In most instances a urethral stricture is a narrowing of the calibre of the urethra caused by the presence of a scar consequent on infection or injury. However, not all strictures are caused by scarring, if a scar is taken to mean only fibrosis. So-called congenital strictures of the bulbar urethra have a substantial smooth muscle content  and are thought to arise from failure of normal canalization rather than fibrotic restriction of a normal-calibre urethra . There is also a difference between a partial constriction of the urethra caused by, e.g. inflammation (in which the epithelial lining is retained) and a fibrotic obliteration between the distracted ends of a ruptured urethra. Thus not all strictures are the same.
However simple the pathology might appear to be, the aetiology of many strictures is unknown. When most strictures were gonococcal in origin it was nonetheless clear that a stricture did not necessarily follow an attack of gonorrhoea, and when it did the time lag could be considerable. Thus an episode of infection or trauma is not necessarily a direct cause of a stricture. In an interesting review of the anatomy of urethral stricture disease, Chambers et al. noted that the first identifiable change in urethral stricture disease was a change in the nature of the urethral epithelium from a pseudo-stratified columnar epithelium to a columnar epithelium that lacks the waterproofing quality of the pseudo-stratified variant. Consequently, they hypothesized that urine could extravasate and lead to fibrosis. The particular association of inflammatory stricture disease with the mid-bulbar and distal penile segments of the urethra was explained by Singh and Blandy  on the basis that most urethral glands were at those sites. Thus inflammatory urethral strictures were the consequence of fibrosis beginning initially in the corpus spongiosum, caused either directly from the extravasation of urine through the urethral epithelium or indirectly by involvement of the urethral glands at specific sites, and subsequent extension into the spongiosum.
Once initiated, fibrosis within the corpus spongiosum can cause constriction of the urethral lumen and if there is also infection, this can cause micro-abscess formation within the urethral glands, which makes fibrosis worse and may lead to peri-urethral extension of that fibrosis .
With progressive constriction of the urethral lumen and if outlet obstruction develops, secondary changes may arise in the lower urinary tract, and even the upper urinary tract . One consequence of obstruction is a predisposition to recurrent urinary infection and secondary prostatitis and epididymitis are also common. Upper tract complications are less common now than they were 50 years ago, but are still common in certain areas of the world with less than adequate access to urological care.
The traditional treatment for urethral stricture disease is urethral dilatation or urethrotomy. Both have an ancient history but currently urethrotomy tends to be associated with visual urethrotomy, which is of course a recent development; traditional urethrotomy was ‘blind’.
Most experienced authors recognized that neither urethrotomy nor dilatation was curative and that once started on either, the unfortunate patient with a urethral stricture was condemned to such treatment for life. Consequently the patients was prone to false passages, bleeding, septicaemia and ultimately an impassable stricture, which would culminate in complete urinary retention unless the patient developed an external fistula. Thus the ‘watering-can’ perineum, which is currently perceived as a disastrous situation, would have been the only way some patients with strictures would have survived in earlier times. With the advent of sterile instrumentation and antibiotics the infective consequences of a lifetime of urethral dilatation are less apparent than they were, but nonetheless Devereux and Burfield  concluded that there was a considerable morbidity in life-long dilatation which could not be overlooked.
The results of dilatation and urethrotomy were recently reviewed by various authors, who concluded that urethrotomy had no advantage over dilatation in terms of cure rate and that if one urethrotomy or dilatation fails to cure a patient then a second rarely will and a third never does. Furthermore, the only strictures that regularly respond are single short strictures of the bulbar urethra and then only in ≈ 60% of cases. Multiple stricteres, strictures > 1 cm long and strictures in the penile urethra rarely respond [8–10].
Thus for most people with urethral stricture disease, dilatation or urethrotomy are palliative management rather than curative treatment. Various methods have been used to improve the results, e.g. using a laser to divide the stricture rather than a cold knife , using a stent to hold the stricture segment and the urethra open , or using self-dilatation in the form of CISC to hold the urethra open after an initial urethrotomy [13,14].
The results of laser urethrotomy suggest that it has no advantage over cold-knife urethrotomy  although theoretically the vaporization of a superficial annular intraluminal ring constriction might work, where a laser urethrotomy performed in the same way as the cold-knife urethrotomy might not. Intraurethral stents may be effective when spongiofibrosis is not too extensive. When it is, the fibrotic tissue tends to grow through into the lumen of the stent, particularly in traumatic strictures and those with periurethral fibrosis . Moreover, even when successful, there is a significant morbidity with intraurethral stents because the urine pools, causing postvoid dribbling and irritative symptoms, if not frank infection . There is also pooling of semen after intercourse, interfering with ejaculation and its perception. Some patients have discomfort if not frank pain at the site of a stent. Likewise CISC is not accepted by all patients, who like to have their diseases treated by doctors rather than by themselves. Even those who are prepared to try CISC tend to discontinue subsequent CISC programmes because of difficulties, if only the social inconvenience of the technique . Thus the only curative treatment for most patients with urethral stricture disease is urethroplasty.
Principles of urethroplasty
The results of urethroplasty show that the best success rate (i.e. the lowest re-stricture rate and the lowest complication rate) is achieved by excising the stricture and end-to-end anastomosis of the two healthy ends on either side [19–22]. Thus the first principle of urethroplasty is that an end-to-end anastomotic repair should be performed whenever possible. In practice this is not applicable to strictures of the penile urethra because the shortening of the urethra would interfere with erection thereafter. Short strictures of the bulbar urethra (< 2 cm) can be excised and repaired end-to-end without interfering with erection significantly, although the angle of the erect penis against the anterior abdominal wall may be slightly more of a right angle than the usual acute angle. For appropriate strictures of the membranous urethra end-to-end repair is almost always possible irrespective of the length of the stricture.
When an end-to-end anastomotic repair is not possible, i.e. when the stricture is in the penile urethra or for strictures of the bulbar urethra that are > 2 cm long, then some form of substitution urethroplasty will be necessary. Several materials have been tried but essentially they fall into two groups, i.e. vascularized flaps of local genital skin or free grafts. Because they carry their own blood supply, skin flaps have traditionally been the most popular form of substitution urethroplasty but they seem more prone to complications than free grafts. Recently, particularly since the advent of the buccal mucosal free graft, there has been a tendency towards the use of free grafts but not necessarily using buccal mucosa. Full-thickness grafts of penile shaft skin are as effective as buccal mucosal free grafts, although the patient may not be as satisfied with the consequent scarring at the donor site. Free grafts may be taken from other sites but as a general rule extragenital skin is only suitable when it is taken from above the jaw line and here, with the exception of buccal mucosa and postauricular skin, donor site scarring would be too conspicuous. Thus, in general, free grafts of either penile shaft skin or buccal mucosa are used for most instances of substitution urethroplasty unless the circumstances demand the use of a skin flap, e.g. when there is infection or poor vascularity sufficient in either case to impair graft take.
Bladder neck and prostatic urethral stricture disease
Bladder neck strictures are generally the result of TURP or retropubic prostatectomy in a patient in whom dyssynergic bladder neck obstruction was the actual pathology and in whom the bladder neck obstruction persists with fibrotic stricturing in addition . This situation is perhaps made worse by secondary infection. In such circumstances the treatment is a bladder neck incision right through the full circumference of the bladder neck and into fat.
Strictures of the prostatic urethra are rare; they occasionally follow some of the ‘new technology’ treatments for BOO, e.g. laser surgery or cryosurgery. When they occur they can be extremely difficult to treat . Some form of ‘cut to the light’ treatment may be necessary to provide a channel that can then be enlarged by transurethral resection, but ultimately excision of the scarred tissue, which is essentially an anatomical prostatectomy, is necessary, with end-to-end anastomosis of the bladder neck onto the urethra below. In such circumstances it is common to find that both sphincter mechanisms are totally destroyed and the patient is therefore incontinent; an artificial sphincter may then also be necessary. The alternative in the unfit patient would be some form of supravesical urinary diversion. Similar strictures sometimes follow radical prostatectomy for prostate cancer, particularly after subsequent radiotherapy . These strictures can be particularly difficult to treat and, if instrumentation fails, revision of the anastomosis is rarely possible.
Membranous urethral strictures
Membranous urethral strictures may occur as a result of transurethral resection, in which case the urethra is still in continuity, the lumen is usually still patent, the stricture is usually superficial and the real problem is when sphincter function is compromised by fibrosis. Such patients are best treated by urethral dilatation to preserve sphincter function .
Less commonly but more dramatically the urethra may be ruptured in association with a pelvic fracture; this not so much a stricture in most patients but a distraction defect . There is no continuity of the urethra, the lumen is (usually) completely obliterated and there is simply a block of fibrous tissue between the ends of the urethra.
There is considerable controversy about the immediate management of pelvic fracture injuries of the posterior urethra. There are probably as many who advocate suprapubic catheterization and then a urethroplasty ≈ 3 months later (e.g. [27,28]) as those advocating early so-called ‘realignment’. The latter used to be achieved by the open surgical procedure of ‘railroading’ (e.g. [29,30]) but currently it is more commonly performed endoscopically by a variety of endourological manoeuvres, of which the commonest is probably to pass a flexible cystoscope through a suprapubic tract, a rigid cystoscope up the urethra and then a guidewire between them, so that a catheter can be passed over the guidewire . Whatever this achieves it certainly is not a reduction in the stricture rate; the rate is the same whichever of these two approaches is used [27,32]. Open ‘railroading’ was associated with a higher incidence of complications such as impotence, incontinence and bleeding, but this may simply have reflected the selection of patients undergoing this procedure . Certainly suprapubic catheterization and delayed urethroplasty causes the least harm in the early stage of management, and in practice produces a 10-year stricture-free survival of > 90% [19–22]. It is against this value that all other methods of managing these injuries must be judged.
Developing endourological alternatives can be expected in due course to supplant open urethroplasty after 3 months of suprapubic catheterization, but it will be some time before these techniques are shown to be effective. Most are aggressive variants of urethrotomy but achieved with more sophisticated instrumentation  and there seems little reason to suppose that these will ever give results as good as those with open urethroplasty. However, some are a little more imaginative, such as Naude's endoscopic urethroplasty , in which a full-thickness skin graft is mounted on a carrier and fixed in place of the urethra, with a covering suprapubic catheter until the graft has (hopefully) taken.
The key to anastomotic urethroplasty lies in two anatomical points: first, that the bulbar urethra is elastic and can be stretched for 2–4 cm to overcome a defect and allow an overlapping spatulated anastomosis; and second, that the natural course of the bulbar urethra is nearly semicircular so that by straightening out the natural curve even longer defects can be bridged than by elasticity alone. The elasticity of the bulbar urethra can be exploited simply by mobilizing it from its attachments within the bulbospongiosus muscle. This means dividing the bulbar arteries (if they are still patent) and the other smaller arteries that supply it, knowing that there is a second blood supply to the bulbar urethra, retrogradely from the corpora cavernosa through the glans and back up the corpus spongiosum. This sustains the viability of the bulbar urethra and the mobilized segment of the corpus spongiosum, as long as it is not placed under too much tension. Some 2–4 cm of elastic lengthening can be gained by bulbar urethral mobilization, but 1 cm will be lost from this because of the need to spatulate the end of the urethra for anastomosis to the similarly spatulated prostatic urethra. This spatulation is important because it means that even if there is some contraction of the anastomosis after surgery, as would be expected, the calibre will remain adequate.
For longer defects (which are the majority) bulbar urethral mobilization will not be enough and the urethra will require straightening ( Fig. 1a). The anatomical structures that produce the normal curved course of the bulbar urethra are the fusion of the crura of the penis and the underlying inferior pubic arch. Fortunately, the proximal 5–7 cm of the fused crura can be separated in an avascular plane before (more distally) this plane dissipates and the corporal bodies merge together on the shaft of the penis. If splitting the corpora ( Fig. 1b) to allow the urethra to lie between them is not sufficient to allow the two ends of the urethra to come together without tension, then a wedge of the inferior pubic arch can be taken out to straighten the course of the bulbar urethra further ( Fig. 1c). If these two manoeuvres together still fail to allow the bulbar urethra to reach the prostatic urethra above the level of the distraction defect, then the urethra can be re-routed around the shaft of the penis. If the crura of the penis were separate structures for more than the normal 5–7 cm and they could therefore be separated higher than is possible, then re-routing the urethra around the penis would not be necessary. The separation is achieved by creating a channel between one corporal body and the inferior pubic ramus ( Fig. 1d) to which it is attached, and then chiselling sufficient bone from the ramus so that the urethra will not be compressed and thereby devascularized as it lies in its new position ( Fig. 1e).
Each of these manoeuvres (mobilization of the urethra to develop its elasticity and then crural separation, inferior pubectomy and re-routing of the urethra) can gain about 2 cm (and sometimes more) towards bridging a distraction defect. It is almost always possible to bridge any length of defect by a combination of these techniques, as long as the bulbar urethra and its blood supply are normal and have not been compromised by previous surgery or disease.
Anastomotic urethroplasty gives 10-year stricture-free survival rates of > 90% [19–22]. The complications of such surgery, other than failure manifest by re-stricturing, are largely the complications of the original injury, i.e. urgency and stress incontinence with a full bladder because of damage to the urethral sphincter mechanism, and impotence caused by damage to the neurovascular bundles as they run along the inferior pubic rami. Occasionally impotence is the result of surgery but some authorities deny this [20,22]. The neurovascular bundles are at their most vulnerable as they course laterally alongside the membranous urethra to run into the penile hilum. Thus the dissection associated with an inferior pubectomy is particularly liable to damage the neurovascular structures.
Bulbar urethral strictures
Bulbar urethral trauma can arise from ‘fall-astride’ injuries typical in bike riders, farmers and construction workers, but is more commonly the result of endoscopic trauma, including prolonged urethral catheterization (although such trauma is milder). Infection is another common cause, although the association is rather loose and non-specific urethritis is currently a much commoner cause than gonorrhoea. The third equally large category of bulbar urethral strictures occur for no apparent reason; they typically occur in young men at the junction between the proximal and middle thirds of the bulbar urethra, contain a surprisingly high content of smooth muscle on biopsy , and are generally designated ‘congenital’. These have been linked with so-called Cobb's collar [36,37] and it may be that for some reason these strictures develop at the point of fusion of the urethra from its two separate embryological origins. A narrowing at this site of the urethra is a common radiological observation which does not necessarily mean that a stricture is present . A visible endoscopic ‘arching’ at this point is also a common finding that does not mean a stricture. For both these reasons the incidence of congenital strictures is hard to define but strictures at that particular point of the bulbar urethra in young men (sometimes teenagers, rarely younger children) with no history of trauma or infection comprise about a third of all patients currently requiring bulbar urethroplasty.
Congenital and traumatic strictures of the bulbar urethra are generally short and can be treated by anastomotic urethroplasty. This is usually achieved simply by bulbar urethral mobilization with no need for any of the ‘straightening’ procedures mentioned for membranous urethral strictures; 2–4 cm of elastic lengthening can be gained from bulbar urethral mobilization. After excising the stricture, each end needs to be spatulated, which takes 1 cm on either side. Thus the overlapping spatulated anastomosis (which gives a sufficient calibre to the anastomosis, so that even allowing for some long-term contraction there will still be a normal calibre urethra) uses up 2 cm. As a result the maximum length of defect that can be overcome, or maximum length of strictured urethra that can be excised, is 2 cm. Infective strictures are generally longer and if not apparently longer have compromised urethra on each side, either because the urothelium is unhealthy or because there is fibrosis within the corpus spongiosum adjacent to the stricture, even if the adjacent area is not actually strictured. Some authorities have sought to define this area of adjacent spongiofibrosis by preoperative ultrasonography . This provides good imaging of adjacent spongiofibrosis but most surgeons feel that this is equally visible during surgery and can in any case generally be anticipated; therefore such imaging is viewed as not particularly helpful.
For longer strictures of the bulbar urethra, excision and an overlapping end-to-end anastomosis would fail because of tension between the mobilized ends of the urethra and recurrent ischaemic stricture as a consequence. In some circumstances it might be possible to excise the diseased urethra and bring the two ends together, but not with an overlapping anastomosis. In these circumstances it might be possible to anastomose half of the circumference of the two ends of the urethra, leaving both ends spatulated on the other half of their circumference. This spatulated defect (which would be ≈ 2 cm long) could then be closed with a patch. (This is sometimes known as the augmented roof strip procedure). Otherwise the stricture is simply opened throughout its length and for 1–2 cm into healthy urethra on either side, and this ‘stricturotomy’ then held open with a patch. The question is then what to use as a patch. Historically, free grafts were used initially but experience showed that only full-thickness skin grafts of genital skin fared well. Extragenital skin and split-thickness skin grafts have an unacceptably high failure rate ; extragenital full-thickness skin grafts fail more commonly than genital skin grafts because of a much higher failure to take; split skin grafts are less satisfactory than full-thickness skin grafts because they contract.
Vascularized flaps became increasingly popular during the 1970s, in part as a historical development from earlier two-stage procedures and in part because of the instinctive feeling that a patch that carried its own blood supply must logically work better than a patch that did not. Scrotal skin was the material initially used for patch urethroplasty using flaps, but this was subsequently shown to be less successful than penile or preputial flaps . However, although these were better than scrotal flaps they still caused the same problems of ballooning of the flap at the site of the patch, giving rise to pooling of urine and infection in some, and more generalized irritative symptoms in many more patients. These complications were partly the consequence of making a ventral stricturotomy, partly because the flap was much thinner than the natural urethra (supported as the latter is by the corpus spongiosum) and partly because of difficulty in sizing the patch at the time of urethroplasty. More recently the introduction of the buccal mucosal full-thickness graft, with its ease of harvesting, has displaced genital skin flaps as the preferred type of patch urethroplasty in many surgeons' practice . Another popular recent introduction is the Barbagli procedure [43,44] which places the stricturotomy and the subsequent graft dorsally ( Fig. 2a). This has found particular favour because this means that the graft is supported by the underlying corporal bodies, which virtually eliminates ‘out-pouching’ of the graft as a consequence ( Fig. 2b). Urine still pools in the reconstructed urethra and there is also pooling of seminal fluid; some patients find the resulting postvoid dribbling and reduced force of ejaculation significantly irritating. However, the more serious consequences of ballooning of the graft appear to be less common.
Thus, in general, the patch-graft procedure using a buccal mucosal free graft is rapidly becoming the procedure of choice for strictures of the bulbar urethra that are not amenable to excision and end-to-end anastomotic repair. This does not mean that a buccal mucosal or genital full-thickness graft is any better than a flap in managing the stricture (indeed, the results suggest that they are the same, with an 85% cure rate ), but full-thickness grafts are quicker and easier to perform than flap procedures, and buccal mucosal grafts leave no visible scar as do genital full thickness skin grafts.
Occasionally, a one-stage urethroplasty using a full-thickness graft is not possible. In the presence of gross infection or gross scarring or a poor blood supply (e.g. after irradiation) a flap is necessary because it is the only way of securing the survival of the patch. Sometimes even this is not possible, as when a large segment of the urethra has been lost by disease, or when a urethral stent has failed and has to be removed with its surrounding segment of urethra. In these instances the only alternatives are a one-stage tube graft or a two-stage repair. One-stage tube grafts have a disappointingly high failure rate however they are constructed  and so a two-stage reconstruction with an interval urethrostomy is the safest way of securing a long-term cure.
There are four main types of penile urethral strictures. i.e. infective, traumatic (usually after instrumentation), balanitis xerotica obliterans (BXO) and failed hypospadias surgery. BXO is the genital form of lichen sclerosus and in dermatological and gynaecological circles lichen sclero-sus is the preferred terminology . The aetiology is poorly understood but the disease affects the genitalia more often than it affects extragenital skin, and particularly so in women. The disease normally starts on the prepuce and the glans. In many instances circumcision for phimosis (of which BXO is the commonest cause) is curative, presumably by exposing the glans to the air. However, in some patients the disease persists as itchy white plaques on the glans which may constrict the meatus and cause overt obstructive stenosis . In more severe cases the normal anatomical features of the corona and glans are obliterated and the meatus retracts, particularly after a meatotomy; in still worse cases oedema of the penile skin compounds the problem. Extension of the disease down the urethra is more common than reports suggest , and may produce a palpable and uncomfortable anterior urethra. For reasons that are not clear the stricture, if there is an anterior urethral stricture, normally stops abruptly at the level of the bulbar urethra.
Failed hypospadias surgery may be anything from meatal stenosis to a full-length anterior urethral stricture with urethrocutaneous fistulae, depending on the scale of the original surgery.
Simple urethral strictures (those not caused by BXO or failed hypospadias surgery) are easily treated by a patch urethroplasty using a skin flap, such as described by Jordan  for the meatus and by Orandi  for the rest of the penile urethra. These simple techniques give reliable results as long as the stricturotomy is adequate. Strictures caused by BXO and failed hypospadias surgery generally require excision of the diseased segment of the urethra and a circumferential repair , although this view is controversial . As mentioned earlier, tube grafts do not fare well and therefore a circumferential repair demands a two-stage reconstruction . The problem here is that in BXO genital skin cannot be used for this repair, because the disease is of the genital skin and will recur in the reconstructed segment ; therefore extragenital material must be used. Indeed, as it is a disease of skin, extragenital mucosa is better and therefore buccal mucosal free grafts are the material of choice for a two-stage reconstruction . In the first stage a patch is put in place and in the second this is rolled into a tube. The problem differs in revision surgery for hypospadias; here genital skin can be used but it is often in short supply because of the scarring from previous surgery. Thus, if genital skin is unavailable, extragenital skin must be used, although it is not very satisfactory for urethroplasty. However, the exception is skin from above the jaw-line; such skin takes well, has little subcutaneous fat, is thin and its characteristics are well preserved after transplantation. Unfortunately, only excision of the postauricular skin is cosmetically acceptable and this restricts the amount of material available, although this is usually adequate for reconstructing the penile urethra . Again, the postauricular Wolfe graft is transferred as a patch in the first stage and rolled into the tube in the second stage.
Whenever the glandular urethra is reconstructed care must be taken to ensure that the final result produces a glandular urethra and meatus that is at least as wide as the rest of the urethra. If at the time of the second-stage closure the meatus is significantly narrower than the rest of the urethra, then this will simply recreate the stricture and all of the proximal complications associated with it. Thus the glans must be widely split open and grafted with enough material to ensure that this does not happen. Unfortunately, even a deeply incised glans cleft tends to heal by becoming shallower in the early postoperative period, so great care must be taken with this. Indeed, the interval care of any so-called two-stage urethroplasty often means that the procedure becomes a three- or four-stage urethroplasty, but such care is always important to ensure a satisfactory end result.
The end-results of anastomotic repair are currently so good that it is difficult to imagine at present how any alternative could be better. However, whereas anastomotic repair carries a > 90% success rate which is sustained, the results of substitution urethroplasty are considerably worse. Here the average success rate is initially 85% (at 1–3 years) , deteriorating at 3–5% per year, so that by 10–15 years about half the patients have developed recurrent strictures either in relation to their previous urethroplasty or elsewhere in the urethra . This is perhaps acceptable given that the original problem in many patients was an impassable stricture, but nonetheless leaves considerable room for improvement. Technical improvements seem less likely to be successful than the development of urethral autografts derived from cultured natural tissue. These are becoming a realistic proposition in the laboratory and are likely to be the way that substitution urethroplasty will develop in the future. There is no sign that urethral stricture disease is becoming less common; indeed, with more road traffic accidents, increasing urethral instrumentation for diagnostic and therapeutic purposes, and no overall decrease in the incidence of urethral inflammatory disease, the incidence of strictures seems set to rise, if anything, in the foreseeable future.
Our thanks go to Philip Wilson who drew the illustrations and to Ellen Olds who prepared the typscript.