Is There a Rationale for Penile Rehabilitation Following Radical Prostatectomy?

Abstract

Erectile function recovery after radical prostatectomy (RP) is an increasingly prominent quality-of-life outcome following surgery. Following RP many men, despite the advent of cavernous nerve-sparing surgical technique, have moderately or significantly impaired erectile function (EF). The term penile rehabilitation (PR) is used to define interventions that maintain the health of erectile tissue in the context of nervous, vascular, and structural tissue injury. The goal of PR is to regain, as closely re-approximate, preoperative erectile function. PR is based on an increasing volume of preclinical and clinical data, but conclusive evidence of efficacy has not been established, and therefore the concept of PR remains controversial. The optimal PR regimen has not been established, but all strategies rely on one or more erectile dysfunction treatments to be administered on a regular basis regardless of actual use for sexual activity. This review highlights recent studies and evidence related to PR.

Introduction

Prostate cancer is a common diagnosis, with an estimated annual incidence of 190,000 cases in the United States in 2009 (“What are the key statistics about prostate cancer,” 2012). Since the prostate specific antigen test was introduced more than 20 years ago, serum screening and increased awareness of prostate cancer have changed the nature of the disease for the vast majority of affected patients. Long-term functional outcomes, namely, sexual and urinary function, are increasingly important to patients as oncological outcomes have improved.

The mainstays of treatment for localized prostate cancer are radical prostatectomy (RP) and radiation therapy (RT). Both therapeutic modalities are associated with a high risk of erectile dysfunction (ED). According to a recent meta-analysis, ED prevalence rates in contemporary RP series are approximately 42% overall and 50% in studies in which nerve-sparing is described (Tal, Alphs, Krebs, Nelson, & Mulhall, 2009). The prevalence of ED following RT varies widely in the literature. A recent review of studies published since 1995 reported ED rates ranging from 6% to 51% after brachytherapy monotherapy and 25% to 89% in men who underwent combined brachytherapy and external beam radiation therapy (Stember & Mulhall, 2012).

Sexual function has been demonstrated to be an extremely important quality of life factor. In a survey of men without prostate cancer, more than 67% of respondents reported that they would exchange a 10% advantage in 5-year survival to maintain erectile function (Singer et al., 1991).

Penile rehabilitation is defined as medical intervention to maintain penile structure and function after radical pelvic surgery or RT. This concept has been gaining significant traction worldwide, even in the absence of definitive data in its favor. Appreciation of the strategy of penile rehabilitation requires an understanding of a number of principles regarding the pathophysiology of ED following RP. The aim of this review is to summarize the evidence related to penile rehabilitation in men who have undergone RP.

Concepts Underlying the Strategy of Penile Rehabilitation

Normal Erectile Function

Penile erection is a neurovascular event that depends on intact cavernous nerve signal, arterial blood flow, and erectile tissue structure. In the flaccid state, cavernosal spaces are constricted and venous vessels are noncompressed. Tumescense is mediated by cavernosal nerve release of nitric oxide (NO) and other mediators to effect production of cyclic guanosine monophosphate (cGMP). The action of this second messenger, cGMP, causes smooth muscle (SM) relaxation in trabeculae and arterial vasculature. Another vasodilator, prostaglandin E1, also causes relaxation of SM in the same structures by increasing the concentration of cyclic adenosine monophosphate (cAMP; Williams, Liu, & Deyoung, 2005).

Both cGMP and cAMP activate SM relaxation via a decrease in intracellular calcium. As SM relaxes in the corporal sinusoids, they rapidly dilate as they become engorged with blood. SM relaxation in cavernosal arteries, with concomitant dilation and increased rate of flow, facilitates this process. Pressure increases in the penis because the engorged sinusoids become compressed within the relatively noncompliant tunica albuginea layer. As the venous plexus becomes compressed under this layer, venous flow significantly decreases, thereby resulting in a maintained rigid erection.

All the factors that are necessary for natural physiologic erection—that is, cavernous nerve signal, sufficient arterial blood flow, elasticity of erectile tissue, and the veno-occlusive mechanism of blood-trapping—are potentially compromised by RP.

Causes of ED after RP

Nerve Damage

In the flaccid state, cavernosal spaces are constricted, and venous vessels are noncompressed. In the erect state, SM relaxation in trabeculae and arterial vasculature results in increased blood flow, rapid filling, and dilation of the cavernosae. The result is a compression of the emissary veins against the internal surface of the tunica albuginea.

Despite improvements in nerve-sparing techniques during radical prostatectomy (RP), the prevalence of postoperative ED reportedly ranges from 14% to 90% and represents a significant detriment to the health-related quality of life of affected patients (Defade, Carson, & Kennelly, 2011). Damage to cavernosal nerves acutely results in transient or permanent loss of endothelial nitric oxide synthase (eNOS) and NO to the penis.

Surgical technique is critically important for EF preservation or recovery following RP. Complete impairment of EF was historically considered inevitable after surgical treatment RP. Since Walsh and Donker (1982) published anatomical studies that identified the course of the neurovascular bundle along the prostatic capsule, however, intraoperative changes have allowed for postoperative recovery of EF in many patients (Walsh & Donker, 1982). In a further demonstration of the principle, bilateral nerve-sparing technique yields improved erectile functional outcomes relative to unilateral nerve-sparing (Greco et al., 2011).

Although transection or thermal injury by electrocautery is widely recognized as the major threat to the cavernosal nerves during RP, nonmacroscopic factors can also cause nerve injury. Masterson, Serio, Mulhall, Vickers, and Eastham (2008) prospectively evaluated 275 men with prostate cancer to determine whether a modified surgical technique could affect postoperative EF. Specifically, the modified technique (MT) involved avoiding traction on the urinary catheter until the neurovascular bundles have been completely excised from the lateral aspects of the prostate. EF recovery of patients who underwent RP with the MT was compared with a cohort that underwent RP with the standard technique (ST). The ST differed from the MT in that it involved traction on the catheter to allow for easier isolation of the lateral vesicle pedicles. At 6 months after RP, 67% of men in the MT group had good EF recovery versus only 40% of men in the ST group (Masterson et al., 2008). These results demonstrate that traction alone, even in the absence of transection or thermal injury, may lead to cavernosal nerve injury/neuropraxia.

In addition to traction, it is possible that minimal manipulation or even exposure without any manipulation can affect the cavernous nerves. A study using a rat model of cavernous nerve injury demonstrated these findings: Even minor neural trauma subsequently impaired intracavernosal pressure relative to mean arterial pressure (ICP–MAP ratio) after cavernosal nerve stimulation. Furthermore, simple exposure of the cavernosal nerves during laparotomy, without any direct pelvic organ or nerve manipulation, also caused decreased ICP–MAP ratio (Mullerad, Donohue, Li, Scardino, & Mulhall, 2006).

The implication of these findings is that nerve-sparing prostatectomy, as visualized and defined by the surgeon, may not account for functional impairment of cavernous nerves due to traction or minimal manipulation. Neuropraxia during RP is important because it leads to three subsequent events that threaten EF: upregulation of fibrogenic cytokines (predominantly TGF-beta), SM and endothelial apoptosis, and production of collagen excess. These conditions play a role in the development of corporovenocclusive dysfunction (CVOD).

Structural Tissue Damage

In response to NO stimulation (or intracavernosal injection of erectogenic medication) the erectile tissue expands in a three-dimensional fashion. Existence of a subalbugineal venous plexus was first demonstrated by electron microscopy in cadaveric penile tissue in the 1980s, and compression of these venules during erection in a canine model was described in the same report (Fournier, Juenemann, Lue, & Tanagho, 1987).

Structural changes to erectile tissue such as fibrosis can, in an advanced state, render the erectile tissue incapable of expanding sufficiently to compress the subtunical venules. Because the veins remain patent and continue to drain the penis, even in the presence of NO or intracavernosal injection of erectogenic medication, pressure rises only moderately, and penile rigidity fails to occur. This principle was demonstrated by Nehra et al. (1996), who performed dynamic infusion cavernosometry (DIC) on men with ED who were scheduled for insertion of penile prosthesis. Cavernosal tissue was biopsied at the time of prosthesis implantation surgery. The hemodynamic parameters identified on DIC were correlated with ultrastructural description of the penile tissue for each patient. The authors reported that as SM percentage on biopsy decreased to 40% to 45%, DIC parameters were consistent with CVOD. The lower the percentage of SM, the higher the magnitude of CVOD (Nehra et al., 1996).

In a study of 19 men with normal EF (as defined by Rigiscan testing) who underwent RP for localized prostate cancer, Iacono et al. (2005) investigated structural changes to erectile tissue at various time points. The authors performed penile biopsy at the time of RP as well as 2 and 12 months postoperatively. Patients did not receive treatment for post-RP ED. Histologic assessment demonstrated that 2 months following RP, all patients had SM fibers largely replaced by trabecular collagen fibers. These effects were even more pronounced at 12 months (Iacono et al., 2005).

Mulhall et al. (2002) published a study that evaluated the incidence of penile hemodynamic abnormalities following RP. Study subjects included men with excellent partner-corroborated EF prior to surgery. All men underwent penile duplex Doppler ultrasonagraphy at various time points postoperatively. The incidence of venous leak (defined as end-diastolic velocity >5 cm/s) following RP was 10% at 4 months, 35% between 8 and 12 months, and 50% after 12 months. These findings demonstrate that the onset of venous leak is a duration-dependent phenomenon and are consistent with the concept of progressive onset of corporal fibrosis (Mulhall et al., 2002).

Cavernosal Oxygenation

In the flaccid state, the penile blood has oxygen saturation equal to venous blood (PO₂ = 40 mmHg). In the erect state, on the other hand, the penile oxygen saturation mirrors that of arteries (PO₂ = 100 mmHg). Beginning with the onset of puberty, men typically achieve 3 to 5 erections nightly. Even sexually abstinent men, therefore, achieve multiple erections daily. Although the precise physiologic role of nocturnal erections has not been defined, data suggest that regularly occurring erections are necessary to preserve cavernosal structural homeostasis. This concept, indeed, is the basis of penile rehabilitation.

Moreland initially proposed that penile oxygen tension, which is present at increased levels during erection, is a critical factor for modulating erectile function. His theory was based on in vitro evidence of upregulated fibrogenic cytokines in response to low cavernosal oxygen and elevated endogenous prostanoids (which reduce fibrogenic cytokines) in response to normal oxygen levels (Moreland RB 1998).

Using a rabbit model, researchers from Boston University demonstrated that atherosclerosis-induced ischemia leads to deleterious changes in cavernosal NO and NO synthase (Azadzoi, Master, & Siroky, 2004). In a separate study, the same group later demonstrated that cavernosal ischemia causes increased production of connective tissue along with decreased SM (Wang, Soker, Atala, Siroky, & Azadzoi, 2004).

In further support of the protective role of cavernosal oxygenation, Müller et al. (2008) revealed that rats exposed to hyperbaric oxygen therapy (HBOT) had recovery of intracavernosal pressure following cavernous nerve crush injury compared with those who had cavernous nerve crush injury but no HBOT (55% vs. 31%, respectively; Müller et al., 2008).

Although it is generally accepted that the cavernosa are fully oxygenated during rigid erections, Tal, Mueller, and Mulhall (2009) demonstrated that significant increases in oxygenation also occur during partial erections. To the extent that oxygenation may protect erectile tissue from injury due to hypoxia, these findings suggest that even partial erections following RP may be beneficial (Tal, Mueller, et al., 2009).

Arterial Damage

Accessory pudendal arteries (APA) are super-diaphragmatic vessels that have been thought to play a vascular role in erectile function, although the importance of the contribution is unclear. There is significant variation in pelvic anatomy and the incidence of APA presence ranges from 4% to 70% (Blander, Broderick, Malkowicz, VanArsdalen, & Wein, 1999).

Rogers, Trock, and Walsh (2004) retrospectively reviewed 52 men who underwent bilateral nerve-sparing RP with identification and preservation of APA. A control population of men without APA but who were otherwise matched for age, stage, and nerve-sparing status was identified. With potency defined as ability to achieve intercourse without the use of PDE5i, the authors revealed that APA more than doubled the likelihood of potency (Rogers et al., 2004).

Droupy et al. (1999) performed transrectal and perineal color Doppler flow imaging on 12 men immediately prior to radical pelvic surgery. Perineal ultrasound visualized APA in all patients and was performed before and during erection induced by intracavernosal injection of papaverine. Significant APA hemodynamic changes were visualized sonographically with the onset of erection. Presence of APA was intraoperatively confirmed for all patients. Droupy et al. (1999) concluded that APA likely play an important role in erectile physiology.

In contrast with the conclusions of these earlier studies, the findings of Box et al. (2010) suggested that APA were irrelevant to post-RP erections. Sexual function questionnaires and distribution of APA were prospectively recorded in 200 men who underwent robotic-assisted laparoscopic RP. Preoperative EF was not associated with presence of APA, which were identified in 80% of the cohort. Of men with APA, there were no significant differences in EF rate or recovery time between those who had APAs spared or sacrificed. Furthermore, no differences existed between men with and without APA. Box et al. (2010) concluded that there is no relationship between presence or absence of APA and preoperative erectile function or recovery of erectile function following RP.

Animal Data Regarding Penile Rehabilitation

Early animal models have helped identify three primary factors responsible for post-RP ED: neural injury, vascular injury, and SM damage (Mulhall et al., 2008). Neuropraxia resulting from crush, traction, or electrocautery injury to the cavernous nerves has been demonstrated to induce structural changes to penile SM and endothelium in rat models. For example, researchers at Northwestern University demonstrated that transaction of the cavernous nerves in rats led to loss of penile weight, significantly increased levels of apoptosis in SM cells, and an increase in extracellular protein consistent with corporeal fibrosis (User, Hairston, & Zelner, 2003). Work with other rat models has similarly linked cavernous nerve injury to SM and endothelial apoptosis, SM collagenization, and endothelial cell retraction (Klein et al., 1997; Leungwattanakij et al., 2003; Mulhall, 2008).

The second factor mentioned, vascular injury, refers to damage to the APA occurring during RP. Such vascular injury likely magnifies the erectile tissue changes resulting from cavernous nerve injury, as illustrated by Müller et al. (2008) using a rat cavernous nerve injury model. In this study, the authors concluded that HBOT following cavernous nerve injury helped preserve erectile function in rats (Müller et al., 2008). These results support the concept that cavernosal oxygenation is key to the recovery of erectile function and explain why vascular injury would be expected to exacerbate the aformentioned effects of neural injury. Thus, neural and vascular injury work together to produce the third major factor responsible for post-RP ED: corporal SM fibrosis.

A decrease in corporal elastic and SM fibers and a progressive increase in collagen content in the corpora postoperatively will ultimately prevent corporal muscle from expanding sufficiently to compress all of the subtunical venules, resulting in corporoveno-occlusive dysfunction (Mulhall et al., 2008).

Multiple animal models of ED after RP have demonstrated a benefit to routine PDE5 inhibitor use and supported a role for penile rehabilitation. In the first experimental demonstration that corporal veno-occlusive dysfunction develops after bilateral cavernosal nerve resection (BCNX), Ferrini et al. (2006) reported that rats undergoing BCNX demonstrated a 60% reduction in the SM/collagen ratio and a threefold increase in intracorporal apoptosis compared with the sham group. However, in this study, rats that were administered vardenafil after BCNX were characterized by increased markers of SM cell replication, an increased SM/collagen ratio, and a normalization of the DIC drop rate (Ferrini et al., 2006).

A second study designed to expand on this finding and to determine whether this effect was common to other PDE5 inhibitors also examined rats undergoing cavernosal nerve resection, this time ± sildenafil administration. As in the prior study, this investigation demonstrated that continuous long-term PDE5 inhibitor use prevented the alterations in corporal histology induced by cavernosal nerve damage (Kovanecz et al., 2008).

A third study of rats undergoing cavernosal nerve resection with or without daily tadalafil presented similar findings. In this study, tadalafil treatment normalized the high drop rate in the intracavernosal pressure measured by cavernosometry after cavernosal nerve resection. Moreover, tadalafil normalized the increase in penile shaft collagen content, the reduction in corporal SM cell content, the SM/collagen ratio, and the increase in apoptotic index. An additional study by Lysiak et al. also examined the effect of tadalafil following cavernous nerve resection and concluded that treatment with Tadalafil significantly decreased the number of apoptotic cells by activating survival-related kinases (Lysiak et al., 2008). Thus, sildenafil, vardenafil, and tadalafil all prevent corporal veno-occlusive disease and the underlying corporal fibrosis caused by cavernosal nerve damage in the rat. Together, these and other animal studies provide a strong rationale for the early use of PDE5 inhibitor use following RP.

Human Data Regarding Penile Rehabilitation

Iacono et al. (2005) evaluated histomorphological changes in cavernous collagen and SM content before and after RP. Corporal biopsies were performed just prior to RP as well as 2 and 12 months following RP. No structured penile rehabilitation was employed for any patients. At the first postoperative biopsy, SM content was significantly decreased (p < .0003), and collagen content was significantly increased (p < .0003) compared with preoperative biopsies. Furthermore, at 12 months post-RP, SM content was again significantly increased (p < .0003), and collagen content was again significantly decreased compared with the first postoperative biopsies.

In an effort to correlate corporeal function with structure, Nehra et al. (1996) studied 24 men with impaired EF who underwent penile prosthesis insertion. Preoperative parameters of erectile function on DIC were correlated with histopathological descriptions of corporal biopsy specimens taken at the time of surgery. SM corporal content below 45% was correlated with DIC flow-to-maintain (FTM) values consistent with veno-occlusive dysfunction. This study demonstrated, for the first time, that lower SM corporal content is associated with a greater degree of venous leak.

While multiple studies have revealed that PDE5 inhibitors are effective in preserving SM and reducing collagen content in rat models, similar histological benefits have also been demonstrated in humans. Schwartz, Wong, and Graydon (2004) investigated the impact of sildenafil on corporal SM in men after RP. Forty men with good baseline EF were randomized after RP to either sildenafil 50 mg or 100 mg every other night for 6 months beginning the day after urethral catheter removal. Corporal biopsies were obtained just prior to RP and again at 6 months postoperatively; tissues were then stained for SM and connective tissue. The authors demonstrated that SM content was significantly preserved at both sildenafil doses. The major limitations of this trial are the use of SM content as a surrogate for potency and the lack of a control group. In the historical context of the aforementioned Iacono and Nehra studies, however, these data assume greater importance.

The first study yielding clinical human data in support of the concept of penile rehabilitation was conducted in Italy by Montorsi et al. (1997). A total of 30 patients who were deemed potent preoperatively underwent bilateral nerve-sparing RP for localized prostate cancer. Subjects were randomized to intracavernosal injection therapy with alprostadil three times a week for 12 weeks versus observation. In this investigation, 67% of men in the treatment group reported the recovery of spontaneous erections (that is, without injections) sufficient for satisfactory sexual intercourse compared with only 20% of the observation group. Although this study lacked a placebo group and was underpowered it represented the first evidence that regular erections after RP are important for erectile function recovery and thus sparked an enormous amount of interest.

In a subsequent trial, Mulhall, Land, Parker, Walters, and Flanigan (2005) compared patients who opted for penile rehabilitation after RP with those who did not. In this study, patients were voluntarily enrolled into a rehabilitation group treated with either oral sildenafil or intracavernosal injection therapy (for sildenafil nonresponders). Data from men who were committed to rehabilitation were compared with those of men who did not enroll in the protocol. At 18 months post-RP, patients in the penile rehabilitation group were more likely to be capable of having medication-unassisted intercourse (52% vs. 19%) and reported higher mean IIEF erectile function domain scores (22 ± 6 vs. 12 ± 14). Moreover, the percentage of patients with normal EF domain scores (22% vs. 6%), the percentage of patients responding to sildenafil (64% vs. 24%), the time to become a sildenafil responder (9 ± 4 vs. 13 ± 3 months), and the percentage of patients responding to intracavernous injection (ICI; 95% vs. 76%) were all improved for the rehabilitation group. The major limitation of this report was its nonrandomized nature; provider bias or patient motivation issues potentially confounded the results. The magnitude of the treatment effect, on the other hand, was extremely large. The study therefore provides further suggestion that pharmacologically-induced erections following RP result in higher rates of subsequent spontaneous erections as well as improved response to erectogenic medications.

A landmark study was sponsored by Pfizer, Inc. and published in 2008 (Padma-Nathan et al., 2008). A total of 76 men with robust preoperative EF (defined as IIEF EF domain score ≥8/10) who underwent bilateral nerve-sparing RP for localized prostate cancer were randomized in a double-blind fashion to one of three arms: sildenafil 50 mg, sildenafil 100 mg, or placebo nightly for 36 weeks beginning 4 weeks after surgery. Intervention was stopped 40 weeks after surgery and medication-unassisted EF was assessed after an 8 week washout period. Positive responders were defined as men who achieved their preoperative IIEF EF domain score without medication assistance. Using this definition, the authors reported that spontaneous EF occurred in only 4% of the placebo group versus 27% of the sildenafil group. There was no statistically significant difference between the different sildenafil dosage groups.

The only other study of penile rehabilitation conducted in a randomized, controlled, double-blind manner was the REINVENT trial (Montorsi et al., 2008). This study, sponsored by Bayer, Inc., enrolled over 600 men who underwent nerve-sparing RP and were randomized 1:1:1 to three groups: vardenafil 10 mg nightly plus placebo on-demand for sex, placebo nightly, and vardenafil (5, 10, or 20 mg) on-demand for sex or placebo nightly with placebo on-demand for sex. After the 9-month double-blind phase, the patients entered a single-blind placebo washout phase and then entered a third phase with open-label vardenafil for 2 months. The primary study end-point at the end of each of the three phases was the percentage of men with an IIEF score ≥22. At the end of the 2-month washout period this endpoint was similar in the three treatment arms (28.9%, 24.1%, and 29.1% in the placebo, nightly vardenafil, and on-demand vardenafil arms, respectively).

Since nightly vardenafil was not demonstrated to be more effective than on-demand vardenafil, the study has been frequently cited by critics of the penile rehabilitation strategy. There are multiple factors that should be considered, however, when interpreting the results. First, the study design was extremely complicated. There were 423 subjects at 87 different centers, and no standardized definition of nerve-sparing RP was employed. No information was provided about the number of surgeons at each center. The study duration of less than 1 year may have been too short. Perhaps most important, no assessment of on-demand vardenafil use was described, so it is theoretically possible that on-demand vardenafil users had similar active drug ingestion as nightly vardenafil users.

Vacuum Erection Device (VED) for Penile Rehabilitation

Vacuum therapy may be used either alone or in combination with the PDE5i to preserve erectile function following RP. This form of therapy generally involves placing a tube over the penis to which a vacuum is applied, producing a negative pressure that draws blood into the penis causing engorgement. When an erection is achieved, a constricting ring may be applied at the base for up to 30 minutes to maintain the erection.

Although PDE5i are the most commonly used agents in the setting of postprostatectomy penile rehabilitation, the VED plays an important role in cases where PDE5i are contraindicated, ineffective, or associated with significant side effects that limit their use. In addition to being cost-effective, the VED has the added benefits of noninvasive application (especially compared to penile injection therapy), increased spontaneity (Pahlajani et al., 2012), efficacy rates ranging from 60% to 80% (Zippe, Nandipati, Agarwal, & Raina, 2006), and an excellent safety profile. For all these reasons, this treatment modality is associated with long-term compliance rates as high as 70% after 1 year of treatment (Zippe et al., 2006). However, limitations to the vacuum device include discomfort caused by the constricting band, instability at the base of the penis, which may result in uncomfortable pivoting, a cool or cold cyanotic erection resulting from impairment of penile blood flow (Pahlajani, Raina, Jones, Ali, & Zippe, 2012), swelling of the glans, and petechiae at the base of the penis (Zippe et al., 2006).

As with PDE5i use, this form of penile rehabilitation exerts its protective effects by providing oxygenation to the cavernosal erectile tissue. In fact, while the efficacy of oral agents in the early postoperative period may be limited by nerve injury that prevents spontaneous erections, the VED draws blood into the penis that is 58% arterial and 42% venous independent of nerve function (Bosshardt, Farwerk, Sikora, Sohn, & Jakse, 1995). Thus, the VED may bypass the early neuropraxic period, dilating cavernosal arteries and protecting against hypoxia even in the absence of nerve function, averting corporal fibrosis that may otherwise develop prior to functional recovery of the cavernous nerves (Yuan et al., 2010).

The VED has been demonstrated to maintain length and girth following RP. In clinical trial by Köhler and colleagues, patients randomized to a daily rehabilitation protocol consisting of 10 min/day using the VED with no constriction ring experienced a significant preservation of stretched penile length (SPL) compared to controls. In fact, SPL was significantly decreased at both 3 and 6 months by approximately 2 cm in controls; however, SPL was preserved in the group undergoing daily rehabilitation with the VED at all sample times (Köhler et al., 2007). In another clinical trial, 109 RP patients were assigned to VED use daily for 9 months versus observation. While 23% of VED users reported a decrease in penile length and circumference at 9 months, 63% of patients in the control group reported decrease in penile length and circumference (Raina et al., 2006).

In addition to minimizing the rate and extent of penile shrinkage, early and consistent use of vacuum erection therapy enhances recovery of erectile function and promotes return to sexual activity during the erectile recovery period. When used without constrictive rings in penile rehabilitation protocols, VEDs have been reported to produce a 60% improvement in spontaneous erections as well as a significant improvement in International Index of Erectile Function scores when used early in the postoperative setting (Hoyland, Vasdev, & Adshead, 2013). In their randomized trial of post-RP patients, Raina et al. (2006) demonstrated that at the end of a 9-month follow-up period, 80% of men assigned to daily use of VED were able to have sexual intercourse using the device. Moreover, 10 patients in the VED group reported erections sufficiently firm for vaginal penetration as compared to only 4 patients in the control group (Raina et al., 2006). More recently, Raina et al. (2006) reported on the long-term potency of 141 sexually active men following RP. In this study, early intervention with nonoral therapy played a key role in maintaining sexual activity and interest, as well as in maintaining the health of cavernosal tissue. Of the 113 sexually active patients in this study at 1 year, 50 patients experienced a return of spontaneous erections, and all 50 of these men were using nonoral standard treatments like VED, ICI, and MUSE. Moreover, of these 50 patients, 60% reported early VED use, and these patients demonstrated greater compliance than with the other nonoral agents (Raina, Pahlajani, Agarwal, Jones, & Zippe, 2010).

Like PDE5i, the VED promotes earlier return of sexual function, enhances patient and partner satisfaction, and preserves sexual interest and comfort following RP. Unlike the oral agents, however, vacuum therapy exerts its effect independently of (and therefore in the absence of) the functional cavernosal nerves. This advantage of the VED, together with its ease of use, cost-effectiveness, and noninvasiveness make it an ideal treatment tool for post-RP penile rehabilitation. Whether it is used alone or in combination with other medical agents, vacuum therapy should be considered as a first-line treatment option to facilitate preservation of erectile tissue and promote recovery of erectile function postoperatively (Brison, Seftel, & Sadeghi-Nejad, 2013).

Benefits of a Structured Penile Rehabilitation Program

The optimal rehabilitation strategy has yet to be defined, and many different protocols are currently used. It is worth noting that nearly 90% of surveyed urologists who are members of the American Urological Association (AUA) recommend rehabilitation practices, and further, there was no statistically significant difference in this regard between sexual medicine and urologic oncology specialists (Tal, Teloken, & Mulhall, 2011).

Despite the absence of specific established protocols, the majority of the published literature suggests that men who undergo RP will benefit from the clinical application of general principles of penile rehabilitation; specifically, achieving erections on a regular basis is physiologic and should be maintained. For some patients PDE5i are sufficient for achieving post-RP erections, whereas many others require ICI to achieve penetration-hardness erections. The ideal frequency, time course, and hardness of erections are unknown. The optimal therapeutic regimen is also unknown, but it is likely that men who do not respond to PDE5i alone may benefit from ICI alone or in combination with PDE5i.

Mulhall (2009) has detailed the algorithm used at Memorial Sloan-Kettering Cancer Center (MSKCC) for use as a guide. In the MSKCC program, men are ideally seen by the sexual medicine specialist prior to undergoing RP. Patients begin taking daily low-dose PDE5i for 2 weeks prior to surgery or beginning the day that the Foley catheter is removed postoperatively. Patients then take daily low-dose PDE5i for the next 2 to 4 weeks, except that they are encouraged to take a full dose of PDE5i on several occasions during this time period. They are then seen at 6 weeks post-RP, at which time it is determined whether they are a PDE5i responder (rigid erections in response to full- or low-dose PDE5i) or PDE5i nonresponder (no rigid erections in response to any dose PDE5i). Responders continue with daily low-dose PDE5i with weekly exceptions for full-dose PDE5i, and this regimen is continued for 24 months with periodic follow-up visits. Nonresponders are immediately started on ICI therapy one to three times per week plus low-dose PDE5i on days that they do not use ICI. Nonresponders are also challenged with high-dose PDE5i each month. Once they begin to respond to high dose PDE5i, injections are discontinued, and they are switched over the protocol for PDE5i responders (Mulhall, 2009).

Conclusion

Animal and human studies have revealed that structural changes occur following cavernous nerve injury. These alterations, particularly the increase of cavernosal collagen with decreasing SM content, have been demonstrated to ultimately result in VOCD. Multiple animal trials have documented that administration of PDE5i attenuates these alterations. Data from human trials indicate that obtaining regular erections after RP, whether achieved with PDE5i, ICI, or VED improves the likelihood of postoperative functional erections compared with men who did not use this strategy. Despite the scientific rationale and growing body of literature supporting rehabilitation, definitive data supporting its routine use has not been demonstrated. The major exception among these studies is the REINVENT trial, which was a negative study but has been criticized for major methodological flaws.

ED is a major quality of life factor that is associated with depression. Urologists should discuss PDE5i therapy with all men interested in maintaining EF after RP to minimize the duration-dependent penile structural changes that can lead to irreversible damage. Although the magnitude of clinical benefit and optimal regimen remain unknown, patients deserve a discussion about penile rehabilitation in the post-RP setting so that they can make their own informed decisions and weigh the potential benefits against the additional costs, inconvenience, and side effects of therapy.