Salvage Radical Prostatectomy for Recurrent Prostate Cancer: A Systematic Review (French ccAFU)

Simple Summary We reviewed the available studies assessing salvage surgery after recurrent prostate cancer with primary non-surgical treatment. While the studies used had the potential for bias, due to their retrospective type, we looked at treatment outcomes and toxicity for men treated with a number of salvage radical prostatectomies for recurrent prostate cancer. We demonstrated that SRP can be considered a suitable treatment option for selected patients. Abstract The aim of this study was to systematically review the current evidence regarding the oncological and functional outcomes of salvage radical prostatectomy (sRP) for recurrent prostate cancer. A systematic review was conducted throughout September 2022 using the PubMed, Science Direct, Scopus, and Embase databases. Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines were followed to identify eligible studies. A total of 55 studies (3836 patients) met our eligibility criteria. The vast majority of men included had radiation therapy (including brachytherapy) as their first-line treatment (n = 3240, 84%). Other first-line treatments included HIFU (n = 338, 9%), electroporation (n = 59, 2%), proton beam therapy (n = 54, 1.5%), cryotherapy (n = 34, 1%), focal vascular targeted photodynamic therapy (n = 22, 0.6%), and transurethral ultrasound ablation (n = 19, 0.5%). Median preoperative PSA, at the time of recurrence, ranged from 1.5 to 14.4 ng/mL. The surgical approach was open in 2300 (60%) cases, robotic in 1465 (38%) cases, and laparoscopic in 71 (2%) cases. Since 2019, there has been a clear increase in robotic versus conventional surgery (1245 versus 525 cases, respectively). The median operative time and blood loss ranged from 80 to 297 min and 75 to 914 mL, respectively. Concomitant lymph node dissection was performed in 2587 cases (79%). The overall complication rate was 34%, with a majority of Clavien grade I or II complications. Clavien ≥ 3 complications ranged from 0 to 64%. Positive surgical margins were noted in 792 cases (32%). The median follow-up ranged from 4.6 to 94 months. Biochemical recurrence after sRP ranged from 8% to 51.5% at 12 months, from 0% to 66% at 22 months, and from 48% to 59% at 60 months. The specific and overall survival rates ranged from 13.4 to 98% and 62 to 100% at 5 years, respectively. Urinary continence was maintained in 52.1% of cases. sRP demonstrated acceptable oncological outcomes. These results, after sRP, are influenced by several factors, and above all by pre-treatment assessment, including imaging, with the development of mpMRI and metabolic imaging. Our results demonstrated that SRP can be considered a suitable treatment option for selected patients, but the level of evidence remains low.


Introduction
Prostate cancer (PCa) is the most commonly diagnosed cancer in men, with an estimated 1.4 million diagnoses recorded worldwide in 2020 [1].Although active surveillance is increasingly used, most PCa patients undergo definitive local treatment, followed by prostate-specific antigen (PSA) monitoring [2].However, it is estimated that 27% to 53% of all patients undergoing radical prostatectomy (RP) or radiation therapy (RT) develop biochemical recurrence (BCR) [3].While there is a standard treatment pathway for post-RP BCR, there is no widely adopted treatment paradigm for BCR after primary nonsurgical treatment.In addition, there have been no randomized trials comparing the oncological outcomes of available salvage therapies, and thus, there is no clear consensus regarding the best treatment option.As such, many patients with BCR after primary nonsurgical treatment receive androgen deprivation therapy (ADT), which denies them any chance of curative therapy [4].
Salvage radical prostatectomy (sRP) is a challenging procedure that is rarely performed, although it represents a guidelines-validated option for BCR after primary nonsurgical treatment.The historical series of sRP with frequent major complications, such as rectal injury and poor functional outcomes [5], have played a major role in the low use of this option in a salvage situation.However, minimally invasive approaches may provide significant improvements, which could lead to improved functional outcomes and reduced complications [6,7].With the renewed interest in sRP, identifying patients who would benefit most from sRP is crucial to avoid overtreatment and limit treatment-related toxicities.
In this study, we aimed to systematically review the current evidence regarding the oncological and functional outcomes of sRP for recurrent PCa after primary nonsurgical treatment.

Search Strategy
We conducted a systematic review in line with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [8].
This protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (Registration Number: CRD42022378227).We conducted a literature search in PubMed/Medline, Embase, and Science Direct databases, to identify reports published through September 2022, which addressed the oncological and functional outcomes of sRP.The search strategy included the following MeSH terms: prostatectomy, Prostate Cancer, Neoplasm Recurrences, treatment, Local, Radiation Therapy, Cryotherapies, Salvage Treatment, Robot-Assisted Surgery, and Surgical Procedure.Initial screening was independently performed by two investigators (A.S. and G.P.) based on the titles and abstracts of the articles to identify ineligible reports.Reasons for exclusions were noted.Potentially relevant reports were subjected to a full-text review, and the relevance of the reports was confirmed after the data extraction process.Disagreements were resolved by consultation with a third co-author (M.B.).

Study Selection
Studies were deemed eligible if they included men with recurrent PCa after primary nonsurgical treatment (patients), managed with sRP (intervention), and if they assessed oncological and/or functional outcomes (outcome) in randomized controlled trials, nonrandomized prospective studies, and retrospective studies (study design).In case of duplicate publications, either the higher-quality or the most recent publication was selected.Reviews, meta-analyses, editorials, commentaries, authors' replies, meeting abstracts of unpublished studies, and case reports were excluded, but the reference section was checked for relevant articles.No restriction on the publication language was applied.We searched reports published between January 2008 to September 2022 (Supplementary Material Table S1).

Data Extraction
Data on studies, patients, treatment, and follow-up were independently extracted by two authors (A.S. and G.P.).We extracted the following variables from the included studies: first author's name, publication year, sample size, age, pre-sRP PSA, pre-and post-sRP TNM stage, International Society of Urological Pathology (ISUP) score at pre-sRP biopsy, surgical approach, operative time, estimated blood loss, rate and severity of postoperative complications according to the Clavien-Dindo classification, rate of urinary continence, follow-up data, BCR rates, cancer-specific survival, and overall survival.

Assessment of Methodological Quality
Two authors (A.S. and G.P.) independently assessed the quality of the studies and the risk of bias.The risk of bias was assessed according to EAU recommendations for performing systematic reviews and meta-analysis [9].The Quality Appraisal tool for case series using a Modified Delphi technique was used for retrospective studies [10].

Study Selection and Characteristics
The study selection process is outlined in the PRISMA flow diagram (Figure 1).A total of 403 full-text articles were assessed for eligibility and 55 met our inclusion criteria [7,.

Perioperative Results
The perioperative data are presented in Table 2. Regarding the surgical approach, an open approach was used in 2300 cases (60%), a robotic approach in 1465 cases (38%), and a laparoscopic approach in 71 cases (2%); but since 2019, there has been more frequent use of robotic versus conventional surgery (1245 versus 525 of cases, respectively).NA: not available, sRP: salvage radical prostatectomy, PSM: positive surgical margins, RT: radiation therapy.* According to classification of Clavien-Dindo.
A total of 45 studies reported data on concomitant lymph node dissection.The median number of nodes yielded was reported in 14 studies and ranged from 6 to 17, including 593 (20.5%) patients which were staged pN+ at final pathology.
Regarding pathological features, stage ≥pT3, positive surgical margins, and pN+ status ranged from 5 to 75%, 25 to 82%, and 3 to 60%, respectively.The pathological Gleason score was ≥8 in 6 to 67% of cases.These data were missing in nine studies.

Complications and Functional Results
The reported postoperative complications are summarized in Table 2.The overall complication rate was 34%; with a majority of Clavien grade I or II complications.Clavien grade ≥3 complications ranged from 0 to 64%.The complete urinary continence rate (no pad use) was 52.1% (Table 3).The rates of urinary continence were 56% and 47%, respectively, in minimally invasive (i.e., laparoscopic and robotic) and open approaches.
The urinary continence rate in the primary non-radiation-treatment group (HIFU, electroporation, proton beam therapy, cryotherapy, focal vascular targeted photodynamic therapy, and transurethral ultrasound ablation) was 67% versus 55% in patients formerly treated by RT.
The rates of BCR were 20%, 27%, and 47%, respectively, in laparoscopic, robotic, and open approaches.Overall survival was 100% and 98% in the laparoscopic and robotic groups, respectively, and 74% in the open surgery group.The rates of BCR were 36% and 21% in the group of patients treated by non-radiation therapy and RT, respectively.Overall survival was 98% in the group of other primary treatments and 85% for patients treated by RT.

Discussion
sRP for recurrent PCa after primary non-surgical treatment failure is challenging for urologists due to its aggressive features and technical demands.The majority of PCa patients who present recurrent disease after RT are therefore treated with palliative ADT while a salvage treatment initiated early may change the disease course.As a result, only 1% of the patients recurring after RT indeed undergo salvage surgery [65].
In the present systematic review, we found that sRP may represent a good alternative that can be provided to carefully selected patients.It may lead to a durable response if initiated early and may delay progression and use for systemic therapies.
The introduction of minimally invasive approaches regarding sRP could be associated with many advantages, such as decreasing the rates of overall and high-grade complications (i.e., Clavien > 2).The robotic approach has been also associated with lower rates of blood loss, rectal injury, anastomotic stricture, and postoperative incontinence [66].Recently, it has been suggested that the Retzius-sparing approach could also be interesting as it allows a meticulous dissection near the often fibrotic rectal plane.Using this approach, Madi et al.only noted one intraoperative urine leak in their salvage Retzius-sparing (SRS) group [12].Taken together, the implementation of a minimally invasive approach in sRP has led to a renewed interest in this option for managing recurrence after the primary nonsurgical management of PCa.
One of the major limitations attributed to sRP is the poor functional outcomes regarding the urinary continence associated with this option.Thus, we found an overall complication rate of 34%, including rectal wounds, ureteral complications, rectourethral fistula, lymphoceles, anastomotic leakage, and urinary tract infections, which is in line with a previous report from Matei et al., who reported a Clavien > 2 complication rate of 0-33% [66].
However, the functional results widely differed between the studies included in this systematic review.Continence rates reported after sRP ranged from 10 to 100%.This heterogeneity could be explained, once again, by the surgical approach used.Roboticassisted sRP appeared to improve the early return to continence, compared to open surgery series.This is thought to be due to the support of the surrounding ligaments to the anterior urethra, which helps to maintain sphincteric integrity after SRS [67,68].Mason et al. suggested that continence outcomes were significantly improved in the SRS group for the treatment of radioresistant prostate cancer [69].
Oncological outcomes after sRP are influenced by several factors and may vary depending on the patient/tumor characteristics, type of initial treatment, surgical approach used, length of follow-up, and, above all, pre-treatment assessment (including imaging, with the development of MRI and metabolic imaging).At the mid-term follow-up, we found that the oncological outcomes were acceptable, as a significant proportion of men were disease-free after five years (i.e., the BCR-free survival rate ranged from 48% to 59% at 60 months).In addition, cancer-specific survival and overall survival rates ranged from 13.4 to 98% and 62 to 100% at 5 years, respectively.However, long-term data remain poorly reported in the literature.Two series showed a 10-year BCR-free survival of 31% and 37%, respectively [25,36].We therefore encourage further studies evaluating long-term oncological outcomes in these patients.
Our study has several strengths, including the important number of studies/patients included, with a variety of nonsurgical primary treatments with a clear distinction between them, the inclusion of most updated data, and their careful review for study inclusion.Some limitations must be acknowledged.The main limitation is the significant risk of bias, as all included studies were retrospective, which prevented us from reaching a high level of evidence and from providing clear recommendations.Finally, the heterogeneity regarding the surgical approach used, the type of initial local, and the functional erectile results, which are not reported in our review, are important limitations to notice.Of note, although we performed a systematic review, a meta-analysis was not possible given the heterogeneity of the studies in terms of the initial treatment proposed and the surgical approach.

Conclusions
sRP appears to be feasible with acceptable morbidity in well-selected PCa patients who recur after primary non-operative surgical treatment.The development of a minimally invasive approach and the improvement of surgical techniques are considered to be two key factors in improving perioperative outcomes.However, the level of evidence remains low as comparative and long-term data are lacking.

Table 1 .
Pre-operative characteristics of the included studies.

Table 2 .
Intra-operative parameters and pathological features of the overall cohort.

Table 3 .
Oncological and functional outcomes of the overall cohort.