European Urology

3.1.1. Prevention is better than cure: claim for intelligible discussion with the patient about the true prevalence of postoperative erectile dysfunction

The conviction that clinicians should comprehensively discuss with the patient—and, whenever possible, with his partner [8]—the locally and internationally published ED prevalence rates at the earliest point in the decision-making process (LE: 4) [7] continues to grow. At the same time, the prevention of a poor functional outcome passes through the objective description of the limitations of current scientific literature in seriously defining postoperative ED prevalence rates [7], [9], [10], and [11].

Recent rigorous reviews about postoperative ED prevalence [9], [10], and [11] already exist, and a detailed meta-analysis of the most updated postoperative ED rates is not the focus of this review. However, to understand how to prevent and promote recovery of satisfactory EF after RP, we consider it relevant to discuss some key points in terms of prevalence and predictors of postoperative ED. In this sense, it is important to present data in an effort to communicate how difficult it is to approach this issue in a systematic way. Overall, the incidence of post-RP ED varies between 14% and 90% [9] and [10]. As a whole, the incidence of reported postoperative ED is extremely discrepant among series because of a great variation in the nature of the populations studied and the modality for data collection and reporting [9], [10], and [11]. The crucial point in this direction is the lack of an objective, reliable, and shareable definition of both baseline EF and postoperative EF. This means that there is great inconsistency in what is considered normal EF before surgery and what one may consider a normal erection after RP [9] and [11]. The results of the meta-analysis conducted by Tal et al. [11] clearly stated that most of the published literature does not meet strict criteria for reporting post-EF recovery. The same meta-analysis found as many as 22 different definitions of favourable EF outcome. In this context, the overall fixed-effects EF recovery rate was 58% (95% confidence interval, 56–60%), with significant heterogeneity among effect impacts [11].

Similar results were discussed by Burnett et al. [9] as a side analysis of outcomes assessed as part of an update of the American Urological Association PCa guidelines. They reported how, in many instances, only qualitative and subjective determinations were used, thus interpretable at best as erectile ability insufficient for sexual intercourse (complete ED), somewhat functional for sexual intercourse (partial EF), or regularly sufficient for sexual intercourse (intact EF). Moreover, it has been well observed that EF data from the largest series often refer only to a small fraction of the total RP population, and the evaluable number of patients represents but a fraction of the total exposed to surgery [11]. These are the most significant obstacles to the accurate interpretation of the incidence and prevalence of post-RP ED, when considering all study types and all technical approaches [7], [9], [10], and [11].

To our knowledge, there are no randomised prospective well-designed head-to-head comparative studies that prove EF outcomes are significantly different (namely, better) between open and laparoscopic or robotic RP (LE: 4) [10], [11], [12], and [13]. In this context, laparoscopic and, even more, many large series of robot-assisted approaches are mature enough and have demonstrated that EF outcomes are at least equivalent to those obtained with open retropubic RP (LE: 2B, 3, 4) [10], [11], [12], [13], [14], [15], [16], [17], and [18]. Given the huge discrepancy in the definition of the outcome in question, Tal et al. reported that differences were found in EF recovery among open (57%), laparoscopic (58%), and robotic RP (73%) [11]. A more thorough and critical discussion of this aspect is reported across this review.

Accordingly, the words of the ICSM committee sound even wiser (LE: 4): “[P]atients should be given individualised outcomes based on surgical technique, patient, and surgeon factors, thus including accurate data on EF recovery from their own patient population”[7]. This last aspect, which seems polemic against a certain type of behaviour, actually allows clinicians to bypass the errors of the scientific literature, eventually providing patients with more realistic expectations.

3.1.2. Prevention is better than cure: defining what is meant by (adequate) postoperative erectile function passes through the objective assessment of the baseline condition

It has been widely demonstrated that preoperative EF is an important predictor of EF recovery after RP [19], [20], [21], [22], [23], [24], [25], [26], and [27]. To this specific purpose, it is mandatory to verify how the definition of baseline EF was created; indeed, numerous means of assessment have been reported in the literature, including patients’ subjective self-reports, partner-corroborated function, and validated psychometric instruments. In a recent critical reappraisal of the literature, Mulhall reported that a clear mention of baseline EF was only given in 16 (66%) of 24 studies considered as representative of the experience of large-volume worldwide centres [10].

It was a further recommendation of the ICSM committee that clinicians use a validated psychometric instrument with recognised cut-offs for normalcy and severity during the pre- and postoperative evaluation of their patients [7]. Among the other available psychometric tools, the International Index of Erectile Function (IIEF) [28], with a dedicated EF domain, or the Sexual Health Inventory for Men (SHIM) [29]—with cut-off scores for “normal” EF of 26 and 21, respectively [28], [29], and [30]—continue to be highly recommended and widely used to define EF.

Baseline EF assessment is still problematic, and the most reliable timing of this assessment is even more poorly defined [7]. Indeed, on the one hand, some patients might overestimate their previous EF; on the other hand, proximity to the surgery may reduce the sexual activity or desire of the patient, his partner, or the couple as a whole [20] and [22]. PCa diagnosis and cancer-related or treatment-associated psychological distress may also significantly impair the real-time assessment of EF immediately prior to RP; therefore, this “late” real-time evaluation might not be fully representative of the patient's true sexual functioning [7], [20], and [22]. Kim et al, for instance, recently reported that to accurately assess the prediagnostic baseline EF in candidates for RP, the psychometric tools should be administered before prostate biopsy rather than before robot-assisted laparoscopic RP (RARP) because cancer diagnosis–related symptoms and depression can ultimately affect sexual function and index scores [31].

The US National Institutes of Health (NIH) consensus conference defined ED as “the consistent inability to obtain and/or maintain an erection sufficient for satisfactory sexual performance”[32]. This definition was not designed to be applied to a specific population and even less for a complex population—that is, patients undergoing RP. This is of particular importance today because such a widely used definition does not include any mention of the role or use of erectogenic aids or other assistance. In this context, data from clinical studies show that phosphodiesterase type 5 inhibitors (PDE5-Is) are efficacious in the treatment of post-RP ED [7] and [33]; even more importantly, in any series reporting postoperative EF outcomes, the proportion of patients using PDE5-Is—or second-line therapies—is of great relevance [1], [7], and [33]. Likewise, in many reported series, it is extremely difficult to define whether the patients are having spontaneous, unassisted, or PDE5-I–assisted erections when assessing both preoperative EF and post-RP EF recovery. Thus it must be certainly emphasised that the definition of adequate EF in post-RP studies is unique for the ED literature because most studies include PDE5-I responders as those with adequate EF [33] and [34].

Adequate knowledge of the potential use of any preoperative or post-RP erectogenic aid and what proportion of data in any given series was collected from patients using these agents is of major importance in allowing an accurate interpretation of the findings. These aspects make giving a proper definition of postoperative EF extremely difficult in addition to accurately defining a rate of post-RP ED [7], [9], [10], and [11].

The ICSM committee interestingly outlined a further crucial aspect: It is not always possible to take EF “back to baseline” postoperatively [7]. This last aspect is considered the keystone of the whole matter of prevention and maintenance of postoperative EF. Overall, a significantly small proportion of preoperative potent men appear to return spontaneously to baseline EF after surgery [9], [10], [11], and [35]. To the best of our knowledge, this last sentence could easily be applied to open RP series, but the item has scarcely been investigated for laparoscopic RP [11], [34], and [36] and probably even less so for RARP [11]. So, with the clear intention of limiting false patient expectations, it is important that clinicians comprehensively discuss the objective of regaining an erection equivalent to that prior to RP, especially using data from their own RP population. If we start from the NIH definition of ED, the focal point would become the fact that adequate postoperative erections may be “at least” sufficient for satisfactory sexual performance [32]. Mulhall [10] reported that adequate EF was defined as the “ability to have successful intercourse by patient self-report” in 42% out of the series. Unfortunately, overall, 37% of the series did not even report how the authors define adequate erections[10]. Satisfactory sexual relations is a complex issue; simplistically, it can be considered an expression of the relationship of several factors, including patient age, preoperative sexual activity, pre- and post-RP libido, baseline (namely, immediately preoperative) and previous EF (that is, remote from the diagnosis of disease) [22], preoperative and post-RP use of erectogenic aids, penile morphology issues, OF, mental health of the patient, continence, and micturition disorders [10].

Certainly, the important aspects of cancer control outcomes cannot be overlooked; therefore, we cannot overlook the possibility that the patient might also receive androgen-deprivation therapy or radiation therapy as an adjuvant treatment, with all the negative implications that these therapies may have on the sexual functioning of the patient. No less important is the impact of the partner and the couple's relationship [8], [10], [37], [38], and [39].

In an attempt to combine the two concepts of getting EF “back to baseline” and being able to achieve erections sufficient for satisfactory sexual intercourses as defined by the NIH [32], it becomes important to consider the potential misinterpretation of the established cut-offs for normalcy as well as the grading of ED severity (mild, moderate, and severe) of both the IIEF-EF domain and SHIM. Indeed, using the usual cut-offs for normal function, which are generally considered for the broad-spectrum ED population, might be overly stringent for post-RP populations; indeed, there are patients who still consider themselves fully functional while having lower scores [9], [40], and [41]. To try to define a cut-off value as close as possible to the real-life setting, Briganti et al. [41] considered a relatively small cohort of preoperatively fully potent patients (IIEF-EF ≥26) treated with retropubic BNSRP at a single institution. All patients reached an IIEF-EF ≥17 following surgery, which should be usually considered as mild to moderate ED. Moreover, the authors considered intercourse satisfaction (IS) and overall satisfaction (OS) as assessed by the corresponding domains of the IIEF administered prior to and after surgery; after a 26.7-mo mean follow-up, findings demonstrated that mean postoperative IIEF-OS and IIEF-IS domain scores, assessed at the time of EF recovery, were comparable for patients reaching an IIEF-EF of 22–25 points and for those men scoring postoperatively ≥26. Conversely, mean IIEF-OS and IIEF-IS domain scores of these patients were significantly higher compared with patients reaching an IIEF-EF domain score <22 after surgery. Therefore, based on the results of their analysis, the authors concluded that a cut-off of IIEF-EF ≥22 may represent a reliable score for defining EF recovery after BNSRP (LE: 4) [41].

3.1.3. Prevention is better than cure: defining time to recovery and postoperative quality of erection is a compulsory task for the clinician

Facing the various aspects of EF recovery with the candidate for RP or with the patient after his operation, the issue of the chronology of events has to be addressed. Burnett et al. [9] rightly commented that in the modern era of RP, most men usually achieve resumption of all physical activities, recovery of urinary control, and normalisation of bowel function within a few months after surgery. In contrast, postoperative EF continues to improve over time, at least up to 24 mo and in some series up to 48 mo post-RP [9], [10], [19], [42], [43], [44], and [45]. In this context, studies limiting follow-up assessment to <24 mo likely underestimate EF recovery [10]. Tal et al. [11] showed that studies reporting ≥18-mo follow-up outlined somewhat higher EF recovery (60%) rates compared with studies with <18-mo follow-up (56%) (p = 0.02). Overall, RARP seems to promote a more rapid EF recovery as compared with open RP (LE: 4) [34]. A growing amount of data are published almost daily to reinforce and better specify this type of widespread feeling.

Time of erection recovery does not uniformly occur in all cases, and several predictors of EF recovery have been identified, including patients’ age at surgery (ie, the younger, the better), better preoperative EF, extent of neurovascular bundle (NVB) preservation, and erectile haemodynamic changes after surgery [7]. In this context, surgery (ie, type, quality, surgical volume, and the actual nerve-sparing approach) probably emerges as the most compelling aspect. For instance, Tal et al. reported that BNSRP was certainly associated with higher EF recovery (namely, 60% vs 47%; p = 0.01) than UNSRP [11]. However, as perfectly stated by the ICSM committee [7], when dealing with preservation of the NVBs, most patients—and, unfortunately, a number of clinicians—do not have an adequate understanding of the concept of nerve sparing. Indeed, they think that nerve sparing always leads to complete preservation of the nerves and, consequently, to the absence of any transient postoperative ED. Even when surgeons believe that they have achieved complete bilateral cavernous nerve preservation, there is inevitably some trauma to the nerves. As outlined, to prevent false and unrealistic expectations, clinicians must provide patients with a realistic time frame for EF recovery; experts suggest that a potential period of 6–36 mo would be necessary, although most individuals experience functional recovery within 12–24 mo after RP [7].

Interestingly enough, Katz et al. showed that the recovery of functional erections in the early postoperative phase, especially without the need for PDE5-Is, is a good prognostic indicator for EF at 12-mo assessment [46]. However, a distinct cohort of men lose functional erections within 6 mo after surgery, with a time to nadir in postoperative EF not immediately after RP but more likely around 3 mo after surgery [46]. In terms of postoperative management, it is important to inform patients of this possibility, as it has an impact on their QoL and, potentially, on their compliance with post-RP therapy for ED. Similarly, it should be stated that the ability to have either a spontaneous or a pharmacologically assisted functional erection (namely, PDE5-I sustained) within 3 mo of RP is an excellent prognostic indicator [10].

Overall, the fact that it can take a long time until the first spontaneous erections occurs should not lead the physician to wait inactively. Conversely, the patient should start with supportive medication therapy for EF recovery as early as possible [7], [33], and [47].

As a second major aspect, the clinician should debate issues concerning the quality of erection. Unfortunately, science has only a few parameters that may help to characterise this aspect, for which the patient's actual perception is always a key point, although often not clearly decipherable. In this sense, capacity of recovering an erection sufficient for successful intercourse has caused the neglect of two important concepts: the quality of the erection itself, otherwise defined as erection hardness, and the consistency of functional erections.

3.1.4. Prevention is better than cure: choosing the right patient at the right time for the right surgery

The complexity of the issues we have touched on throughout this review leads us to suggest that prevention and management of EF in patients treated with RP necessarily is heavily predicated on a careful choice of the correct patient for the correct type of surgery. This fits perfectly with what has recently been highlighted by the ICSM committee [7]: Clinicians should comprehensively discuss the recognised predictors of EF recovery with any candidate for RP. Although this would seem implicit, it is of paramount importance to note that this discussion should be taken into account both considering and regardless of the oncologic indication [1]—in other words, all patients with PCa who might benefit from RP as a curative treatment should also receive appropriate counselling regarding the possibility of being subjected to RP (see, in this context, the most updated EAU guidelines [1]) and the possibility of being subjected to NSRP or, conversely, the need to be treated with non-NSRP, according to the baseline oncologic condition [1]. Interestingly, Imbimbo et al. [50] investigated factors related to patients’ desire to preserve post-RP sexual activity and those determinants for surgeons’ final decision to eventually perform NSRP. Out of a multicentre cohort of 2408 men, 1667 (69%) were interested preoperatively in preserving their sexual activity. Of the entire cohort, 13% were not interested but suitable for BNSRP, 18% were neither interested nor suitable, 39% were both interested and suitable, but up to 31% were interested but not suitable. Age (p = 0.001) and normal preoperative sexual function parameters (all p≤ 0.008) emerged as independent determinants of patients’ desire to preserve postoperative sexual functioning. Overall, 318 (13%) underwent UNSRP and 856 (36%) BNSRP. Along with oncologic indications, age (p = 0.026) and patients’ desire to preserve sexual activity (p < 0.001) were among the main independent determinants of surgeon's final decision for NSRP. Overall, findings from this study underlined the existence of discrepancies among patient's desire to preserve postoperative sexual function, guideline indications for NSRP, and surgeons’ final decision for a nerve-sparing approach [50].

Once it has been established that the patient can receive NSRP, counselling should focus on the patient's baseline functional situation along with the potential predictors of postoperative EF recovery. Preoperatively, these factors necessarily include age at surgery, baseline EF, and comorbidities. Likewise, type of surgery, coupled with surgical volume and surgical skill, and nerve-sparing status achieve the role of significant contributors to EF recovery [7].

As seen, baseline EF is related to several different variables, including age and comorbid conditions status. In this context, Rabbani et al. [19], for instance, demonstrated that at 36-mo follow-up, approximately 70%, 45%, and 30% of patients had recovered EF when ≤60 yr of age, between 60 and 65 yr of age, and >65 yr of age, respectively. More recently, Penson et al. [51] confirmed those previous findings using data from a cohort of 1288 patients from the Prostate Cancer Outcomes Study, with a postoperative 5-yr follow-up. The authors observed a significant trend effect of age on sexual function following BNSRP, with 61% of men ≤55 yr of age compared with 49% of those between 55 and 59 yr of age, 44% of those 60–64 yr of age, and 18% of those >65 yr of age having functional erections (p < 0.001) [51]. As a point of criticism, both of the previous two studies clearly detailed a long list of comorbid conditions at baseline, but they did not consider them during any multivariate analysis. In contrast, Solomon et al. [25] recently reported that baseline age and comorbidity-dependent ED affects up to 48% of all patients prior to open RP. The importance of the result is that although comorbidities were not considered for their actual weight as obtainable using a validated index, age, body mass index, type 2 diabetes mellitus, and depression emerged as independent predictors of baseline ED (all p≤ 0.005) [25].

Overall, it is certainly true that the presence, number, and potentially severity of those comorbidities (eg, vascular risk factors such as dyslipidemia, diabetes mellitus, hypertension, coronary artery disease) afflicting the patient well before the RP will likely afflict that man even after surgery, adversely affecting EF recovery, just as usually happens in the general population [52].

Recently, Briganti et al. [47] developed a preoperative risk-stratification tool aimed at assessing the probability of EF recovery after open BNSRP. They used routinely available baseline data such as patient age and preoperative EF as psychometrically objectified with the IIEF; moreover, as a proxy for general health status, the authors scored health-significant comorbidities using the Charlson Comorbidity Index (CCI) [53]. For the specific purpose of the analysis, CCI was categorised as a score of 0, 1, or ≥2. The resulting tool was able to stratify patients into three groups according to the relative risk of ED after surgery: low risk (≤65 yr of age, IIEF-EF ≥26, CCI ≤1), intermediate risk (66–69 yr of age or IIEF-EF 11–25, CCI ≤1), and high risk (≥70 yr of age or IIEF-EF ≤10 or CCI≥ 2). Overall, the EF recovery rate was 58% at 3-yr follow-up; however, the 3-yr EF recovery rate significantly differed among the three groups, being 85%, 59%, and 37% in patients at low, intermediate, and high risk of postoperative ED, respectively (p < 0.001). More recently, Novara et al. [26] applied the same risk-stratification categories to a cohort of 208 patients treated with bilateral nerve-sparing RARP who were assessed at a minimum 12-mo follow-up. Data showed that 62% of the entire cohort was potent 12 mo after surgery, with age at surgery (hazard ratio [HR]: 2.8; p < 0.001), CCI (HR: 2.9; p = 0.007), and baseline EF (HR: 0.8; p < 0.001) achieving independent predictor status for potency recovery. According to the risk group stratification proposed by Briganti et al, the 12-mo EF recovery rate was 82%, 57%, and 29% in the low-risk, intermediate-risk, and high-risk groups, respectively (p < 0.001) [26].

Overall, prevention and management of postoperative EF necessarily pass through rigorous selection of the patients who may benefit from RP; this should not be seen as a kind of discrimination against some patients but rather a strong awareness by the clinician and the most respectful form of aid that the patient can receive. Data in hand so as to avoid generating false expectations [54] and [55], the clinician must clearly say to the patients that NSRP does not ensure the recovery of erections at all and, because it may be potentially subjected to a number of sequelae, a nerve-sparing approach should be reserved for young patients without significant comorbidities and an adequate preoperative EF [7], [19], [26], [47], [25], [56], and [57].

In this context, it is certainly important to remember that there are literature data that seem to suggest that bilateral nerve-sparing surgery could also significantly contribute to the preservation of postoperative urinary continence. Indeed, Burkhard et al. [58] reported that in their series of 536 patients who underwent open RRP, the incidence of postoperative urinary incontinence was lower and continence was highly associated with a nerve-sparing technique. Therefore, they concluded that nerve sparing should be attempted in all patients if the principles of oncologic surgery are not compromised [58]. However, whether this association equals causation is unclear. It is possible that nerve-sparing surgery is a surrogate for meticulous apical dissection, which may be the pathway to continence preservation.

In the general sense, however, for patients who cannot undergo a nerve-sparing approach, a technically flawless surgery without preservation of NVBs has to be suggested, with the specific aim of allowing them to have a satisfactory sexual activity assisted by second- and third-line therapies [7], [59], [60], [61], and [62].

3.1.5. Prevention is better than cure: choosing the right cancer for the right surgery

Since NSRP was introduced, it has continuously been accompanied by the discussion about the right indication for this surgical modification [63] and [64]. Nerve sparing should not compromise the effectiveness of RP by producing artificial surgical margins that could have been avoided if a non–nerve-sparing approach had been chosen. Therefore, an estimation of organ confinement of the cancer is crucial, with the idea that in such cancers, nerve sparing could be safely indicated. Tailoring the NSRP approach according to the cancer extension of each subject is important for saving as many NVBs as possible to improve outcome with respect to sexual function.

Despite recent improvements, the reliability of imaging techniques to securely identify microscopic extracapsular extension is still limited [65]. In this context, Graefen et al. [66] published the first nomogram that specifically estimated the likelihood of extracapsular cancer growth to aid the indication of nerve sparing. Since then, several other statistical models have been created incorporating readily available information of the cancer to estimate organ confinement [67] and [68].

However, all available models deliver only a likelihood of organ confinement while leaving the patient and the surgeon with a degree of uncertainty. Therefore, several institutions have introduced intraoperative frozen sections to identify cancer tissue at the resection margin during NSRP [69]. This approach may allow surgeons to tailor the surgical approach during the procedure (eg, secondary resection of an initially preserved NVB). Even though this procedure seems to be the safest approach for indicating nerve sparing without the risk of an artificial surgical margin, it certainly remains time consuming. It is recognised that the positive predictive value of frozen-section analysis for positive surgical margins (PSM) is high; however, the sensitivity is too low to expect that a policy of routine frozen-section analysis will be offered outside specialised centres [70]. However, to preserve as many NVBs as possible, it is crucial that a reproducible approach indicating whether a nerve bundle will be preserved be used to give the patient a clear idea of the sexual function he can expect postoperatively.

3.2.1. Open nerve-sparing radical prostatectomy is still common

Open NSRP—mainly retropubic RP—is still a common surgical approach for operable PCa [79], [80], [81], and [82]. Budäus et al. reported the most updated description of open intrafascial NVB dissection coupled with the 12-mo postoperative functional outcome [82]. They used the abridged five-item version of the IIEF to assess EF, and patients were asked to complete the tool before surgery and 12 mo after RP [25] and [82]. The analyses excluded those men with a preoperative IIEF score <19, indicating some degree of ED, and the patients who stated that they had not had had sexual stimulation within the past 6 mo, according to question 2 of the IIEF-5 questionnaire. Defining as potent those men who reported erection sufficient for penetration after sexual stimulation, EF recovery rates varied from 84% to 92% in men who underwent BNSRP and from 58% to 70% following UNSRP. The authors also reported that the use of PDE5-Is was left to the patient's discretion, and 80% of men who completed the questionnaire did not use such medications, thus ultimately meaning spontaneous recovery of postoperative EF. Moreover, Budäus rigorously outlined that if a “strict” definition of EF was applied (namely, a postoperative IIEF score >19), the corresponding potency rates were 25–59% [82]. One must presume that if higher EF domain scores were used as a cut-off, EF recovery would be even lower.

3.2.2. Laparoscopic nerve-sparing radical prostatectomy can be performed extraperitoneally or transperitoneally

Laparoscopic or endoscopic NSRP can be performed extraperitoneally or transperitoneally [83] and [84], and this further heterogeneity is responsible for an even more complex objective assessment of the functional outcomes across the literature. Greco et al. [85] reported the results of a retrospective, parallel-arms study comparing the 12-mo follow-up functional outcomes in patients undergoing intrafascial retropubic and laparoscopic NSRP, both performed by high-volume surgeons. At the time of assessment, 66% of patients in the laparoscopic group and 51% in the open group reported being able to engage in sexual intercourse, respectively (p < 0.05). A comparison between intrafascial and interfascial laparoscopic extraperitoneal BNSRP in preoperatively potent patients was recently reported by Stolzenburg et al. [86]. Considering as potent those men responding positively to several specific questions on the sexual encounter profile diary, 12-mo postoperative potency rates were 94% (<55 yr of age), 83% (55–65 yr of age), and 60% (>65 yr of age) in the intrafascial group compared with 77%, 50%, and 40%, respectively, for men who underwent interfascial BNSRP according to the same age stratification. The authors thus concluded that intrafascial laparoscopic BNSRP overall provides significantly better EF recovery rates [86].

More recently, using the dichotomous definition of erections satisfactory or unsatisfactory for intercourse, Stewart et al. confirmed this significant difference between the two dissections, also highlighting that men <60 yr of age may certainly achieve the highest functional results with the intrafascial laparoscopic approach [87] and [88]. Conversely, Neill et al. [89] did not find a significant difference in terms of either continence or EF recovery rate after intrafascial versus interfascial extraperitoneal laparoscopic NSRP. Overall, the rates of men who actually have spontaneous erections were not even considered in most of the previous studies. This does not permit us to provide any clear-cut conclusion either concerning the type of surgery (namely, laparoscopic RP) or the appropriateness of the selected anatomic dissection.

3.2.3. Intrafascial vs interfascial bilateral nerve-sparing radical retropubic prostatectomy outcomes have also been considered for robot-assisted laparoscopic radical prostatectomy

The outcomes of intrafascial versus interfascial BNSRP techniques have more recently be considered for RARP, as well. Potdevin et al. [90] reported that EF recovery rates at 3, 6, and 9 mo in the intrafascial group were 24%, 82%, and 91%, respectively, whereas in the interfascial group, they were 17%, 44%, and 67%, respectively, at the same follow-up time points. Erectile function was defined as being able to achieve erection adequate for penetration more than half of the time (namely, Expanded Prostate Cancer Index Composite [91] questions 18 and 19) with or without PDE5-Is. As a major point, the authors outlined that the intrafascial technique was associated with higher PSM rates in patients with pT3 disease, although no significant differences emerged in pT2 disease [90].

Xylinas et al. recently supported the idea that a formal robot-assisted intrafascial approach provided early (ie, 1-mo assessment) satisfactory functional results with respect to postoperative continence and potency [92]. Interestingly, these authors outlined that 60% of their 50 consecutive patients exhibited erection sufficient for intercourse; of these latter patients, 50% did not need to use any PDE5-I while 50% required an oral compound [92]. Asimakopoulos et al. [36] reported the results of a prospective randomised study aimed at comparing the functional results of 128 consecutive patients treated by a single experienced surgeon with either laparoscopic RP or RARP with the intent of using intrafascial bilateral nerve sparing in all cases. The 12-mo evaluation of patients’ EF—defined as capability for sexual intercourse—showed that RARP scored significantly better (p < 0.0001) than laparoscopy. Men in the RARP group also showed a faster time to capability for intercourse (p = 0.0001) and a higher rate of return to baseline IIEF-EF domain scores (p = 0.0002) than those operated on with laparoscopic RP [36]. Conversely, a similar rate of patients with functional erections did use PDE5-Is within the two groups (55% vs 47%, respectively; p = 0.58).

Despite the belief that there are no sufficient prospective randomised and rigorously conducted clinical trials to confirm this unequivocally—especially with external validation of the data already published—there is an increasing school of thought that RARP eventually may be shown to be the most potency-protective surgical approach [13] and [36]. In this context, the circumferential extent of fascia preservation—otherwise known as intrafascial surgery—and patient age [93] as well as preoperative EF and patient comorbidity [26] emerged as the best independent predictors of postoperative EF even for RARP.

3.2.4. The importance of intraoperative preservation of erectile haemodynamics

Accumulating evidence suggests that a certain amount of men have vascular abnormalities after RP [75]. These abnormalities can generally be traced to two main forms: arterial insufficiency and venous leakage [75], [76], and [94]. In this context, Mulhall et al. clearly showed that in a cohort of men who underwent open BNSRP, had ED, and never received any pharmacotherapy, up to 75% had some form of vascular alteration. Overall, 59% of the patients had arterial insufficiency, and 26% had venous leakage [75].

Arterial insufficiency has been attributed to the trauma to one of two types of accessory pudendal arteries: a lateral or aberrant pudendal artery that runs along the fascial tendinous arch of the pelvis in the groove between the bladder, prostate, and pelvic sidewall, and an apical accessory pudendal artery that is located inferiorly and laterally to the Denonvilliers’ fascia, in proximity to the anterolateral aspect of the prostate apex [74]. Between 4% and 75% of all men have accessory or aberrant pudendal arteries [74], [77], [95], and [96], depending on which type of series is reviewed (cadaveric, angiographic, or operative). These aberrant arteries may originate from the internal or external iliac or obturator arteries (see, in this context, Walz et al. [73] for a more comprehensive review). Although still controversial [24], [78], and [96] when reconsidering the way we surgically perform BNSRP, the importance of preserving accessory pudendal arteries comes from several observations that support the concept that they could be solely responsible for arterial blood supply to the corpora cavernosa [75] and [77]; if this were the case, preservation of these arteries during any surgical approach (open, laparoscopic, or robot assisted) would be compulsory to avoid ED secondary to penile arterial insufficiency [74] and [76]. Ohebshalom et al, indeed, clearly showed a significant difference in terms of self-reported erectile rigidity (p = 0.013) and IIEF-EF domain scores (p = 0.025) at 18-mo follow-up between those patients who had normal postoperative penile haemodynamics and those with abnormal haemodynamics [97]. Likewise, the percentage of patients with normal EF domain scores (p < 0.01), the percentage of patients positively responding to sildenafil (p = 0.03), and the percentage of patients with functional erections permitting pharmacologically unassisted sexual intercourse (p = 0.018) were significantly higher in men with normal haemodynamics [97]. Overall, published data support the concept that accessory pudendal arteries may have a role in preserving post-RP EF as well as in promoting postoperative recovery of EF [76]. Surgical approaches with a clear intrafascial goal have been postulated to eventually allow a more stringent preservation of any accessory pudendal artery [98].

Failure to recover EF after RP may also result from venous leakage as a sequela of neuropraxia-induced and possibly the absence of cavernosal oxygenation-associated erectile tissue damage [78]. Overall, data seem to suggest a strong clinical correlation between postoperative time and interval of EF recovery and the incidence of venous leakage [75] and [76], with previous findings showing an incidence of 14% at <4 mo postoperatively that becomes 35% at 9–12 mo post-RP [75]. As a whole, these reports suggest that venous leakage portends a poor prognosis for EF recovery, with a high probability of never recovering preoperative EF, to a poor response to PDE5-Is and even an unsatisfactory outcome with intracavernosal injection therapy (ICI) [78] and [97]. All of these observations outline the importance of performing nerve-sparing surgery whenever possible. Therefore, surgeons should not only consider those men who might either preserve or recover spontaneous, pharmacologically unassisted postoperative erections but also all of those individuals who are destined to need second-line erectogenic pharmacotherapy, who still deserve preservation of the cavernous tissue to ensure continued excellent response to ICI [78]. This ultimately would mean that, with the specific aim of preserving viable neural tissue while reducing the risk of long-term venous leakage, a bilateral approach should always be attempted when PCa characteristics allow BNSRP without compromising cancer control. Conversely, UNSRP should be at least provided when the oncologic characteristics do not allow a bilateral nerve-sparing surgery [78].

Interestingly, a recent suggestion for more precisely predicting postoperative EF outcomes comes from the use of a subjectively assigned intraoperative nerve-sparing score that allows the surgeon to self-define the quality of intraoperative nerve sparing (1 = complete preservation, 4 = complete resection) for each patient [99]. In this context, in multivariate analyses, lower nerve-sparing scores achieved independent predictor status for 24-mo EF recovery (p = 0.001), as did age (p = 0.001) and baseline IIEF-EF domain score (p = 0.02) [99].