The use of periurethral injectables in the treatment of genuine stress incontinence

Up to 45% of all women admit to some degree of incontinence1 with genuine stress incontinence accounting for about 50% of all cases1,2. Many women do not seek medical help, due to the fact that incontinence is not always perceived as abnormal3 and because women have a low expectation of benefit from treatment4. Women may be too embarrassed to consult their general practitioner and may fear that surgery is needed to cure the problem5. However, physiotherapy has been shown to provide effective relief for some women who are motivated6. Oestrogen therapy may be useful in postmenopausal women7 especially if progestogens are not needed8. Women who do not respond to physiotherapy may have to endure their symptoms or undergo major surgery. Many elderly women are not fit for surgery under general anaesthesia, although a number of procedures can now be performed with regional anaesthetics9. Women who are unfit or unwilling to undergo major surgery have previously been condemned to a lifetime of incontinence with its economic costs10 and psychological morbidity11. Genuine stress incontinence is the involuntary loss of urine when the intravesical pressure exceeds the maximum urethral pressure in the absence of detrusor activity12. Since clinical diagnosis is accurate in only 65% of cases urodynamic investigations are essential13. In women there are two main types of sphincter abnormality14. urethral hypermobility and intrinsic sphincter deficiency. However, both conditions may exist in the same woman, especially when subjective loss of urine is high15. In urethral hypermobility raised intra-abdominal pressure causes descent of the poorly supported urethra to a position below the pelvic floor and outside of the abdominal cavity. There is unequal transmission of pressure to the bladder neck and proximal urethra resulting in a higher pressure than in the urethra, with subsequent involuntary urine loss. Urodynamic investigations often show a high abdominal leak point pressure16. Operations which reposition the bladder neck will have a good success rate17–19. With intrinsic sphincter deficiency there is poor urethral closure which may be due to scarring from previous surgery (such as anterior colporrhaphy), sympathetic neural injury or myelodysplasia20,21. The abdominal leak point pressure or maximal urethral closure pressure may be low16. Periurethral injectables are able to coapt the urethral mucosa and resist raised abdominal pressure. Other suitable treatments include the insertion of an artificial urinary sphincter22,23, a suburethral sling24,25 or even urinary diversion26. Operations to reposition the bladder neck may have a higher failure rate27–28. The ideal injectable would be inert, cheap, easily administered, free from complications and provide long lasting continence with a single injection. Procedures that are used in stress incontinence may be judged against operations, such as colposuspension, which offers cure rates in excess of 80% at five years19. This could be an appropriate comparison for primary procedures in fit healthy women, but not in women who are frail or who have had previous surgery. In 1938 Murless29 described the injection of a sclerosing solution into the anterior vaginal wall in 20 women. The resultant scarring caused compression of the incompetent urethra and 12 women (60%) were cured. Despite this success the technique did not become established. Since then a variety of substances have been used with varying degrees of success. In this review of periurethral injectables I conducted a computerised MEDLINE search using the medical subject heading ‘urinary incontinence, stress’. A secondary hand search was performed using the references produced by the initial MEDLINE search. This has the advantage of being readily available, bio-compatible with tissues and inexpensive. Fat is obtained by suction from the anterior abdominal wall and then injected periurethrally30. Studies have been small with short term follow up30–33. Although women with genuine stress incontinence tend to be overweight34, gynaecologists are generally unfamiliar with the methods used to obtain fat. However, the main problem with autologous fat is that it is quickly phagocytosed. Experience of its injection in other sites of the body shows that little autologous fat is still present after one year35. As with other injectables, the initial results were very good33 but after only 12 months, the failure rate reached 77%31. Despite this, some clinical studies of the use of autologous fat have been more promising, but success is not always judged by continence: a reduction in the number of pads used or the number of incontinent episodes may be judged as a successful outcome32. Judging success by different subjective outcomes makes comparison of different injectables difficult but in general autologous fat has not given good, sustained objective cure rates. Although autologous fat should be compatible with the tissues, complications and reactions (such as the formation of urethral pseudolipomata) can still occur36. Because of its low rate of cure in thelong term, autologous fat is not widely used. Longer term studies are available on the use of teflon (polytetrafluoroethylene). The longest follow up is described by Lopez et al.37 who reported a series of 128 women treated between 1964 and 1991, with a mean follow up of 3 1 months and a cure rate of 55% increasing to 61 % as experience improved. In a mixed group of women with moderate or severe incontinence who were a poor anaesthetic risk they injected a mean of 18.9 cc with a mean of 1.5 sittings. They estimated the probability of continence at one, five and ten years as 0.82,0.61 and 0.45, respectively. However, Harrison et al. reported only four women out of 36 (11 %) cured after five years38. Other studies have had shorter follow up (< 18 months) or fewer than 30 women39–48, in which cure rates vary widely from 7%39 to 70%40. In some of the smaller studies many of the surgeons were inexperienced37, and there is evidence that training will give better results45,46. These smaller studies may underestimate the usefulness of the technique; even then cure rates of up to 70% were reported40. Some of the more modest success rates should be considered useful in women who are old or infirm or who have had previous surgery. However, the use of teflon has been limited due to concern over particle migration and local tissue response. The term ‘teflonoma’ has been used to describe the granulomatous reaction that is sometimes encountered49. Enlarging granulomas have occurred in the vagina50 and have resulted in urinary obstruction nine years after injection51. The tendency for teflon to migrate from the injection site to lymph nodes52, lung and brain53,54 is worrying. Despite these few reports most large case series have not observed significant complications. Politano55 did not find any significant complications in more than 700 periurethral injections in men with incontinence after prostatectomy. The risk of malignancy induced by teflon is probably extremely A diagnostic and therapeutic technology assessment concluded that periurethral injection of teflon was reasonably effective and technically easy to perform but stated that the safety of the product remained uncertain57. The larger studies of teflon with long term follow up show the potential of periurethral injectables. However, the disadvantages associated with teflon have led to the development of alternative materials that aim to combine high success rates with minimal local and distant tissue reaction. There is some evidence that women with genuine stress incontinence are collagen-difficient and therefore in theory, the injection of collagen may be beneficial58,59. Bovine collagen is combined with glutaraldehyde to make it more resistant to denaturation and degradation by proteolytic enzymes (GAX collagen). GAX collagen is marketed as Contigen (Bard, Crawley, UK). Human studies in vivo have demonstrated that this preparation encourages the host to replace the collagen implant with endogenous collagen60 which may contribute to its sustained clinical effect. Some authors report the maintenance of continence if it is initially obtained61,62, but others report a loss of effect over time63,64. Repeated injections are often necessary to achieve sustained continence. The optimum timing and frequency of the injections is not known, but two or three injections are often needed65. GAX collagen has a low viscosity and is easy to inject through a spinal needle. Four weeks before the first periurethral injection 0.1 mL of GAX collagen is injected intradermally into the volar surface of the forearm to ensure that the woman is not allergic to the material. There is a single report of a woman becoming pregnant after treatment with GAX-collagen66. She was cured by this therapy before her pregnancy and interestingly remained cured after delivery. This experience may justify the use of injectables before the completion of childbearing. The use of GAX collagen was first reported in 198967 so there is a shorter experience with this material than with teflon. Since collagen implants have been placed intradermally to correct body contour defects with minimal local or systemic reactions, it appeared to be an ideal substance to correct defects of the urethral sphincter60. An initial relatively low success rate was expected due to the selection of women unsuitable for other forms of treatment or who had failed other treatment68. Cure rates in published case series vary from 7%63 to 83%69, with most studies reporting cures in the region of 40–60%62,65,70–77 (Table 1). Although it would be possible to combine this information in a metanalysis, this is inappropriate as the series contain different types of participant. As was the case with teflon, some of the studies had short follow up and small numbers. Also, the cure rates reported were a mixture of subjective and objective definitions of cure, with subjective cure rates being higher than objective measures78. In some studies the association between objective and subjective cures was poor which makes the comparison of different studies even more difficult70. The requirement for complete cure may be inappropriate since many women will be satisfied with improvement short of a complete cure75. When women have previously been incontinent without effort, change of position or activity, the use of one pad daily should be regarded as a significant success73 even though it might not be considered a cure. Success rates (often quoted as cure and improvement) vary from 68%65 to 90%71. The report by Monga et al.65 is notable due to the fact that it has a larger sample than other studies (60 women), a follow up period of two years and results expressed in terms of objective and subjective cures65. They reported subjective success rates of 86% at three months, 77% at 12 months and 68% at 24 months. Objective cure rates were 6 1% at three months, 54% at 12 months and 48% at 24 months. The mean duration of incontinence was 11.6 years, with the women having undergone an average of 1.6 previous incontinence procedures. The authors achieved these results with a mean of 1.6 injections. The majority of the women (n= 54) were followed up for one year but less than half for the full two year period. The longest follow up comes from Richardson et al.79 who reported 42 women with a mean follow up of 46 months (range 10–66). The average volume of collagen injected was 28.3 cc, and the average number of injections was two; 40% of the women were cured and a further 43% improved. In view of the recent outbreak of bovine spongiform encephalopathy women may be concerned about the use of bovine products. However, women should have no fear because Contigen is manufactured in the United States (Bard, Covington, Georgia, USA), where there have been no reported cases. It is produced from hides of steers living in a closed herd and fed on grain products only. This is the most recently developed injectable material and consists of vulcanised silicone rubber particles suspended in a non silicone carrier gel80. Silicone is marketed as Macroplastique (Uroplasty, Reading, UK). This has the theoretical advantage of larger particle size making migration and displacement less likely81. The silicone particles are designed to act as a bulking agent with the eventual removal of the carrier gel, encapsulation in fibrin and its replacement with collagen fibres14. Silicone is a viscous material and therefore needs a special ratchet gun for injection and a lubricant to prime the equipment. The gun is unwieldy and requires practice. Information concerning the usefulness of this material has been published as unreviewed abstracts and not as peer review main articles. Adile et al. reported a 60% cure rate at one82 and two years follow up83. Harriss et al.84 reported a series of 40 women treated with a single injection (3–7 mL), with 40% dry at three months and three years, suggesting the possibility of long term cure in some women with a single injection. However, Buckley et al. reported a 70% cure rate after 17 months85 which fell to 31% after 39 month86. As with collagen, silicone injections can be repeated when necessary. There has been only one, nonrandomised, trial to compare collagen and silicone87. The characteristics of the different injectables are summarised in Table 2. Randomised comparisons of these treatments are required. Other materials are under development88. An increased failure rate occurs with detrusor instability in some studies44,45,71 but not others65. Overall, women with coexistent detrusor instability are likely to do less well than women with genuine stress incontinence alone. Although most studies have focused on women with intrinsic sphincter deficiency, some have included women without previous surgery. A review by Jarvis78 suggested that previous surgery made no difference to the results of treatment with periurethral injection. Women with a mobile bladder neck seem to have similar success rates to those with intrinsic sphincter deficiency65,71,77, but the numbers so far reported are small, making firm conclusions difficult. In other studies it is impossible to determine the cause of the genuine stress incontinence63. Contraindications to periurethral injection include untreated urinary infection, and hypersensitivity to bovine collagen61. Caution is suggested in the use of collagen in women with connective tissue diseases. The mechanism of continence after periurethral injections is unclear but a better urethral seal achieved by submucosal bulking is probably important62,65. Some authors suspect the mode of action is obstructive75, with a fall in urine flow rate and an increase in maximum detrusor pressure72. However, others show no alteration in flow rates62 or report the only significant change to be an increase in functional urethral length65,89. Improvement in continence is associated with an increase in the abdominal leak point pressure69,71. Each surgeon may want to experiment with different methods of injection and use the one with which he feels most comfortable. Injectables can be administered either periurethrally or transurethrally; neither method seems to give better results. Transurethral injection is performed through a needle placed directly through a cystoscope, while periurethral injection is performed by inserting a spinal needle percutaneously adjacent to the urethra and under cystoscopic control. These techniques are described by Appell64. The method of injection is probably less important than precise placement of the material and its retention at this site. The accuracy of the injection may be improved by first injecting a small volume (0.25 to 1.0 cc) of a mixture of methylene blue and lignocaine to demarcate the injection site90. The depth of injection is very important; the material must deform the urethral mucosa so that it closes the urethral lumen. Too deep an injection is a waste of material and is not effective91. There is also a suggestion that injecting at the 12 o'clock position is of no benefit89. Transvaginal ultrasound can also be used to see if the placement of the material is correct92. However, follow up magnetic resonance imaging scans 12 months after injection show that the position and volume of collagen are not predictive of clinical outcome93. Intra-operative assessment of the degree of closure of the urethral lumen may give some idea of the likelihood of success. If there is good closure at the time of surgery there is probably a better chance of success65. Either local or general anaesthesia can be used. General anaesthesia has certain attractions while a surgeon is learning the technique, but once it is mastered local anaesthesia is generally preferable. One of the advantages of using local anaesthesia is that the woman can be asked to cough at the end of the procedure to see if continence has been attained. If leakage is detected more material can be injected. Local anaesthesia also enables the procedure to be performed in the outpatient department. The procedure takes twenty minutes if the surgeon is fully trained65,73. Some women will not be able to tolerate the operation under local anaesthesia73. Antibiotics may be given around the time of the procedure. Complications are uncommon and short-lasting. The manufacturers of GAX collagen report urinary retention in 9% of women (Contigen data sheet). This is usually temporary but may require continuous catheterisation for 24 hours65 or intermittent catheterisation for 48 or 72 hours72. Slight bleeding and discomfort at the injection site are common and transient frank haematuria may occasionally Urinary infection may occur in about 20% of women (Contigen data sheet), justifying the use of prophylactic antibiotics. Dysuria is very common and may be universal70. Most studies report that post-operative detrusor instability is rare65 but others report a rate of 39%77. No woman has required long term catheterisation. With the newer materials there has been no instance of significant migration of the material or systemic reaction to it. Approximately 3% of women will be unsuitable for collagen therapy due to hyper sensitivity diagnosed on pre-operative skin testing64. This has been poorly researched. Neither silicone nor collagen is cheap, the cost being £250 to £325 per 2.5 mL. Collagen is supplied in a pack which contains needles and pre-loaded syringes, while silicone requires a needle and ratchet gun supplied separately. Due to the viscous nature of silicone it is also necessary to purchase a lubricant to prime the equipment. However, these expenses may be fairly modest compared with the cost of open surgery and inpatient care. Usually only 5 mL of material are needed but volumes up to 38.7 mL have been used77. Inexperienced surgeons use larger volumes; the volume can fall by two-thirds after 40 women have been treated65. Periurethral injections may be useful in the treatment of genuine stress incontinence. They have been shown to be effective in the treatment of intrinsic sphincter deficiency, a condition which is difficult to treat. Periurethral injections may be especially useful in frail elderly women with genuine stress incontinence, and even if cure is not always possible substantial improvement in continence will improve the quality of life. Women can resume their normal activities quickly and side effects are much less than those of major surgery. Repeat injections may be necessary, but if they fail periurethral injection does not make further surgery more difficult. The major uncertainty with periurethral injection is its effectiveness, for as this review has shown, experience has been confined to small uncontrolled case series with varying results. Firstly, periurethral injections should be compared with a standard operation for genuine stress incontinence. The standard treatment is Burch colposuspension which is a major operation, and the experimental treatment is periurethral injection, which is an outpatient procedure and which does not complicate a future Burch colposuspension should it fail. The cure rate of Burch colposuspension is approximately 85%; if the cure rate is < 60% with periurethral injection we can conclude that periurethral injection is so much worse than Burch colposuspension that we should recommend the Burch operation as the primary treatment. If the cure rate with periurethral injection is > 60% we recommend that periurethral injection should be tried in the first instance; if it fails, Burch colposuspension should be carried out. In order to show this difference between treatments the number of women required is 100 (α= 0.05; 1 - β= 0.8). If both GAX collagen and silicone are compared with Burch colposuspension the number of participants should be 173. We need to perform this trial.