Feasibility of robot-assisted radical prostatectomy in men at senior age ≥75 years: perioperative, functional, and oncological outcomes of a high-volume center

Abstract Objectives The aim of this study was to assess whether age ≥75 years impairs surgical, functional, and oncological outcomes after robot-assisted radical prostatectomy (RARP). Materials and methods Patients with prostate cancer (PCa) were stratified in ≥75(n = 669) vs. <70 years(n = 8,268). Multivariable cox regression analyses (MVA) tested for effect of senior age on erectile function-, urinary continence-recovery, biochemical recurrence (BCR), and metastatic progression (MP). Results RARP duration, blood loss, and 30d complication rates were similar between groups. For patients ≥75 vs. <70 years, rates of erectile function after 36 and urinary continence after 12 months were 27 vs. 56% (p < 0.001) and 85 vs. 86% (p = 0.99), respectively. Mean quality of life (QoL) score after 12 months improved in both groups (p = 0.9). At 48 months, BCR- and MP-free rates were 77 vs. 85% (p < 0.001) and 97 vs. 98% (p = 0.3), respectively. MVA confirmed the negative effect of senior age on erectile function but no significant effect on urinary continence, BCR or MP, before and after propensity score matching. Conclusion Apart from erectile function, senior age has no significant effect on urinary continence recovery, BCR- or MP-free rates after RARP. Post-RARP QoL improved even in senior patients. Modern therapy of senior PCa patients should be based on individual counseling than just age.


Introduction
Prostate cancer (PCa) represents the second most common cancer in men, particularly in industrialized countries, where healthcare improvements contribute to an extended life expectancy (LE) [1]. Moreover, PCa is a highly variable disease but has a generally protracted natural history in comparison to many other cancerous diseases. As both the development of PCa and the average LE converge, improved overall survival rates and advanced age will result in a greater proportion of senior men who present with advanced PCa stages, which requires local treatment [2]. To a certain degree, these trends are accompanied by evolving surgical techniques and expertise of robotassisted radical prostatectomy (RARP) [3]. Several reports confirmed safety and a low perioperative morbidity of RARP[4,5] even in locally advanced or oligometastatic PCa [6]. Such surgical features suggest that RARP is also suited for senior patients. However, for RARP eligibility, an estimated LE of at least 10 years is usually recommended [7]. Moreover, LE may often be underestimated due to demographic rather than individual estimates. In consequence, many urologists are reluctant to perform RARP in men aged �75 years. Instead, options with non-curative intent [8] or radiotherapy [9] are often preferred, but still subject to substantial side effects. Moreover, restraint to perform surgery in senior PCa patients is supported by the notion that advanced age appears to be associated with worse biochemical-recurrence(BCR)-and metastatic progression(MP)-free survival rates [10], which compromises the risk-benefit ratio of RARP. Although senior PCa patients represent a public health concern, it is of note that data on oncological outcomes for RARP patients aged �75 years are quite scarce. Methodological limitations apply to most studies on that topic. They had limited sample size [5,11], did not account for comorbidity [5,9,10], did not fully adjusted for tumor characteristics [5,9,10,12,13] or only focused on functional or perioperative outcomes [4,5,14,15].
Thus, we assessed whether advanced age �75 years in appropriately selected RARP patients impaired perioperative, surgical, functional, quality of life (QoL) and oncological outcomes compared to younger patients <70 years.

Materials and methods
Overall, 12,277 consecutive PCa patients were identified, who had complete pathological data and received a RARP at the Prostate Cancer Center Northwest, Gronau, Germany, between 5/2006 and 12/ 2018. A simultaneous extended pelvic lymph node dissection was performed during RARP in all of the patients. Charlson Comorbidity Index (CCI) was based on patient health records [16]. Accordingly, the PCa diagnosis was not included in the CCI. The assessment of frailty and biological age was based on standardized clinical interview before RARP with rigorous anesthesiologic evaluation at our institution, including cardiovascular examinations if deemed medically necessary by the surgeon or anesthesiologists. Patients with suspected metastases or previous radiotherapy or local endoscopic therapy of the prostate (e.g. HIFU or TURP) were excluded. Patients were stratified according to age groups <70 years vs. �75 years, omitting the age group 70-75 years to provide more age-specific results.
Combination of International Index of Erectile Function (IIEF-5) score �18 and/or score �3 at the second question ("When you had erections with sexual stimulation, how often were your erection hard enough for penetration") were used to define preoperative potency and recovery of erectile function [19].
Recovery of urinary continence was defined as fulfilling the following criteria: Up to one pad usage within 24 h (safety pad) or score of �2 at the International Consultation on Incontinence Questionnaire-Urinary Incontinence Short Form Questions 1 and 2 ("How often do you leak urine?" and "How much urine do you usually leak (whether you wear protection or not)?") [20] or finally, the International Continence Society male questionnaire score of �1 at each of the three questions I2, I3 and I4. For preoperative continence, no preoperative pad usage was allowed.
Oncological outcomes were evaluated based on BCR, defined as PSA levels �0.2 ng/ml, and radiographic MP.

Propensity score matching
We used propensity score matching (PSM) at a 4:1 ratio to create similar cohorts between RARP patients <70 vs. �75 yrs. In order to ensure that matching improved the balance between aforementioned age groups, the standardized mean differences were checked. PSM variables consisted of the same variables that were used for the multivariable cox regression analyses (MVA). Unless otherwise stated, the corresponding variables were preoperative neoadjuvant ADT, CCI (0 vs. 1 vs. 2 vs. �3), year of surgery, surgical expertise, preoperative serum PSA, pathological ISUP grade group, tumor stage, nodal stage and surgical nerve-sparing procedure (none vs. unilateral vs. bilateral).

Statistical analyses
Chi-square test was used for categorical and t-test for continuous variables. For graphical display we used cumulative incidence plots or Kaplan-Meier curves. The log-rank tested for differences of urinary continence recovery, erectile function recovery, BCR, and MP, respectively, between RARP patients <70 vs. �75 yrs.
MVA assessed effect of age on time to urinary continence recovery, erectile function recovery, BCR or MP. MVA of urinary continence recovery were restricted to preoperatively continent men and additionally adjusted for preoperative International Prostate Symptom Score (IPSS). Within the same sub cohorts, mean difference between preoperative vs. 12 months post-RARP IPSS/QoL was compared. Second, MVA of erectile function recovery were performed only by preoperatively potent men with at least unilateral nerve sparing. Finally, to account for low number of MP events, we used age, the cancer of the prostate risk assessment (CAPRA-S) postsurgical score and BCR and postoperative radiation therapy, which were both modeled as time-dependent covariates [21]. MVA were repeated after PSM.
All tests were two-sided with a statistical significance set at p < 0.05. Analyses were performed with the statistical package for R (R Foundation for Statistical Computing, version 3.2.2).

Surgical outcomes and 30d complications
Surgical expertise was high in both groups, i.e. 921 (IQR 186-3626) vs. 712 (175-2297) for �75 < 70 years, respectively. Median RARP duration and estimated blood loss were virtually identical between groups, 150 min and 200 ml, respectively (Table 2). Most 30d complication categories did not significantly differ or differences were below 2%, such as the miscellaneous surgical complication category with complication rates of 3.9 vs. 2.3% (p ¼ 0.01). No rectum injury occurred.

Discussion
Previous series suggested that senior age was an independent predictor for worse oncological and functional outcomes after radical prostatectomy (RP) [10]. This conclusion is of utmost importance since patient counseling largely focuses on proposed treatmentrelated survival benefit [22]. Consequently, senior patients might be excluded from RARP despite current demographic trends towards an increasing number of senior PCa patients with unfavorable tumor characteristics [2]. This trend is globally set to grow due to a COVID-19 related observed decrease of over 50% in PCa diagnosis in senior men �70 years, which in turn leads to postponed diagnosis of advanced stage PCa and delayed treatment [23]. Furthermore, vital considerations are perioperative morbidity, functional outcomes and postoperative QoL, all of which must be weighed against treatment-related risks.
Our study showed several important findings: First, we confirmed trends of growing proportions of �75 yrs. men in high-volume centers [10]. This is in line with inverse stage migration observations and indicates a paradigm shift towards extending the indication for RARP and offering the operation to suitable senior patients [2]. It is of note that at the same time we observe a trend toward more aggressive PCa characteristics in the patients undergoing RP [24].
Second and consistent with the previous series, we demonstrated greater proportions of unfavorable pathological characteristics in senior vs. younger patients [2, 10,13,25,26]. Possible explanations for such differences between age groups include prostate Figure 1. (a) Cumulative incidence rates of erectile function recovery after robot-assisted radical prostatectomy, stratified according to age groups �75 vs. <70 years. (b) Cumulative incidence rates of urinary continence recovery after robot-assisted radical prostatectomy, stratified according to age groups �75 vs. <70 years. enlargement, associated under-grading at biopsy and potentially delayed PCa screening and treatment. Moreover, hypogonadism in aging men may also be associated with adverse pathology features [27][28][29]. Even with limited LE, unfavorable PCa represents a significant oncological risk for senior men, who do not receive local treatment such as RARP. Specifically, two large series reported 27-33% 15-years cancer-specific mortality rates in �75 years patients with high-risk PCa features and non-curative management [30,31]. When applying the same definitions to our cohort, 33 to 59% of our senior cohort fulfill those criteria, which reaffirms proper candidate selection for RARP. Third, despite advanced age and the associated increase of comorbidity burden, we could not identify differences of surgical duration, estimated blood loss and perioperative/30d complication rates, before and after PSM. Our findings are in line with previous RP series, which demonstrated a weak association of CCI with perioperative complications [32]. Moreover, high surgical expertise and rigorous anesthesiologic assessment at our institution, particularly cardiovascular   [4,5,14,15]. Specifically, several series, which mostly relied on animal models, clearly demonstrated that increasing age negatively affects the regeneration and reinnervation of peripheral nerve fibers after injury, being delayed and less effective [33][34][35]. Moreover, with increasing age the neuroimmune profile usually becomes pro-inflammatory, potentially further contributing to nerve decline after RARP [36]. Thus, even at the same baseline erectile function and after matching, our MVA findings are highly conceivable, i.e. that younger PCa patients have a significantly faster and/or better recovery of erectile function after RARP compared to senior patients. This is in line with senior age as the strongest predictor of erectile dysfunction and impaired nerve regeneration even beyond the context of RARP [37].
Fifth, there was no effect of age on urinary continence recovery, which was >85% in both groups at 12 months. These rates are consistent with high-volume centers, albeit contradicting their MVA findings [14]. Here, differences in sample size, definition of urinary continence, surgical approach, i.e. less than 27 RARP patients �75 years, surgical expertise and underlying tumor characteristics might explain opposite findings. To our knowledge, we are the first to also adjust for preoperative IPSS. It should be considered as imperative to account for higher burden of preoperative urinary symptoms in senior patients, which represents a great potential confounder. QoL comparisons revealed an overall postoperative improvement even at senior age.
Sixth, contrary to previous reports, age did not significantly influence BCR before and after PSM in our study. This is consistent with large series by Kumar et al. and Gurung et al, who reported comparable BCR and cancer-specific survival (CSS) rates in patients �70 yrs [12,13]. We cannot confirm previous reports from a large high-volume center that senior age is associated with greater hazard of BCR and MP [10]. Despite similar clinicopathological adjustment variables this might be explained by our additional adjustment for CCI, year of surgery, surgical expertise and higher granularity of pathological variables. We do not believe that the RARP approach attributed to survival-related findings compared to open RP. Our BCR findings translate to MP, i.e. age is no contributor within MVA.
Our study has limitations. First, patients were treated by specifically trained or high volume RARP surgeons in a high-volume center, therefore limiting the generalizability of our findings. However, based on large series that compare RARP with ORP and/or laparoscopic RP, we do not expect oncological differences between surgical methods [38]. However, senior patients might particularly benefit from lower blood loss and lower perioperative morbidity if treated with RARP [3]. Second, as with all other studies on this topic, our study is retrospective and originates from a single center. Third, the impact of patient preference vs. recommendation of the treating urologist/physician to choose RARP is unknown. Furthermore, only those men who were considered fit enough and screened for potential cardiovascular conditions preoperatively were treated with RARP, resulting in a certain preselection bias. Fourth, our study relies on CCI but does not utilize a dedicated geriatric assessment tool such as G-8 geriatric screening tool or the Comprehensive Geriatric Assessment tool, to better determine physical and cognitive fitness [22,39].

Conclusion
Our study refutes several preconceptions of senior age thresholds and functional as well as oncological outcomes after RARP. We were able to demonstrate high rates of urinary continence recovery and improvement of QoL even at senior age. Moreover, we show that age itself does not significantly impair BCR-or MP-free survival after RARP, if pathological characteristics are taken into account. We do confirm however that advanced age is associated with more unfavorable PCa characteristics. In consequence, contemporary therapy of senior PCa patients should be based on an individually tailored treatment counseling with focus on tumor-and functional characteristics than just chronological age. Overall, our findings suggest that offering RARP to fit and selected senior PCa patients is a suitable treatment option regarding oncological and functional outcomes.

Ethical approval
The institutional review board at the St. Antonius-Hospital, Gronau, approved the retrospective study design and access to the patients' medical records. All methods were carried out in accordance with the Declaration of Helsinki. Written informed consent was obtained from individual participants in the study. The authors declare that they have no conflict of interest according to the current manuscript. All authors of this research paper have directly participated in the planning, execution, or analysis of the study. All authors of this paper have read and approved the final version submitted. The contents of this manuscript have not been copyrighted or published previously. The contents of this manuscript are not under consideration for publication elsewhere.

Author contributions
SRLB had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: JHW, SRLB, and CW.

Funding
This study is supported by an Intuitive Surgical Clinical Research Grant ("Pathological, surgical, functional and oncological outcomes of prostate cancer patients treated with robot-assisted radical prostatectomy, who represent great surgical challenges", ID: 22047, Attestation Date 18 September 2020).

Data availability statement
Full availability.