Introduction

Learning curve (LC) is considered the period during which there is an improvement of surgical performance, referring to a process of gaining knowledge/improving skills in performing specific surgical tasks/procedures, and is generally quantified by the minimum number of cases needed for a surgeon to perform an intervention in a proficient way [1]. Nevertheless, there is no evidence-based definition or standardized platform to measure LCs of the various surgical procedures and there is a great reporting variability, depending among others on the setting of the evaluating studies [2, 3]. LC is related to several surgeon- and patient-dependent factors and, therefore, specific LCs may vary among surgeons and may also be affected by the investigated outcome measure [1, 2]. High-volume compared to low-volume centers are generally associated with more favorable surgical outcomes [4, 5], but outcomes can be also different among high-volume surgeons within the same hospital since proficiency level is not similar among all surgeons [6] and even experienced ones may repeat similar errors [7]. Studies investigating surgical LCs are becoming increasingly important, since LCs may have substantial impact on surgical metrics, clinical outcomes and cost–benefit decisions [8]. In line with other surgical specialties, LCs of urological procedures such as laparoscopic and robotic-assisted prostate surgery have been a matter of continuous debate and there appears to be no standard definition of the minimum number of cases to achieve optimal outcomes [1, 2, 8].

Systematic reviews addressing LCs of laparoscopic and robotic-assisted prostate surgery are scarce [1, 8]. The aim of the present work is to provide a comprehensive overview of LCs in laparoscopic and robot-assisted prostate surgery based on a systematic search and review of the available literature.

Methods

Medline Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) 1946 to January 30, 2021 were systematically searched to detect all relevant studies in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement [9], based on the following literature search strategy: (laparoscopy OR robot) AND (prostate OR prostatectomy) AND (learning curve). The reference lists of selected studies, systematic reviews and meta-analyses were screened for other potentially eligible studies. After excluding citations in abstract form and non-English citations, titles/abstracts of full papers were screened for relevance, defined as original research studies reporting on the LC in laparoscopic radical prostatectomy (LRP), laparoscopic simple prostatectomy (LSP), robot-assisted radical prostatectomy (RARP) and robot-assisted simple prostatectomy (RSP). Review articles, editorial letters comments, bulletins and studies describing non-technical skills were excluded. Two review authors (NG and IZ) independently scanned the titles, abstracts or both of every record retrieved, to determine which studies should be further assessed and extracted all data. Disagreements were resolved through consensus or after consultation with a third review author (CM).

Results

In total, 47 studies were included for qualitative synthesis (Fig. 1) evaluating a single technique (LRP, RARP, LSP, RSP; 45 studies) or two techniques (LRP and RARP; 2 studies). Seventeen studies (four prospective and 13 retrospective) including 19,681 patients investigated the LC in LRP (Table 1) [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26]. Thirty studies (13 prospective and 17 retrospective) including 30,822 patients investigated the LC in RARP (Table 2) [23, 25, 27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54]. The study outcomes based on which LC was evaluated included peri-operative outcomes (operative time, estimated blood loss (EBL), length of hospital stay (LOS)), oncological outcomes (positive surgical margin rate (PSM), biochemical recurrence rate (BCR)), safety (overall complication rate) and functional outcomes (urinary continence and potency rates).

Fig. 1
figure 1

Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) flowchart

Full size image
Table 1 Main characteristics and results of studies evaluating LC of LRP
Full size table
Table 2 Main characteristics and results of studies evaluating LC of RARP
Full size table

Laparoscopic radical prostatectomy

Regarding peri-operative outcomes, it was reported that operative time reduces significantly after 100–200 cases with a reported plateau at 40–250 cases. Similarly, EBL was found to be significantly reduced after 100–200 cases with a plateau ranging between 80 and 250 cases. Nevertheless, a single series of 110 cases reported that no EBL plateau was reached. LOS was significantly reduced after 58–200 cases, while a plateau was reported to be reached at 100 cases. Regarding oncological outcomes, significantly lower PSM rates were reported after 50–60 cases, reaching a plateau at 150–350 cases. BCR rate was reported to decrease after 110 cases with a plateau reached at 350 cases. Prior laparoscopic experience decreased the BCR rate by 8% according to a single study. Regarding safety and functional outcomes, it was reported that the complication rate was significantly lower after 55–200 cases with a plateau reached at 100–250 cases. Nevertheless, a single series of 110 cases reported that no plateau was reached for overall complication rate. A significant increase in 12-month urinary continence rate was reported after 70 cases (78.2% vs 65.6%), with a plateau reached at 70–350 cases. No study reported on LC based on potency rate.

Robot-assisted laparoscopic radical prostatectomy

Regarding peri-operative outcomes, it was reported that operative time reduces significantly after 20–300 cases and reached a plateau at 16–300 cases. Similarly, EBL was significantly reduced after 20–300 cases with a plateau reached at 90–250 cases. No statistically significant difference between the case groups were reported in four studies. LOS was found to be significantly reduced after 30–200 cases and according to the results of a single study it reached a plateau at > 25 cases. No statistically significant difference between the case groups were reported in two studies. Regarding oncological outcomes, a significant decrease in PSM rate was reported after 67–80 cases with a plateau reached at 50–400 cases. No statistically significant difference between the case groups were reported by eight studies. BCR rate was reported to decrease after 40 cases and the plateau was 100 cases. After salvage RARP, BCR rate was increased after 90 cases compared to the initial 30 cases (36% vs 23%) according to one study. Regarding safety, it was reported that the complication rate was significantly lower after 20–125 cases and reached a plateau at 50–250 cases. Nevertheless, a single series of 120 cases reported that no plateau was reached for overall complication rate. No statistically significant difference between the case groups were reported in two studies. Regarding functional outcomes, a significant increase in urinary continence rate was reported after 30–70 cases. The plateau for early and 12-month urinary continence was reached at 100 and 100–200 cases, respectively. A large series of 1477 cases reported that no plateau was reached for 12-month urinary continence rate. No statistically significant difference between the case groups were reported by two studies. A significant increase in 12-month potency rates was reported after 80 cases (76.6% vs 60.5%) in one study. Following salvage RARP, an increase in 12-month potency rate from 3.3% in the first 30 cases to 23% after 90 cases was reported in one study. No statistically significant difference between the case groups were reported in two studies.

Laparoscopic retropubic simple prostatectomy and robotic simple prostatectomy

A single retrospective study evaluating the LC of LSP on 11 cases with benign prostate hyperplasia was detected [55]. The evaluation of LC was based on conversion-to-open rate, operative time and EBL. No conversion was performed. Operative time and EBL were reported to be significantly reduced with increasing experience. Similarly, a single retrospective study evaluating the LC of RSP on 120 cases with benign prostate hyperplasia was detected [56]. Operative time was significantly reduced from 162 min in the first 10 cases to 134 min in the last 10 cases. For the first ten cases, hematocrit drop had logarithmic improvement, showing linear transitioning thereafter. Regarding overall complication rate and catheterization time no statistically significant difference between the case groups were reported.

Discussion

To the best of our knowledge, this is the first review providing a comprehensive overview of LCs focusing on laparoscopic and robot-assisted prostate surgery based on a systematic search of the available literature. It can therefore provide a useful guide for urologists starting with these procedures regarding the number of cases needed for reaching proficiency.

RARP was found to be the most widely investigated technique (30 studies), followed by LRP (17 studies), LSP (1 study), and RSP (1 study). According to our findings, the definition and assessment method of LC for laparoscopic and robot-assisted prostate surgery is not well defined with various metrics used among studies. Nevertheless, LCs appear to be steep and continuous with the required number of cases for reaching proficiency showing intra-technical variability and outcome-dependent inter-technical variability. In LRP, the reported LC based on operative time; EBL; LOS; PSM; BCR; overall complication rate and urinary continence rate ranged 40–250, 80–250, 58–200, 50–350, 110–350, 55–250, 70–350 cases, respectively. In RARP, the corresponding ranges were 16–300, 20–300, 25–200, 50–400, 40–100, 20–250, 30–200, while LC for potency rates was 80–90 cases.

The outcomes of radical prostatectomy have become a great topic for discussion. Initially, the outcomes focused on cancer control, urinary continence and potency rates, the so- called trifecta. But such reporting system has failed to meet the increasing expectations of patients that need to undergo radical prostatectomy, especially those that choose less invasive and more advanced treatment options such as LRP and RARP. Therefore, the pentafecta entity has been suggested. This incorporates the trifecta outcomes (BCR, urinary continence and potency), peri-operative complications and surgical margin status. At present, this comprehensive reporting method is utilized to counsel patients prior to offering treatment options for prostate cancer, to estimate the efficiency of the modality and patients’ satisfaction [57].

The most commonly used metrics both in LC-LRP and RARP studies include PSM rate (12 and 18 studies, respectively), operative time (10 and 16 studies, respectively) and EBL (8 and 16 studies, respectively). This may be attributed to the fact that the above outcomes are more easily recorded. Nevertheless, it should be noted that PSM rate can be affected by several factors including nerve preservation and extracapsular disease, while a shorter operative time does not necessarily mean proficiency, especially if not combined with favorable oncological outcomes. Five LRP studies reported on urinary continence outcomes, while urinary continence and potency outcomes were reported in ten and three RARP studies, respectively. Despite the fact that it is generally more difficult to record functional outcomes since they necessitate a longer follow-up, they are considered equally important since they reflect the patient’s quality of life. A recent systematic review focusing on LC in robotic surgery confirmed the outcome reporting heterogeneity among studies, showing that the majority of the studies uses surgical rather than patient-related metrics [2].

Abboudi et al. performed the first systematic review investigating the LCs in various urological procedures and reported a mean operative time plateau after 50–200 cases and PSM rates plateau at 50–1600 cases for surgeons with unknown experience level on RARP [1]. They also investigated experienced laparoscopic surgeons’ outcomes, who achieved acceptable PSM rates after 100–300 cases. In another systematic review investigating LCs in robot-assisted surgery, early urinary continence rate and 12-month urinary continence rate plateau were reported after 100 and 112–541 cases, respectively. BCR and PSM rate plateau was reported to be reached at 100 cases [8]. A literature-based analysis on LC-LRP studies reported diverse results for EBL, operative time and complication rates; PSM and BCR rates were reported to reach a plateau at 200–250 and 150 cases, respectively; urinary continence rates were reported to reach a plateau at 250 cases; no plateau was reported to be reached for potency rates [58].

LC can be affected by various factors (Table 3). Anatomical factors such as prostate volume and pelvis diameter have been reported to be significantly correlated with operative time and EBL in RARP [59]. Furthermore, the complexity of each case and the dexterity of the surgeon play a key role [60]. Most LC studies do not take into account the low case volume urologists who usually have a limited operation time available and are forced to achieve LC with fewer cases. Case volume, institutional resources and technical factors such as instruments with improved dexterity in LRP and double console in RARP may affect LC [45]. Other factors that may affect LC include anesthetists’ and surgical nurses’ level of experience, as well as residents’/fellows’ previous training [61, 62]. Cimen et al. reported that the higher level of experience of the bedside assistants in RARP can significantly shorten operative time, although it does not affect the oncological outcomes [63].

Table 3 Factors affecting LC metrics
Full size table

Furthermore, the role of surgical simulation should be considered and we should compare the LC training with traditional wet- and dry-laboratory methods, because a combination of training environments could result in most effective training [64]. An RCT reported that the tasks performed by residents in the da Vinci Surgical Skills Simulator are correlated with bladder mobilization and urethro-vesical anastomosis during RARP [65]. Three-dimensional printing has also been shown to be feasible in phase II RARP trials [66]. Nevertheless, the elasticity of normal tissue has not been replicated with these models and therefore cost-effectiveness warrants further testing [66]. A systematic review focusing on robotic surgery in urology has shown that mentored training and non-structured pathways combined with traditional laboratory training favorably affect LCs [67]. Surgical curricula including training on technical and non-technical skills, such as surgical cognitive and social skills have also been applied to improve outcomes in laparoscopic and robot-assisted surgery [68]. The role of a training program under supervision appears to result in shortening of the LC and minimization of adverse outcome rates during this period [67, 68]. Nevertheless, LRP seems to involve skills not translated well from radical retropubic prostatectomy [69] and therefore outcome improvements appear to be achieved more slowly in laparoscopic compared to open surgery [12].

A limitation of the present review is the general low quality of the available studies in the international literature and were therefore included. In addition, there was a variety of different metrics making comparison among studies difficult. Furthermore, most studies were performed in high-volume centers by surgeons with various previous level of experience. Last but not least, substantial information on the complexity of each step to overcome the beginning of the LC was lacking in many studies.

Based on the above findings, we recommend standardized data collection and outcome reporting for each procedure to have more accurate estimation of the LCs. Moreover, simulator and mentored training should be applied to shorten the LC, especially in conventional laparoscopy in which the LC is longer compared to robot-assisted procedures. Case selection can also shorten the LC, avoiding demanding nerve-sparing procedures and complex cases such as bulky prostates or cases with extracapsular disease. Finally, experienced surgical assistants can contribute to improved intra-operative outcomes during the LC.

Conclusions

LC for laparoscopic and robot-assisted prostate surgery is not well defined with various metrics used among studies. Nevertheless, LCs appear to be steep and continuous. Implementation of training programs/standardization of the techniques is necessary to improve outcomes.