Neural Trauma

It is recognized that a macroscopic injury to the cavernous nerves during RP, such as transection or thermal damage, will result in a permanent loss of erectile function after surgery. Consequently, the recovery of erectile function has been attributed largely to the success of the macroscopic preservation of the neurovascular bundle at the time of RP. It has been well demonstrated that the nerve-sparing status of a RP greatly influences erectile function recovery. Bilateral nerve-sparing is associated with an increased spontaneous and phosphodiesterase type 5 (PDE5) inhibitor (PDE5i)-assisted recovery of erectile function compared with unilateral nerve sparing; both lead to a better erectile function recovery than non–nerve-sparing surgery.

Nevertheless, the macroscopic integrity of the cavernous nerves at surgery does not guarantee their future function. In a recent study, Moskovic and colleagues showed that the subjective characterization of nerve sparing, using a nerve-sparing grading system, represented an accurate means to predict erectile function recovery after RP. In this study, a significant proportion of men who underwent RP with bilateral nerve sparing assessed as complete, still experienced a significant impairment of their erectile function, with a mean reduction in their International Index of Erectile Function (IIEF): Erectile Function Domain score of 7.2 points at 24 months after surgery.

Indeed, as was shown by Masterson and colleagues, even minimal traction trauma of the cavernous nerves, is likely to cause long-term damage. In this study, a cohort of 275 RP patients (74%) operated on using a standard technique had their erectile function recovery compared with a simultaneous cohort of 97 patients (26%) treated with a technique avoiding the use of a Foley catheter as a tool to stretch the lateral pedicles. At 6 months after RP, 67% of the patients treated with the “no stretch” technique had recovery of functional erections compared with 45% in the standard technique group.

However, the ability of the surgeon to preserve the cavernous nerves at the time of surgery is not the only factor that influences erectile function outcomes after RP. As highlighted in a study by Katz and colleagues, 25% of the patients who had functional erections with or without PDE5i within the first 3 months after surgery were nonfunctional by 6 months. These data suggest that postoperative factors such as edema and inflammation may lead to ongoing postoperative wallerian degeneration, which may in turn be responsible for delayed damage to the cavernous nerves.

The underlying pathophysiological mechanisms that lead from neuropraxia or neurotomy to ED have been well documented in animal studies and imply consequences on both smooth muscle cells and endothelial cells.

Corporal Smooth Muscle Alterations

Corporal smooth muscle alterations are the result of two mechanisms, each leading to increased deposition of collagen with decreased penile distensibility, which translates into venoocclusive dysfunction.

First, after RP, in the absence of erection due to neuropraxia, cavernosal oxygenation is diminished and, therefore, smooth muscle cells are exposed to a prolonged environment of decreased oxygen tension. This results in the inhibition of prostaglandin E1 (PGE1), which suspends its inhibiting effect on profibrotic substances, such as, transforming growth factor-β1 (TGF-β1) and TGF-β1–dependent endothelin-1 (ET-1). Consequently, TGF-β1 is permitted to promote connective tissue synthesis (especially collagen I and III), with the subsequent replacement of trabecular smooth muscle. These data have been consistently demonstrated in the penile tissue of denervated rat models, in which a significant increase in collagen content and a decrease in the smooth muscle-collagen ratio compared with controls have been seen.

Corporal smooth muscle alterations also happen through the apoptosis induced by the proapoptotic cytokines released by the damaged nerve axons. This was first shown by Klein and colleagues. In their study, the penile tissue of 15 rats with bilateral cavernous neurotomy was compared with 15 sham-operated controls. In the glans penis and the cavernous tissue of the denervated rats elevated sulfated glycoprotein-2, intranuclear DNA fragmentation, and apoptotic nuclei were seen, all of which are characteristic of apoptotic cells. These elements were not present in the controls. In this study, however, the specific cell subtypes undergoing apoptosis were not identified.

In a subsequent study, User and colleagues operated on male rats that were randomized to unilateral or bilateral cavernous nerve transection. Changes in wet penile weight, DNA, protein, and apoptotic cells of the penis were measured at different time points after injury. Penile wet weight was significantly decreased at all time points after bilateral neurotomy ( P <.001). Unilateral neurotomy allowed much greater preservation of penile weight. DNA content was also significantly decreased in bilaterally denervated penes. Bilateral neurotomy induced significant apoptosis, whereas unilateral surgery caused significantly less apoptosis. Interestingly, the study showed that, in both cases, the apoptotic process affected mostly the smooth muscle cells located beneath the tunica albuginea. It was proposed by the investigators that this sudden and massive apoptosis of smooth muscle cells in the subtunical area might be a mechanism for the development of venoocclusive dysfunction observed after RP.

However, it was later shown by other investigators that the apoptotic process not only involved smooth muscle cells but also endothelial cells. More precisely, in a study by Lysiak and colleagues, mice were subjected to cavernous nerve resection or sham surgery and their penes were processed for the identification of apoptotic cells, changes in phosphorylation of several protein kinases, and immunolocalization of specific kinases. An increase in apoptotic cavernous smooth muscle and endothelial cells was evident by 2 weeks, which further increased by 4 and 6 weeks after cavernous nerve resection. Apoptosis coincided with an increase in the phosphorylation of c-jun N-terminal kinase and p38 mitogen activated protein kinase. Phospho-c-jun N-terminal kinase was immunolocalized to endothelial and smooth muscle cells. These findings are consistent with earlier findings by Podlasek and colleagues.

Various types of cavernous nerve injury have been studied in an attempt to model RP-associated cavernous nerve injury using rodent models, including crush, freezing, transection, and excision of the cavernous nerve either unilaterally or bilaterally depending on the underlying aims of each particular experiment. It was shown by many investigators that the apoptotic process in both smooth muscle and endothelium may be different according to the model used. The process was indeed shown to occur in a more delayed fashion in nerve crush injury model compared with the neurotomy model in a study by Mulhall and colleagues. In another study by Jin and colleagues, performed on four groups of 36 mice (control, sham operation, bilateral cavernous nerve crush, and bilateral cavernous neurotomy), erectile function was significantly less in the cavernous nerve crushing and neurotomy group than in the control or sham group. This difference was observed at the earliest time point assayed (day 3) and persisted up to 4 weeks after nerve crushing and to 12 weeks after neurotomy.

A study by Iacono and colleagues showed that the same process is likely to happen in the corpora cavernosa of men after RP. Indeed, in their study, the preoperative histologic data of the corpora cavernosa biopsies of 19 men undergoing RP were compared with the data of the same biopsies performed at 2 and 12 months after surgery. Trabecular elastic fibers ( P <.0003) and smooth muscle fibers were decreased and collagen content was significantly increased ( P <.0003) compared with preoperative biopsies. One year after surgery elastic fibers ( P <.0003) and smooth muscle fibers were decreased and collagen content was significantly increased ( P <.001) compared with the first postoperative biopsy.

Many theories have been proposed to explain smooth muscle alterations and increased corpus cavernosum collagenization after RP. It is likely that these changes result from the combination of suppression of growth factor production by damaged cavernosal nerves as well as the subsequent production of proapoptotic and profibrotic factors within the corpora cavernosa.

Arterial Injury

Accessory pudendal arteries (APAs) are vessels that arise from a source above the levator ani (usually obturator, femoral, vesical, or iliac arteries) and travels caudal toward the anterior perineum. The fact that they run in the periprostatic region puts them at risk for injury during RP. More precisely, as shown by Secin and colleagues, APAs can be divided into two distinct varieties. There are APAs that course along the lateral aspect of the prostate, termed lateral APAs, and APAs that emerge through the levator ani fibers near the apical region of the prostate, termed apical APAs.

Literature reports a variable prevalence of APAs, ranging from 4% to 75% depending upon their identification method: open surgery, laparoscopic prostatectomy, angiographically during an internal pudendal or internal iliac artery flush, or during cadaveric studies. APAs have been found to provide a significant blood supply to the corpora cavernosa in many studies. In a study by Rosen and colleagues, where selective bilateral internal pudendal angiography was performed in 195 men suspected of having arteriogenic ED, APAs were the major source of inflow to the penis in all cases where they were identified (7%). In a cadaveric dissection study by Breza and colleagues, APAs were identified in 7 cadavers out of 10. APAs were found to be the major source of arterial inflow to the corpora cavernosa in four cases, and the only source of erectile arterial inflow in one case. In a study by Benoit and colleagues, where a total of 33 APAs were identified in 20 cadavers, in 15% of cases the penile arterial inflow originated exclusively from APAs.

It was later demonstrated by Droupy and colleagues, that APAs are functional. In their study, preoperative transrectal color Doppler ultrasound was performed in 12 patients with normal erectile function who were undergoing radical pelvic surgery. APAs were identified in 75% of the patients. Pharmacologically-assisted erections (obtained after intracavernosal injection of papaverine) induced hemodynamic changes in APAs similar to those described in cavernous arteries, suggesting the functional role of APAs in penile erection.

In a study performed at the Johns Hopkins Hospital, with 52 patients with identified APAs undergoing bilateral nerve-sparing surgery between 1987 and 1994, the effect of artery preservation was shown to increase the likelihood of erectile function recovery more than twofold and significantly shorten median time to erectile function recovery—6 versus 12 months.

Nevertheless, these conclusions have not been confirmed by other investigators. In a study by Box and colleagues, for example, in a cohort of 200 patients undergoing robot-assisted laparoscopic RP, of which 80 patients (40%) were found to have APAs, multivariate analysis showed no significant correlation between the presence or absence of APAs and preoperative sexual function. Sacrificing APAs did not correlate with time of erectile function recovery, quality of postoperative erections, or mean IIEF-5 score.

Cavernosal Oxygenation

Neurapraxia may not be the only process by which smooth muscle integrity is affected. It was proposed that the absence of cavernosal oxygenation might play an important role as well. In the flaccid state, the penis acts as a large vein and during the erect state, as an artery. In the flaccid state, the P o ₂ is approximately 35 to 40 mm Hg, which has a propensity to upregulate fibrogenic cytokines such as TGF-β, involved in collagen production and therefore in the genesis of fibrosis and venous leak. During oxygenation, P o ₂ rises to 75 to 100 mm Hg, which instead upregulates production of endogenous prostanoids as well as cyclic adenosine monophosphate (cAMP).

Moreland and colleagues, have shown in a series of in vitro experiments that exposure of cultured corporal smooth muscle cells to low oxygen levels suppresses PGE1 and cAMP production. Upon returning oxygen tension to normoxic levels, measured levels of both PGE1 and cAMP are normalized. In a further series of experiments, the same investigators showed that, in the in vitro setting, prostanoids inhibit TGF-β activity and thus reduce collagen production.

Therefore, in a healthy male it is plausible that the alternation between the flaccid and erect states, as long as it occurs with a certain frequency, allows preservation of erectile tissue. After RP, however, in a state of unantagonized flaccidity, the balance between flaccid and erect states may be shifted in favor of fibrogenic cytokine production, leading to structural changes and venous leak development.

In a study by Muller and colleagues, in a rat cavernous-nerve-injury model, the potential benefit of the use of hyperbaric oxygen therapy (HBOT) on erectile function and cavernosal tissue has been demonstrated. Animals with bilateral nerve crush were divided into two groups: exposure to a 10 day course of 90 minute treatments with HBOT (3 atm) beginning the day of the cavernous nerve injury compared with animals exposed to room air within an identical chamber. Ten days after bilateral nerve crush, the animals underwent cavernous nerve stimulation measuring the maximal intracavernosal pressure (ICP) to mean arterial pressure (MAP) ratios. Rats exposed to HBOT had significantly higher ICP to MAP ratio recovery compared with the control group (55% vs 31%, P <.01). These data suggest that cavernosal oxygenation may be a significant factor in erectile tissue health and to recovery of erectile function.

Nevertheless, in a study by Vignozzi and colleagues, in rats with bilateral cavernous neurotomy, sildenafil treatment was capable of counteracting penile hypo-oxygenation, and the over-expression of the profibrotic ET-1 type B receptor (associated with a 3 month period of hypoxia) with its effect being more evident the earlier it was administered. The threshold of oxygenation beyond which the balance is shifted in the penis from profibrogenic to the synthesis of prostanoids that inhibits TGF-β activity remains unclear, as well as the optimal level of erectile rigidity for cavernosal oxygenation.

Confirming a study by Kim and colleagues, Tal and colleagues have shown, in a study population of 13 patients undergoing cavernosometry, with blood specimens collected at various intracavernosal pressures, that significant increases in cavernosal oxygenation occur in the earliest stages of erection at relatively low ICP. In their study, blood specimens were collected at an ICP range of 6 to 90 mm Hg. Mean plus or minus SD P o ₂ was 39 mm Hg at ICP less than 10 mm Hg, 87 at ICP 11 to 20 mm Hg, 89 at ICP 21 to 45 mm Hg, and 96 at ICP greater than 45 mm Hg. These findings suggest that partial erections may be sufficient to oxygenate erectile tissue and protect it from prolonged hypoxia-induced damage.

Reproducing in human subjects a study previously done on monkeys by Lue and colleagues, Knispel and Andresen have investigated the dynamic evolution of oxygen pressure within the penis during the flaccid state and the erection state. In their study, 34 patients with ED had their cavernous oxygen tension monitored, with an oxygen-sensitive Eppendorf needle electrode. The mean cavernous oxygen tension of 38 mm Hg during flaccidity was shown to undergo a continuous and gradual increase to 61 mm Hg a minute after injection of vasoactive agent 30 to 60 seconds lasting up to 8 minutes after the injection. In this study, the duplex Doppler ultrasound peak arterial flow was shown to correlate with maximal cavernosal P o ₂ in 71% of cases. However, no ICP was monitored to examine the correlation between the cavernous oxygen pressure and the degree of rigidity.

Finally, it is possible that each PDE5i does not have the same kinetics, tissue selectivity, and impact on tissue oxygenation. In a study by Ghofrani and colleagues, 60 consecutive patients with pulmonary artery hypertension were assigned to oral intake of 50 mg sildenafil (n = 19), 10 mg (n = 7), or 20 mg (n = 9) vardenafil, or 20 mg (n = 9), 40 mg (n = 8), or 60 mg (n = 8) tadalafil. Maximum effects on pulmonary artery vasorelaxation was obtained more rapidly for vardenafil (40–45 minutes) compared with 60 and 75 to 90 minutes for sildenafil and tadalafil respectively. Only sildenafil and tadalafil (but not vardenafil) allowed significant reduction in the pulmonary to systemic vascular resistance ratio. Significant improvement in arterial oxygenation (equally to nitric oxide [NO] inhalation) was only seen with sildenafil.

Venous Leak

The hemodynamic alterations that frequently underlie the development of long-term ED after RP is venous leak (corporovenocclusive dysfunction [CVOD]). During the erection process, as the smooth muscle expands in a three-dimensional fashion under NO control, it induces the compression of the subtunical venules that are positioned externally between the tunica albuginea and the corporal smooth muscle. Conditions in which the muscle fails to expand adequately leave subtunical venules in an noncompressed state, leading to venous leak. The two things that lead to failure of the corporal smooth muscle to expand are adrenaline and structural changes such as fibrosis.

Nehra and colleagues have shown, in human corporal tissue biopsy specimens taken at the time of cavernosometry, that when smooth muscle content in the penis drops below 40%, venous leak occurs. Indeed, the further this figure drops below 40%, the greater the magnitude of the leak. Iacono and colleagues have shown that as early as 2 months after RP in an untreated man, there is a marked increase in collagen deposition and a marked increase in elastic fiber content in erectile tissue. This is in keeping with the finding of the animal data outlined above, which suggests that structural changes occur even in the earliest stages after cavernous nerve injury. Mulhall and Graydon have shown, in a series of patients who had preoperative and postoperative hemodynamic assessment, that more than half of the men developed venous leak after surgery. In a more recent analysis by Mulhall and colleagues, of men who had partner-corroborated excellent erectile function before surgery and who underwent duplex Doppler penile ultrasound after surgery, venous leak (based on elevated end-diastolic velocities) developed in a time-dependent fashion after RP. The incidence of venous leak less than 4 months after surgery was approximately 10% and rose to 35% between 8 to 12 months after surgery and 50% after 12 months. The importance of this information is that, in the same series, men with normal erectile hemodynamics were more likely to have recovery of natural erectile function. However, only 8% of men who had venous leak had recovery of naturally functioning erections (capable of sexual intercourse) after surgery. It is also known that men with venous leak are far less likely to respond to a PDE5i than men with arterial insufficiency.