Comparative assessment of the learning curve of retropubic, laparoscopic, perineal, and robot-assisted radical prostatectomy

Introduction

According to the World Health Organization, prostate cancer ranked third in the structure of the most significant oncological world diseases in 2020 [1]. In Russia, prostate cancer ranks second (14.9%) in the structure of oncological diseases in men after tumors of the trachea, bronchi, and lungs (16.9%) [2]. Radical prostatectomy (RP) in various modifications (retropubic (rpRP), perineal (plRP), laparoscopic (LRP), and robot-assisted (RARP)) remains the most common method of treatment, despite the availability of alternatives in the form of active observation, hormonal and radiation therapy, cryoablation, and others [3]. However, the achievement of optimal oncological and functional results directly depends on the experience and manual skills of the surgeon [4][5]. A prospective multicenter study of LAPPRO (2018) demonstrated satisfactory functional and oncological indicators of surgical treatment for prostate cancer in surgeons who performed more than 100 interventions [6]. Currently, there is no consensus on assessment parameters, as well as the duration of the learning curve (LC) for the four main RP techniques: retropubic, perineal, laparoscopic, and robot-assisted ones [7][8]. When analyzing scientific publications, specialists concluded that there is a shortage of studies devoted to the problems of mastering this surgery, and most of them are devoted to the robotic technique [9][10]. The attempts to calculate metric indicators are complicated by the questionable quality of the reported results, the impossibility of statistical processing, and the lack of consensus on terminology [11].

The study aimed to analyze the nature and duration of the radical prostatectomy learning curve performed by retropubic, perineal, laparoscopic, and robot-assisted techniques.

Materials and methods

The study included 400 patients (4 groups of 100 patients each) with locally advanced prostate cancer (T1c-T3bN0M0) who underwent LRP, rpRP, plRP, or RARP performed by a single surgeon during the period 2011-2021. The paper evaluated perioperative (surgery time, volume of blood loss, duration of catheterization, duration of hospitalization, complications), functional (continence and erectile function), and early oncological (positive surgical margin (PSM), survival) results.

To evaluate the LC, each group was divided into 4 equal subgroups according to the number of operations performed in chronological order (A, B, C, D). For example, in the context of rpRP, it looked like this: A — the 1^st 25 surgeries (1 – 25), B — the 2^nd 25 surgeries (26 – 50), C — the 3^rd 25 surgeries (51 – 75), D — the 4^th 25 surgeries (76 – 100). The results were evaluated in total for the whole group and separately for each subgroup. This approach is used to calculate the number of operations required performed, ensuring the achievement of satisfactory perioperative, functional, and early oncological results with a minimum frequency of intraoperative and postoperative complications (Clavien-Dindo I – II).

Statistical analysis. Statistical data processing was implemented using the Microsoft Office Excel application (Microsoft Corp., Redmond, WA, USA) and the Prizm for Windows ver. 9.1.0 program (GraphPad Software Inc., Graphpad Holdings, LLC, San Diego, CA, USA). Quantitative data were necessarily checked for compliance with the Shapiro-Wilk and Kolmogorov-Smirnov criteria. In the case of compliance, parameters such as the mean value (M) ± standard deviation (a) were used as a criterion. If the test result was negative, the median [Me] and the quartiles [Q25; Q75] were used. Qualitative data were characterized using the frequency index (%). Differences were considered statistically significant at p < 0.05 (95% significance level). The Kaplan-Meier method was used to assess overall and relapse-free survival.

Results

According to the main indicators (age, body mass index, gland volume, urodynamic, and oncological characteristics), there were no statistical differences between the groups (p > 0.05). In the plRP group, patients with less aggressive TNM stages and a low risk of progression according to the D'Amico classification prevailed (71.0%). The indication for expanded pelvic lymph node dissection was the risk of invasion >5.0%, according to the Briganti nomogram. By restoring continence, the researchers meant the use of one pad per day, restoration of erectile function, sufficient rigidity for introjection with/without phosphodiesterase-5 inhibitors.

The average duration of surgery in the rpRP group was 172 ± 42 minutes, in the LRP group — 177 ± 50 minutes, in the plRP group — 101 ± 23 minutes, and in the RARP group — 119 ± 43 minutes. The stabilization of indicators in the rpRP group required 50 – 75 interventions, in the LRP group — 75 – 100, in the plRP group — 25 – 50, and in the RARP group — 25 – 50. The median blood loss in the rpRP group was 410 [ 162; 500] ml, in the LRP group 145 [ 100; 200] ml, in the plRP group — 80 [ 50; 100] ml, and in the RARP group — 60 [ 50;100] ml. The plateau was reached with rpRP and LRP after 50 – 75 surgeries, with plRP and RARP — after 25 – 50 surgeries. The median duration of urethral catheter drainage in rpRP was 14 [ 12; 14] days, in LRP — 14 [ 12; 14] days, in plRP — 12.5 [ 11.0; 14.0] days, and in RARP – 7 [ 7; 10] days. Stabilization of this indicator was observed in the rpRP, LRP, and plRP groups after 50 – 75 interventions, in RARP – 25–50. The median hospital stay in the rpRP group was 13 [ 11; 14] days, in LRP — 7 [ 5; 8] days, in plRP — 10 [ 8; 11] days, and in RARP — 4.5 [ 3; 6] days. Reaching the plateau was observed in the rpRP, LRP, and plRP groups after mastering 50–75 surgeries, and in the RARP group — after 25 – 50 surgeries.

In the rpRP group, complications developed in 21 (21.0%) patients, in LRP — in 9 (9.0%), in plRP — in 6 (6.0%), in RARP — in 4 (4.0%). The highest number of complications were observed in groups I and II (60.0 – 86.7%). Stabilization of this indicator was observed after 25–50 operations in the LRP, plRP, and RARP groups, after 50 – 75 with rpRP. The function of urinary retention by the end of the 12^th month of the year after surgery was restored in the rpRP, LRP, plRP, and RARP groups in 80.0%, 84.0%, 88.0%, and 96.0%, respectively. Stabilization of the results was observed with rpRP, LRP, and plRP after 50 – 75 interventions, with 25 – 50 in the RARP group. Erectile function recovery was achieved in the rpRP, LRP, and RARP groups in 75.0%, 79.0%, and 92.0%, respectively. The plateau was achieved after 50 – 75 interventions in the rpRP and LRP groups, with RARP — 25 – 50. The attempts to perform nerve-sparing in the plRP group were unsuccessful due to the technical features of perineal access.

The PSM frequency in the rpRP group was 29.0%, LRP — 23.0%, plRP — 14.0%, and RARP — 18.0%. The plateau was reached with rpRP after 50 – 75 interventions, with LRP, plRP, and RARP — after 25 – 50. The 24-month relapse-free survival rates in the rpRP, LRP, plRP, and RARP groups were 91.4%, 88.6%, 98.5%, and 93.9%, respectively. The cancer-specific 24-month survival rate in all groups was 100.0%; the overall survival rate was 97.3% for LRP, 98.6% for plRP, 98.9% for rpRP, and 100.0% for RARP. In total, the results of the evaluation of the LC of RP by four main techniques are presented in Table 1.

Discussion

There are currently four main techniques that are actively used: retropubic, perineal, laparoscopic, and robot-assisted. The latter, according to a large systematic review and meta-analysis conducted by Du et al. (2018), has advantages in the context of the blood loss volume, frequency of blood transfusion, nerve-sparing, hospital stay, restoration of urinary retention functions, and potency [12]. Moreover, RARP is characterized by the shortest duration of LC and may soon become the "gold standard" for surgical treatment of prostate cancer [13]. Next, it is necessary to observe the dynamics of the main metric indicators during the development of surgical skills.

Surgery duration. The duration of the surgery certainly depends on the surgeon's experience and decreases significantly as the technique is mastered. RARP not only demonstrates advantages in the context of this indicator, but also a short LC in 25 – 50 surgeries. In contrast, LRP is the most difficult in terms of manual skills, which naturally leads to an increase in surgical intervention time, as well as the duration of LC of 75 – 100 surgeries. Similar data were obtained in a large systematic review by Groves et al. (2022): the surgery time in the context of LRP varies from 100 to 250 cases [13].

Blood loss volume and frequency of blood transfusion. Obviously, rpRP is the most dangerous method from the point of view of massive bleeding [14]. In this study, plateau output was observed in the rpRP and LRP groups after performing 50 – 75 surgeries, in the plRP and RARP groups — 25 – 50 surgeries. At the same time, it is necessary to consider the absence of nerve-sparing and pelvic lymphadenectomy in plRP, which, in the authors’ opinion, puts RARP in the first place in the rank of LC duration. Rasner et al. (2016) also demonstrate the advantages of the robotic technique in minimizing blood loss, and stabilization of the parameter is observed after the first 50 surgeries [15].

Duration of urethral catheter drainage. The issue of bladder drainage is inextricably related to the quality of vesicourethral anastomosis. This stage is considered one of the most difficult, especially with LRP, and its reliability directly determines the pace of recovery from continence [16, 17]. The results of the current study show the stabilization of this indicator with rpRP, plRP, and LRP after 50 – 75 surgeries. On the other hand, the ability to perform apical dissection with proximal urethra allocation as delicately and "sighted" as possible, a robotic imitation of the human wrist, as well as 100.0% use of self-fixing suture material with RARP allowed reaching a plateau after 25 – 50 interventions. The works by Nyushko et al. (2021) and van Poppel et al. (2019) report similar data [18][19].

Hospital stay. In this study, the minimum period of hospitalization was noted at RARP as 4.5 days. The data obtained correlate with the results of Russian and foreign specialists. In the work by Simsir et al. (2021), the average number of hospital bed-days was 3.5 [20]. Stabilization of this indicator was observed in the rpRP group after 50 – 75 surgeries, in LRP — after 50 – 75 surgeries, in plRP — after 50 – 75 surgeries, and in RARP — after 25 – 50 surgeries. On the other hand, according to Slyusarenko et al. (2020), the LC of RARP in the context of this indicator is quite flat: 12 for the first 49 cases, 11 for 50 – 88, and 10 for 89 – 145 surgeries [21].

Complications. The number of surgeries performed directly correlates with the risk of serious complications. Frequency stabilization is possible, according to the authors’ data, after 25 –50 surgeries with plRP, LRP, and RARP. The retropubic technique requires 50–75 interventions, which has a few explanations: deep location of the gland, lack of necessary enlargement, less precise work with the dorsal venous complex, etc. Slyusarenko et al. showed the dynamics of a decrease in the frequency of complications during the LC of RARP: 26.5% in groups I – II and 16.3% in group III during the first 49 surgeries; 50 – 88 surgeries: 23.1% in groups I – II and 5.1% in group III; 89 – 145 surgeries: 7.0% in groups I – II, and 3.5% in group III, respectively [21]. On the other hand, at the stages of mastering perineal modification, it is necessary to remember the high risk of injury to the intestine. RARP is the safest technique, which, in the authors’ opinion, is due to previous experience in open and laparoscopic surgery.

Continence. The retention function is provided by three critically important structures: the pudendal nerve, the innervating external rhabdomyosphincter, the autonomic nerve fibers of the internal sphincter, and the cavernous nerves of the neurovascular bundle, innervating the membranous urethra. In the authors’ opinion, the precision of apical dissection, the precision of work with stress incontinence and in the cervical region, as well as the quality of the anastomosis are factors that depend primarily on the restoration of this function. From the point of view of this metric, RARP demonstrates the longest training: 25 – 50 surgeries, rpRP, LRP, and plRP — 50 – 75 surgeries, respectively. At the same time, the total experience of the clinic of more than 50 surgeries per year demonstrates an advantage in the context of restoring this function [22].

Erectile function. The nerve-sparing technique and penile rehabilitation make it possible to achieve high rates of erectile function recovery by the end of first year after RP, regardless of the technique. In this work, RARP showed the most optimal rates for reaching the plateau of satisfactory results: 25 – 50 surgeries, rpRP and LRP — 50 – 75 surgeries. An attempt to save nerves in 12 patients in the plRP group was unsuccessful due to technical difficulties caused by access. A systematic review made by Ju et al. (2021) showed that the annual experience of the surgeon >25 surgeries and the total experience of the clinic >1000 RP demonstrated higher indicators of erectile function [23]. In contrast, Ucar et al. (2019) with bilateral nerve-sparing during RARP showed a 12-month erectile recovery rate of 33.8% during the first and second 45 surgeries [24].

Positive surgical margin. This indicator is one of the most important when discussing the oncological effectiveness of surgical treatment of prostate cancer. The rpRP modification was the most difficult in this context in terms of the duration of the LC — reaching a plateau of satisfactory results was observed after 50 – 75 operations: 48.0%, 40.0%, 16.0%, 12.0% for the 1^st, 2^nd, 3^rd, and 4^th 25 surgeries. Also, a relatively high PSM frequency was noted at LRP as 23.0% with an LC in 25–50 surgeries, which is confirmed in the work of Mita et al. (2021), where PSM at LRP was 30.0% [25]. On the contrary, plRP and RARP were associated with the lowest PSM rates of 14.0% and 18.0%, respectively, and the LC is 25 – 50 interventions.

Survival rate. In this study, the 24-month relapse-free survival rates in the rpRP, LRP, and RARP groups were 91.4%, 88.6%, and 93.9%, respectively. Maximum values were observed in plRP of 98.5%, which is due to the prevalence (67.0%) of patients with a low risk of progression. The cancer-specific 24-month survival rate in all groups was 100.0%; the overall survival rate was 97.3% for LRP, 98.6% for plRP, 98.9% for rpRP, and 100.0% for RARP. Similar data can be seen in the work of Song et al. (2017): 5-year cancer-specific survival rate for plRP was 99.0%, rpRP — 98.7%, LRP — 100.0%, and RARP — 99.8%, respectively [26].

Thus, this study describes for the first time the duration and nature of the LC of the four main techniques of RP by the example of one surgeon. The dynamics of basic perioperative, functional, and oncological indicators with the development of surgical techniques are studied. On the other hand, the multiplicity and heterogeneity of the metrics used are an obstacle to the formation of a conclusion. In the authors’ opinion, it is necessary to standardize the methodology of this kind of research and form a unified register of the studied parameters.

Conclusion

The development of RP should begin in large clinics with a large volume of operations performed annually and the presence of an experienced mentor; rpRP is associated with a high frequency of complications and blood transfusions, which dictates the need to use magnifying optics. Technical difficulties with perineal access limit the performance of energy saving and pelvic lymph dissection in the initial stages. Minimally invasive techniques (LRP and RARP) require mandatory modulated training, involving the stages of mastering in parallel with increasing complexity (working with a camera, assistance, and transition in independent work from simple stages to complex ones) and working on simulators. The expansion of state guarantee programs can ensure the growth of minimally invasive high-tech operations, and the higher cost of RARP is offset by a reduction in the cost of correcting complications and functional disorders, as well as a shorter LC.