Retzius-sparing robotic prostatectomy is associated with higher positive surgical margin rate in anterior tumors, but not in posterior tumors, compared to conventional anterior robotic prostatectomy

Abstract

Objectives

The analysis of the oncological outcomes and postoperative continence recovery between conventional robotic-assisted radical prostatectomy (cRARP) and Retzius-sparing RARP (rsRARP), and the effect of the tumor location on them.

Materials and methods

A total of 317 patients who underwent cRARP (n = 228) or rsRARP (n = 89) from August 2017 to July 2020 were assessed. Patients were categorized into groups based on the tumor location by pathology. Positive surgical margin (PSM) rates and biochemical recurrence (BCR)-free survivals and continence recovery were compared between the two procedures.

Results

Patient age, prostate-specific antigen (PSA) levels, clinical stages, and Gleason score were not significantly different between the two groups. There was no significant difference in PSM rates (25.8% vs. 33.7%, p = 0.13) or BCR-free survivals (p = 0.28) between cRARP and rsRARP in patients. When tumor was located in the anterior lesion in the prostatectomy pathology, rsRARP was associated with significantly higher PSM rates than cRARP (53.3% in rsRARP vs. 27.0% in cRARP, p = 0.0086), while BCR-free survival did not vary significantly (hazard ratio: 2.15, p = 0.11). When tumors were identified in the posterior in prostatectomy pathology, PSM rates (28.8% in rsRARP vs. 24.7% in cRARP, p = 0.59) or BCR-free survivals (hazard ratio: 0.78, p = 0.51) did not differ significantly between the two groups. rsRARP yielded superior continence recovery in all time points compared to cRARP, which was not affected by the pathological tumor location.

Conclusion

In posterior tumors, rsRARP results in similar oncological outcomes as cRARP with superior continence recovery, while in anterior tumors, rsRARP may associate with higher PSM rate, combined with improved continence recovery.

1. Introduction

Prostate cancer (PCa) is one of the most commonly diagnosed cancers worldwide, with 1,414,259 new cases (7.3% of the total) and 375,304 deaths (3.8% of the total) in 2020.¹ Radical prostatectomy is one of the main treatments for localized PCa, and its frequency has been rapidly increasing and evolving with the increase of early stage PCa due to the extensive use of prostate-specific antigen (PSA) screening.^2,3 Claude Abbou and Jochen Binder introduced robot-assisted radical prostatectomy (RARP) in the early 2000s using the da Vinci Surgical System®,^4,5 which is now performed worldwide as a minimally invasive procedure.

Quality of life outcomes is one of the major concerns for patients after RARP.⁶ Postoperative urinary continence, in particular, is a major quality of life concern in patients that undergo RARP. Incontinence rate, defined as the use of at least 1 pad/day, was reported to be 4%–31% one year after RARP.⁷ The etiology of postoperative incontinence is multifactorial with surgical factors playing a key role. Surgical procedures have been improved based on anatomical landmarks in order to yield the highest functional results without compromising oncological safety.⁸

One of the latest relevance developments in terms of continence recovery is the Retzius-sparing RARP (rsRARP), as described by Galfano et al in 2010.⁹ This approach can evade dissection of the endopelvic fascia, the dorsal vascular complex, and Retzius space in contrast to conventional anterior RARP (cRARP), while rsRARP has been demonstrated to provide superior postoperative urinary continence recovery compared to cRARP.^10,11 Oncological safety of rsRARP on the other hand, compared to cRARP is controversial. Although some studies have showed that biochemical recurrence (BCR)-free survival in rsRARP was not inferior compared to cRARP, other studies reported that rsRARP offers a higher positive surgical margin (PSM) than cRARP, especially for anterior tumors.12, 13, 14, 15, 16 It was therefore aimed to analyze whether oncological outcomes differed between cRARP and rsRARP, and how tumor location affects the comparison of the oncological outcomes between the two procedures.

2. Materials and methods

2.1. Study design, inclusion, and exclusion criteria

This retrospective observational study was approved by the institutional review board of Jichi Medical University Saitama Medical Center (RinS20-058). Patients who underwent RARP at our institute from August 2017 to July 2020 were included in the study. All RARP were performed using a da Vinci Si® Surgical System (Intuitive Surgical Inc.) via a transperitoneal approach. Two different approaches were employed: the cRARP and the rsRARP. rsRARP was initiated in August 2017 and consequently, this series included the first case and the following for rsRARP in our institute. A total of 366 patients were included and 49 patients were excluded for the reasons shown in supplementary Table 1. Overall, we evaluated 317 patients in the present study, among which 228 and 89 patients underwent cRARP and rsRARP, respectively.

2.2. Indication and surgical techniques

rsRARP was recommended in patients that were interested in early urinary continence recovery. Two moderately experienced surgeons, having completed more than 100 cases of RARP, mainly performed rsRARP while a total of seven surgeons performed cRARP. The procedures described by Galfano et al.⁹ for rsRARP. cRARP were initially performed via a posterior approach, followed by dissection of the Retzius space through an anterior approach (the details are shown in supplementary materials). The nerve sparing technique was mainly recommended in patients who had clinical T1 and T2 PCaand maintained sexual function. Nerve sparing was performed at the side where no tumor was observed in the latero-posterior lesion on MRI or a Gleason score >7 was indicated in the prostate biopsy. The pelvic lymph node dissection (PLND) was performed only in cases exhibiting swelling of lymph nodes in the pelvis on MRI and/or CT as PLND in radical prostatectomy failed to improve the oncological outcomes.¹⁷

2.3. Assessment of pathological findings and BCR

The pathological report and tumor mapping were retrospectively reviewed in each patient, and data were collected from the pathological findings, including T and N stage, the presence of PSM, and the location of tumors with maximum diameter ≥5 mm. Patients were categorized into three groups with respect to the location of index tumor: i.e., “anterior,” “posterior,” and “anterior + posterior” tumors according to the criteria shown in Supplementary Fig. 1A. BCR was defined as PSA levels >0.2 ng/ml or the initiation of androgen deprivation therapy following RARP. When PSA <0.2 ng/ml was not achieved, the date in BCR was set at the date of the surgery.

2.4. Prostate MRI

Biparametric MRI (T2 weighted image and diffusion weighted image) was typically performed in our institute before RARP. Cases with the interval from prostate MRI undergoing RARP more than 12 months were excluded from MRI analysis. Patients were categorized into two groups with respect to the location of index tumors on MRI. When the index tumor was indicated in the anterior lesion (or both anterior and posterior lesions), it was categorized into MRI anterior tumors, while the tumors limited to the posterior lesion were categorized into MRI posterior tumors (supplementary Fig. 1B).

2.5. Objectives of this study

The primary objective of this study was to compare PSM rates and BCR-free survivals between cRARP and rsRARP. The tumor location, anterior versus posterior tumors, were in the focus of the current study, and it was assessed how the tumor location affected these oncological outcomes. Oncological outcomes between the two procedures according to the MRI tumor location were also compared. Additionally, urinary continence recovery and perioperative outcomes were compared among the two groups. We defined two levels of postoperative urinary continence, a completely pad-free (0 pad/day) and one safety pad or less. Complications were graded according to Clavien-Dindo classification.¹⁸

2.6. Statistical analysis

All statistical analyses were performed using the GraphPad Prism software version 9.0 (GraphPad Software, La Jolla, CA, USA). Data were compared using Student's t-test, Mann–Whitney U test, or the χ² test. Kaplan-Meier curves were constructed to evaluate survival and continence recovery and the log-rank test was employed to compare the curves. Hazard ratio (HR) was computed by log-rank test. All data are presented as means with Standard Deviation, unless otherwise indicated. A P < 0.05 was considered statistically significant difference.

3. Results

3.1. Patients’ characteristics

There was no significant difference in patient characteristics between cRARP and rsRARP, including age (70.1 ± 5.5 vs. 69.9 ± 5.9, p = 0.71), BMI (24.1 ± 2.8 vs. 24.2 ± 3.0, p = 0.68), PSA levels (10.9 ± 12.1 vs. 8.6 ± 5.4 ng/ml, p = 0.08), clinical stage (p > 0.99), International Society of Urological Pathology Gleason grade (p = 0.22), and National of Comprehensive Cancer Network risk (p = 0.92) (Table 1). Nerve sparing and PLND was performed in 56 (62.9%) and 0 (0%) patients in rsRARP, and 27 (11.8%) and 6 (2.6%) patients in cRARP.

Table 1.

Patients’ characteristics

	cRARP	rsRARP	P value
	(n = 228)	(n = 89)
Age, mean ± SD	70.1 ± 5.5	69.9 ± 5.9	0.71
Body mass index(kg/m2), mean ± SD	24.1 ± 2.8	24.2 ± 3.0	0.68
Serum PSA(ng/mL), mean ± SD	10.9 ± 12.1	8.6 ± 5.4	0.08
Clinical stage, n (%)			>0.99
T1-T2	197 (86.4%)	77 (86.5%)
T3-T4	31 (13.6%)	12 (13.5%)
Biopsy ISUP grade, n (%)			0.22
1	18 (7.9%)	4 (4.5%)
2	102 (44.7%)	42 (47.2%)
3	41 (18.0%)	21 (23.6%)
4	25 (11.0%)	13 (14.6%)
5	42 (18.4%)	9 (10.1%)
NCCN Risk classification, n (%)			0.92
Low	11 (4.8%)	3 (3.4%)
Intermediate-Favorable	99 (43.4%)	43 (48.3%)
Intermediate-Unfavorable	31 (13.6%)	12 (13.5%)
High	65 (28.5%)	24 (27.0%)
Very High	22 (9.6%)	7 (7.9%)

cRARP, conventional robot-assisted radical prostatectomy; ISUP, International society of urological pathology; NCCN, National comprehensive cancer network; rsRARP, Retzius-sparing robot-assisted radical prostatectomy; PSA, prostate-specific antigen; .

3.2. Perioperative outcomes and complications

There was no significant difference in perioperative outcomes between cRARP and rsRARP, including operation time (199.9 ± 52.5 vs. 190.1 ± 41.1 ml, p = 0.08), console time (162.1 ± 49.0 vs. 154.3 ± 40.9 min, p = 0.19), estimated blood loss (74.4 ± 86.2 vs. 75.4 ± 75.9 ml, p = 0.92), catheter placement time (7.2 ± 5.2 vs. 7.6 ± 5.8 days, p = 0.61), hospitalization (7.5 ± 2.2 vs. 7.1 ± 0.6 days, p = 0.17), and complications (p = 0.18) (Table 2).

Table 2.

Comparison of intraoperative and postoperative outcomes

	cRARP	rsRARP	P value
	(n = 228)	(n = 89)
Intraoperative outcomes
Operation time(minutes), mean ± SD	199.9 ± 52.5	190.1 ± 41.1	0.08
Console time(minutes), mean ± SD	162.1 ± 49.0	154.3 ± 40.9	0.19
Estimated blood loss (ml), mean ± SD	74.4 ± 86.2	75.4 ± 75.9	0.92
Postoperative outcomes
Catheter placement (days), mean ± SD	7.2 ± 5.2	7.6 ± 5.8	0.61
Hospitalization (days), mean ± SD	7.5 ± 2.2	7.1 ± 0.6	0.17
Complicationsaduring 30 days, n (%)
Grade 0	160 (90.9%)	57 (83.8%)	0.18
Grade 1	7 (4.0%)	3 (4.4%)
Grade 2	9 (5.1%)	8 (11.8%)

^aClavien-Dindo classification.

3.3. Oncological outcomes

There was no significant difference in PSM rate between the two groups in the entire cohort (33.7% in rsRARP vs. 25.8% in cRARP, p = 0.13) (Fig. 1A). However, when assessing tumors involved in anterior lesions (pathological anterior tumors or anterior + posterior tumors), PSM rate was significantly higher in rsRARP than cRARP (53.3% in rsRARP vs. 27.0% in cRARP, p = 0.0086) (Fig. 1B). In contrast, there was no significant difference in PSM rate between the two groups in pathological posterior tumor (23.7% in rsRARP vs. 21.3% in cRARP, p = 0.84) (Fig. 1C).

Figure 1.

Comparison of PSM rate in entire cohort (A) and by tumor location (B), (C) PSM rates in cRARP versus rsRARP in the entire cohort (A) and by tumor location (B, C) are presented. Blue bars and red bars indicate cRARP and rsRARP, respectively. cRARP, conventional anterior robot-assisted radical prostatectomy; PSM, positive surgical margin; rsRARP, Retzius-sparing robot-assisted radical prostatectomy.

BCR-free survival did not differ significantly between cRARP and rsRARP in the entire cohort (1-year and 3-year BCR-free survival of 89.3% in cRARP vs. 93.2% in rsRARP; 88.4% in cRARP vs. 83.0% in rsRARP, p = 0.28) (Fig. 2A). In pathological “anterior” or “anterior + posterior” tumors, rsRARP presented a lower BCR-free survival compared to cRARP although the difference was not statistically significant (HR: 2.15, 95% confidence interval [CI]: 0.66–6.96, p = 0.11) (Fig. 2B). On the other hand, the BCR-free survival in pathological posterior tumors was similar between the two groups (HR: 0.93 [95% CI: 0.39–2.24], p = 0.87) (Fig. 2C).

Figure 2.

Comparison of BCR-free survivals in the entire cohort (A) and by tumor location (B), (C) BCR-free survivals in cRARP versus rsRARP in the entire cohort (A) and by tumor location (B, C) are presented. Blue lines and red lines indicate cRARP and rsRARP, respectively. BCR, biochemical recurrence; cRARP, conventional anterior robot-assisted radical prostatectomy; rsRARP, Retzius-sparing robot-assisted radical prostatectomy.

3.4. MRI tumor location stratifying PSM rate and BCR-free survival between cRARP and rsRARP

It was then assessed whether tumor location on MRI can facilitate the stratification of PSM rate and BCR-free survival in cRARP and rsRARP. PSM rate was significantly higher in rsRARP than cRARP in MRI anterior tumor (55.6% in rsRARP vs. 28.4% in cRARP, p = 0.030) (Fig. 3A). Nevertheless, there was no significant difference in PSM rate between the two groups in MRI posterior tumor (28.8% in rsRARP vs. 24.7% in cRARP, p = 0.59) (Fig. 3B). The BCR-free survival was significantly lower in rsRARP than cRARP in MRI anterior tumors (HR: 3.34 [95% CI: 0.63–17.8], p = 0.040, Fig. 3C), while there was no significant difference in MRI posterior tumors (HR: 0.78 [95% CI: 0.37–1.62], p = 0.51, Fig. 3D). The location of index tumors on MRI was identical with that in the pathological specimen in 294 of 310 cases (94.8%) in which tumors were visible on MRI.

Figure 3.

Comparison of PSM rates and BCR-free survivals by MRI tumor location. In MRI anterior tumor, PSM rate and BCR probability are significantly higher in rsRARP than in cRARP (A, C). In MRI posterior tumor, PSM rate and BCR probability did not differ significantly between two groups (B, D). Blue bars/lines and red bars/lines indicate cRARP and rsRARP, respectively.

PSM, positive surgical margin; BCR, biochemical recurrence; cRARP, conventional anterior robot-assisted radical prostatectomy; rsRARP, Retzius-sparing robot-assisted radical prostatectomy.

3.5. Continence recovery

Compared with cRARP, rsRARP was associated with better urinary continence recovery in all time points (1 safety pad or less: 27.3% vs. 77.9% at one month, 46.8% vs 84.1% at 3 months, and 86.5% vs. 97.7% at 12 months, p < 0.0001; and 0 pad: 3.8% vs. 39.0% at one month, 10.4% vs. 54.5% at 3 months, and 41.5% vs. 75.3% at 12 months, p < 0.0001) (Fig. 4A and B). Better urinary continence recovery in rsRARP was observed irrespective with the tumor location (Fig. 4C–F). There was no significant difference in the number of patients who took medication for postoperative incontinence within the first six months after surgery (8.8% in cRARP vs. 12.4% in rsRARP, p = 0.40, supplementary Table 2).

Figure 4.

Comparison of continence recovery in the entire cohort (A), (B) and by tumor location (C–F). Continence recovery in terms of 0 pad use (A: entire cohort, C: pathological “anterior” or “anterior + posterior” tumor, E: pathological posterior tumor) and 0–1 pad use for safety (B: entire cohort, D: pathological “anterior” or “anterior + posterior” tumor, F: pathological posterior tumor) are shown. Blue lines and red lines indicate cRARP and rsRARP, respectively. cRARP, conventional anterior robot-assisted radical prostatectomy; rsRARP, Retzius-sparing robot-assisted radical prostatectomy.

4. Discussion

The present study demonstrated that rsRARP was associated with a higher PSM rate in anterior tumors but not in posterior tumors compared to cRARP. Furthermore, when index tumor was located in the anterior side on MRI rsRARP it caused a higher probability of PSM and BCR than cRARP. Continence recovery was superior in rsRARP than cRARP irrespective of the tumor location.

4.1. Comparison of PSM rates

A control trial comparing oncological outcomes between rsRARP and cRARP has been previously reported,¹⁹ showing that non-focal PSM rate was not significantly different (11.7% in rsRARP vs. 8.3% in cRARP). Nevertheless, several studies have shown that rsRARP has a higher PSM rate than cRARP.^16,20 A recent meta-analysis discovered a nonsignificant difference in PSMs between rsRARP and cRARP, although the rsRARP group presented significantly higher PSM rates in the anterior site (OR 2.25).²¹ In addition, Edgan J showed that locations of PSMs were different between rsRARP and cRARP: i.e., anterior was predominant in rsRARP (54.2%) versus posterior that was predominant in cRARP (70.6%).²² In accordance with these prior studies, the present study demonstrated that PSM rate was significantly higher in rsRARP when the tumor was observed in the anterior lesion (Fig. 1B) but not when tumor was limited in the posterior lesion (Fig. 1C). MRI reveals tumor location and tumor aggressiveness,²³ which may help decision for an optimal surgical method. When index tumor was located in the anterior side, but not in the posterior side, on MRI rsRARP it caused a higher probability of PSM than cRARP (Fig. 3A and B), suggesting MRI can assist such a decision making.

4.2. BCR

Several previous studies have presented BCR data in rsRARP.^19,24,25 Raheem AA et al. showed 1- and 3-year BCR-free survival of 92.3% and 80.9% in rsRARP.²⁴ Menon M et al. compared the BCR-free survival between rsRARP and cRARP, demonstrating that 1- and 1.5-year BCR-free rate was 91.4% and 83.8% in rsRARP versus 92.7% and 92.7% in cRARP, respectively.¹⁹ The 1- and 3-year BCR-free survival of 93.2% and 83.0% in rsRARP versus 89.3% and 88.4%, respectively, in cRARP in the present study, similar to the previously reported studies. Interestingly, no studies comparing BCR-free survival stratified according to the tumor location have been reported. The present study demonstrated rsRARP was associated with a higher BCR probability than cRARP when tumors were involved in anterior lesions in pathology or on MRI (Figure 2, Figure 3C) whereas BCR-free survival was fairly similar between the two groups when tumors were limited to posterior lesions (Figure 2, Figure 3D).

4.3. Comparison of continence recovery

Various previous studies reported that early urine continence (1 month) was significantly improved in rsRARP compared to cRARP.^15,16,20,25 The meta-analysis showed that rsRARP was significantly improved compared to cRARP in early (1-month) urine continence (OR 5.72) as well as in 12-month continence (OR 2.37).²¹ In the present study, rsRARP was superior compared to cRARP in the early point as well as in the late point in the entire cohort (Fig. 4A and B), which was observed irrespective to the tumor location (Fig. 4C–F). Chang et al. reported that the 0–1 safety pad rate for rsRARP after 1 month and 1 year was 91.0% and 100%, respectively, and that of cRARP was 30.0% and 93.3%, respectively.²⁵ The continence recovery in the present study in rsRARP as well as in cRARP appeared to be inferior compared to the study by Chan et al. This could be justified by the difference in surgeons' experience for RARP. The early series were included in rsRARP and seven surgeons with various degrees of experience performed cRARP in the present study while only one surgeon performed rsRARP as well as cRARP in the study by Chang et al. Other favorable potentials in rsRARP include less incidence of postoperative inguinal hernia, shortening of penis, and Peyronie's disease in rsRARP compared to cRARP.²⁶

4.4. The mechanisms mediating anterior tumors association with poor oncological outcomes in rsRARP

The difference in oncological outcomes between rsRARP and cRARP in anterior tumors may result from the differences in surgical approaches between the two procedures. rsRARP typically preserves the Santorini plexus and detrusor apron more efficiently than cRARP. In anterior tumors, the distance between tumor edge and resection margin could become smaller in rsRARP, as the PSM status and BCR may be affected. In posterior tumors, the margin status is affected by sparing of tissues at the posterolateral of prostate such as the neurovascular bundle but probably not of the Santorini plexus and the detrusor apron. The indication and procedure of sparing of these lateral tissues are similar between rsRARP and cRARP, resulting in similar oncological outcomes.

4.5. Limitations

Initially, this was a retrospective analysis and as such, patient selection bias was occurred. Especially, rsRARP was recommended to patients who were interested in early continence recovery and such patients may perform more postoperative exercise and effort for continence recovery, which may have an effect on the different study outcomes in continence recovery between the two procedures. Moreover, surgeons' experience for rsRARP was more limited than that for cRARP, which might have an effect on the different study outcomes such as oncological outcomes and continence recovery. Furthermore, rsRARPs were primarily performed by two moderately experienced surgeons while cRARPs were performed by seven surgeons, which had various degrees RARP experience. The variation among surgeons’ experience could have an effect on the outcomes. Patients undergoing nerve sparing were less in cRARP compared to rsRARP, which might affect the outcomes. Lastly, the number of each cohort was relatively small especially in rsRARP and long-term follow-up outcomes are missing in the present study.

5. Conclusions

In conclusion, in posterior PCa tumors, rsRARP can provide superior continence recovery without compromising oncological outcomes while in anterior tumors, rsRARP may be associated with higher PSM rate and lower BCR-free survival with superior continence recovery. This information may assist the decision making process for the optimal selection among rsRARP and cRARP.

Conflicts of interest

The authors have no conflict of interest to declare.

Funding statement

No funding.

Authorship Confirmation Statement

S. Washino conceived of the presented idea. M. Oshima analyzed the data. M. Oshima and S. Washino took the lead in writing the manuscript. T. Miyagawa supervised the project. All authors provided critical feedback and helped shape the research, analysis and manuscript.

Appendix A. Supplementary data

The following is the Supplementary data to this article: