.gov (www.clinicaltrials.gov/)

• World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) search portal (apps.who.int/trialsearch/)

If we detected additional relevant keywords during any of the electronic or other searches, we modified the electronic search strategies to incorporate these terms and documented the changes.

Searching other resources

We tried to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, reviews, meta‐analyses, and health technology assessment reports. We also contacted study authors of included trials to identify any further studies that we may have missed. We contacted drug/device manufacturers for ongoing or unpublished trials.

We performed no handsearching for abstract proceedings, as all relevant meetings such as the American Urological Association, European Association of Urology, and Society of Urologic Oncology of the last three years (2018 to 2020) are published electronically and therefore were captured in our electronic search.

Data collection and analysis

Selection of studies

We used the reference management software Endnote to identify and remove potential duplicate records. Two review authors (JR, HL, or SL) independently scanned the abstract or title (or both) of remaining records retrieved, to determine which studies should be assessed further. Two review authors (JR, HL, or SL) investigated all potentially relevant records as full text; they mapped records to studies and classified studies as included, excluded, awaiting classification, or ongoing, in accordance with the criteria for each provided in the most recent Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). We resolved any discrepancies through consensus or recourse to a third review author (JHJ or PD). If resolution of a disagreement was not possible, we designated the study as 'awaiting classification' and contacted the study authors for clarification. We documented reasons for exclusion of studies that may have reasonably been expected to be included in the review in the Characteristics of excluded studies table. We present a PRISMA flow diagram in Figure 1, showing the process of study selection (Liberati 2009).

1.

Data extraction and management

We developed a dedicated data abstraction form that we pilot tested ahead of time. For studies that fulfilled our inclusion criteria, two review authors (JR, HL, or SL) independently abstracted the following information, which we provide in the Characteristics of included studies table.

• Study design

• Study dates (if dates are not available then this is noted as such)

• Study settings and country

• Participant inclusion and exclusion criteria (including tumor stage, PSA values, magnetic resonance imaging (MRI) findings)

• Participant details, baseline demographics (age, comorbidities, body mass index)

• Surgeons' characteristics (surgical experience)

• The number of participants by study and by study arm

• Details of relevant posterior musculofascial reconstruction RALP and no posterior reconstruction RALP interventions (as applicable), e.g. type and extent of nerve‐sparing technique, bladder neck reconstruction, etc.

• Definitions of relevant outcomes, method and timing of outcome measurement, any relevant subgroups

• Study funding sources

• Declarations of interest by primary investigators

We extracted outcome data relevant to this Cochrane Review as needed for calculation of summary statistics and measures of variance. For dichotomous outcomes, we attempted to obtain numbers of events and totals in order to populate a 2 × 2 table, as well as summary statistics with corresponding measures of variance. For continuous outcomes, we attempted to obtain means and standard deviations or data necessary to calculate this information. For time‐to‐event outcomes, we attempted to obtain hazard ratios (HRs) with corresponding measures of variance or data necessary to calculate this information.

We resolved any disagreements by discussion, or by consultation with a third review author (JHJ or PD), if required. We provided information, including trial identifier, about potentially relevant ongoing studies in the Characteristics of ongoing studies. We attempted to contact authors of included studies to obtain key missing data as needed.

Dealing with duplicate and companion publications

In the event of duplicate publications, companion documents or multiple reports of a primary study, we maximized yield of information by mapping all publications to unique studies and collating all available data. We used the most complete dataset, aggregated across all known publications. In case of doubt, we gave priority to the publication reporting the longest follow‐up associated with our primary or secondary outcomes.

Assessment of risk of bias in included studies

Two review authors (JR, ZE, HL, or SL) independently assessed the risk of bias of each included study. We resolved disagreements by consensus, or by consultation with a third review author (JHJ or PD).

We assessed risk of bias using Cochrane's risk of bias assessment tool (Higgins 2017). We assessed the following domains.

• Random sequence generation (selection bias)

• Allocation concealment (selection bias)

• Blinding of participants and personnel (performance bias)

• Blinding of outcome assessment (detection bias)

• Incomplete outcome data (attrition bias)

• Selective reporting (reporting bias)

• Other sources of bias

For each study, we judged the risk of bias for each domain as low, high, or unclear, using the guidance described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). We presented risk of bias summary figures to illustrate these findings (Figure 2; Figure 3).

2.

Footnote: Empty cells indicate studies that did not report the outcome of interest. 

3.

Footnote: Empty cells indicate studies that did not report the outcome of interest. 

For performance bias (blinding of participants and personnel) and detection bias (blinding of outcome assessment), we evaluated the risk of bias separately for each outcome, and grouped outcomes in the risk of bias table according to whether they were measured subjectively or objectively. We considered all outcomes as being similarly susceptible to performance bias. We judged the following endpoints as being susceptible to detection bias (i.e. subjective outcomes), thereby making blinding important.

• Urinary continence (at various time points)

• Serious adverse events

• Potency

• Urinary function scores on quality of life scale

• Sexual function scores on quality of life scale

We judged the following endpoints as not being susceptible to detection bias (i.e. objective outcomes), thereby making blinding unimportant.

• Positive surgical margins

• Biochemical recurrence‐free survival

We also assessed attrition bias (incomplete outcome data) on an outcome‐specific basis and presented the judgment for each outcome separately when reporting our findings in the risk of bias tables.

We further summarized the risk of bias across domains for each outcome in each included study, as well as across studies and domains for each outcome, in accordance with the approach for summary assessments of the risk of bias presented in the most recent Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019).

Measures of treatment effect

We expressed dichotomous data as risk ratios (RR) with 95% confidence intervals (CIs). We expressed continuous data as mean differences (MDs) with 95% CIs, unless different studies used different measures to assess the same outcome, in which case we planned to express data as standardized mean differences (SMDs) with 95% CIs. We planned to express any time‐to‐event data as HRs with 95% CIs.

Unit of analysis issues

The unit of analysis was the individual participant. If studies had multiple treatment arms, we had planned to present any/all treatments that included posterior reconstruction RALP versus non‐posterior reconstruction RALP.

Dealing with missing data

We obtained missing data from study authors, when feasible, and performed intention‐to‐treat (ITT) analyses if data were available; we otherwise performed available‐case analyses. We investigated attrition rates, for example, dropouts, losses to follow‐up, and withdrawals, and critically appraised issues of missing data. We did not impute missing data.

Assessment of heterogeneity

We only performed meta‐analysis where this was meaningful, that is, the treatments, participants, and outcomes were similar enough. In the event of excessive heterogeneity unexplained by subgroup analyses, we did not report outcome results as the pooled effect estimate in a meta‐analysis but provided a narrative description of the results of each study.

We identified heterogeneity (inconsistency) through visual inspection of the forest plots to assess the amount of overlap of CIs, and by using the I² statistic, which quantifies inconsistency across studies to assess the impact of heterogeneity on the meta‐analysis (Deeks 2019; Higgins 2003). We interpreted the I² statistic as follows (Deeks 2019).

• 0% to 40%: may not be important

• 30% to 60%: may indicate moderate heterogeneity

• 50% to 90%: may indicate substantial heterogeneity

• 75% to 100%: considerable heterogeneity

When we identified heterogeneity, we attempted to determine possible reasons for it by examining individual study and subgroup characteristics.

Assessment of reporting biases

We attempted to obtain study protocols to assess for selective outcome reporting. If we had included 10 studies or more investigating a particular outcome, we would have used funnel plots to assess small‐study effects. Several explanations can be offered for the asymmetry of a funnel plot, including true heterogeneity of effect with respect to trial size, poor methodologic design (and hence bias of small trials), and publication bias. Therefore, we interpreted results carefully. We worked in line with recommendations within the Cochrane Handbook (Page 2019).

Data synthesis

Unless we found good evidence for homogeneous effects across studies, we summarized data using a random‐effects model. We interpreted random‐effects meta‐analyses with due consideration of the whole distribution of effects. In addition, we performed statistical analyses according to the statistical guidelines contained in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). For dichotomous outcomes, we used the Mantel‐Haenszel method; for continuous outcomes, we used the generic inverse variance method. We used Review Manager 5 software to perform the analyses (Review Manager 2019).

Subgroup analysis and investigation of heterogeneity

We expected the following characteristics to introduce clinical heterogeneity and planned to carry out subgroup analyses with investigation of interactions.

• Participant age (less than 65 years versus 65 years and above). The decision to perform this subgroup analysis was based on studies suggesting the prognostic importance of age on continence recovery (Lavigueur‐Blouin 2015).

• Nerve‐sparing approach (complete or partial nerve‐sparing versus non‐nerve‐sparing). The decision to perform this subgroup analysis was based on studies suggesting the prognostic importance of nerve‐sparing status on continence recovery (Sridhar 2017).

• Clinical stage (cT1 versus cT2). The decision to perform this subgroup analysis was due to the fact that outcomes, especially PSM and BCRFS, can vary based on clinical stage (Retèl 2014).

• Anterior reconstruction technique when combined with posterior reconstruction versus only posterior reconstruction. This was based on studies suggesting the prognostic importance of anterior reconstruction on continence recovery (Arroyo 2019).

• Posterior musculofascial reconstruction with one layer versus two or more layers. This is based on studies suggesting that further reinforcement creates less tension on the vesicourethral anastomosis, which could lead to earlier recovery of urinary continence (Ogawa 2017).

We used the test for subgroup differences in Review Manager 2019 to compare subgroup analyses when we had sufficient studies.

Sensitivity analysis

We planned to perform sensitivity analyses to explore the influence of the following factors (when applicable) on effect sizes.

• Restricting the definition of continence to no pad use.

• Restricting the analysis by taking into account risk of bias, by excluding studies at high or unclear risk of bias.

Summary of findings and assessment of the certainty of the evidence

We presented the overall certainty of the evidence for each outcome according to the GRADE approach, which takes into account criteria related to internal validity (risk of bias, inconsistency, imprecision, publication bias) and external validity (such as directness of results) (Guyatt 2008). For each comparison, two review authors (JR, ZE, HL, or SL) independently rated the certainty of evidence for each outcome as 'high', 'moderate', 'low', or 'very low' using GRADEpro GDT. We resolved any discrepancies by consensus, or, if needed, with arbitration by a third review author (JHJ or PD). For each comparison, we presented a summary of the evidence for the main outcomes in a summary of findings table. The table provides key information about the best estimate of the magnitude of the effect in relative terms and absolute differences (with corresponding confidence intervals presented in brackets) for our single comparison; numbers of participants and studies addressing each important outcome; and the rating of the overall confidence in effect estimates for each outcome (Guyatt 2011; Schünemann 2011). If we had not been able to conduct a meta‐analysis, we would have presented results in a narrative summary of findings table. Review findings were interpreted using a minimally contextualized approach (Hultcrantz 2017) using GRADE‐approved narrative wording (Santesso 2020).

Main outcomes for summary of findings table

The summary of findings table reports the following outcomes, listed according to priority.

1. Urinary continence within one week after catheter removal

2. Urinary continence three months after surgery

3. Serious adverse events

4. Urinary continence 12 months after surgery

5. Potency 12 months after surgery

6. Positive surgical margins

7. Biochemical recurrence‐free survival

Results

Description of studies

Our search of multiple electronic databases yielded 528 references. We found no records through other sources.

Results of the search

We screened 520 references at the title/abstract stage. Of these, 13 full‐text articles were assessed for eligibility. Five studies entering full‐text review stage were excluded, and we summarized reasons for exclusion at the full‐text stage in the PRIMSA flow diagram (Figure 1). We ultimately included eight studies in the quantitative analyses.

Included studies

Source of data

We included eight studies: six published studies (Hurtes 2011; Jeong 2015; Koliakos 2010; Menon 2008; Salazar 2021; Sutherland 2010); and two abstract proceedings (Hoogenes 2018; Pushkar 2016). We contacted the authors and received relevant study data from Hoogenes 2018 and Salazar 2021. All studies were published in English. Study characteristics are described in the Characteristics of included studies table.

Study design and settings

Seven of the eight studies were parallel single center randomized controlled trials (RCTs); the study by Hurtes 2011 was a parallel RCT undertaken at three tertiary referral centers. Study locations included Belgium (Koliakos 2010), Canada (Hoogenes 2018), France (Hurtes 2011), Italy (Salazar 2021), Korea (Jeong 2015), Russia (Pushkar 2016), and the USA (Menon 2008; Sutherland 2010). Accrual periods ranged from 2007 to 2019.

Participants

We included 1085 randomized participants, of whom 963 completed the trials (88.8%). Mean age ranged from 60 to 67 years old. A common inclusion criterion was clinically localized prostate cancer, with the exception of the study by Jeong 2015, which included participants with stage cT3a cancer or less. Hoogenes 2018 did not require participants to have preoperative urinary continence to be included, but 98.6% of the intervention group and 100% of the control group had preoperative continence. Baseline characteristics are summarized in Table 2.

1. Study Characteristics.

Study Name	Age (years)	PSA (ng/mL)	Surgeon experience	Number of participants in each arm	Number of participants receiving nerve sparing	Primary Outcome	Secondary Outcomes	Time Points Measured	Duration of Follow‐up
Menon 2008	PR: 60.1 No PR: 59.2	PR: 6.1 No PR: 6.4	1 of 2 surgeons with 2444 cases and 811 RALP cases, respectively	PR: 59 No PR: 57	Veil PR: 37% No PR: 46%   Standard PR: 63% No PR: 31%	Continence recovery	Urinary continence, median urinary loss, rate of contrast extravasation in cystograms	1, 2, 7, 30 days after catheter removal	31 days after surgery
Sutherland 2010	PR: 59.9 No PR: 60.4	PR: 4.9 No PR: 5.6	1 of 3 surgeons with > 60 RALP cases each	PR: 47 No PR: 47	PR: 81% No PR: 75%	Continence recovery at postoperative month 3	Self‐reported daily leaks, self‐perception of urinary function, IPSS, 24‐hour pad weight	3 days, 6 weeks, 3 months	3 months after surgery
Koliakos 2010	PR: 60.96 ± 6.56 No PR: 61.75 ± 5.96	PR: 11.3 ± 11.1 No PR: 10.47 ± 2.22	N/a	PR: 50 No PR: 24	N/a	Continence at catheter removal and 7 weeks after catheter removal	Number of days participant was incontinent, perceived frequency and amount of leaking, pad usage, quality of life during the period of incontinence	1, 2, 7, 30 days after catheter removal	7 weeks after surgery
Hurtes 2011	PR: 62.5 ± 6.8 No PR: 62.4 ± 5	PR: 6.4 (2.7 – 16.6) No PR: 7.9 (3.6 – 23.6)	> 100 RALP cases	PR: 39 No PR: 33	PR: 76.3% No PR: 60.6%	Continence recovery at postoperative month 3	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.	15 days, 1, 3, and 6 months after surgery	6 months after surgery
Jeong 2015	PR: 64.3 ± 6.8 No PR: 66.9 ± 7.0	PR: 11.3 ± 11.1 No PR: 14.9 ± 20.9	> 700 RALP cases	PR: 50 No PR: 50	Bilateral PR: 98% No PR: 91.1%   Unilateral: PR: 2% No PR: 6.7%	Continence at 6 months	Social continence recovery (0 to 1 pad/day), continence score, urinary leak, QoL, total operative time, blood loss	1 day, 2 weeks, 1 month, 3 months, 6 months	6 months after surgery
Hoogenes 2018	PR: 63.1 ± 9.7 No PR: 63.3 ± 6.3	PR: 7.6 ± 3.8 No PR: 8.4 ± 7.8	Single high‐volume robotic surgeon	PR: 84 No PR: 80	Bilateral intrafascial PR: 31.3% No PR: 54.3%   Unilateral: PR: 17.9% No PR: 13.8%	Continence recovery	Frequency of urinary leak, quantity of pad use, subjective urinary control, overall bother, erectile function. (erectile function as demonstrated by patient self‐report on items within the sexual function domain of the EPIC‐26 at baseline and at each follow‐up point).	2, 3, 4, 6, 8, 12 months after surgery	12 months after surgery
Pushkar 2016	PR: N/a No PR: N/a	PR: N/a No PR: N/a	N/a	PR: 201 No PR: 197	Bilateral (participants with IIEF‐5 > 17 prior to surgery): PR: 82.5% No PR: 82.2%	Continence recovery	Potency recovery (IIEF‐5 scores at 6 and 12 months), biochemical recurrence rates at 6 and 12 months, positive surgical margins anastomotic time, serious adverse events	1, 2, 3, 6, and 12 months after surgery	12 months after surgery
Salazar 2021	PR: 64.5 No PR: 64.05	PR: 7.42 No PR: 6.78	> 700 RALP cases	PR: 81 No PR: 72	N/a	Continence at 1 and 6 months after prostatectomy	Urinary continence measured in grams, urinary continence at 12 months, potency quality of life, biochemical recurrence, positive surgical margins, pathological pTNM in prostatectomy specimens	1 week, 1 month, 3 months, 6 months, 12 months	12 months after surgery

N/a: not applicable
IPSS: International Prostate Symptom Score
QoL: quality of life
PR: posterior reconstruction
pTNM: pathological tumor, node and metastasis stage
RALP: robot‐assisted laparoscopic prostatectomy

Interventions

All trials performed a variation of posterior reconstruction, and we described these difference in Table 3.

2. Comparison of different urinary continence measurement and surgical techniques employed.

	Urinary continence measurement (in pads per day)	Inclusion of Denonvilliers' fascia	Suspension of posterior bladder wall to posterior musculofascial plate	Anterior reconstruction	Layers of posterior reconstruction	Comments
Menon 2008	0‐1	Yes	No	Yes	1	Did not join the reconstructed sphincter to the posterior wall of the bladder
Sutherland 2010	0‐1	Yes	Yes	No	1	Suspended posterior bladder wall to posterior musculofascial plate
Koliakos 2010	0‐1	Yes	No	Yes	1
Hurtes 2011	0‐1	Yes	No	Yes*	1	*Anterior suspension technique
Jeong 2015	0‐1	No	Yes	Yes	1	Novel 1‐step reconstruction not including Denonvilliers' fascia
Hoogenes 2018	0‐1	Yes	Yes	No	1
Pushkar 2016	0	N/a	N/a	N/a	N/a	Abstract only; did not describe technique
Salazar 2021	0‐1	Yes	Yes	No	2

N/a: not applicable

Hoogenes 2018 and Salazar 2021 performed a two‐layer posterior rhabdosphincter reconstruction, but first approximated the free edge of Denonvilliers' fascia to the posterior aspect of the rhabdosphincter with a continuous suture. In the second layer, the posterior lip of the bladder neck was sutured to the posterior urethral edge. They then performed the urethrovesical anastomosis.

Hurtes 2011, Koliakos 2010, and Menon 2008 joined the Denonvilliers' fascia and the posterior wall of the striated rhabdosphincter, but did not join the reconstructed sphincter to the posterior wall of the bladder. Rocco 2012 exemplified the importance of this step because it lengthens the urethral sphincter, increasing the function length of the urethral sphincteric complex and repositioning it into the most feasible anatomic position. Rocco 2012 also performed anterior reconstruction by suturing the puboprostatic ligament to the anterior pubovesical collar. Sammon 2010 was a two‐year follow‐up of the Menon 2008 study, with 86.5% of participants followed up. Hurtes 2011 and Koliakos 2010 performed the same technique as Menon 2008. As with the Menon 2008 study, they did not mention joining the reconstructed sphincter to the posterior wall of the bladder, a critical step according to Rocco 2012. Koliakos 2010 performed anterior reconstruction, while Hurtes 2011 used a peri‐urethral suspension technique.

Jeong 2015 had a unique and novel way of performing the posterior reconstruction. They called it a one‐step technique, where “reconstruction was done between the posterior part of the rhabdosphincter, including the median dorsal fibrinous raphe only to the posterior detrusor apron without including the Denonvilliers' fascia.” They also used an anterior reconstruction technique in this trial for both the experimental and control arms.

Pushkar 2016 was an abstract, in which they state they used the “classic Rocco technique”. We believe that this was the same two‐step technique used by Hoogenes 2018 and Salazar 2021.

Sutherland 2010 explained their method by stating that “the surgical principles described by Rocco et al. were followed,” but did not describe this surgical technique in detail. As we know, there are many variations of the Rocco technique. Additionally, their control group also ended up being a partial reconstruction technique because they placed posterior anastomotic sutures through the posterior urethra and “the underlying thick layer of Denonvilliers' fascia”. However, this “thick layer of Denonvilliers' fascia” is extremely close anatomically to the posterior rhabdosphincter (possibly indistinguishable from it). Thus, this could be considered a partial posterior reconstruction, which may influence the results of early recovery of urinary continence in the control arm (Coelho 2011; Rocco 2001). They did not perform anterior reconstruction.

Comparator

All trials performed standard RALP as a comparator, creating a conventional urethrovesicular anastomosis with a continuous running technique in a semicircular manner. 

Outcomes

Five trials reported urinary continence within one week of catheter removal (Jeong 2015; Koliakos 2010; Menon 2008; Salazar 2021; Sutherland 2010), and six reported urinary continence three months after surgery (Hoogenes 2018; Hurtes 2011; Jeong 2015; Menon 2008; Salazar 2021; Sutherland 2010). Five trials reported serious adverse events (Hoogenes 2018; Hurtes 2011; Jeong 2015; Menon 2008; Salazar 2021), and five reported continence at six months after surgery (Hoogenes 2018; Hurtes 2011; Jeong 2015; Pushkar 2016; Salazar 2021). Only three trials reported continence 12 months after surgery (Hoogenes 2018; Pushkar 2016; Salazar 2021). Hoogenes 2018 and Hurtes 2011 reported data on positive surgical margins, while only Hoogenes 2018 reported on potency recovery 12 months after surgery.

Pushkar 2016 and Salazar 2021 reported on biochemical recurrence 12 months after surgery. Pushkar 2016 defined biochemical recurrence as PSA > 0.2 ng/mL, while we were unsure how Salazar 2021 defined this outcome. 

Menon 2008 reported on urinary or sexual function, and no trials reported on sexual function quality of life.

Four trials used the Expanded Prostate Cancer Index Composite (EPIC) urinary domain to determine continence (Hoogenes 2018; Jeong 2015; Pushkar 2016; Sutherland 2010). Hurtes 2011 used the UCLA Prostate Cancer Index (UCLA‐PCI) self administered questionnaire to measure urinary continence. Menon 2008 measured the proportion of men using no pad or one pad (30 gm or less leakage) per day. Koliakos 2010 and Salazar 2021 measured continence via a telephone interview, using the International Consultation on Incontinence Questionnaire‐Short Form (ICIQ‐SF) and pad usage per day.

We defined serious adverse events using the Dindo‐Clavien system score ≥ 3 and measured at up to 12 months after surgery (Dindo 2004). Menon 2008 reported that there were two men in the control group with bladder neck contractures that needed dilation. There were 50 men in the control group and 46 in the intervention group who were present for the long‐term follow‐up at two years. Pushkar 2016 measured serious adverse events using the Dindo‐Clavien system score ≥ 3 and measured within 30 days after surgery (Dindo 2004). The analysis included 159 of the 201 men (79.1%) randomized to the experimental group and 156 of the 197 men (79.2%) randomized to the control group.

Hoogenes 2018 measured recovery of potency using the EPIC‐26 survey. The study defined recovery of erectile function as a score of 'good/very good ability to have an erection'; 45.8% of men in the intervention group and 42.2% in the control group had baseline ability to have an erection. Pushkar 2016 included men with International index of erectile function (IIEF) > 17 prior to surgery in their analysis of potency recovery 12 months after surgery. Prior to surgery, 103 men in the posterior reconstruction group and 101 men in the no posterior reconstruction group had IIEF > 17, and the investigators measured IIEF > 17 at one year after surgery.

We are unsure how positive surgical margins were defined in the other three trials (Hoogenes 2018; Hurtes 2011; Pushkar 2016).

Urinary function quality of life scores were determined by the quality of life question on the International Prostate Symptom Score (IPSS) questionnaire. The response to the question “If you were to spend the rest of your life with your urinary condition just the way it is now, how would you feel about that?” ranged from 0 ('delighted') to 6 ('terrible'), with 3 being 'mixed'. We treated this as continuous data (from 0 through 6), measuring the mean difference from the baseline quality of life score. Menon 2008 was the only study to report this. Koliakos 2010 measured urinary function quality of life but measured this using the International Consultation on Incontinence Questionnaire (ICIQ). Three trials used the EPIC questionnaire (Hoogenes 2018; Jeong 2015; Sutherland 2010), but they did not break this down into measuring urinary function quality of life. Sutherland 2010 also used the total International Prostate Symptom Score (IPSS) questionnaire but did not report the urinary function quality of life question separately, so the data were not usable.

Funding sources and conflicts of interest

Hoogenes 2018 received funding from The Masonic Foundation of Ontario. Menon 2008 received funding from the Vattikuti Urology Institute and reported a conflict of interest due to their relationship with Intuitive Surgical. Sutherland 2010 reported a conflict of interest with Vivus and TIMM Medical. No other trials reported funding or conflicts of interest.

Excluded studies

We excluded five studies at the full‐text review stage. Anceschi 2013 and Salvaggio 2009 did not perform RALP. The other three studies were not RCTs (Boylu 2009; Checcucci 2019a; Coelho 2011). 

Studies awaiting classification and ongoing trials

We identified no studies awaiting classification and no ongoing trials.

Risk of bias in included studies

For details, please refer to the Characteristics of included studies section, the risk of bias graph (Figure 2), and the risk of bias summary (Figure 3).

Allocation

Random sequence generation

All studies reported an appropriate method of random sequence generation, so we rated them as low risk.

Allocation concealment

Five studies failed to document an appropriate method of concealing allocation, and we rated them as having an unclear risk (Hoogenes 2018; Hurtes 2011; Koliakos 2010; Pushkar 2016; Salazar 2021). The remaining three studies reported an appropriate method of concealing allocation, and we rated them as low risk.

Blinding

Blinding of participants and personnel

It is not feasible to blind the surgeon or operating room (OR) personal, so we judged all eight studies to be at high risk. Hoogenes 2018; Jeong 2015; Menon 2008; Salazar 2021; Sutherland 2010 blinded their participants and post‐op caregivers; we judged these to have a low risk of bias. Hurtes 2011 and Koliakos 2010 were unclear whether they blinded their post‐op caregivers; we judged these to have an unclear risk of bias.  Pushkar 2016 did not blind post‐operative caregivers or data assessors and was rated as high risk of bias. 

Blinding of outcome assessment

Subjective outcomes were: urinary continence (at various time points), potency, urinary function, and sexual function quality of life. All studies were rated to be at low risk for participant self‐assessed outcome assessment because they blinded the participants. We rated the studies that did not address blinding of non‐OR personal as unclear risk for investigator‐adjudicated outcome assessment (Hurtes 2011; Koliakos 2010) and the one study that did not blind post‐operative caregivers or data assessors as high risk of bias (Pushkar 2016). We judged five studies to be at low risk for investigator‐adjudicated outcome assessment because they blinded their participants and postoperative caregivers (Hoogenes 2018; Jeong 2015; Menon 2008; Salazar 2021; Sutherland 2010).

Objective outcomes were: positive surgical margin and freedom from biochemical recurrence. We rated all studies to be at low risk for bias for this domain.

Incomplete outcome data

• Urinary continence: Five studies reported low levels of attrition that permitted a low risk of bias judgment. Two studies reported high levels of attrition, so we rated these as high risk of bias. One study reported intermediate levels of attrition, so we rated this as having an unclear risk of bias

• Serious adverse events: Three studies reported low levels of attrition that permitted a low risk of bias judgment. One study reported a high level of attrition, so we considered this to have a high risk of bias. Two studies reported intermediate levels of attrition, so we rated these as having an unclear risk of bias. The remaining two studies did not present this data.

• Potency: Two studies reported intermediate levels of attrition, so we rated these as having an unclear risk of bias. The remaining studies did not present this data.

• Oncologic outcomes: Two studies reported low levels of attrition that permitted a low risk of bias judgment. Two studies reported a high level of attrition, and we rated this as having a high risk of bias. The remaining studies did not present this data.

Selective reporting

Three of the eight studies reported outcomes analyses consistent with an a priori, registered protocol; therefore we rated these studies as having low risk of bias (Hoogenes 2018; Jeong 2015; Salazar 2021). We could not find an a priori written protocol for any of the remaining five studies; therefore we rated them as having unclear risk of bias, because we have no assurance that all measured outcomes were reported and analyzed as intended.

Other potential sources of bias

For the study by Menon 2008, we realized that the interim analysis was done when 58 men were entered into the study, with the intention of terminating the study early. While the study reported non‐significant results and the authors used low P value thresholds at the interim analysis, multiple observations of the data may create a source of bias. We therefore gave this a high risk of bias rating for this domain. We found no other sources of bias for five of the remaining included studies, and we rated them as low risk. Two of the studies never published a finalized manuscript and were abstract only so these were rated as unclear risk of bias (Hoogenes 2018; Pushkar 2016). 

Effects of interventions

See: Table 1

The review had only one comparison: posterior reconstruction versus no reconstruction during RALP. Key findings are presented in Table 1. 

Primary outcomes

Urinary continence within one week after catheter removal

Posterior reconstruction RALP (PR‐RALP) may improve urinary continence one week after catheter removal compared to no posterior reconstruction during RALP (risk ratio (RR) 1.25, 95% confidence interval (CI) 0.90 to 1.73; I² = 42%; studies = 5, participants = 498; low certainty evidence (CoE); Analysis 1.1) although the CI also includes the possibility of no effect. Assuming 335 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 84 more men per 1000 (33 fewer to 244 more) reporting urinary continence recovery. We downgraded the CoE once for serious study limitations and again for serious inconsistency (for an I² of 67%). We also observed considerable imprecision but did not rate down further.

1.1. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 1: Urinary continence within 1 week after catheter removal

Urinary continence three months after surgery

PR‐RALP may have little to no effect on urinary continence three months after surgery compared to RALP with no reconstruction (RR 0.98, 95% CI 0.84 to 1.14; I² = 67%; studies = 6, participants = 842; low CoE; Analysis 1.2). Assuming 701 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 14 fewer men per 1000 (112 fewer to 98 more) reporting urinary continence recovery. We downgraded the CoE once for serious study limitations and again for serious, clinically relevant inconsistency (for an I² of 42%). We also observed considerable imprecision but did not rate down further since we attributed this mainly to inconsistency.

1.2. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 2: Urinary continence 3 months after surgery

Serious adverse events

PR‐RALP probably results in little to no difference in serious adverse events compared to RALP with no reconstruction (RR 0.75, 95% CI 0.29 to 1.92; I² = 0%; studies = 6, participants = 835; moderate CoE; Analysis 1.3). Assuming 25 per 1000 men undergoing standard RALP have experienced a serious adverse event at this time point, this corresponds to six fewer men per 1000 (17 fewer to 23 more) reporting serious adverse events. We downgraded the CoE once for serious study limitations. 

1.3. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 3: Serious adverse events

Secondary outcomes

Urinary continence six months after surgery

PR‐RALP probably results in little to no difference in recovery of continence six months after surgery compared to RALP with no reconstruction (RR 1.01, 95% CI 0.97 to 1.05; I² = 0%; studies = 5, participants = 741; moderate CoE; Analysis 1.4). Assuming 817 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 16 more men per 1000 (26 fewer to 44 more) reporting urinary continence recovery. We downgraded the CoE for serious study limitations.

1.4. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 4: Urinary continence 6 months after surgery

Urinary continence 12 months after surgery

PR‐RALP may result in little to no difference in recovery of continence 12 months after surgery compared to RALP with no reconstruction (RR 1.02, 95% CI 0.98 to 1.07; I² = 25%; studies = 3, participants = 602; low CoE; Analysis 1.5). Assuming 918 per 1000 men undergoing standard RALP are continent at this time point, this corresponds to 18 more men per 1000 (18 fewer to 64 more) reporting urinary continence recovery.  We downgraded the CoE once for serious study limitations and again for serious imprecision given a confidence interval that included the possibility of worsened continence. 

1.5. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 5: Urinary continence 12 months after surgery

Potency 12 months after surgery

We are very uncertain about the effects of PR‐RALP on potency recovery 12 months after surgery compared to RALP with no reconstruction (RR 1.02, 95% CI 0.82 to 1.26; I² = 3%; studies = 2, participants = 308; very low CoE; Analysis 1.6). Assuming 433 per 1000 men undergoing standard RALP are potent at this time point, this corresponds to nine more men per 1000 (78 fewer to 113 more) reporting potency recovery. We downgraded the CoE for serious study limitations and twice for very serious imprecision.

1.6. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 6: Potency recovery 12 months after surgery

Positive surgical margins

PR‐RALP may result in little to no difference in positive surgical margins compared to RALP with no reconstruction (RR 1.24, 95% CI 0.65 to 2.33; I² = 50%; studies = 3, participants = 517; low CoE; Analysis 1.7). Assuming 130 per 1000 men undergoing standard RALP have a positive surgical margin, this corresponds to 31 more men per 1000 (46 fewer to 173 more) reporting positive surgical margins. We downgraded the CoE for serious study limitations and serious, clinically relevant inconsistency (I² = 50%).

1.7. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 7: Positive surgical margins

Biochemical recurrence at 12 months after surgery

PR‐RALP may result in little to no difference in biochemical recurrence compared to RALP with no reconstruction (RR 1.36, 95% CI 0.74 to 2.52; I² = 0%; studies = 2, participants = 468; low CoE; Analysis 1.8). Assuming 70 per 1000 men undergoing standard RALP have experienced biochemical recurrence by this time point, this corresponds to 25 more men per 1000 (18 fewer to 107 more) reporting biochemical recurrence. We downgraded the CoE for once for serious study limitations and again for serious imprecision.

1.8. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 8: Biochemical recurrence at 12 months

Urinary function quality of life

PR‐RALP may result in little to no difference in urinary function quality of life compared to standard RALP using the IPSS quality of life scale (0 to 6; higher value reflect more bother/worse quality of life) (MD –0.40, 95% CI –1.09 to 0.29; studies = 1, participants = 94; low CoE; Analysis 1.9). We downgraded the CoE for serious study limitations and serious imprecision given that the 95% CI crossed the assumed threshold of a 0.5 point change in IPSS‐Quality of Life.

1.9. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 9: Urinary function quality of life two years after surgery

Sexual function quality of life

We found no evidence for this outcome.

Subgroup analyses

Anterior reconstruction technique when combined with posterior reconstruction versus only posterior reconstruction.

Urinary continence within one week after catheter removal

Comparing the technique with anterior reconstruction versus that without anterior reconstruction, we found an RR of 1.40 (95% CI 0.90 to 2.17) with anterior reconstruction versus an RR of 0.92 (95% CI 0.38 to 2.22) without anterior reconstruction. The test for interaction showed no evidence of a difference between subgroups (P = 0.40, I² = 0%) (Analysis 1.1).

Urinary continence three months after surgery

We found an RR of 1.61 (95% CI 0.35 to 7.46) with anterior reconstruction versus an RR of 0.94 (95% CI 0.77 to 1.14) without anterior reconstruction. The test for interaction showed no evidence of a difference between subgroups (P = 0.49, I² = 0%) (Analysis 1.2).

Serious adverse events

We found an RR of 0.22 (95% CI 0.01 to 4.40) with anterior reconstruction versus an RR of 0.86 (95% CI 0.32 to 2.31) without anterior reconstruction. The test for interaction showed no evidence of a difference between subgroups (P = 0.40, I² = 0%) (Analysis 1.3).

Urinary continence six months after surgery

We found an RR of 1.03 (95% CI 0.96 to 1.11) with anterior reconstruction versus an RR of 1.00 (95% CI 0.95 to 1.05) without anterior reconstruction. The test for interaction showed no evidence of a difference between subgroups (P = 0.52, I² = 0%) (Analysis 1.4).

Positive surgical margins

We found an RR of 1.74 (95% CI 0.57 to 5.25) with anterior reconstruction versus an RR of 1.11 (95% CI 0.45 to 2.70) without anterior reconstruction. The test for interaction showed no evidence of a difference between subgroups (P = 0.53, I² = 0%) (Analysis 1.7).

We were unable to conduct the preplanned subgroup analyses based on participant age, nerve‐sparing status, or clinical stage, due to lack of relevant data in the included studies. We were also unable to conduct subgroup analyses based on the number of layers of posterior reconstruction, due to lack of relevant data in the included studies.

Sensitivity analyses

Based on definition of continence of no pad use

Urinary continence within one week after catheter removal

This sensitivity analysis included three studies (Jeong 2015; Menon 2008; Salazar 2021). The RR for this sensitivity analysis was 1.14 (95% CI 0.76 to 1.72; participants = 359; studies = 3; I² = 0%; Analysis 1.10), which did not alter the treatment effect compared to the main analysis (RR 1.25, 95% CI 0.90 to 1.73).

1.10. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 10: Sensitivity analysis: Urinary continence within 1 week after catheter removal (0 pads)

Urinary continence three months after surgery

We included four studies in this sensitivity analysis (Jeong 2015; Pushkar 2016; Salazar 2021; Sutherland 2010). The RR for this sensitivity analysis was 1.09 (95% CI 0.94 to 1.26; participants = 645; studies = 4; I² = 22%), which was similar to the treatment effect seen in the main analysis (RR 0.98, 95% CI 0.84 to 1.14; Analysis 1.11).

1.11. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 11: Sensitivity analysis: Urinary continence 3 months after surgery (0 pads)

Urinary continence six months after surgery

We included two studies in this sensitivity analysis (Jeong 2015; Salazar 2021). The RR for this sensitivity analysis was 1.11 (95% CI 0.96 to 1.28; participants = 241; studies = 2; I² = 0%), which did not alter the treatment effect compared to the main analysis (RR 1.01, 95% CI 0.97 to 1.05; Analysis 1.12).

1.12. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 12: Sensitivity analysis: Urinary continence 6 months after surgery (0 pads)

Urinary continence 12 months after surgery

The RR for this sensitivity analysis, which contained two studies (Pushkar 2016; Salazar 2021), was 1.01 (95% CI 0.96 to 1.06; participants = 462; studies = 2, I² = 0%). This did not alter the treatment effect compared to the main analysis (RR 1.02, 95% CI 0.98 to 1.07; Analysis 1.13).

1.13. Analysis.

Comparison 1: Posterior reconstruction vs standard RALP, Outcome 13: Sensitivity analysis: Urinary continence 12 months after surgery (0 pads)

Based on risk of bias

We were unable to conduct the preplanned sensitivity analyses by excluding studies at high or unclear risk of bias, given the paucity of studies and the similar risk of bias profiles.

Discussion

Summary of main results

Findings of this study are based on eight RCTs that randomized 1085 men with clinically localized prostate cancer. The mean age of participants in each RCT ranged from 60 to 67 years old. All participants had either cT1c, cT2 or cT3a disease, with a mean PSA of 8.15 ng/mL. Six studies were available as full text (Hurtes 2011; Jeong 2015; Koliakos 2010; Menon 2008; Salazar 2021; Sutherland 2010), and two as abstract only (Hoogenes 2018; Pushkar 2016). Three studies had an a priori registered protocol (Hoogenes 2018; Jeong 2015; Salazar 2021).

We found that posterior reconstruction may result in improved continence one week after catheter removal (although the CI also includes the possibility of no effect), but may result in little to no difference after three and 12 months after surgery (low‐certainty evidence) and also probably does not six months after surgery (moderate‐certainty evidence).

Based on the available evidence, posterior reconstruction probably results in little to no difference in serious adverse events compared to no posterior reconstruction  (moderate‐certainty evidence). Posterior reconstruction may also result in little to no difference in positive surgical margins and biochemical recurrence at twelve months after surgery compared to no posterior reconstruction (low‐certainty evidence). We are very uncertain about the effects of posterior reconstruction RALP on potency recovery (very‐low certainty evidence). There may be little to no difference in urinary function quality of life at two years after surgery (measured by the IPSS‐Quality of Life item; low‐certainty evidence).

This review was unable to determine if or how these findings may be impacted by the man's age, nerve‐sparing status, or clinical stage. Subgroup analysis did not identify any significant difference in outcomes for anterior suspension/reconstruction with posterior reconstruction compared with posterior reconstruction only. Restricting continence definition to no pad use did not alter the treatment effect compared to the main analysis.  

Overall completeness and applicability of evidence

This systematic review represents the most rigorous and up‐to‐date review on the question of posterior reconstruction RALP. Although we perceive this body of evidence to be broadly applicable to current clinical practice, the following issues deserve mention.

• The trials informing this review originate from many different countries, which increases the applicability of the review's findings. Additionally, most comparisons (besides potency recovery) were informed by multiple trials, therefore increasing reproducibility.

• The included trials used different ways of assessing continence as summarized in this review, which we explored as potential source of heterogeneity.

• Several unique methods of posterior musculofascial reconstruction have been developed to maximize posterior reinforcement and preserve structures related to continence. We included any of these in our review, as our main focus was on posterior musculofascial reconstruction versus no construction. The differences between the techniques are detailed in Table 3. We performed a subgroup analysis on any surgical technique performing anterior reconstruction or suspension to see if performing it in addition to posterior reconstruction has an impact on continence rates.

• Only one trial was multicenter (Hurtes 2011). Surgical experience varies between each study, which may be an important predictor of outcomes, thus limiting the applicability of this review's findings (Vickers 2009). There is also concern about performance bias, which ideally would be addressed by an expertise‐based trial (Devereaux 2005; Scholtes 2012), in which participants are not only randomized to a given procedure but also to an expert surgeon.

• A number of modifications to the standard RALP have been explored to improve continence outcomes. The most notable is the Retzius‐sparing technique, which a companion Cochrane Review has assessed (Rosenberg 2020b). Recent innovations have focused on anterior preservation by virtue of a modified apical dissection and lateral prostatic fascia preservation (Moschovasa 2020), and the detrusor apron sparing hood (DASH) technique (Cumarasamy 2019). These modifications were not applied in the standard RALP arms of studies included in this review (neither has their value been rigorously evaluated in the context of RCTs).

Quality of the evidence

We rated the certainty of evidence as moderate to very low. Reasons for rating down were as follows.

• Study limitations: all of the studies blinded participants (low risk). We found frequent high risk of performance bias and unclear risk of selective reporting, which prompted us to downgrade the certainty of evidence for all outcomes. For subjective, self‐reported outcomes, there is the additional, related concern over detection bias. We recognize that blinding of surgeons is never possible and that blinding of other involved personnel would have been extremely challenging, but we believe that outcome assessors could be blinded; therefore concerns over bias exist.

• Inconsistency: for three outcomes, we downgraded the certainty of evidence for considerable unexplained heterogeneity that was clinically relevant in the context of clinical decision‐making in this setting.

• Imprecision: the finding of wide confidence intervals that crossed the threshold of clinical relevance led to us downgrading the evidence. When observed imprecision could plausibly be explained by inconsistency, which prompted downgrading, we did not downgrade further.

• Selective reporting bias: six of the included studies were not prospectively registered, raising concerns about omission of information and post hoc changes in the analytical approach that may have influenced the results in a systematic manner.

Potential biases in the review process

The study was performed based on rigorous Cochrane standards, which included a published protocol. Nevertheless, certain issues could be a source of bias.

• We performed a comprehensive literature search for eligible studies, irrespective of language and publication status. Nevertheless, we may have missed studies, in particular 'negative' studies. Given the paucity of studies, we were unable to formally assess for publication bias. 

• This review included eight randomized controlled trials, of which six were available as full‐text reports (Hurtes 2011; Jeong 2015; Koliakos 2010; Menon 2008; Salazar 2021; Sutherland 2010). Two were only available as abstracts (Hoogenes 2018; Pushkar 2016). Nevertheless, we recognize the importance of selective reporting bias. Three studies had an a priori registered protocol (Hoogenes 2018; Jeong 2015; Salazar 2021).

• Included studies reported participants' urinary continence at different time points. To provide meaningful summary data that might be helpful for clinicians and men undergoing surgery, we grouped available data by four time periods: within one week after catheter removal, three months after surgery, six months after surgery and one year after surgery. We established these categories with input from expert clinicians before writing the protocol or extracting the study data, and before any quantitative analysis was performed. Nevertheless, findings for these outcomes are potentially sensitively to the specific time ranges we chose, and this may be viewed as a potential source of bias.

• Our interpretations of the magnitudes of effects for both desirable and undesirable outcomes, as well as judgments about the certainty of evidence when it comes to the domains of imprecision and inconsistency, hinge on assumptions about the MCID. This is in the context of a minimally contextualized approach, based on GRADE guidance (Hultcrantz 2017). Unless the trials used validated instruments and there were published MCIDs available, the thresholds we used were based on input by the clinical co‐authors, and we have sought to do so transparently. We recognize that the use of different thresholds would somewhat alter the conclusions of this review.

Agreements and disagreements with other studies or reviews

To date, no review has applied the rigorous Cochrane methodology to this topic. Defining characteristics of this review include an a priori protocol (Rosenberg 2020), a comprehensive literature search irrespective of language and publication status, a focus on patient‐centered outcomes, and the application of GRADE methodology. Furthermore, our interpretation focuses on clinically relevant (rather than statistically significant) findings and provides absolute effect size estimates for all dichotomous outcomes. This is also the most up‐to‐date review that included data from studies that are only available in their abstract form, and unpublished data that we sought from the trial authors (Hoogenes 2018; Pushkar 2016; Salazar 2021). 

With regard to the findings of three of the more methodologically rigorous and recently published reviews, our findings compare as follows.

The latest systematic review included 21 studies with three RCTs comparing posterior musculofascial reconstruction after radical prostatectomy (Grasso 2016). Risk of bias was assessed on a study, not outcome, level and the study failed to provide any certainty of evidence rating to place their findings into perspective. A subgroup of only robotic RCTs showed the pooled RR at: three to seven days (RR 1.07, 95% CI 0.75 to 1.51), 30 days (RR 1.69, 95% CI 0.46 to 6.27), 90 days (RR 1.48, 95% CI 0.41 to 5.32) and 180 days (RR 1.13, 95% CI 0.70 to 1.82). Pooled meta‐analysis found no difference in PSM. Pooled meta‐analysis including six studies found no difference in the rate of acute urinary retention. Our findings on return of urinary continence and positive surgical margins correspond to the review by Grasso 2016.

Aside from several non‐RCTs, Rocco 2012 only included the two randomized controlled trials that were published at the time, neither of which showed an improvement in urinary continence when performing posterior musculofascial reconstruction after RALP. Risk of bias was not assessed on a study or outcome level, and the study failed to provide any certainty of evidence rating to place their findings into perspective. They found no RCT evidence for PSM, but three non‐randomized studies showed no difference in PSM rate. Of the 1260 participants, 23 had acute urinary retention (1.8%).

None of the existing systematic reviews provided a certainty of evidence rating, which we consider critical to any systematic review.

Authors' conclusions

Implications for practice.

Findings of this review suggest a possible, transient positive impact on very early urinary continence (although the CI also includes the possibility of no effect) but do not support improved continence three to 12 months later. This finding is in contrast to the Retzius‐sparing, posterior approach to RALP, which may result in improved continence one week after catheter removal as well as three and six months thereafter, albeit at the risk of possibly more positive surgical margins (Rosenberg 2020a; Rosenberg 2021). Therefore, there appears to be a very limited continence benefit to posterior reconstruction. 

However, we did not observe any potential downsides to posterior reconstruction, which should only prolong operative time to a very small degree. Posterior reconstruction may be useful for indications other than continence recovery that were outside the scope of this review, namely providing some anchorage for the posterior bladder neck in the pelvis and facilitating a tension‐free and watertight vesicourethral anastomosis. 

Implications for research.

A number of important research implications can be drawn from this review.

• Various observational studies suggest posterior reconstruction during RALP improves continence recovery compared to no reconstruction. This may be due to the reconstruction technique of the individual surgeon, while in our review many of the reconstruction techniques varied. Large randomized controlled trials with a standardized posterior reconstruction technique may give a clearer picture of the benefits or harms of posterior reconstruction. 

• Description and video compilations describe and compare many of the posterior, anterior, and peri‐urethral surgical reconstruction techniques (Vis 2019). However, randomized controlled trials comparing and combining these techniques need to be assessed against each other to make definite conclusions.

• All surgical trials preclude the possibility of blinding the surgeon, but an expertise‐based randomized controlled trial design appears feasible, thereby limiting the risk of performance bias (Cook 2015; Devereaux 2005). In such a trial, participants would not only be randomized to a given approach but also to a surgeon who is a recognized expert in that procedure.

• We performed a subgroup analysis on anterior suspension/construction combined with posterior reconstruction. We were unable to conduct multiple preplanned subgroup analyses, given the lack of appropriate data. However, it is possible that the effects of age, neurovascular bundle sparing, multiple layered reconstruction and other factors affected our findings. Specifically designed and adequately powered, high‐quality studies are necessary to elucidate these issues.

• Given the importance of quality of life as an outcome, assessment of prostate cancer treatment‐related domains (such as urinary and sexual function) should be consistently collected and reported using a validated, prostate cancer‐specific instrument, such as the Expanded Prostate Cancer Index Composite (EPIC) tool developed by Wei 2000.

What's new

Date	Event	Description
13 December 2022	Amended	Minor error correction: Forest plot scale (Analysis 1.4).

History

Protocol first published: Issue 7, 2020
Review first published: Issue 8, 2021

Notes

We have based parts of the Methods section of this protocol on a standard template developed by the Cochrane Metabolic and Endocrine Disorders Group, which has been modified and adapted for use by the Cochrane Urology Group. We acknowledge that some sections of this protocol are very similar or intentionally identical to the parallel, closely related review title on Retzius‐sparing RALP (Rosenberg 2020b).

Appendices

Appendix 1. MEDLINE search strategy

Ovid MEDLINE(R) and Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Daily and Versions(R) (1946 to present)

1. exp Prostatectomy/

2. exp Prostatic Neoplasms/

3. (prostat* adj3 (neoplas* or cancer* or tumor* or tumour*)).mp.

4. or/1‐3

5. exp Robotic Surgical Procedures/

6. exp Robotics/

7. 5 or 6

8. 4 and 7

9. ((robot* OR radical OR laparoscopic) adj5 prostat*).mp

10. (RARP or RALP).mp.

11. or/8‐10

12. ((posterior OR musculofascial or rhabdosphincter) adj5 reconstruct*).mp.

13. PMPR.mp

14. or/12‐13

15. 11 and 14

16. ..dedup 15

Appendix 2. Embase search strategy

1. exp robot‐assisted prostatectomy/

2. exp prostatectomy/

3. exp prostate tumor/

4. (prostat* adj3 (neoplas* or cancer* or tumor* or tumour*)).mp.

5. or/2‐4

6. exp robotic surgical procedure/

7. exp robotics/

8. or/6‐7

9. 5 and 8

10. ((robot* OR radical OR laparoscopic) adj5 prostat*).mp

11. (RARP or RALP).mp.

12. or/1,9‐11

13. ((posterior OR musculofascial or rhabdosphincter) adj5 reconstruct*).mp.

14. PMPR.mp.

15. or/13‐14

16. 12 and 15

17. ..dedup 16

Appendix 3. Cochrane Library search strategy

1. MeSH descriptor: [Prostatectomy] explode all trees

2. MeSH descriptor: [Prostatic Neoplasms] explode all trees

3. (prostat* near/3 (neoplas* or cancer* or tumor* or tumour*)):ti,ab,kw

4. {OR #1‐#3}

5. MeSH descriptor: [Robotic Surgical Procedures] explode all trees

6. MeSH descriptor: [Robotics] explode all trees

7. {OR #5‐#6}

8. #4 AND #7

9. ((robot* OR radical OR laparoscopic) near/5 prostat*):ti,ab,kw

10. (RARP or RALP):ti,ab,kw

11. {OR #8‐#10}

12. ((posterior OR musculofascial or rhabdosphincter) near/5 reconstruct*):ti,ab,kw

13. (PMPR):ti,ab,kw

14. {OR #12‐#13}

15. #11 AND #14

Appendix 4. PubMed search strategy

((("Prostatectomy"[Mesh] OR "Prostatic Neoplasms"[Mesh] OR (prostat*[tw] AND (neoplas*[tw] OR cancer*[tw] OR tumor*[tw] OR tumour*[tw]))) AND ("Robotic Surgical Procedures"[Mesh] OR "Robotics"[Mesh])) OR (((robot*[tw] OR radical[tw] OR laparoscopic[tw]) AND prostat*[tw]) OR RARP[tw] OR RALP[tw])) AND (((posterior*[tw] OR musculofascial[tw] OR rhabdosphincter[tw]) AND reconstruct*[tw]) OR PMPR[tw])

Appendix 5. Scopus search strategy

(INDEXTERMS("robot‐assisted prostatectomy") OR (INDEXTERMS("prostatectomy" OR "prostatic neoplasms" OR "prostate tumor") OR TITLE‐ABS‐KEY(prostat* W/3 (neoplas* or cancer* or tumor* or tumour*)) AND INDEXTERMS("robotic surgical procedure" OR "robotic surgical procedures" OR "robotics")) OR TITLE‐ABS‐KEY(((robot* OR radical OR laparoscopic) W/5 prostat*) OR RARP OR RALP)) AND TITLE‐ABS‐KEY(((posterior OR musculofascial or rhabdosphincter) W/5 reconstruct*) OR PMPR)

Appendix 6. Web of Science search strategy

1. TS=(((robot* OR radical OR laparoscopic) near/5 prostat*) OR RALP OR RARP)

2. TS=(((posterior OR musculofascial or rhabdosphincter) near/5 reconstruct*) OR PMPR)

3. #1 AND #2

Appendix 7. WHO ICTRP search strategy

rhabdosphincter AND reconstruction

Appendix 8. ClinicalTrials.gov search strategy

(musculofascial OR rhabdosphincter OR posterior) AND (reconstruct OR reconstruction OR reconstructing)

Appendix 9. Global Index Medicus search strategy

(( mh:("Prostatectomy")) OR ( mh:("Prostatic Neoplasms")) OR (tw:(prostat* AND (neoplas* OR cancer* OR tumor* OR tumour*)))) AND (( mh:("Robotic Surgical Procedures")) OR ( mh:("Robotics")) OR ( tw:(((robot* OR radical OR laparoscopic) AND prostat*) OR RARP OR RALP))) AND (tw:(((posterior* OR musculofascial OR rhabdosphincter) AND reconstruct*) OR PMPR))

Data and analyses

Comparison 1. Posterior reconstruction vs standard RALP.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Urinary continence within 1 week after catheter removal	5	498	Risk Ratio (M‐H, Random, 95% CI)	1.25 [0.90, 1.73]
1.1.1 With anterior reconstruction	3	258	Risk Ratio (M‐H, Random, 95% CI)	1.40 [0.90, 2.17]
1.1.2 Without anterior reconstruction	2	240	Risk Ratio (M‐H, Random, 95% CI)	0.92 [0.38, 2.22]
1.2 Urinary continence 3 months after surgery	6	842	Risk Ratio (M‐H, Random, 95% CI)	0.98 [0.84, 1.14]
1.2.1 With anterior reconstruction	2	152	Risk Ratio (M‐H, Random, 95% CI)	1.61 [0.35, 7.46]
1.2.2 Without anterior reconstruction	4	690	Risk Ratio (M‐H, Random, 95% CI)	0.94 [0.77, 1.14]
1.3 Serious adverse events	6	835	Risk Ratio (M‐H, Random, 95% CI)	0.75 [0.29, 1.92]
1.3.1 With anterior reconstruction	3	236	Risk Ratio (M‐H, Random, 95% CI)	0.22 [0.01, 4.40]
1.3.2 Without anterior reconstruction	3	599	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.32, 2.31]
1.4 Urinary continence 6 months after surgery	5	741	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.97, 1.05]
1.4.1 With anterior reconstruction	2	140	Risk Ratio (M‐H, Random, 95% CI)	1.03 [0.96, 1.11]
1.4.2 Without anterior reconstruction	3	601	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.95, 1.05]
1.5 Urinary continence 12 months after surgery	3	602	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.98, 1.07]
1.5.1 Without anterior reconstruction	3	602	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.98, 1.07]
1.6 Potency recovery 12 months after surgery	2	308	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.82, 1.26]
1.6.1 Without anterior reconstruction	2	308	Risk Ratio (M‐H, Random, 95% CI)	1.02 [0.82, 1.26]
1.7 Positive surgical margins	3	517	Risk Ratio (M‐H, Random, 95% CI)	1.24 [0.65, 2.33]
1.7.1 With anterior reconstruction	1	71	Risk Ratio (M‐H, Random, 95% CI)	1.74 [0.57, 5.25]
1.7.2 Without anterior reconstruction	2	446	Risk Ratio (M‐H, Random, 95% CI)	1.11 [0.45, 2.70]
1.8 Biochemical recurrence at 12 months	2	468	Risk Ratio (M‐H, Random, 95% CI)	1.36 [0.74, 2.52]
1.8.1 Without anterior reconstruction	2	468	Risk Ratio (M‐H, Random, 95% CI)	1.36 [0.74, 2.52]
1.9 Urinary function quality of life two years after surgery	1		Mean Difference (IV, Random, 95% CI)	Totals not selected
1.9.1 With anterior reconstruction	1		Mean Difference (IV, Random, 95% CI)	Totals not selected
1.10 Sensitivity analysis: Urinary continence within 1 week after catheter removal (0 pads)	3	359	Risk Ratio (M‐H, Random, 95% CI)	1.14 [0.76, 1.72]
1.11 Sensitivity analysis: Urinary continence 3 months after surgery (0 pads)	4	645	Risk Ratio (M‐H, Random, 95% CI)	1.09 [0.94, 1.26]
1.12 Sensitivity analysis: Urinary continence 6 months after surgery (0 pads)	2	241	Risk Ratio (M‐H, Random, 95% CI)	1.11 [0.96, 1.28]
1.13 Sensitivity analysis: Urinary continence 12 months after surgery (0 pads)	2	462	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.96, 1.06]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Hoogenes 2018.

Study characteristics
Methods	Study design: randomized controlled study Number of centers/Setting: single academic‐based tertiary healthcare center, single surgeon Country: Canada Dates of the study: April 2014 to July 2015
Participants	Inclusion criteria: over age 18 (with the ability to give informed consent) with localized prostate cancer (cT1–2, N0, M0). Exclusion criteria: history of previous prostatic or urethral surgery, or both; a known history of a disease or comorbidity that could affect continence, such as insulin‐dependent diabetes or urethral stenosis; and the presence of a urinary catheter preventing preoperative evaluation of continence. Total number of participants randomly assigned:164 Group A Number of all participants randomly assigned: 84 Age (years) (mean, SD): 63.1 ± 9.7 PSA (mean, SD): 7.6 ± 3.8 ng/mL Comorbidity: ASA I: 3; ASA II: 24; ASA III: 41; ASA IV: 4 Gleason score: T2a: 4, T2b‐c: 34; T3: 19 cStage: median T1c (T1, T3 range) Day of catheter removal: n/a Group B Number of all participants randomly assigned: 80 Age (years) (mean, SD): 63.3 ± 6.3 PSA (mean, SD): 8.4 ± 7.8 ng/mL Comorbidity: ASA I: 4; ASA II: 20; ASA III: 41; ASA IV: 2 Gleason score: T2a: 10, T2b‐c: 65; T3: 40 cStage: median T1c Day of catheter removal: n/a
Interventions	Group A: posterior reconstruction of the rhabdosphincter Group B: standard RARP Follow‐up (including intervention duration): up to 12 months after treatment
Outcomes	Primary outcome: return to urinary continence as demonstrated by participant self‐report on items within the urinary domain of the EPIC‐26 at baseline and each follow‐up point How measured: EPIC‐26 questionnaire 0 pads per day Time points measured: 2, 3, 4, 6, 8, 12 months after surgery Time points reported: 2, 3, 4, 6, 8, 12 months after surgery Secondary outcomes: frequency of urinary leak, quantity of pad use, subjective urinary control, overall bother, erectile function (erectile function as demonstrated by participant self‐report on items within the sexual function domain of the EPIC‐26 at baseline and at each follow‐up point). How measured: EPIC‐26 questionnaire, Clavien‐Dindo Classification of Surgical Complications scores. Time point measured: 2, 3, 4, 6, 8, 12 months after surgery Time points reported: 2, 3, 4, 6, 8, 12 months after surgery Subgroup: none
Funding sources	The Masonic Foundation of Ontario
Declarations of interest	None
Notes	Abstract only Protocol: #BFCRS‐RP‐U‐01 in the Canadian Cancer Clinical Trials Network Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “random number generator via GraphPad QuickCalcs software.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Unclear risk	Quote: “study coordinator informed the surgeon of the randomization outcome immediately prior to surgery.” Judgement: not described.
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “Patients were blinded to allocation status, while the surgeon was notified immediately prior to the case.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Low risk	Quote: “The post‐operative care time was blinded to assessors. The outcome assessors and data collectors were all blinded to allocation status. Patients were blinded to allocation status and were offered to be unblinded at the 12‐month mark following the final interview.” Judgement: non‐OR personnel were blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “The post‐operative care time was blinded to assessors. The outcome assessors and data collectors were all blinded to allocation status. Patients were blinded to allocation status and were offered to be unblinded at the 12‐month mark following the final interview.” Judgement: outcome assessors blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Low risk	Quote: “The post‐operative care time was blinded to assessors. The outcome assessors and data collectors were all blinded to allocation status. Patients were blinded to allocation status and were offered to be unblinded at the 12‐month mark following the final interview.”
Incomplete outcome data (attrition bias) Urinary continence	Unclear risk	Judgement: 73/84 (86.9%) and 67/80 (83.8%) participants in intervention and standard group at 12 months follow‐up, respectively, included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	Unclear risk	Judgement: 67/84 (79.8%) and 65/80 (81.3%) participants in intervention and standard group at 12 months follow‐up, respectively, included in analysis.
Incomplete outcome data (attrition bias) Potency	Unclear risk	Judgement: 73/84 (86.9%) and 67/80 (83.8%) participants in intervention and standard group at 12 months follow‐up, respectively included in analysis.
Incomplete outcome data (attrition bias) Oncological/QOL outcomes	High risk	Judgement: 67/84 (79.8%) and 64/80 (80.0%) participants in intervention and standard group at 12 months follow‐up, respectively, included in analysis.
Selective reporting (reporting bias)	Low risk	Judgement: protocol was published and study outcomes were well predefined and described.
Other bias	Unclear risk	Judgement: abstract only.

Hurtes 2011.

Study characteristics
Methods	Study design: randomized controlled study Number of centers: three tertiary referral centers Country: France Dates of the study: July 2009 to July 2010
Participants	Inclusion criteria: histologically proven prostate cancer, clinical stage cT2 or less, and had chosen surgery Exclusion criteria: 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. Total number of participants randomly assigned: 72 Group A Number of all participants randomly assigned: 39 Age (years) (mean, SD): 62.5 ± 6.8 PSA (mean, SD): 6.4 (2.7 to 16.6) ng/mL Comorbidity: ASA 1: 14; ASA 2: 22; ASA 3: 1 Gleason score: < 7: 24; 7: 11; >7: 4 cStage: T1: 32; T2: 5; T3: 0 Day of catheter removal median (range): 6 (5 to 40) Group B Number of all participants randomly assigned: 33 Age (years) (mean, SD): 62.4 ± 5 PSA (mean, SD): 7.9 (3.6 – 23.6) ng/mL Comorbidity: ASA 1: 20; ASA 2: 11; ASA 3: 0 Gleason score: <7: 18; 7: 11; >7: 4 cStage: T1: 25; T2: 7; T3: 1 Day of catheter removal median (range): 7 (5 to 20)
Interventions	Group A: posterior reconstruction Group B: standard RALP Follow‐up (including intervention duration): 6 months
Outcomes	Primary outcome: continence recovery at postoperative month 3 How measured: self‐administered questionnaire (UCLA‐PCI) Time points measured: 15 days, 1, 3, and 6 months after surgery Time points reported: 15 days, 1, 3, and 6 months after surgery Secondary outcomes: 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. Time point measured: n/a Time points reported: n/a Subgroup: None
Funding sources	None declared
Declarations of interest	None declared
Notes	Protocol: none found Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “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.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Unclear risk	Quote: “This list, which was generated using dedicated software of Tours University Statistics Laboratory, was known only to the study coordinator.” Judgement: not described.
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “After enrollment 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.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Unclear risk	Quote: “patients were blinded to their allocated group” Judgement: unclear if non‐OR personnel were blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “patients were blinded to their allocated group” Judgement: Patients were blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Unclear risk	Quote: “patients were blinded to their allocated group” Judgement: blinding of outcome assessors not addressed.
Incomplete outcome data (attrition bias) Urinary continence	High risk	Judgement: 26/39 (66.7%) and 19/33 (57.6%) participants in intervention and standard group at 6 months follow‐up, respectively included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	Low risk	Judgement: all participants who were randomized were included in analysis.
Incomplete outcome data (attrition bias) Oncological/QOL outcomes	Low risk	Judgement: 33/33 (100%) and 38/39 (97.4%) participants randomized in experimental and control group were included in the analysis, respectively.
Selective reporting (reporting bias)	Unclear risk	Judgement: study outcomes were well predefined and described, but protocol was not found.
Other bias	Low risk	Judgement: not detected.

Jeong 2015.

Study characteristics
Methods	Study design: prospective, single‐blinded, parallel randomized controlled study Number of centers: single center; single surgeon Country: Republic of Korea Dates of the study: October 2012 to August 2013
Participants	Inclusion criteria: cT3a or less pathologically proven prostate cancer; intent to undergo RARP Exclusion criteria: prior hormone therapy, prior radiation treatment to the prostate or pelvis, preoperative urinary incontinence; refusal to participate Total number of participants randomly assigned: 100 Group A Number of all participants randomly assigned: 50 Age (years) (mean, SD): 64.3 ± 6.8 PSA (mean, SD): 11.3 ± 11.1 ng/mL Comorbidity: DM 8 ± 16.0; HTN 10 ± 20; Cerebrovascular disease 3 ± 6; CVD 0 Gleason score: 6: 4; 7: 42; 8 to 9: 3 cStage: T2: 36; T3a: 8; T3b: 6 Day of catheter removal: 8.5 ± 1.1 Group B Number of all participants randomly assigned: 50 Age (years) (mean, SD): 66.9 ± 7.0 PSA (mean, SD): 14.9 ± 20.9 ng/mL Comorbidity: DM: 6 ± 13.3; HTN: 17 ± 37.8; cerebrovascular disease: 2 ± 4.4; CVD: 1 ± 2.2 Gleason score: 6: 5; 7: 34; 8 to 9: 6 cStage: T2: 35; T3a: 6; T3b: 4 Day of catheter removal: 8.2 ± 0.6
Interventions	Group A: posterior reconstruction of the median dorsal raphe solely to the posterior counterpart of the detrusor apron and anterior reconstruction, which involved opposing the anterior detrusor apron to the remaining puboprostatic ligaments and dorsal vascular complex Group B: standard RARP with anterior reconstruction as above Follow‐up (including intervention duration): 6 months
Outcomes	Primary outcome: urinary continence at 6 months How measured: EPIC questionnaire 5 (no pad use) Time points measured: screening, 1 day, 2 weeks, 1 month, 3 months, 6 months Time points reported: 1 day, 2 weeks, 1 month, 3 months, 6 months Secondary outcomes: social continence recovery (0 to 1 pad/day); measured by EPIC question 5; continence score; measured by EPIC question 5; urinary leak; measured by EPIC question 1; QoL; measured by EPIC question 7; total operative time; measured by console time; blood loss. Time points measured: 1 and 2 at screening, 1 day, 2 weeks, 1 month, 3 months, 6 months; 3 and 4 at screening, 2 weeks, 1 month, 3 months, 6 months; 5 and 6 at operative time. Time points reported: 1 at screening, 1 day, 2 weeks, 1 month, 3 months, 6 months; 2 and 4 at screening, 2 weeks, 1 month, 3 months, 6 months; 5 and 6 at operative time. Subgroup: none
Funding sources	None
Declarations of interest	None
Notes	Protocol: NCT01714219 Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “Registered residents in our department screened and enrolled study participants, who were assigned to the intervention or control group by a computer generated randomization sequence. The sequence, which was created using SAS 9.2 with a 1:1 allocation and randomly varied block size, was performed at the Medical Research Collaboration Center at our institution.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Low risk	Quote: “The allocation number was concealed in sequentially numbered, opaque, 2‐layer sealed envelopes. The surgeon opened each envelope just before performing the urethrovesical anastomosis so that most of the surgical procedure was not influenced by the allocation.” Judgement: appropriate method of allocation concealment used.
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “The surgeon opened each envelope just before performing the urethrovesical anastomosis so that most of the surgical procedure was not influenced by the allocation. Patients and caregivers were blinded until the completion of the followup evaluation at 6 months.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Low risk	Quote: “Patients and caregivers were blinded until the completion of the followup evaluation at 6 months. A blinded, independent clinical research nurse collected and documented the data.” Judgement: non‐OR personnel were blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “Patients and caregivers were blinded until the completion of the followup evaluation at 6 months. A blinded, independent clinical research nurse collected and documented the data.” Judgement: outcome assessor was blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Low risk	Quote: “Patients and caregivers were blinded until the completion of the followup evaluation at 6 months. A blinded, independent clinical research nurse collected and documented the data.” Judgement: outcome assessor was blinded.
Incomplete outcome data (attrition bias) Urinary continence	Low risk	Judgement: 50/50 (100%) and 45/50 (90.0%) participants in intervention and standard group at 6 months follow‐up, respectively included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	Low risk	Judgement: 50/50 (100%) and 45/50 (90.0%) participants randomized in experimental and control group were included in the analysis, respectively.
Selective reporting (reporting bias)	Low risk	Judgement: protocol was published and study outcomes were well predefined and described.
Other bias	Low risk	Judgement: not detected.

Koliakos 2010.

Study characteristics
Methods	Study design: randomized controlled study Number of centers: single center Country: Belgium Dates of the study: not reported
Participants	Inclusion criteria: people with cT1 or cT2 prostate cancer undergoing RALP Exclusion criteria: history of previous pelvic surgery, previous treatment of their prostate cancer, people with diabetes or neurological disease, symptoms of urinary incontinence before RALP Total number of participants randomly assigned: 50 Group A Number of all participants randomly assigned: 26 Age (years) (mean, SD): 60.96 ± 6.56 PSA (mean, SD): 10.43 ± 3.37 ng/mL Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: n/a Group B Number of all participants randomly assigned: 24 Age (years) (mean, SD): 61.75 ± 5.96 PSA (mean, SD): 10.47 ± 2.22 ng/mL Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: n/a
Interventions	Group A: posterior reconstruction Group B: standard RALP Follow‐up (including intervention duration): 6 months
Outcomes	Primary outcome: urinary continence at catheter removal and 7 weeks after catheter removal How measured: telephone interview using ICIQ‐SF Time points measured: day 0, 7 weeks from catheter removal Time points reported: day 0, 7 weeks Secondary outcomes: number of days participant was incontinent, perceived frequency and amount of leaking, pad usage, quality of life during the period of incontinence How measured: telephone interview using ICIQ‐SF Time points measured: n/a Time points reported: 7 weeks Subgroup: none
Funding sources	None
Declarations of interest	None
Notes	Protocol: none found Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “The patients were randomized into two groups using a random number table.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Unclear risk	Judgement: not described
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “Each patient was unaware of the technique used during their operation.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Unclear risk	Quote: “Each patient was unaware of the technique used during their operation.” Judgement: unclear whether non‐OR personnel were blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “Each patient was unaware of the technique used during their operation.” Judgement: Patients were blinded
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Unclear risk	Quote: “Each patient was unaware of the technique used during their operation.” Judgement: blinding of outcome assessors not addressed.
Incomplete outcome data (attrition bias) Urinary continence	Low risk	Judgement: 23/24 (95.8%) and 24/26 (92.3%) participants in intervention and standard group at 7 weeks follow‐up, respectively, included in analysis.
Selective reporting (reporting bias)	Unclear risk	Judgement: study outcomes were well predefined and described, but protocol was not found.
Other bias	Low risk	Judgement: not detected

Menon 2008.

Study characteristics
Methods	Study design: parallel randomized controlled study Number of centers: 1 center, 2 surgeons Country: USA Dates of the study: September 2007 to December 2007
Participants	Inclusion criteria: signed consent form, undergoing RARP for prostate cancer Exclusion criteria: people requiring bladder neck reconstruction (more than 1 suture placed to narrow the bladder neck) and international patients for whom follow‐up would be difficult. Total number of participants randomly assigned: 116 Group A Number of all participants randomly assigned: 59 Age (years) (mean): 60.1 PSA: 6.1 ng/mL Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: 7 Group B Number of all participants randomly assigned: 57 Age (years) (mean): 59.2 PSA: 6.4 ng/mL Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: 7
Interventions	Group A: posterior reconstruction Group B: standard RALP Follow‐up (including intervention duration): 31 days
Outcomes	Primary outcome: urinary continence measured as 0 to 1 pad How measured: proportion of participants using 0 to 1 pad (30 gm or less leak) per day Time points measured: 1, 2, 7, 30 days after catheter removal Time points reported: 1, 2, 7, 31 days after catheter removal Secondary outcomes: urinary continence; measured as the percentage of participants with no leakage or 0 pads per day measured by pad weight; median urinary loss (gm/day); rate of contrast extravasation in cystograms. Time points measured: 1, 2, 7, 30 days after catheter removal; 1, 2, 7, 30 days after catheter removal; at catheter removal Time points reported: 1, 2, 7, 31 days after catheter removal; 1, 2, 7, 30 days after catheter removal; n/a Subgroup: none
Funding sources	Vattikuti Urology Institute
Declarations of interest	Intuitive Surgical
Notes	Protocol: none found Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “Randomization was performed using a remotely located random number generator.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Low risk	Quote: “A telephone call to the study coordinator was made by the operating room nurse after urethral transection to learn the type of anastomosis to be performed.” Judgement: appropriate method of allocation concealment was used.
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “The patients and the independent database manager were blinded to the urethrovesical anastomosis technique. The surgeon was blinded to the technique until the urethral transection and the support staff involved in postoperative care was blinded regarding the treatment.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Low risk	Quote: “The patients and the independent database manager were blinded to the urethrovesical anastomosis technique. The surgeon was blinded to the technique until the urethral transection and the support staff involved in postoperative care was blinded regarding the treatment.” Judgement: participants, database manager, and postoperative care team blinded to treatment.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “The patients and the independent database manager were blinded to the urethrovesical anastomosis technique. The surgeon was blinded to the technique until the urethral transection and the support staff involved in postoperative care was blinded regarding the treatment.” Judgement: outcome assessor was blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Low risk	Quote: “The data were collected by an independent database manager who had no knowledge of the anastomotic technique.” Judgement: outcome assessor was blinded.
Incomplete outcome data (attrition bias) Urinary continence	Low risk	Judgement: 57/57 (100%) and 59/59 (100%) participants in intervention and standard group at 1 month follow‐up, respectively, included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	Unclear risk	Judgement: 46/57 (80.7%) and 50/59 (84.7%) participants in intervention and standard group at 1 month follow‐up, respectively, included in analysis.
Selective reporting (reporting bias)	Unclear risk	Judgement: study outcomes were well predefined and described, but protocol was not found.
Other bias	High risk	Quote: “Interim analysis was done when 58 patients were entered into the study with the intention of terminating the study early.” Judgement: while the study reported non‐significant results and the author used a low P value threshold at the interim analysis, multiple views of the data may represent another source of bias.

Pushkar 2016.

Study characteristics
Methods	Study design: prospective, single‐blinded, parallel randomized controlled study Number of centers: single center Country: Italy Dates of the study: January 2012 to June 2014
Participants	Inclusion criteria: cT1c‐cTcNoM0 pathologically proven prostate cancer; ≥ 40 and ≤ 75 intent to undergo RARP; informed consent signed; no urinary incontinence pre‐surgery Exclusion criteria: prior hormone therapy, prior TURP, salvage prostatectomy Total number of participants randomly assigned: 398 Group A Number of all participants randomly assigned: 201 Age (years): n/a PSA: n/a Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: n/a Group B Number of all participants randomly assigned: 197 Age (years): n/a PSA: n/a Comorbidity: n/a Gleason score: n/a cStage: n/a Day of catheter removal: n/a
Interventions	Group A: posterior reconstruction of the rhabdosphincter Group B: standard RARP Follow‐up (including intervention duration): up to 12 months after treatment
Outcomes	Primary outcomes: urinary continence How measured: EPIC questionnaire 5 (no pad use) Time points measured: 1, 2, 3, 6, and 12 months after surgery Time points reported: 1, 2, 3, 6, and 12 months after surgery Secondary outcomes: potency recovery (IIEF‐5 scores); biochemical recurrence 12 months; positive surgical margins; serious adverse events; anastomotic time. Time points measured and reported: 12 months after surgery; 12 months after surgery; within one week after surgery; up to 30 days after surgery; n/a. Subgroup: none
Funding sources	None
Declarations of interest	None
Notes	Abstract only Protocol: none found Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “computer generated sequence” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Unclear risk	Judgement: not described.
Blinding of participants and OR personnel (performance bias)	High risk	Judgement: not feasible to blind surgeon and OR personnel.
Blinding of non‐OR personnel (performance bias)	High risk	Judgement: author reply: participants were blinded, outcome assessors were not blinded. Judgement: non‐OR personnel were not blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Judgement: patients were blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely to be affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	High risk	Judgement: outcome assessor were not blinded
Incomplete outcome data (attrition bias) Urinary continence	High risk	Judgement: 159/201 (79.1%) and 156/197 (79.2%) participants in intervention and standard group at 12 months follow‐up, respectively, included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	High risk	Judgement: 159/201 (79.1%) and 156/197 (79.2%) participants randomized in experimental and control group were included in analysis, respectively.
Incomplete outcome data (attrition bias) Potency	Unclear risk	Judgement: 85/103 (82.5%) and 83/101 (82.2%) participants randomized in experimental and control group were included in analysis, respectively.
Incomplete outcome data (attrition bias) Oncological/QOL outcomes	High risk	Judgement: 159/201 (79.1%) and 156/197 (79.2%) participants randomized in experimental and control group were included in analysis, respectively
Selective reporting (reporting bias)	Unclear risk	Judgement: study outcomes were well predefined and described, but protocol was not found.
Other bias	Unclear risk	Judgement: abstract only; author reply: never published finalized manuscript.

Salazar 2021.

Study characteristics
Methods	Study design: parallel designed randomized controlled study Number of centers: single center Country: Spain Dates of the study: January 2017 to December 2019
Participants	Inclusion criteria: Age ≥ 18 and ≤ 80 years; histological confirmation of prostate cancer; localized or locally advanced prostate cancer; informed consent signed Exclusion criteria: presence of urinary incontinence prior to the procedure; previous radiation therapy of the prostate or pelvis; presence of any prostatic surgery prior to the procedure; prior medical history of psychiatric diseases or drug addiction; any condition that contraindicates a radical prostatectomy. Total number of participants randomly assigned: 148 Group A Number of all participants randomly assigned: 81 Age (years): 64.5 PSA: 7.42 ng/mL Comorbidity: BMI: 25.93 (2.97) Gleason score: n/a cStage: T1c: 60; T2a‐b: 11; T2c: 1; T3a: 0 Day of catheter removal: n/a Group B Number of all participants randomly assigned: 72 Age (years): 64.05 PSA: 6.78 ng/mL Comorbidity: BMI: 26.57 (3.06) Gleason score: n/a cStage: T1c: 63; T2a‐b: 15; T2c: 0; T3a: 3 Day of catheter removal: n/a
Interventions	Group A: posterior reconstruction of the rhabdosphincter Group B: standard RARP Follow‐up (including intervention duration): up to 12 months after treatment
Outcomes	Primary outcomes: urinary continence at 1 and 6 months after prostatectomy How measured: participant response of 0 to question 3 of the EPIC26 questionnaire urinary domain Time points measured: 1 week, 1 month, 3 months, 6 months, 12 months Time points reported: 1 week, 1 month, 3 months, 6 months, 12 months Secondary outcomes: Urinary continence measured in grams, urinary continence at 12 months, potency quality of life, biochemical recurrence, positive surgical margins, pathological pTNM in prostatectomy specimens How measured: response of 0 to question 3 of the EPIC26 questionnaire urinary domain; SHIM assessment; (EPIC‐26, ICIQ‐SF, IPSS) Time points measured: 1, 6, 12 months Time points reported: 1, 6, 12 months Subgroup: none stated
Funding sources	None
Declarations of interest	None
Notes	Protocol: NCT03302169 Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “computer randomization (random.org)” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Unclear risk	Judgement: not described.
Blinding of participants and OR personnel (performance bias)	High risk	Judgement: not feasible to blind surgeon and OR personnel.
Blinding of non‐OR personnel (performance bias)	Low risk	Quote: “Patients and data gatherers were blinded to treatment allocation.” Judgement: non‐OR personnel was blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “Patients and data gatherers were blinded to treatment allocation.” Judgement: Patients were blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely to be affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Low risk	Quote: “Patients and data gatherers were blinded to treatment allocation.” Judgement: outcome assessor was blinded.
Incomplete outcome data (attrition bias) Urinary continence	Low risk	Judgement: 77/81 (95.1%) and 70/72 (97.2%) participants in intervention and standard group at 12 months follow‐up, respectively included in analysis.
Incomplete outcome data (attrition bias) Serious adverse events	Low risk	Judgement: all participants who were randomized were included in analysis.
Incomplete outcome data (attrition bias) Oncological/QOL outcomes	Low risk	Judgement: all participants who were randomized were included in analysis.
Selective reporting (reporting bias)	Low risk	Judgement: protocol was published and study outcomes were well predefined and described.
Other bias	Low risk	Judgement: not detected

Sutherland 2010.

Study characteristics
Methods	Study design: parallel randomized controlled study Number of centers: 1 center, 3 surgeons Country: USA Dates of the study: January 2008 to June 2008
Participants	Inclusion criteria: clinically localized T1– T2NxMx or N0M0 biopsy proven prostate cancer, Eastern Cooperative Oncology Group 0 performance status, life expectancy that exceeded 10 years Exclusion criteria: Inability to give informed consent, prostate volume greater than 200 gm on trans‐rectal ultra‐sonography, prior history of urinary incontinence or bladder dysfunction (e.g. interstitial cystitis, overactive bladder, detrusor areflexia etc.), prior pelvic radiation, cryoablation or hormonal deprivation Total number of participants randomly assigned: 94 Group A Number of all participants randomly assigned: 47 Age (years): 59.9 Comorbidity: low: 23; intermediate: 18; high: 6 PSA: 4.9 ng/mL Gleason score: 6 = 1.5%; 7 (3 + 4) = 68.7%; 7 (4 + 3) = 14.9%; 8 = 9.0%; 9 to 10 = 6.0% cStage: T1c: 39; T2: 8 Day of catheter removal: n/a Group B Number of all participants randomly assigned: 47 Age (years): 60.4 Comorbidity: low: 19; intermediate: 19; high: 9 PSA: 5.6 ng/mL Gleason score: 6 = 3.1%; 7 (3 + 4) = 64.1%; 7 (4 + 3) = 26.6%; 8 = 3.1%; 9 to 10 = 3.1% cStage: T1c: 37; T2: 10 Day of catheter removal: n/a
Interventions	Group A: posterior reconstruction RARP Group B: standard RARP Follow‐up (including intervention duration): 3 months
Outcomes	Primary outcome: continence recovery at postoperative month 3 How measured: a score of 0 or 1 in question 5 of the EPIC urinary domain Time points measured: 3 months post surgery Time points reported: 3 months post surgery Secondary outcomes: self‐reported daily leaks; self‐perception of urinary function; IPSS; 24‐hour pad weight. How measured: EPIC question 1; EPIC question 5; IPSS; 24‐hour pad weight. Time points measured: 3 days, 6 weeks, 3 months Time points reported: 3 days, 6 weeks, 3 months Subgroup: none
Funding sources	None stated
Declarations of interest	Vivus and TIMM Medical
Notes	Protocol: none found Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: “Computer randomization was achieved through the link, https://www.random.org/integers/ that generates an integer between 1 and 2 for each subject.” Judgement: appropriate method of random sequence generation used.
Allocation concealment (selection bias)	Low risk	Quote: “surgeon was notified of patient randomization after completion of the prostatectomy.” Judgement: appropriate method of random sequence generation used.
Blinding of participants and OR personnel (performance bias)	High risk	Quote: “A transperitoneal RARP approach was used in each case and the surgeon was notified of patient randomization after completion of the prostatectomy... Patients blinded to group allocation.” Judgement: OR personnel were not blinded.
Blinding of non‐OR personnel (performance bias)	Low risk	Quote: “Patients blinded to group allocation” and author reply: nurses were blinded to allocation status. Judgement: caregivers were blinded.
Blinding of outcome assessment (detection bias) Subjective outcomes; participant self‐assessed	Low risk	Quote: “patients were blinded to group allocation.” Judgement: Patients were blinded.
Blinding of outcome assessment (detection bias) Objective outcomes	Low risk	Judgement: objective outcomes are unlikely affected by lack of blinding.
Blinding of outcome assessment (detection bias) Subjective outcomes; investigator‐adjudicated	Low risk	Quote: “information was collected and recorded by the nursing staff...” Judgement: author reply: nurses were blinded to allocation status.
Incomplete outcome data (attrition bias) Urinary continence	Low risk	Judgement: 46/47 (97.8%) and 41/47 (87.2%) participants in intervention and standard group at 3 months follow‐up, respectively, included in analysis.
Selective reporting (reporting bias)	Unclear risk	Judgement: study outcomes were well predefined and described, but protocol was not found.
Other bias	Low risk	Judgement: not detected.

ASA: Anesthesiologists Physical Status Score
BMI: Body mass index
CVD: Cardiovascular disease
DM: Diabetes mellitus 
HTN: Hypertension 
ICIQ‐SF: International Consultation on Incontinence Questionnaire‐ Short Form
IPSS: International prostate symptom score 
n/a: not applicable
OR: operating room
PSA: Prostate specific antigen 
pTNM: Pathological Tumor‐Node‐Metastasis
QoL: quality of life
RARP: Robot‐Assisted Radical Prostatectomy 
SHIM: Sexual Health Inventory for Men 
TURP: Transurethral resection of the prostate 
UCLA‐PCI: UCLA Prostate Cancer Index

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Anceschi 2013	Route of administration does not meet inclusion criteria.
Boylu 2009	Study design does not meet inclusion criteria.
Checcucci 2019a	Study design does not meet inclusion criteria.
Coelho 2011	Study design does not meet inclusion criteria.
Salvaggio 2009	Route of administration does not meet inclusion criteria.

Differences between protocol and review

We were unable to obtain time‐to‐event type data for the outcome of biochemical recurrence‐free survival. We therefore performed this analysis using the reported risk ratios as effect size measure.  

The published protocol of this review (Rosenberg 2020) mistakenly identified eight (instead of the recommended maximum of seven) outcomes to be included in the summary of findings table. The outcome of biochemical recurrence‐free survival (which is a surrogate outcome) was since removed.

Contributions of authors

Joel Rosenberg (JR): protocol drafting, search strategy development, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation, review drafting, future review updates.

Jae Hung Jung (JHJ): protocol drafting, search strategy development, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation, review drafting, future review updates.

Hunju Lee (HL): protocol drafting, search strategy development, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation.

Solam Lee (SL): protocol drafting, search strategy development, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation.

Caitlin Bakker (CB): search strategy development.

Phillip Dahm (PD): protocol drafting, search strategy development, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation, review drafting, future review updates.

Sources of support

Internal sources

• Minneapolis VAMC , USA

  Salary support for Philipp Dahm

• University of Minnesota , USA

  Salary support for Caitlin Bakker

External sources

• None, USA

  No sources of support

Declarations of interest

JR: none known

JHJ: none known

HL: none known

SL: none known

CB: none known

PD: none known

Edited (no change to conclusions)