Robotic vs. Retropubic radical prostatectomy in prostate cancer: A systematic review and a meta-analysis update

CONTEXT The safety and feasibility of robotic-assisted radical prostatectomy (RARP) compared with retropubic radical prostatectomy(RRP) is debated. Recently, a number of large-scale and high-quality studies have been conducted. OBJECTIVE To obtain a more valid assessment, we update the meta-analysis of RARP compared with RRP to assessed its safety and feasibility in treatment of prostate cancer. METHODS A systematic search of Medline, Embase, Pubmed, and the Cochrane Library was performed to identify studies that compared RARP with RRP. Outcomes of interest included perioperative, pathologic variables and complications. RESULTS 78 studies assessing RARP vs. RRP were included for meta-analysis. Although patients underwent RRP have shorter operative time than RARP (WMD: 39.85 minutes; P < 0.001), patients underwent RARP have less intraoperative blood loss (WMD = -507.67ml; P < 0.001), lower blood transfusion rates (OR = 0.13; P < 0.001), shorter time to remove catheter (WMD = -3.04day; P < 0.001), shorter hospital stay (WMD = -1.62day; P < 0.001), lower PSM rates (OR:0.88; P = 0.04), fewer positive lymph nodes (OR:0.45;P < 0.001), fewer overall complications (OR:0.43; P < 0.001), higher 3- and 12-mo potent recovery rate (OR:3.19;P = 0.02; OR:2.37; P = 0.005, respectively), and lower readmission rate (OR:0.70, P = 0.03). The biochemical recurrence free survival of RARP is better than RRP (OR:1.33, P = 0.04). All the other calculated results are similar between the two groups. CONCLUSIONS Our results indicate that RARP appears to be safe and effective to its counterpart RRP in selected patients.


INTRODUCTION
Prostate cancer (PCa) is the most common cancer in the worldwide and its morbidity,mortality is the first and second common cancer in men, respectively [1]. RP is the standard therapy for patients with localized PCa [2]. However, open retropubic radical prostatectomy (RRP) is associated with higher overall complications, including estimated blood loss (EBL), wound infections. With the development of surgical techniques, laparoscopic techniques and robot assisted surgeries have become a very popular procedure for the management of urological disease throughout the world [3]. Compared with RRP, the advantages of laparoscopic radical prostatectomy (LRP) are less EBL, fewer complications, better cosmetic effect and shorter hospital stay [4]. The disadvantages of LRP is lack of 3D visualization and poor ergonomics.
As alternatives to open surgery, RARP has became a predominant procedure for the treatment the localized prostate cancer in the world [5]. Assessing of the robotic Clinical Research Paper www.impactjournals.com/oncotarget surgery by expert indicate better ergonomics and quicker learning curve, but its shortage is high cost of the robotic surgery system.
In recent years, many experts have reported on comparative study of RARP and open RRP. And some meta-analysis were performed to evaluate the advantages and disadvantages of two approaches, including perioperative outcomes, oncologic outcomes [5]. Their early experience showed that the outcomes of this approach with fewer overall complications, quicker convalescence, and lower EBL and transfusion [5][6][7]. However, the outcomes of RARP compared with RRP have not been fully evaluated, and no conclusive results are available. Therefore, a systematic review and metaanalysis of the included published studies was performed to compare RARP with RRP.

Characteristics of eligible studies
According to search strategy, the included 78 studies [4, assessing RARP vs. RRP met the inclusion criteria and were applied to perform this meta-analysis ( Figure 1). Those studies include forty-three retrospective and thirty-five prospective studies and were listed in Table  1.
Quality of the studies and level of evidence (Table 1) In this meat-analysis, the Newcastle-Ottawa Scale quality assessment method of the observational studies [86], and the US Preventive Services Task Force grading system [87] were applied to evaluate the quality of included studies. Twenty studies scored seven stars and were evaluated as the high quality studies. Additionally, The clinical variables of RARP and RRP were extracted independently from included literatures ( Table 1).

DISCUSSION
The incidence of prostate cancer and its mortality is the first and the second common cancer in man, respectively [1]. Our results indicated that RARP seemed to have an younger age (WMD: -1.00; P < 0.001), and to have the lower level of pre-PSA (WMD: -0.93; P < 0.001) than RRP group, and that these differences are primarily due to surgeon's preference for surgical modality. Another reason is that the younger is more easier to choose new approach. However, there is no difference on BMI and prostate volume between the two groups. Sensitivity analysis showed that there was no change in the significance of any other outcomes except that readmission rate(P = 0.002 vs P = 0.13) and BCR for free survival(P = 0.04 vs. P = 0.55).It demonstrated that selection bias of demographic and clinical data of patients is small between two groups.
Novara G et al [6] evaluated oncologic outcomes of RARP and RRP, and the results indicated that RARP had less EBL and transfusion rate than RRP. Their results presented similar results and strengthened our results. The  other analyzed parameters operative time and complication rate were similar. However, in our meta-analysis, RARP had longer operative time than RRP(WMD:39.85min, P<0.001), which likely reflects the early learning curve with RARP. But the learning curve indicated that operative time was decreased with growing operative experience and it won't influenced operative outcomes [88].
With regard to the pathologic outcomes, patients underwent RARP had more pathological Gleason score = 7, less pathological Gleason score ≥8, higher lymph node   yield and fewer positive lymph node than RRP. However, the pathological T stage is no significant difference between the RARP and RRP group. LN yield was deemed an indicator of surgical quality by many surgeons [89]. RARP had a higher LN yield than RRP, the reason is that RARP has meticulous dissection with 3D vision and decrease the intraoperative blood loss which made the surgeon have more time and patience to acquiring higher LN yield. Therefore, the oncological outcomes in terms of PSM for T3 is higher in RARP than RRP. With the results that BCR free survival was higher in RARP than RRP. Some studies showed that the predictors of BCR were preoperative PSA. Gleason score, pathological stage, and PSM [49].
The experts suggested that patient outcomes and surgical approach were mainly required to improve for an accurate characterization of complications [90]. In our meta-analysis, Patients underwent RARP had fewer overall complications than RRP. The possible reason may be associated with lower EBL and less transfusion rate in RARP. Then a comprehensive classification of complications indicated that RRP had a higher incidence of rectal injury, pulmonary embolism, wound infections, bladder neck contracture, urinary retention, deep venous thrombosis, urinary leakage, lymphocele, and obturator nerve injury. There were no significant differences with regard to ileus and UTI between two groups.
Ficarra V et al [91] compared RARP with RRP with respect to 12-mo urinary continence. Their results indicated that RARP had a better 12-mo urinary continence recovery than RRP(OR:1.53; P = 0.03). However, our results indicated that there were no statistical differences with regard to 3-mo and 12-mo urinary continence in two groups. The urinary continence receiving RP is influenced by preoperative patient characteristics, surgical techniques, and so on. Some studies found that patient age [92,93], BMI [94], comorbidity index [95], and prostate volume [96,97] were also the potential predictors of urinary incontinence. Increasing age, higher BMI, and large prostate volume are correlated with high risk of urinary incontinence who underwent RP. However, the 3-and 12-mo potent recovery rate of RARP was also better than RRP group, respectively. Analysis of predictors indicated that peroperative parameters might influence potency results. Relevant predictors included age at surgery, baseline erectile function, and comorbidities [98]. Other authors also confirmed that age and baseline erectile function of patients were affected the potent recovery in nerve-sparing RARP [93,99].
On the other hand, we found better BCR free survival and lower readmission rate in RARP group in the original analysis. The reason is that meticulous dissection, lower blood loss and complications might provide patients better oncologic prognosis in RARP group. However, we observed no statistical differences between RARP and RRP in sensitivity analysis. Therefore, multicenter, large sample, long follow-up RCTs are required to prove our findings.
Nevertheless, there were several limitations when analyzing and interpreting results in our meta-analysis. The major limitation is lack of well designed prospective, randomized control studies in our meta-analysis. Secondly, there existed heterogeneities of studies, especially in the comparing of the continuous data such as the length of hospital stay, operative time. whereas these parameters were influenced by the heterogeneities of patients' conditions, surgeon's surgical skills and the sample size of studies. In addition, short follow-up duration may have an influence on the confidence of outcomes. In the future, well-designed, prospective, multicenter randomized control studies are required to help us better demonstrate the advantages as well as drawbacks of this novel approach.

Literature search strategy
To update previous systematic review [5-7, 91, 98, 100, 101], a systematic review of published literature was performed according to the Cochrane Handbook recommendations [102]. No ethic issues get involved in this article. A systematic dissertion was conducted using Medline, Embase, Pubmed, CNKI, and all relevant studies had been identified by the Cochrane Library. The following key words were used: "comparative studies", "retropubic", "open", "radical prostatectomy ", "Da Vinci", "robot-assisted", and "prostate cancer".

Data extraction and outcomes of interest
Two of the authors(JKH and TK) extracted data from the selected studies including: author identification, country, publication year, study design, age, No. of patients, operative approaches were mentioned previously, and results of intervention. All disagreements about eligibility were reached a consensus through authors discussion. Perioperative outcomes including operative time, EBL, LOS, overall complications, and oncological outcomes were compared between the two methods from all the studies that were finally selected. Overall complications were graded on the basis of the Clavien-Dindo system [103].

Inclusion criteria and exclusion criteria
Studies should satisfy the following requirements: (1) to compare RARP with RRP, (2) to display on outcome of two approaches, (3) to document the surgery as RARP or RRP, (4) to clearly document indications for prostatectomy with prostate cancer. Studies will be excluded if (1) the study was not satisfied inclusion criteria or (2) the outcomes of literature were not mentioned or the parameters were impossible to analysis for either RARP or RRP from the published findings and (3) studies focusing on pure robot surgery system and/or on singlesite techniques.

Study quality assessment and level of evidence
In accordance with the criteria of Centre for Evidence-Based Medicine in Oxford, we evaluated the level of evidence(LOE) of included sixteen studies. The Jaded Score was applied to evaluated the methodological quality of RCTs [104]. The Newcastle-Ottawa Scale(NOS) was applied to assessed the methodological quality of non-RCTs observational studies [86,105]. Two authors(JKH and TK) evaluated the quality of the studies and discrepancies were rechecked by the third reviewer(CZQ) and consensus was achieved by discussion.

Statistical analysis
All meta-analysis were conducted by Review Manger 5.3(Cochrane Collaboration, Oxford, UK). Continuous and dichotomous variables were calculated by weighted mean differences (WMDs) and odds ratios(ORs). All analysis results were reported with 95% confidence intervals(CIs). I 2 test and chi-square-based Q test were applied to evaluated the quantity of heterogeneity, and when I 2 > 50%, the evidence was considered to have substantial heterogeneity, the random-effects(RE) model would be applied, otherwise, the fixed effects(FE) model was applied. The presence of publication bias was evaluated by Egger's test and funnel plot. Sensitivity analysis was used to estimate the influence of studies with a high risk of bias on the overall effect.

CONFLICTS OF INTEREST
The authors have no conflict of interest to disclose.