What's known on the subject? and What does the study add?
Removing of prostate for the treatment of localized prostate cancer is associated with a variable loss of erectile function due to injury of the nerves of erection during operation. Some researchers have reported that after nerve-sparing radical prostatectomy (RP), the natural recovery time of erectile function is at least 2 years. Factors such as thermal damage, ischaemic injury, mechanically induced nerve stretching and the local inflammatory effects of surgical trauma may also impair the cavernous nerves during RP. The concept of penile rehabilitation was first studied by Montorsi et al. in 1997. They showed that the use of any drug or device at or after RP could maximize the recovery of erectile function. Penile rehabilitation programmes (PRPs) with vasoactive agents, such as oral phosphodiesterase-5 inhibitors (PDE5Is), intraurethral and intracavernosal vasoactive agents, and vacuum erection devices (VEDs) can protect erectile tissue integrity and prevent corporal smooth muscle atrophy and diminish collagen formation.
The present findings are consistent with previous reports that PRPs have a significant beneficial effect on early erectile function recovery and that preoperative erectile function is one of the important predictors of erectile function after RP. Patients can be referred for penile rehabilitation if they have any degree of erectile function (mild, moderate or normal) before operation. We also showed that the combination of PDE5Is and VEDs for PRPs offers the shortest erectile function recovery period.
To define the optimal penile rehabilitation programme (PRP) based on preoperative Sexual Health Inventory for Men (SHIM) scores after robot-assisted radical prostatectomy (RARP).
Patients and Methods
The medical records of 203 patients who underwent bilateral nerve-sparing RARP between 2007 and 2011 were reviewed for the present retrospective study.
According to patients' preoperative erection status, group 1 (SHIM = 8–16), group 2 (SHIM = 17–21) and group 3 (SHIM = 22–25) were defined.
After bilateral nerve-sparing RARP, phosphodiesterase-5 inhibitors (PDE5Is), a vacuum erection device (VED), the combination of PDE5Is and a VED, or none of them were utilized by all patients for penile rehabilitation.
Treatment success was defined as a rigid erection suitable for successful sexual intercourse.
The numbers of patients in groups 1, 2 and 3, respectively, were 9, 22 and 73, and the mean erectile function recovery periods (EFRPs) were 15.44 ± 7.73, 12.31 ± 8.12 and 8.73 ± 5.67 months (P < 0.05).
Group 3 offered the best results for EFRP. Only PDE5Is or the combination of PDE5Is and VED use had a beneficial effect on EFRP (P < 0.05).
Using PDE5Is and VED together provided the best result, but there was no difference between PDE5Is and a VED (P ≥ 0.05).
After bilateral nerve-sparing RARP, PRP with PDE5Is, including the combination of PDE5Is and VED, has a beneficial effect on erectile function recovery across all levels of baseline erectile function.
Further large randomized control studies are needed to validate these findings.
Prostate cancer has a significant impact on men's health and quality of life in the United States, with more than 230 000 new cases diagnosed each year and more than 30 000 deaths attributed to this malignancy every year . Radical prostatectomy (RP) is a standard treatment for patients with organ-confined prostate cancer and offers a life expectancy of more than 10 years . This operation is associated with a variable loss of erectile function due to injury to the autonomic cavernous nerves. The exact time it takes for patients to regain full erectile function after nerve-sparing RP is still uncertain. In terms of regaining potency after RP, the most important prognostic factors are the quality of the spared neurovascular bundles, the age of the patient, and the baseline sexual function status [3-6]. It has been reported that maximal erectile recovery is not seen until a mean period of 18 months after bilateral nerve-sparing RP . On the other hand, other researchers have reported that after nerve-sparing RP, the natural time for recovery of erectile function is at least 2 years .
The International Index of Erectile Function (IIEF) questionnaire is a multidimensional validated questionnaire with 15 questions in the five domains of sexual function (erectile and orgasmic functions, sexual desire, satisfaction with intercourse and overall sexual satisfaction) approved by the National Institutes of Health (NIH) . More recently, to simplify the IIEF, an abridged five-item version (IIEF-5, SHIM) was developed as a diagnostic tool for erectile dysfunction (ED) . It consists of five selected items to distinguish clearly between subjects with and without ED, as well as to address the NIH  definition of this condition.
The concept of penile rehabilitation was first studied by Montorsi et al.  in 1997. They demonstrated that the use of any drug or device at or after RP could maximize the recovery of erectile function. Penile rehabilitation programs (PRPs) with vasoactive agents, such as oral phosphodiesterase-5 inhibitors (PDE5Is), intraurethral and intracavernosal vasoactive agents, and vacuum erection devices (VEDs) can protect erectile tissue integrity and prevent corporal smooth muscle atrophy and diminish collagen formation. Nowadays, robot-assisted radical prostatectomy (RARP) is a well-established method for treating localized prostatic cancer, and potency outcome has become a central interest to researchers . In the present study, we tried to define the optimum PRP based on preoperative IIEF-5 scores after bilateral nerve-sparing RARP.
Material and Methods
Patients from the Columbia University Comprehensive Urologic Oncology Database, approved by the institutional review board, were used for analysis. The medical records of 203 patients who underwent bilateral nerve-sparing RARP by a single surgeon (KKB) between 2007 and 2011 were retrospectively reviewed for this retrospective uncontrolled study. Patients with clinically organ-confined prostate cancer who had IIEF-5 scores ≥ 8 were included in the study. Patients were excluded from the present study if they had received androgen deprivation therapy or pelvic radiotherapy before or after surgery.
Patient age, co-morbidities (diabetes mellitus, hypertension, hyperlipidaemia and cardiovascular diseases), preoperative SHIM scores, PRP (oral PDE5Is, VED, PDE5Is and VED together, or none of these), erection strength (ES) after surgery and the erectile function recovery period (EFRP) were obtained from patients' medical records. ES based on a four-point scoring system: full baseline rigidity (four points), hard enough for penetration but not back to baseline (three points), partial erection not suitable for penetration (two points), and penile fullness without rigidity (one point). ES 3 and 4 were defined as successful recovery of potency in the present study .
All patients were evaluated in three groups as follows: group 1, patients with moderate erectile dysfunction (SHIM = 8–16); group 2, patients with mild erectile dysfunction (SHIM = 17–21); and group 3, patients with normal erectile function (SHIM = 22–25).
If PRP was planned, it was started when the patients' catheters were removed. Patients used oral PDE5Is (recommended on an every-other-day basis) and/or VED (recommended on a daily basis). After surgery, both PDE5I and VED were offered to all patients.
Two hundred and three patients with complete data were included in the present study. Groups 1, 2 and 3 comprised 48, 48 and 107 patients, respectively. Patients' demographic characteristics are summarized in Table 1. All patients had clinically localized prostate cancer (cT1–T2c) and the mean follow-up period was 17.40 ± 1.14 months.
Table 1. Demographic specialities of patients [n (%)].
In group 1, the mean patient age, PSA level, body mass index (BMI), specimen weight, operation time and estimated blood loss were 61.92 ± 6.59 years, 7.75 ± 10.26 ng/mL, 29.00 ± 4.42 kg/cm2, 53.94 ± 16.16 g, 187.66 ± 41.37 min and 167.43 ± 77.92 mL, respectively (Table 2). In this group, 17, three, two and 26 patients, respectively, used only PDE5Is, only VED, both PDE5Is and VED and none of them. At a mean follow-up of 30.31 ± 2.57 months, the success rate was 18.8% (n = 9) and the mean EFRP was 15.44 ± 7.73 months (Table 3). For this group, using PDE5Is or PDE5Is and VED together resulted in the greatest recovery of erectile function and the shortest recovery period (P < 0.05), as most patients did not choose to use the VED (Fig. 1A).
*Number of patients. †Difference between treated and non-treated groups; P < 0.05 is considered to indicate statistical significance.
15.44 ± 7.73
30.31 ± 2.57
12.31 ± 8.12
18.40 ± 14.93
8.73 ± 5.67
11.08 ± 0.75
10.07 ± 6.74
17.40 ± 1.14
In group 2, the mean patient age, PSA level, BMI, specimen weight, operation time and estimated blood loss were 58.56 ± 7.30 years, 6.48 ± 7.85 ng/mL, 27.41 ± 4.37 kg/cm2, 52.55 ± 38.31 g, 200.78 ± 52.58 min and 355.00 ± 88.89 mL, respectively (Table 2). In this group, 22, one, six and 19 patients, respectively, used only PDE5I, only VED, both PDE5I and VED and none of them for rehabilitation. At a mean follow-up period of 18.40 ± 14.93 months, the success rate was 45.8% (n = 22) and the mean EFRP was 12.31 ± 8.12 months (Table 3). For this group, both the use of PDE5Is and the use of PDE5Is and VED together resulted in the greatest recovery of erectile function and the shortest recovery period (P < 0.05; Fig. 1B).
In group 3, the mean patient age, PSA, BMI, specimen weight, operation time and estimated blood loss were 58.77 ± 7.43 years, 6.38 ± 7.03 ng/mL, 27.43 ± 4.33 kg/cm2, 52.27 ± 41.97 g, 201.87 ± 55.13 min and 148.90 ± 91.16 mL, respectively (Table 2). In this group, 50, two, 28 and 27 patients, respectively, used PDE5I, VED, the combination of PDE5I and VED and none of them. At a mean follow-up period of 11.08 ± 0.75 months, the success rate was 68.2% (n = 73) and the mean EFRP was 8.73 ± 5.67 months (Table 3). For this group, using PDE5I or PDE5I and VED together resulted in the greatest recovery of erectile function and the shortest recovery period (P < 0.05; Fig. 1C).
Race, marital status, existence of hypertension, hyperlipidaemia or cardiovascular disease are not statistically significant different between groups. The median age of patients decreased while the ES increased, and prevalence of diabetes mellitus was lower among patients in group 3. In patients undergoing PRP, there was a shorter EFRP in all groups compared with patients who did not receive PRP (P < 0.05). The shortest EFRP was seen in group 3 (P < 0.05; Table 2, Fig. 1D). In all groups, using PDE5Is or PDE5Is and VED together resulted in the greatest recovery of erectile function and the shortest recovery period (P < 0.05). However, the EFRP was shorter in the patients using PDE5Is and VED in combination than the patients using PDE5I alone.
Radical prostatectomy is the oldest and the most frequent treatment modality for patients with an organ-confined prostate cancer, with a life expectancy of more than 10 years . The challenge for the urologist treating patients with prostate cancer is cancer control with the preservation of continence and erectile function. Since initiation of penile erection is dependent on nerves, preservation of the cavernous nerves during RP is the most important factor for the recovery of erectile function . Even after the original description of the neural pathways by Walsh & Donker  in 1982, modified surgical techniques for RP and the advent of nerve-sparing procedures, a significant percentage of patients still continue to develop ED after RP. The reported incidence of ED after RP is tremendously variable. The CaPSURE study  revealed that only 20% of patients returned to their preoperative baseline potency at 1 year after RP. Similarly, the Prostate Cancer Outcomes Study  measured the changes in urinary and sexual function in 1291 patients who had undergone RP for clinically localized prostate cancer. At 18 months or more after RP, 59.9% of patients self-reported having ED. In a recent population-based cohort study  in a sample of 1288 patients with localized prostate cancer, temporal changes in urinary and sexual function were noted for up to 5 years after RP. At 60 months, 28% of the patients reported erections firm enough for intercourse, compared with only 22% at 24 months.
Robot-assisted RP has become an established method for treating localized prostatic cancer. Various degrees of nerve-sparing techniques have been described during RARP [16, 18, 19]. Regardless of the technique, the removal of the prostate appears to result in an obligate period of neuropraxia with secondary end-organ damage and varying degrees of ED . In the literature, the mean (range) potency rate of 19% (11–40%) probably reflects the results of sexual function after RRP in a general urological practice irrespective of operation technique .
Even after bilateral nerve-sparing RP, an apparent ‘neuropraxia’ resulting from relative trauma to the neurovascular bundles during surgical dissection is manifested as impaired erectile function in the early postoperative period. Factors such as thermal damage, ischaemic injury, mechanically induced nerve stretching and the local inflammatory effects of surgical trauma might also impair the cavernous nerves during RP . Kim et al.  showed that penile blood flow is decreased after RP, especially on the side where the nerve bundle has been sacrificed. Chronic hypoxia and denervation cause ED by initiating the apoptosis process and increasing the deposition of connective tissue [24-26]. Numerous investigations support the concept that ED after prostatectomy diminishes the circulation of arterial blood through the corpora cavernosa, causing reduced oxygen content and increased production of transforming growth factor-β, all of which accelerate corporeal fibrosis and subsequent penile shortening [27, 28]. Since the cavernous nerves might be functionally inactive after RP, studies have reported that maximal erectile recovery is not witnessed until a mean period of 18–24 months after bilateral nerve-sparing RP, even if the nerve-sparing technique is correctly executed [22, 29-31]. The major determinants of ED in this population are cavernous nerve injury and corporal smooth muscle structural changes [31-34].
In recognition of the neurogenic basis of ED after RP, new strategies have been devised to initiate the rehabilitation process . In 1997, Montorsi et al.  introduced the idea of penile rehabilitation after RP. The clinical strategy of postoperative penile rehabilitation after RP arose from the idea that induced early sexual stimulation and augmented blood flow to the penis would facilitate the return of natural erectile function and resumption of medically unassisted sexual activity. They showed that early use of intracavernosal alprostadil injections improved the recovery of spontaneous erections when compared with an untreated group. It is hypothesized that penile rehabilitation prevents erectile tissue damage secondary to neural damage which occurs by cavernous nerve traction during RP and is founded on the concept of cavernosal oxygenation, endothelial protection and cavernous nerve regeneration. Additionally, penile rehabilitation could result in the prevention of cavernosal smooth muscle fibrosis caused by the absence of cavernosal oxygenation. For this reason, any penile rehabilitation effort, in theory, is likely to diminish the interval for natural recovery of erectile function. Current penile rehabilitation strategies use treatment approaches such as PDE5I, intraurethral suppositories, vacuum devices and intracavernosal injection therapy [35-37].
Early postoperative treatment regimens using oral PDE5Is are the most frequently recommended strategies to increase nocturnal erections and oxygenation of the cavernosal tissues. In a prospective, randomized, double-blind, placebo-controlled study, the daily use of sildenafil at bedtime (50 or 100 mg) compared with placebo after bilateral nerve-sparing RP resulted in a statistically significantly increase in the proportion of patients returning to baseline erectile function . In the same study 27% of patients using sildenafil recovered erections similar to their baseline erections, compared with 4% of those in the placebo group. In addition to preserving intracorporal smooth muscle, other possible beneficial effects of PDE5Is in penile rehabilitation after RP are the promotion of neurogenesis, the protection of cavernosal endothelium and antifibrinogenetic activity . The main limitations of PDE5Is for penile rehabilitation are high cost and lack of immediate efficiency, leading to poor compliance and high discontinuation rates after RP .
In some studies, patients successfully used a VED after RP and confirmed its safety and tolerability with success rates 84–95% [39-42]. Columbo et al.  reported a series of 52 patients in whom daily use of the VED with no constriction ring, unrelated to intercourse, led to an improvement in spontaneous erections in 31 patients (60%). Raina et al.  showed that the use of VED after RP (with and without nerve preservation) improved the IIEF scores and patient-reported preservation of penile length, and aided in the early return of spontaneous erections. VED helps to inhibit abnormal collagen or scar formation in the hypoxic penile conditions after RP, perhaps prompting a faster return of nocturnal tumescence, which would improve penile oxygenation . By contrast with other alternatives for penile rehabilitation, VED might be more cost-effective, with a low risk of systemic side-effects and the added benefit of empowerment through active involvement of the patient and his partner in rehabilitation and recuperation.
There is no consensus about when to initiate penile rehabilitation after surgery, treatment duration or treatment regimen. Some authors recommend starting pharmacological erectile therapy during the postoperative second month to achieve optimal recovery of erections at 2 years . Others have suggested that at least 6 months using an intracavernous vasoactive agent is needed for recovery of spontaneous erections . In a study by Iacono et al. , it was shown that structural changes including increased collagenization and loss of elastin occur as early as 2 months after RP. Cavernous neurotomy models in animals prove that smooth muscle apoptosis occurs at 1 day after injury, and peaks within the first week of nerve injury . To improve cavernosal oxygenation and prevent hypoxia-induced corporal fibrosis, early postoperative treatment regimens are currently recommended [37, 46]. In addition, patient compliance is higher when therapy is started earlier after surgery .
While a temporary period of hypoxia is inevitable event in nerve-sparing surgery, the period required for recovery is 6–24 months [48-50]. Currently, sildenafil is the commonest pharmacological treatment prescribed to provide erections sufficient for intercourse in patients who have had RP. Sildenafil appears to work in those patients with intact neurovascular bundles [51, 52]. Since an interval of ≥12 months after RP appears to be required for patients to respond to sildenafil, this can discourage them from continuing treatment [53, 54]. On the other hand, the action mechanism of VED does not require intact neural transmission. VED can be recommended as a first-line option for the early postoperative period . When patients on combined PRP with PDE5Is and VED realize they can have erections, this is likely to encourage them to maintain treatment. Furthermore, using combined treatment could also improve a patient's PRP compliance. For the combined treatment, when patients cannot use either PDE5Is or a VED for a period of time, because of contraindications or health or social problems, the integrity of the PRP can be provided by using the other agent. In the present study, we showed that patients' erectile function recovered early using PDE5Is or a combination of PDE5Is and VED in all groups (P < 0.05). There was also a shorter EFRP in patients using the combined treatment than in those using PDE5Is alone, but the difference was not statistically significant (P ≥ 0.05), probably because of the small numbers choosing to use VED.
In the present study, we tried to define the optimum PRP strategy based on preoperative SHIM scores after bilateral nerve-sparing RARP. We started PRP soon after RARP, within 7–10 days, as referenced in the literature. The present findings are consistent with previous reports that PRPs have a significant beneficial effect on early recovery of erectile function, and preoperative erectile function is one of the important predictors of erectile function after RP. We also showed that the combination of PDE5Is and VED for PRP offers the shortest EFRP.
In conclusion, PRP with PDE5Is, including the combination of PDE5Is and VED, after bilateral nerve-sparing RARP has a significant beneficial effect on erectile function recovery across all levels of baseline erection function. The present study has some major limitations, including the fact that it was not pre-planned, the numbers were small, there was a selection bias, a lack of a control group and the groups were not balanced. However, in spite of these major limitations, the present study is clinically important because it reflects the real-life outcomes in one centre with one surgeon and is relevant to current practice. Therefore, further large, randomized controlled studies are needed to validate these findings.