Male sexual dysfunction after fracture of the pelvis is more common than previously supposed with rates as high as 30% reported when the complaint is specifically sought. With the increase in survival from major injuries, the long-term consequences of trauma are being seen with increasing frequency. This review addresses the current understanding of the incidence, mechanism of injury, pathophysiology, acute management, subsequent investigation and long-term treatment of erectile dysfunction after injury to the pelvis.

The intimate relationship of the soft tissues and the bony pelvic ring result in a high risk of concomitant local injury associated with fractures of the pelvis.1 The bladder and urethra are particularly vulnerable, with reported rates of injury of 5% to 10%.1–,4 Even without severe urological injury, damage to the delicate vascular and nervous tissues supplying the genitalia can result in sexual dysfunction.5,6 Patients with polytrauma often have prolonged stays in hospital and require multiple operations on the pelvis, putting genitourinary structures at further risk of injury.

Erectile dysfunction has been defined as the inability to achieve or maintain an erection adequate for sexual satisfaction.7 In 3% of patients with sexual dysfunction, this will be secondary to an episode of pelvic or perineal trauma.8 In younger impotent patients, often at the beginning of their sexual lives, the motivation to restore function is very high.

Incidence of erectile dysfunction after pelvic trauma

Few studies have reported the overall incidence of sexual dysfunction after fractures of the pelvis, most concentrating on patients with urethral injury. King,9 in a review of 90 patients published in 1975, found an incidence of 5% of dysfunction in patients without urethral injury, rising to 42% when urethral injury had occurred. Machtens et al,10 in 2001 described an overall incidence of 11.6% in men in a series of 1722 patients who had suffered fractures of the pelvis. Although this is the largest number reviewed to date, the methods used for assessment were not stated. Malavaud et al11 used a previously validated scale to assess erectile dysfunction in 46 patients who had recovered from pelvic fractures. Their patients reported significantly lower scores in sexual satisfaction than normal historical controls, although the latter group was of a higher mean age by 20 years. Eleven patients had significant impairment of sexual function. The only characteristic of the fracture which correlated positively with impaired sexual potency was diastasis of the symphysis pubis. A trend towards impairment after urethral injury was also observed, although this did not reach statistical significance. The rates of incidence which have been reported are summarised in Table I. This suggests that, taking the data as a whole, impotence is seen in about 42% of patients who have a urethral injury. The overall rate of dysfunction after fracture is likely to be higher than the historical figure of 5%,9 lying somewhere between 11%10 and 30%11 depending upon how dysfunction is assessed.

View this table:
Table I.

The incidence of impotence after fracture of the pelvis

Pathogenesis of erectile dysfunction after pelvic trauma

A strong association between urethral disruption and subsequent impotence is well documented. The prostatic urethra is particularly vulnerable to shearing forces, being fixed by the urogenital diaphragm and puboprostatic ligaments to the ischiopubic rami and the symphysis pubis.1,3,12 Upward or posterior displacement of the symphysis will therefore subject the fixed urethra to considerable stress. Direct laceration by bone fragments is rare.

Erectile dysfunction after injury to the pelvis is due to a combination of neurogenic, vascular, corporal and psychogenic injury.9,13–,16 Table II summarises the information available in the literature on the pathogenesis of impotence in these patients.

View this table:
Table II.

Details of studies on the pathogenesis of erectile dysfunction

Neurogenic injury.

Severe neurological injury, particularly involving the lumbosacral plexus, may occur at the time of the initial injury and during subsequent operations. The quoted incidence varies widely from less than 1% to more than 30%.16–,20 Majeed18 found clinical evidence of neurological injury, confirmed by EMG studies, in 33% of patients with unstable fractures of the pelvis, but only 16% were identified on admission. All patients showed at least partial recovery by three to 12 months, continuing for up to 24 months. Of these, 13 patients were impotent; six recovered. The efferent nerve fibres supplying the cavernous bodies lie within the cavernous nerves to the urethra and symphysis pubis can be injured in isolation without other obvious neurological impairment as they leave the bony pelvis. Unfortunately, there is no specific test for neurogenic erectile dysfunction. The diagnosis is suggested by an abnormality in testing for nocturnal tumescence in association with normal haemodynamic studies.8

Vascular injury.

Major vascular damage has been described in patients with impotence after pelvic fracture. The vessels may be lacerated directly or by damage to the intima leading to thrombosis.21 Sharlip22 described obliteration of both the internal pudendal arteries at the level of the urogenital diaphragm on arteriography in three patients. They remained impotent despite formation of collateral vessels and retrograde filling of the dorsal and deep penile arteries. A traumatic arteriovenous fistula between the internal iliac vessels has been described as causing impotence.23 The symptoms resolved after surgical correction.23

Levine et al14 investigated impotent patients, who had suffered blunt perineal or pelvic trauma, using selective internal pudendal arteriography. All of those with a fracture had suffered injury to the posterior urethra. Vascular lesions were identified in both hypogastric-cavernous beds in all patients. Lesions of the pudendal or common penile arteries were seen in 90% of those with a fracture as opposed to 35% of those without. However, 48% of the latter group were more likely to have had an isolated lesion in the cavernous artery compared with only 8% in the group with a fracture. Similar lesions were observed in a control group with vasculogenic impotence but with no history of trauma, although proximal lesions were more common.

Munarriz et al24 investigated a similar group of patients, 42 of whom had suffered fractures of the pelvis. All those with fractures, and 96% overall, had an abnormality of the cavernosal response to pharmacological stimulation, suggesting that there was a haemodynamic component to their pathology; 80% of patients who had sustained trauma to the pelvis had an abnormality of venous outflow and 70% an abnormality of venous occlusion. All the patients with fractures who underwent cavernosography had abnormal findings, 97% of these being proximal, site-specific leaks into local venous structures. Leaks were seen at three or more sites in 60%, the most common being the crural and cavernous veins and the proximal corpus spongiosum. This venous abnormality was more common than in the patients who had not suffered a fracture. Angiography again detected widespread lesions, most frequently in the common penile and cavernous arteries. They concluded that the most likely cause of the widespread venous leak was the direct impact of the fixed proximal corpora against the pubic rami or a shearing effect at the point of fixation. Purely arterial damage was noted in 30% of the patients indicating that they might respond to microsurgical revas-cularisation.8 Veno-occlusive abnormalities in patients with erectile dysfunction in general are unlikely to be related to a venous anomaly but probably represent inadequate function of smooth muscle which may be secondary to arterial disease or neural damage.25 This may also be the case in those with a fracture.

Corporal injury.

While the penis itself is free and therefore relatively protected from the effect of blunt trauma, the corpora are attached to the undersurface of the ischiopubic rami. This makes them vulnerable to shearing forces during fracture. Such injury could induce fibrosis during healing, subsequently impairing the ability to dilate.8 MRI has shown a high incidence of injury to the corpora in patients with a fracture, urethral injury and impotence. Armenakas et al26 demonstrated injuries in 12 of 15 patients examined, including eight avulsions and four fractures. Doppler imaging revealed that 12 patients had primarily vasculogenic impotence with low peak flow velocities, and three had neurogenic impotence. Of the vasculogenic group, five had a significant venous leak as shown by large rates of diastolic flow. These were associated with severe injury to the corpora in four patients and it may be that fibrosis may also interfere with the complex veno-occlusive mechanisms.26 Such injuries may lead to the formation of venous fistulae, allowing abnormal leakage of blood.

Psychogenic injury.

Patients with fractures of the pelvis are likely to suffer psychological problems for prolonged periods during their recovery. Up to 10% of survivors of motor-vehicle accidents may suffer post-traumatic stress disorder.27 Sexual dysfunction has been reported in 80% of patients with established post-traumatic stress disorder.28 Impotence has also been observed in association with medication for post-traumatic anxiety, resolving on withdrawal of the drug.29 When dealing with these patients it is important to keep an open mind about the possible aetiology of their problems. Furthermore, any patient having problems with sexual performance will develop a psychological response to this in addition to any underlying organic pathology.

It is clear that while vascular lesions are common, neurological abnormalities are probably also present when they are specifically sought, given the close association of the neurovascular supply. In the study by Ellison et al15 (Table II) 42 consecutive patients with fractures of the pelvis underwent evaluation of penile blood flow and pelvic innervation very early after injury. Half were found to have abnormal penile brachial pressure indices. Of these, nine had a urethral injury, 13 abnormal bulbocavernosus reflexes and 14 a multisystem injury. However, patients with normal penile blood flow had normal reflexes and no injuries to the urethra. Only two had suffered polytrauma. At follow-up, 11 of the patients who had abnormal preliminary investigations reported sexual dysfunction, but this occurred only in two of six of the normal group.

Early management: prevention of erectile dysfunction in patients with pelvic trauma

The management of urethral injuries during the acute stage after trauma has a profound influence on the development of sexual dysfunction, and can alter the long-term outcome in terms of continence and potency.1,20 There are various options for management and no clear preference has emerged. Immediate primary suturing has fallen from favour because of high rates of subsequent incontinence and impotence. Urethral realignment, historically the most frequent treatment for complete urethral disruption, lost favour in the early 1970s because of the large numbers of patients with these complications.3,4,12,30,31

The initial management of injury to the urethra.

Suprapubic urinary diversion and delayed primary repair have been adopted in most centres.1,30 In order to reduce the rise of iatrogenic injury, early intervention in local soft-tissue disruption is kept to a minimum. Urine is diverted through a standard suprapubic catheter introduced using an open or radiological technique. Subsequently, after a delay of six to eight weeks, antegrade urethrography is performed through the catheter. Some patients will have incomplete lesions which may heal spontaneously. If the patient can void, the tube can be removed and the patient observed for formation of a stricture. If a stricture is present, delayed end-to-end repair can be undertaken after the acute problems have been resolved.2 The timing will depend on the individual patient.32 Pre-operative retrograde and ante-grade urethrography is used to delineate the location and the length of the defect (Fig. 1).

Fig. 1

Acute management of urethral injury using early suprapubic catheterisation (SPC); clinical signs of urethral injury include blood at the urethral meatus, superiorly displaced prostate, severe perineal haematoma, gross haematuria and urinary retention. Early urethral realignment remains an alternative option to SPC if local expertise is available and the patient is stable.

Although definitive fixation of the fracture and operation on the urological injury are usually avoided, Routt et al33 achieved good results using a team approach. Over a period of five years 28 patients underwent open reduction and internal fixation of anterior pelvic fractures together with repair of the bladder or indirect urethral realignment. The procedures were carried out at a mean of 1.9 days (0 to 15) after injury. There were no deaths and no cases of nonunion or early infection. One late deep infection required multiple debridements and exchange of internal fixation. Late impotence was seen in three of the 18 male patients although it is unclear how this distribution related to urethral injury and realignment.33 However, the problems associated with the management of injuries to the bladder are different from those encountered in the urethra. While the rate of infection is relatively low, these procedures will generally be prolonged and complex. With the increasing adoption of an approach of damage control in patients with severe trauma, appropriate timing and careful selection of patients for this type of management are mandatory.34,35

An immediate suprapubic catheter and delayed reconstruction vs primary realignment.

Since the primitive blind techniques for approximating the two ends of the divided urethra have been replaced by methods using fluoroscopic and endoscopic guidance, primary realignment is regaining some favour. Simultaneous antegrade and retrograde urethroscopy through the suprapubic site can be employed.36–,38 Once alignment has been achieved, a urethral catheter is passed to maintain this position and left in situ for four to six weeks while healing occurs.3,36,38–,43 This method of realignment may reduce iatrogenic injury and consequent sexual dysfunction by avoiding gross manipulation of the peri-urethral tissues. However, this approach remains controversial.2 In stabilised patients, given adequate facilities and expertise, techniques of early realignment can be safely employed. However, unstable patients are managed better by suprapubic urinary diversion and delayed realignment or another secondary procedure.3

Koraitim31 published a meta-analysis which compared the results of various methods of repair for urethral injuries in 871 patients over 40 years until 1995. In grouped analysis, primary suture was associated with highest rates of both impotence and incontinence at 56% and 21%, respectively. Impotence was found twice as frequently after primary realignment (36%) than after delayed repair; and incontinence was unusual after either procedure. However, it is clear that patients who were included in the same group in this study were treated in different ways. Modern endoscopic or image-guided techniques of realignment are very different from the open methods used 50 years ago, and methods of assessment of sexual function have evolved greatly during this time. Many of the more recent studies use very strict criteria to determine sexual function and include patients in the impotent group who have noted a decreased quality of erection. The two techniques should not be viewed as competing but complementary, being selected on an individual basis determined by the type and extent of injury. In centres without expert urological assistance suprapubic drainage can be easily achieved and repair delayed until such support is available.30,31

Tables III and IV collate the information relating to the management of injury to the urethra from all identified studies. When recent data are included in the analysis, the differences in the rates of impotence between early and delayed repair are no longer significant. It has been suggested that trauma causing impotence is related to the injury rather than the treatment and this is consistent with these findings. If only those studies carried out after 1975 are included, patients undergoing primary realignment have a significantly lower rate of impotence (22.6% vs 42.9%; Fisher’s exact test, p < 0.0001) than those with delayed repair. Good results may be achievable using a method which avoids disruption of the peri-urethral tissues and prolonged suprapubic catheterisation. However, these techniques require further assessment and analysis of their long-term outcome, particularly as regards the incidence of stricture formation. With an overall incidence of initial stricture of more than 50%, many of these patients will require further intervention when compared with an incidence of secondary stricture after delayed repair of 15%. However, the method used in this analysis is subject to the same limitations as that in the original work of Koraitim.30,31 The different techniques, criteria for inclusion and outcome measures in these studies indicate that the results should be treated with caution. Patients are also likely to be subject to selection bias, those with more severe injuries being less likely to undergo primary realignment. Moreover, in several series, patients who had undergone failed primary realignment were included in the series of delayed repairs. Prospective randomisation would be helpful but given the relatively small number of patients available, and the heterogeneity of their injuries, such an undertaking would be difficult. In the meantime, both techniques have a place in appropriate situations.30,31

View this table:
Table III.

Studies of delayed repair. In studies with incomplete follow-up data, the percentage quoted is of the number finally assessed. Different assessment rates were used in different studies. The most critical results are quoted here. Primary stricture after suprapubic catheterisation and before final treatment. Secondary stricture following a primary definitive procedure in patients requiring further operative intervention

View this table:
Table IV.

Studies of primary realignment. In studies with incomplete follow-up data, the percentage quoted is of the number finally assessed. Different assessment rates were used in different studies. The most critical results are quoted here. Primary stricture after suprapubic catheterisation and before final treatment. Secondary stricture following a primary definitive procedure in patients requiring further operative intervention

Clinical evaluation of patients with erectile dysfunction after pelvic trauma

The assessment of patients reporting sexual dysfunction after pelvic injuries should include a thorough history, physical examination and psychological review. A general assessment should be made to identify any other pathology which may contribute to sexual dysfunction, such as medication, diabetes and neurological or vascular disease. A full sexual history should establish exactly how the patients perceive their problems. It is important to determine the temporal relationship of the dysfunction to the trauma and to its subsequent treatment, the duration of these symptoms, difficulties in achieving or maintaining an erection, the quality of nocturnal or masturbatory erection and the presence of local abnormalities or symptoms. A basic psychological assessment should be performed and enquiries made regarding other aspects of sexual function such as ejaculation. The presence of normal nocturnal or early morning erections is strongly suggestive of a psychogenic aetiology.7,44,45 A thorough assessment of the patient’s medical history, including exposure to drugs, is necessary to identify other possible causes of erectile dysfunction. The blood pressure should be measured and physical examination of secondary sexual characteristics and the external genitalia undertaken, seeking palpable areas of thickening within the penile shaft or the crura.

Investigation.

Basic screening investigations should be performed in all patients, including the fasting blood sugar and lipids, and an assessment of the serum testosterone. Further investigation is indicated only after a failed trial of oral treatment in patients who would be candidates for reconstructive vascular surgery. Initial investigations should include pharmacological testing and vascular assessment using techniques including colour Doppler and duplex ultrasonography, dynamic infusion cavernosometry and arteriography. Neurophysiological testing is rarely performed, largely because the tests involved are not specific for damage to the autonomic nerves. Assessment of nocturnal penile tumescence is reserved only for those cases in which there is serious consideration of a psychogenic aetiology, but no definitive proof. MRI is helpful in the appraisal of injuries to the corporal bodies. A suggested approach to assessment and management is presented in Figure 2.

Fig. 2

Suggested management for patients presenting with sexual dysfunction after fracture of the pelvis (PDE5i, phosphodiesterase-5 inhibitor, e.g. sildenafil).

Testing of nocturnal tumescence.

Nocturnal erection has long been used to differentiate between patients with psychogenic and organic impotence. Testing to specific criteria has been shown to be highly specific for organic impotence. A patient who has achieved even one sustained, good erection is unlikely to have serious underlying organic pathology.45

Pharmacological testing.

An artificial erection can be induced by injection of a pharmacological agent into one of the corpora cavernosa. Usually 10 to 20 μg of prostaglandin E1 is used and a rigid erection ensues within ten to 20 minutes. Examination of the erect penis allows identification of any deformities which may have been caused by the injury. It also enables assessment to be made of the vasculature of the penis, which is most usefully performed when the smooth muscle of the penis is maximally relaxed. In the presence of a fracture of the pelvis an absent response suggests vascular pathology.7,45

Vascular investigations

Colour Doppler and duplex scanning.

Intracavernous-injection-enhanced duplex ultrasonography has been used to investigate the vascular anatomy and physiology of the genitalia. The most important parameters are the peak systolic (PSV) and end-diastolic velocities (EDV) measured in the central penile (cavernosal) arteries. A reduced PSV is indicative of reduced arterial inflow, although there is uncertainty as to the normal levels. Typically, values of less than 25 cm/s are considered to be abnormal, with those greater than 35 cm/s as normal. A raised EDV may indicate veno-occlusive dysfunction which is usually secondary to failure of relaxation of smooth muscle. This is due either to neurogenic injury or to disorders of inflow, and therefore is not usually a helpful measurement in patients with fractures.

Dynamic infusion cavernosometry.

This is a technique designed to assess veno-occlusive function and is used much less often than in the past. In a patient with erectile dysfunction after fracture of the pelvis, its use is only indicated when there is clear evidence of an arterial injury, but some assessment of function of the smooth muscle of the penis is required. It is both time-consuming and invasive.21

Arteriography.

When Doppler scanning has suggested an arterial lesion and revascularisation is considered, selective pudendal arteriography may be indicated. The technique involves cannulation of the internal pudendal artery, usually through a femoral puncture, and examination of the distal vascular tree to identify and localise any vascular lesion. As with Doppler scanning of the penile vasculature, it is most usefully performed when relaxation of the smooth muscle is maximal, produced either by an intracavernosal injection of prostaglandin E1 or by a localised infusion of glyceryltrinitrate.14,23 Arteriography is only of use when vascular surgical intervention is being considered.

Neurophysiological studies

Bulbocavernosal reflex.

This tests the integrity of the sacral nerve roots and the pudendal nerve. The dorsal nerve of the penis is stimulated and latencies measured at the striated muscle of the bulbocavernosus, and urethral or anal sphincters. It has proved to be unreliable in the investigation of male sexual dysfunction since it assesses only somatic nerves. It is no longer used routinely.46,47

Testing of the thermal threshold.

The thresholds of thermal sensitivity have been used as a research tool to determine dysfunction of small nerve fibres which are of the same diameter as those carrying autonomic signals.46,47 Such tests are useful in patients suspected of having a peripheral neuropathy. Markedly abnormal sensitivities, similar to those seen in diabetics have been recorded in patients after trauma to the pelvis.16 However, the major disadvantage of this technique is that it assesses sensory fibres rather than the efferent parasympathetic fibres which stimulate penile erection.

Sympathetic skin response.

Certain events such as a sudden noise or electrical stimulus can be used to evoke an anatomically-mediated change in the electrical conductance of the skin. Measurement of this has been used to assess the autonomic function of the peripheries, including the penis. Some inconsistencies in results have been observed. For example, some diabetic patients with generalised autonomic neuropathy and erectile dysfunction have preserved responses, with a normal vascular response. It is unclear from current studies if this is because their impotence is not related to autonomic dysfunction or because the test is unreliable.46

Electromyography of the corpora cavernosa.

Needle electrodes, later superseded by surface recording, have been used to measure electrical activity within the smooth muscle of the corpora. Patients with pelvic trauma were shown to have markedly abnormal recordings compared with normal controls. Recent work however, has called into question the validity of these recordings and any results should be viewed with caution.46

MRI.

This has been used to define urethral injuries before reconstruction. It is able to visualise accurately the erectile tissue of the penis and detect any injuries present. In difficult cases, greater clarity can be obtained by producing a pharmacological erection with prostaglandin E1 before MRI. While the resolution of these studies does not allow imaging of the complete vascular supply to the penis, it does give useful information about the degree of local disruption of soft tissues, in conjunction with vascular studies, to gain a greater understanding of the pathological problems.26

The management of patients with sexual dysfunction after fractures of the pelvis

In the early stages after a fracture of the pelvis, the patient is unlikely to be considering resumption of sexual activity. However, as recovery occurs, such issues become more important. Various methods can be used to improve sexual function. Figure 2 outlines an approach to the management of these patients. It is important to note that spontaneous return of sexual function at up to 18 months has been reported in up to 20% of patients after these injuries.26,48–,50 It is worth waiting before instituting invasive methods of treatment.

Pharmacological treatment

Phosphodiesterase inhibitors.

The introduction of sildenafil (Viagra) has revolutionised the treatment of male sexual dysfunction. It is an inhibitor of phosphodiesterase (PDE) type 5, which is the enzyme that terminates the action of cyclic guanine monophosphate within the smooth muscle cells in the penis. It potentiates normal erectogenic signals and facilitates normal erection in men with a normal nerve supply and an adequate vasculature. It has been used with excellent effect in many men with erectile dysfunction, regardless of the cause.7,51–,57 The drug has also been shown to be effective in the treatment of sexual dysfunction related to PTSD.28 Given its efficacy in such a diverse range of conditions, sildenafil would be expected to be effective in some patients after fracture of the pelvis. One group in whom it is often ineffective are those who have undergone radical retropubic prostatectomy, in which both cavernosal nerves have been sacrificed in order to maximise the chances of a cure for the cancer.58–,61 It is therefore reasonable to assume that patients with erectile dysfunction secondary to a fracture of the pelvis who respond to sildenafil, have an intact nerve supply and an adequate vasculature. Those who fail to respond may be expected to have a damaged nerve supply and or vascular insufficiency. Unfortunately, no studies have been carried out to investigate this. In recent months, two new PDE5 inhibitors have been launched which are similar in efficacy and which have similar side-effects.62

Intracavernosal injection.

Various smooth muscle relaxants have been injected into the corpora cavernosa to induce erection directly, without neuronal input. The most commonly used agent is prostglandin E1. These drugs are now usually reserved for use in patients who have not responded to oral therapy. The results are generally good, although a high proportion of patients discontinue the treatment because they are dissatisfied with the technique of administration. Side-effects include penile pain after injection and, rarely, a prolonged erection, which may require active treatment. Patients with severe vasculogenic impotence are unlikely to respond, although those with neurological damage will usually find intracavernosal injections to be effective.7,51

Physical treatments

Vacuum constriction device.

A vacuum pump is placed over the penis, drawing blood into the cavernous system, and a band is placed around the base to maintain the erection. Although this method is non-invasive and often effective, patients are seldom satisfied with the results.7

Penile prostheses.

Various surgical implants have been used to restore potency when other methods have failed. That introduced most recently consists of inflatable cylinders placed within the corpora, with a saline reservoir and pump in the scrotum. The devices are reliable and easy to use. Failure and infection have been uncommon. This solution may be less acceptable to young patients until other options have been exhausted.7

Microsurgical penile revascularisation.

Microsurgical reconstruction has been used in arteriogenic impotence with varying results; success rates varying between 16% and 74% have been reported. Many different techniques have been described in this highly specialised field. The procedure to be undertaken should be chosen for the patient once the vascular anatomy has been defined. Correct patient selection is crucial and while the focal nature of vascular injury should make these patients ideal candidates for operation, the presence of veno-occlusive incompetence is a relative contra-indication since it implies permanent damage to the smooth muscle of the penis. The best results are achieved when careful pre-operative assessment is used to identify patients with isolated arterial lesions and an intact veno-occlusive mechanism. Complications include chronic pain, loss of sensation, penile shortening and failure of the graft.21,63

Although natural recovery may take up to 18 months, there is a theoretical danger that loss of normal erectile activity may be associated with relative hypoxia, loss of smooth muscle and end-organ failure, even if any neurapraxia recovers. There is therefore a case to be made for early treatment both to improve the symptoms and also to prevent long-term loss of smooth muscle, as demonstrated in recent studies in men undergoing radical retropubic pro-statectomy.64

Conclusion

Erectile dysfunction following fractures of the pelvis is more common than has been previously supposed. It is important to explore these problems specifically with all patients at follow-up. Whereas the options for treatment have previously been complex and limited, sildenafil offers a well-accepted potential treatment with few side-effects. A small number of patients may be amenable to surgical reconstruction should this fail.

References