- Email: [email protected]
The Long-Term Results of Artificial Sphincters in Children
0022-534 7/91/1462-0396$03.00/0
Vol. 146, 396-399, August 1991
THE JOURNAL OF UROLOGY Copyright© 1991 by AMERICAN UROLOGICAL ASSOCIATION, INC.
Printed in U.S.A.
THE LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN PETER J. BOSCO, STUART B. BAUER,* ARNOLD H. COLODNY, JAMES MANDELL AND ALAN B. RETIK From the Division of Urowgy, The Children's Hospital and Harvard Medical School, Boston, Massachusetts
ABSTRACT
To determine the long-term efficacy of the artificial urinary sphincter, 36 consecutive children in whom a prosthesis was implanted between August 1978 and July 1985 were followed for a minimum of 5 years and re-evaluated within the last 3 months. Of the 36 sphincters 27 (75%) are currently in place; 2 of these 27 sphincters are dysfunctional due to fluid leakage. Of the 25 functioning sphincters 14 have required no further prosthetic operation, 10 needed 1 revision and 1 patient had 5 revisions. Fluid leakage, occurring in 12 patients, was the most common complication requiring reoperation. Twenty patients are completely continent (11 of whom without any medication) and 5 are occasionally damp. The over-all success rate at 2 and 5 years was 84 and 62%, respectively. The mean survival time of the prosthesis was 7.2 years and the mean standard interval was greater than 10.5 years. Patients with higher balloon pressures (81 to 100 cm. water) had more complications, and a decreased rate of success than those with lower pressure balloons (51 to 80 cm. water, p = 0.02). There was no difference in success rates between boys and girls (p = 0.98), although girls who had a prior bladder neck operation tended to have a higher rate of cuff erosion. Sphincters placed after 1980 (model AS800) had a lower reoperation rate than those placed earlier. In selected instances, the artificial urinary sphincter appears to be a viable long-term alternative for manage ment of children with intractable incontinence who have either failed or are unsuitable for other methods of treatment. KEY WORDS:
urinary incontinence, artificial organs, prostheses, urination disorders
Urinary incontinence can have profound psychosocial con sequences for the developing child and his family. During the last 10 years several operations have been devised in an attempt to correct incontinence in children so that they can grow normally and be accepted by their peers. Since its inception in 1973, the artificial sphincter has been used as a method to treat children with incontinence who are refractory to pharmacolog ical therapy or unsuitable for other types of operation. Several prior reports have demonstrated excellent short-term success rates with the artificial urinary sphincter. Less is known, how ever, about the longevity and the long-term efficacy of this device. We evaluate children in whom an artificial urinary sphincter was implanted between 1978 and 1985 who were followed for greater than 5 years. PATIENTS AND METHODS
There were 36 artificial urinary sphincterst implanted from August 1978 through July 1985 into 22 boys and 14 girls 4 to 30 years old (median age 12 years). The sphincters implanted included the AS742 (6 patients), AS792 (17) and AS800 (13) models. The sphincter cuff was initially placed around the bladder neck in all individuals. Of the patients 27 (75%) had myelodysplasia as the etiology of the neurogenic bladder dys function, 2 had sacral agenesis, 2 had lipomeningocele, 1 had cerebral palsy, 3 had exstrophy-epispadias and 1 had iatrogenic dysfunction. An additional operation was done in 24 patients (67%): 17 before and 7 after placement of the artificial sphinc ter. Of the former 17 patients 8 underwent a pre-sphincter operation in an attempt to increase bladder neck resistance, such as bladder neck suspension or a Young-Dees type of Accepted for publication January 9, 1991. Read at annual meeting of American Urological Association, New Orleans, Louisiana, May 13-17, 1990. * Current address: Division of Urology, The Children's Hospital, 300 Longwood Ave., Boston, Massachusetts 02115. t American Medical Systems, Minnetonka, Minnesota.
reconstruction performed alone (5), or with augmentation cys toplasty with (1) or without (2) ureteral reimplantation. In 5 individuals the outlet resistance was decreased by sphincterot omy or Y-V plasty of the bladder neck in preparation for the sphincter. Three patients underwent ureteral reimplantation and 1 underwent augmentation cystoplasty. Seven patients required a post-sphincter augmentation with (4) or without (3) ureteral reimplantation to decrease detrusor hypertonicity or eliminate intractable uninhibited contractions that occurred in the postoperative period. All children were followed with a urine culture every 3 months, a yearly post-void residual urine measurement and plain abdominal film, and an excretory urogram initially, and more recently renal and pelvic ultrasound to evaluate the status of the upper urinary tract, and the size and shape of the balloon reservoir. If a device malfunction was noted or wetness oc curred, each patient underwent fluoroscopy of the sphincter and urethral pressure profilometry as well as cystometrography. When urinary infection recurred a post-void residual urine volume was obtained and then a radionuclide cystogram or voiding cystogram was performed. Cystoscopy was done if an erosion was suspected by history, physical examination or radiological studies. RESULTS
Of the 36 artificial sphincters 27 (75%) are currently in place and 25 of them are working well. Two sphincters are currently dysfunctional due to fluid leakage from within the hydraulic system and these individuals are awaiting revision. Of the 25 patients with a functioning sphincter 14 have required no further prosthetic surgery (39% of the entire series of 36), 10 needed 1 revision and 1 had 5 revisions. Of the 9 patients whose sphincters have been removed 3 have undergone continent urinary diversion with an ileocecal reservoir, 2 have an ileocecal conduit and 4 are using clean intermittent catheterization with
396
LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN
anticholinergic and/or a-sympathomi:metic agents to maintain continence. Of the 25 patients with a functional sphincter in place 20 are completely continent (11 without any medication) and 5 are occasionally damp. Fifteen patients void spontaneously and 10 are using clean intermittent catheterization to facilitate com plete bladder emptying. Of the 36 patients in this series 22 had a complication involving 1 of the sphincter components. Twelve individuals had loss of hydraulic fluid from within the system. In 9 patients the device was removed because of erosion of the cuff (6), leakage of hydraulic fluid (1) or a urethral injury sustained at placement and that never healed (2, see table). Of the 9 patients 5 were girls, 3 of whom had a prior bladder neck operation (Y-V plasty in 2 and Young-Dees reconstruction in 1). In 2 of the 9 patients whose cuff was removed (22%) the sphincter was replaced with the cuff repositioned around the bulbous urethra. Over-all, 14 patients (39% of the entire series) have not had a prosthetic complication. Nine patients had uncomplicated urinary tract infections during followup (3 were symptomatic with 1 of them requiring hospitalization and intravenous antibiotics). The over-all success rate at 2 and 5 years was 84 and 62%, respectively. The mean survival of the prosthesis was 7.2 years, with a mean standard interval of greater than 10.5 years (fig. 1). When balloon pressures were compared to over-all success rate, the patients with a higher pressure balloon (81 to 100 cm. water) had more complications and a decreased rate of success than those with a lower balloon pressure (51 to 80 cm. water, Complications of the prosthesis in 22 patients
Type
Removal
Revised
Dysfunctional
Total
Fluid leak Cuff erosion Urethral injury Infection Pump migration Valve failure Tube kinking Totals
1 6 2
9 2
2
12 8 2 1 1 1 1
1 1 1 1
9
2
15
Succ� r-=�-�---=-----�-
Balloon Pressure 51-SOcm H20 --81-lOOcm H 'JP
20
2
AS 742
25%
78
80 MST= >10.5 Yrs
60
AS 792
Sil%
95% Confidence Interval
14 16
12
Ill Removed
75%
26
10
FIG. 2. Lower balloon pressures (51 to 80 cm. water) have higher rate of success than higher balloon pressures (81 to 100 cm. water, p = 0.02).
(2)
100
8
Years
Rcopcration Rate
Success Rate
Percent Success
Balloon Pressures
PeKent
397
79
(6)*
80
81
(6)
(4)
.. 82 (5)
Revised/Replaced
AS 800
I I
83 (2)
84
85
(7)
(4)
Year Implanted (number)
FIG. 3. Although removal rate has been substantially less since 1982 when model AS800 was first used, revision rate has changed only slightly since then. *, denotes 2 patients in 1979 in whom model AS792 was implanted. t, 1 patient had 5 revisions since initial implantation. **, 2 children had loss of hydraulic fluid but have not undergone revision.
p = 0.02, fig. 2). There was no difference in the success rate between boys and girls (p = 0.98). Sphincters placed after 1980, which included the model AS792 and the newer model AS800, had a considerably lower rate of removal but they still had a definite rate of revision or replacement when compared with those placed before 1980 (fig. 3).
40
DISCUSSION
20
2
4
6
8
Years
10
12
14
16
FIG. 1. Over-all success rate at 2 and 5 years is 84 and 62%, respectively (S). Mean survival is 7.2 years and mean standard interval (----) is greater than 10.5 years. Shaded area represents 2 standard deviations of mean.
The use of a prosthetic device to manage urinary inconti nence in children has provided a reasonable alternative to medical management and/or a protracted operation. The arti ficial sphincter has evolved from a passive compressive device against the bulbous urethra as described by Berry in 19611 and Kaufman in 19732 to a more dynamic, physiological prosthesis as described by Scott et al in 1973.3 Scott et al have continued to improve and adapt the artificial sphincter with the use of sophisticated engineering. The AS800 series introduced in 1982 has provided a mechanically sound device with a relatively simple surgical installation.
398
BOSCO AND ASSOCIATES
As with many surgical endeavors, the ultimate determination of success is dictated by the ease, longevity and durability of the procedure. Several surgeons have reported excellent results with the use of the artificial sphincter from 1 to 51 months after implantation.4-9 We review a group of children in whom the artificial sphincter was implanted more than 5 years ago. Of our patients 75% have an intact functional sphincter with a minimum followup of 5 years. The 2 and 5-year success rates of 84 and 62%, respectively, pertain to the initial placement of a sphincter. Any revision and/or replacement is considered a device failure and has the effect of decreasing the over-all rate of success. The most common complication requiring revision, replace ment or removal was hydraulic fluid loss from the prosthesis, which occurred in 12 individuals. Light et al reported a 95.4% success rate in 87 children with a 5 to 45-month followup.4 A majority of these individuals were followed for only 2 to 3 years. Fluid leakage occurred most often from the cuff and/or pressure regulating balloon, and was the most common complication requiring reoperation (12 of 36 or 33%). The next most common complication was erosion of the cuff through the bladder neck, which occurred in 8 patients. These people frequently presented with infection, incontinence, he maturia or pain and swelling around the pump. This compli cation resulted in removal of the device in 6 patients, while in 2 a new cuff was repositioned around the bulbous urethra shortly thereafter. In 406 adults, with 236 sphincters placed around the bladder neck and 170 around the urethra described by Scott, cuff erosion was the most common cause of sphincter removal.5 Erosion was associated with infection, pressure ne crosis and surgical trauma. Light and Scott reported no erosions in their series and attributed this to the use of lower pressure regulating balloons and the avoidance of iatrogenic injury of the bladder neck or urethra during the implant procedure.6 Scott reported that higher pressure regulating balloons were likely to be associated with a higher incidence of erosion.5 Similarly, we found that children with a higher pressure regu lating balloon (81 to 100 cm. water) had a greater number of complications and a smaller rate of success than those with a lower pressure balloon (51 to 80 cm. water). This finding was statistically significant (p = 0.02). A higher pressure regulating balloon was used more frequently during the early years of our experience with the artificial sphincter. It has been suggested that girls have a higher failure rate than boys after sphincter implantation. Sidi et al reported on 15 patients, 10 boys and 5 girls, with 51 months of followup.8 Continence was achieved in 9 of the 10 boys compared to 1 of 5 girls. We found no significant difference in the success rate between boys and girls in our study except that girls who had a prior bladder neck operation were prone to suffer erosion of the cuff. We believe there is no contraindication or reservation for the use of the artificial sphincter in girls to control urinary incontinence. However, we do not advocate insertion of an artificial sphincter in any girl who has had a prior bladder neck reconstruction because there is a substantial risk of cuff erosion in these patients. During the last 10 years we have not done a single Y-V plasty in any girl before sphincter implantation as originally expounded by Light and Scott6 for this reason. We would prefer to have them perform clean intermittent cathe terization to empty or better yet, to be able to empty with Valsalva's maneuver. Upper urinary tract deterioration and changes in bladder contractility have been a concern to urologists who implant the artificial sphincter into children with myelodysplasia. Elevated filling pressure and detrusor hyperreflexia have been docu mented in previous studies after implantation of the de vice.10-13 Roth et al reported that 7 of 47 children had an increase in bladder tonicity and 4 of them had hydronephrosis.10 In 1986 Bauer et al reported on 6 children who had increased filling pressure and/or uninhibited contractions of the bladder
after placement of the artificial sphincter.11 Three children presented with urinary incontinence while 3 had hydrone phrosis. All 6 affected individuals had myelodysplasia and each incurred the changes in bladder function postpubertally. Three children had tethering of the spinal cord during this followup period, 2 of whom were managed successfully with lumbosacral laminectomy and detethering. The remaining 3 children had an augmentation cystoplasty to relieve the detrusor hyperton icity. Seven individuals in the series, including those 3 reported on previously, underwent bladder augmentation after place ment of the artificial sphincter, 4 of whom had ureteral reim plantation as well because they had vesicoureteral reflux in addition to poor detrusor compliance. Low bladder filling pres sures were achieved in all patients. Factors important for these changes in detrusor dynamics were the position of the cuff, frequency of urinary infections and adherence to bladder emptying schedules. It is unclear whether placement of the sphincter or only the presence of myelodysplasia is the most important factor in the development of changes in detrusor function. Light and Pietro believed that these changes may be secondary to vesicourethral reflexes that are stimulated after placement of a sphincter cuff around the bladder neck.13 Sometimes, however, detrusor hypertonia may be masked by such low outlet resistance that any increase in intravesical pressure leads to an escape of fluid and an inability to measure true detrusor compliance.14 Whatever the mecha nism, it is mandatory that preoperative assessment include cystometrography performed with the urethra occluded to sim ulate the effect of an activated sphincter cuff so that a true compliance factor of the bladder can be determined. In addition, it is essential that all individuals who have a sphincter in place be monitored routinely with upper urinary tract imaging on a long-term basis. A measurement of post-void residual and uro dynamic assessment are needed if a change is noted. As with any surgical endeavor involving a mechanical device, there is a learning curve for the surgeon as well as the manu facturer. Other studies have noted a difference in success rates among the various sphincter models that have evolved during the first 8 years since the sphincter was perfected.1· 5• 8 We found that a sphincter that was placed before 1980 (models AS721 and AS742) had a higher reoperation rate than a sphincter placed after 1980 (models AS792 and AS800). The ability to activate and deactivate the model AS800 sphincter without an additional operation, the use of a lower pressure regulating balloon due to improved efficiency of the cuff, strict attention to and familiarity with the surgical technique, operative sterility and the use of perioperative antibiotics are factors contributing to the more recent higher rate of success. Decreasing the incidence of fluid loss from the hydraulic system of the sphinc ter would appear to be an issue that deserves consideration by the manufacturer and surgeon in the future, since this was the most common complication in our current study. This would certainly impact on the durability of the prosthesis and have a positive effect on the long-term success rate. Currently, our contention is that the ideal candidate for the artificial urinary sphincter is someone without reflux who had neurogenic bladder dysfunction, characterized by a large capac ity, low pressure bladder without uninhibited contractions and low outlet resistance that allows for complete emptying by straining or Crede's maneuver. Although individuals with other types of lower urinary tract dysfunction can be made to fit into this pattern by using augmentation cystoplasty and/or a blad der neck operation, alternatives to the artificial urinary sphinc ter may be more appropriate if one considers our results with these other operations when performed in conjunction with the sphincter. CONCLUSION
The artificial sphincter remains a good modality for provid ing continence in selected instances and is equally effective for
LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN
boys and girls so long as the latter have not had a prior bladder neck or urethral operation. Approximately 70% of the devices will be intact and functional at 5 years and nearly two-thirds of these will not need any revision. High balloon pressures, greater than 80 cm. water, are associated with more complica tions and a higher failure rate than lower pressure regulating balloons, and, thus, should be avoided. Although a third of the individuals with an intact sphincter will need revision at some time within the first 5 years after implantation, the need for revisions has decreased in recent years as the design for the sphincter has improved and the surgical technique is refined. REFERENCES
1. Berry, J. L.: A new procedure for correction of urinary inconti nence: preliminary report. J. Urol., 85: 771, 1961. 2. Kaufman, J. J.: Treatment of post-prostatectomy urinary inconti nence using a silicone gel prosthesis. Brit. J. Urol., 45: 646, 1973. 3. Scott, F. B., Bradley, W. E. and Timm, G. W.: Treatment of urinary incontinence by implantable prosthetic sphincter. Urol ogy, 1: 252, 1973. 4. Light, J. K., Hawila, M. and Scott, F. B.: Treatment of urinary incontinence in children: the artificial sphincter versus other methods. J. Urol., 130: 518, 1983. 5. Scott, F. B.: The artificial urinary sphincter: experience in adults. Urol. Clin. N. Amer., 16: 105, 1989.
399
6. Light, J. K. and Scott, F. B.: Complications of the artificial urinary sphincter in pediatric patients. Urol. Clin. N. Amer., 10: 551, 1983. 7. Barrett, D. M. and Parulkar, B. G.: The artificial sphincter (AS800). Experience in children and young adults. Urol. Clin. N. Amer., 16: 119, 1989. 8. Sidi, A. A., Reinberg, Y. and Gonzalez, R.: Comparison of artificial sphincter implantation and bladder neck reconstruction in pa tients with neurologic urinary incontinence. J. Urol., 138: 1120, 1987. 9. Gonzalez, R. and Sheldon, C. A.: Artificial sphincters in children with neurogenic bladders: long-term results. J. Urol., 128: 1270, 1982. 10. Roth, D. R., Vyas, P. R., Kroovand, R. L. and Perlmutter, A. D.: Urinary tract deterioration associated with the artificial urinary sphincter. J. Urol., 135: 528, 1986. 11. Bauer, S. B., Reda, E. F., Colodny, A.H. and Retik, A. B.: Detrusor instability: a delayed complication in association with the arti ficial sphincter. J. Urol., 135: 1212, 1986. 12. McGuire, E. J., Woodside, J. R. and Borden, T. A.: Upper urinary tract deterioration in patients with myelodysplasia and detrusor hypertonia: a followup study. J. Urol., 129: 823, 1983. 13. Light, J. K. and Pietro, T.: Alteration in detrusor behavior and the effect on renal function following insertion of the artificial urinary sphincter. J. Urol., 136: 632, 1986. 14. Woodside, J. R. and McGuire, E. J.: Technique for detection of detrusor hypertonia in the presence of urethral sphincteric in competence. J. Urol., 127: 740, 1982.
Vol. 146, 396-399, August 1991
THE JOURNAL OF UROLOGY Copyright© 1991 by AMERICAN UROLOGICAL ASSOCIATION, INC.
Printed in U.S.A.
THE LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN PETER J. BOSCO, STUART B. BAUER,* ARNOLD H. COLODNY, JAMES MANDELL AND ALAN B. RETIK From the Division of Urowgy, The Children's Hospital and Harvard Medical School, Boston, Massachusetts
ABSTRACT
To determine the long-term efficacy of the artificial urinary sphincter, 36 consecutive children in whom a prosthesis was implanted between August 1978 and July 1985 were followed for a minimum of 5 years and re-evaluated within the last 3 months. Of the 36 sphincters 27 (75%) are currently in place; 2 of these 27 sphincters are dysfunctional due to fluid leakage. Of the 25 functioning sphincters 14 have required no further prosthetic operation, 10 needed 1 revision and 1 patient had 5 revisions. Fluid leakage, occurring in 12 patients, was the most common complication requiring reoperation. Twenty patients are completely continent (11 of whom without any medication) and 5 are occasionally damp. The over-all success rate at 2 and 5 years was 84 and 62%, respectively. The mean survival time of the prosthesis was 7.2 years and the mean standard interval was greater than 10.5 years. Patients with higher balloon pressures (81 to 100 cm. water) had more complications, and a decreased rate of success than those with lower pressure balloons (51 to 80 cm. water, p = 0.02). There was no difference in success rates between boys and girls (p = 0.98), although girls who had a prior bladder neck operation tended to have a higher rate of cuff erosion. Sphincters placed after 1980 (model AS800) had a lower reoperation rate than those placed earlier. In selected instances, the artificial urinary sphincter appears to be a viable long-term alternative for manage ment of children with intractable incontinence who have either failed or are unsuitable for other methods of treatment. KEY WORDS:
urinary incontinence, artificial organs, prostheses, urination disorders
Urinary incontinence can have profound psychosocial con sequences for the developing child and his family. During the last 10 years several operations have been devised in an attempt to correct incontinence in children so that they can grow normally and be accepted by their peers. Since its inception in 1973, the artificial sphincter has been used as a method to treat children with incontinence who are refractory to pharmacolog ical therapy or unsuitable for other types of operation. Several prior reports have demonstrated excellent short-term success rates with the artificial urinary sphincter. Less is known, how ever, about the longevity and the long-term efficacy of this device. We evaluate children in whom an artificial urinary sphincter was implanted between 1978 and 1985 who were followed for greater than 5 years. PATIENTS AND METHODS
There were 36 artificial urinary sphincterst implanted from August 1978 through July 1985 into 22 boys and 14 girls 4 to 30 years old (median age 12 years). The sphincters implanted included the AS742 (6 patients), AS792 (17) and AS800 (13) models. The sphincter cuff was initially placed around the bladder neck in all individuals. Of the patients 27 (75%) had myelodysplasia as the etiology of the neurogenic bladder dys function, 2 had sacral agenesis, 2 had lipomeningocele, 1 had cerebral palsy, 3 had exstrophy-epispadias and 1 had iatrogenic dysfunction. An additional operation was done in 24 patients (67%): 17 before and 7 after placement of the artificial sphinc ter. Of the former 17 patients 8 underwent a pre-sphincter operation in an attempt to increase bladder neck resistance, such as bladder neck suspension or a Young-Dees type of Accepted for publication January 9, 1991. Read at annual meeting of American Urological Association, New Orleans, Louisiana, May 13-17, 1990. * Current address: Division of Urology, The Children's Hospital, 300 Longwood Ave., Boston, Massachusetts 02115. t American Medical Systems, Minnetonka, Minnesota.
reconstruction performed alone (5), or with augmentation cys toplasty with (1) or without (2) ureteral reimplantation. In 5 individuals the outlet resistance was decreased by sphincterot omy or Y-V plasty of the bladder neck in preparation for the sphincter. Three patients underwent ureteral reimplantation and 1 underwent augmentation cystoplasty. Seven patients required a post-sphincter augmentation with (4) or without (3) ureteral reimplantation to decrease detrusor hypertonicity or eliminate intractable uninhibited contractions that occurred in the postoperative period. All children were followed with a urine culture every 3 months, a yearly post-void residual urine measurement and plain abdominal film, and an excretory urogram initially, and more recently renal and pelvic ultrasound to evaluate the status of the upper urinary tract, and the size and shape of the balloon reservoir. If a device malfunction was noted or wetness oc curred, each patient underwent fluoroscopy of the sphincter and urethral pressure profilometry as well as cystometrography. When urinary infection recurred a post-void residual urine volume was obtained and then a radionuclide cystogram or voiding cystogram was performed. Cystoscopy was done if an erosion was suspected by history, physical examination or radiological studies. RESULTS
Of the 36 artificial sphincters 27 (75%) are currently in place and 25 of them are working well. Two sphincters are currently dysfunctional due to fluid leakage from within the hydraulic system and these individuals are awaiting revision. Of the 25 patients with a functioning sphincter 14 have required no further prosthetic surgery (39% of the entire series of 36), 10 needed 1 revision and 1 had 5 revisions. Of the 9 patients whose sphincters have been removed 3 have undergone continent urinary diversion with an ileocecal reservoir, 2 have an ileocecal conduit and 4 are using clean intermittent catheterization with
396
LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN
anticholinergic and/or a-sympathomi:metic agents to maintain continence. Of the 25 patients with a functional sphincter in place 20 are completely continent (11 without any medication) and 5 are occasionally damp. Fifteen patients void spontaneously and 10 are using clean intermittent catheterization to facilitate com plete bladder emptying. Of the 36 patients in this series 22 had a complication involving 1 of the sphincter components. Twelve individuals had loss of hydraulic fluid from within the system. In 9 patients the device was removed because of erosion of the cuff (6), leakage of hydraulic fluid (1) or a urethral injury sustained at placement and that never healed (2, see table). Of the 9 patients 5 were girls, 3 of whom had a prior bladder neck operation (Y-V plasty in 2 and Young-Dees reconstruction in 1). In 2 of the 9 patients whose cuff was removed (22%) the sphincter was replaced with the cuff repositioned around the bulbous urethra. Over-all, 14 patients (39% of the entire series) have not had a prosthetic complication. Nine patients had uncomplicated urinary tract infections during followup (3 were symptomatic with 1 of them requiring hospitalization and intravenous antibiotics). The over-all success rate at 2 and 5 years was 84 and 62%, respectively. The mean survival of the prosthesis was 7.2 years, with a mean standard interval of greater than 10.5 years (fig. 1). When balloon pressures were compared to over-all success rate, the patients with a higher pressure balloon (81 to 100 cm. water) had more complications and a decreased rate of success than those with a lower balloon pressure (51 to 80 cm. water, Complications of the prosthesis in 22 patients
Type
Removal
Revised
Dysfunctional
Total
Fluid leak Cuff erosion Urethral injury Infection Pump migration Valve failure Tube kinking Totals
1 6 2
9 2
2
12 8 2 1 1 1 1
1 1 1 1
9
2
15
Succ� r-=�-�---=-----�-
Balloon Pressure 51-SOcm H20 --81-lOOcm H 'JP
20
2
AS 742
25%
78
80 MST= >10.5 Yrs
60
AS 792
Sil%
95% Confidence Interval
14 16
12
Ill Removed
75%
26
10
FIG. 2. Lower balloon pressures (51 to 80 cm. water) have higher rate of success than higher balloon pressures (81 to 100 cm. water, p = 0.02).
(2)
100
8
Years
Rcopcration Rate
Success Rate
Percent Success
Balloon Pressures
PeKent
397
79
(6)*
80
81
(6)
(4)
.. 82 (5)
Revised/Replaced
AS 800
I I
83 (2)
84
85
(7)
(4)
Year Implanted (number)
FIG. 3. Although removal rate has been substantially less since 1982 when model AS800 was first used, revision rate has changed only slightly since then. *, denotes 2 patients in 1979 in whom model AS792 was implanted. t, 1 patient had 5 revisions since initial implantation. **, 2 children had loss of hydraulic fluid but have not undergone revision.
p = 0.02, fig. 2). There was no difference in the success rate between boys and girls (p = 0.98). Sphincters placed after 1980, which included the model AS792 and the newer model AS800, had a considerably lower rate of removal but they still had a definite rate of revision or replacement when compared with those placed before 1980 (fig. 3).
40
DISCUSSION
20
2
4
6
8
Years
10
12
14
16
FIG. 1. Over-all success rate at 2 and 5 years is 84 and 62%, respectively (S). Mean survival is 7.2 years and mean standard interval (----) is greater than 10.5 years. Shaded area represents 2 standard deviations of mean.
The use of a prosthetic device to manage urinary inconti nence in children has provided a reasonable alternative to medical management and/or a protracted operation. The arti ficial sphincter has evolved from a passive compressive device against the bulbous urethra as described by Berry in 19611 and Kaufman in 19732 to a more dynamic, physiological prosthesis as described by Scott et al in 1973.3 Scott et al have continued to improve and adapt the artificial sphincter with the use of sophisticated engineering. The AS800 series introduced in 1982 has provided a mechanically sound device with a relatively simple surgical installation.
398
BOSCO AND ASSOCIATES
As with many surgical endeavors, the ultimate determination of success is dictated by the ease, longevity and durability of the procedure. Several surgeons have reported excellent results with the use of the artificial sphincter from 1 to 51 months after implantation.4-9 We review a group of children in whom the artificial sphincter was implanted more than 5 years ago. Of our patients 75% have an intact functional sphincter with a minimum followup of 5 years. The 2 and 5-year success rates of 84 and 62%, respectively, pertain to the initial placement of a sphincter. Any revision and/or replacement is considered a device failure and has the effect of decreasing the over-all rate of success. The most common complication requiring revision, replace ment or removal was hydraulic fluid loss from the prosthesis, which occurred in 12 individuals. Light et al reported a 95.4% success rate in 87 children with a 5 to 45-month followup.4 A majority of these individuals were followed for only 2 to 3 years. Fluid leakage occurred most often from the cuff and/or pressure regulating balloon, and was the most common complication requiring reoperation (12 of 36 or 33%). The next most common complication was erosion of the cuff through the bladder neck, which occurred in 8 patients. These people frequently presented with infection, incontinence, he maturia or pain and swelling around the pump. This compli cation resulted in removal of the device in 6 patients, while in 2 a new cuff was repositioned around the bulbous urethra shortly thereafter. In 406 adults, with 236 sphincters placed around the bladder neck and 170 around the urethra described by Scott, cuff erosion was the most common cause of sphincter removal.5 Erosion was associated with infection, pressure ne crosis and surgical trauma. Light and Scott reported no erosions in their series and attributed this to the use of lower pressure regulating balloons and the avoidance of iatrogenic injury of the bladder neck or urethra during the implant procedure.6 Scott reported that higher pressure regulating balloons were likely to be associated with a higher incidence of erosion.5 Similarly, we found that children with a higher pressure regu lating balloon (81 to 100 cm. water) had a greater number of complications and a smaller rate of success than those with a lower pressure balloon (51 to 80 cm. water). This finding was statistically significant (p = 0.02). A higher pressure regulating balloon was used more frequently during the early years of our experience with the artificial sphincter. It has been suggested that girls have a higher failure rate than boys after sphincter implantation. Sidi et al reported on 15 patients, 10 boys and 5 girls, with 51 months of followup.8 Continence was achieved in 9 of the 10 boys compared to 1 of 5 girls. We found no significant difference in the success rate between boys and girls in our study except that girls who had a prior bladder neck operation were prone to suffer erosion of the cuff. We believe there is no contraindication or reservation for the use of the artificial sphincter in girls to control urinary incontinence. However, we do not advocate insertion of an artificial sphincter in any girl who has had a prior bladder neck reconstruction because there is a substantial risk of cuff erosion in these patients. During the last 10 years we have not done a single Y-V plasty in any girl before sphincter implantation as originally expounded by Light and Scott6 for this reason. We would prefer to have them perform clean intermittent cathe terization to empty or better yet, to be able to empty with Valsalva's maneuver. Upper urinary tract deterioration and changes in bladder contractility have been a concern to urologists who implant the artificial sphincter into children with myelodysplasia. Elevated filling pressure and detrusor hyperreflexia have been docu mented in previous studies after implantation of the de vice.10-13 Roth et al reported that 7 of 47 children had an increase in bladder tonicity and 4 of them had hydronephrosis.10 In 1986 Bauer et al reported on 6 children who had increased filling pressure and/or uninhibited contractions of the bladder
after placement of the artificial sphincter.11 Three children presented with urinary incontinence while 3 had hydrone phrosis. All 6 affected individuals had myelodysplasia and each incurred the changes in bladder function postpubertally. Three children had tethering of the spinal cord during this followup period, 2 of whom were managed successfully with lumbosacral laminectomy and detethering. The remaining 3 children had an augmentation cystoplasty to relieve the detrusor hyperton icity. Seven individuals in the series, including those 3 reported on previously, underwent bladder augmentation after place ment of the artificial sphincter, 4 of whom had ureteral reim plantation as well because they had vesicoureteral reflux in addition to poor detrusor compliance. Low bladder filling pres sures were achieved in all patients. Factors important for these changes in detrusor dynamics were the position of the cuff, frequency of urinary infections and adherence to bladder emptying schedules. It is unclear whether placement of the sphincter or only the presence of myelodysplasia is the most important factor in the development of changes in detrusor function. Light and Pietro believed that these changes may be secondary to vesicourethral reflexes that are stimulated after placement of a sphincter cuff around the bladder neck.13 Sometimes, however, detrusor hypertonia may be masked by such low outlet resistance that any increase in intravesical pressure leads to an escape of fluid and an inability to measure true detrusor compliance.14 Whatever the mecha nism, it is mandatory that preoperative assessment include cystometrography performed with the urethra occluded to sim ulate the effect of an activated sphincter cuff so that a true compliance factor of the bladder can be determined. In addition, it is essential that all individuals who have a sphincter in place be monitored routinely with upper urinary tract imaging on a long-term basis. A measurement of post-void residual and uro dynamic assessment are needed if a change is noted. As with any surgical endeavor involving a mechanical device, there is a learning curve for the surgeon as well as the manu facturer. Other studies have noted a difference in success rates among the various sphincter models that have evolved during the first 8 years since the sphincter was perfected.1· 5• 8 We found that a sphincter that was placed before 1980 (models AS721 and AS742) had a higher reoperation rate than a sphincter placed after 1980 (models AS792 and AS800). The ability to activate and deactivate the model AS800 sphincter without an additional operation, the use of a lower pressure regulating balloon due to improved efficiency of the cuff, strict attention to and familiarity with the surgical technique, operative sterility and the use of perioperative antibiotics are factors contributing to the more recent higher rate of success. Decreasing the incidence of fluid loss from the hydraulic system of the sphinc ter would appear to be an issue that deserves consideration by the manufacturer and surgeon in the future, since this was the most common complication in our current study. This would certainly impact on the durability of the prosthesis and have a positive effect on the long-term success rate. Currently, our contention is that the ideal candidate for the artificial urinary sphincter is someone without reflux who had neurogenic bladder dysfunction, characterized by a large capac ity, low pressure bladder without uninhibited contractions and low outlet resistance that allows for complete emptying by straining or Crede's maneuver. Although individuals with other types of lower urinary tract dysfunction can be made to fit into this pattern by using augmentation cystoplasty and/or a blad der neck operation, alternatives to the artificial urinary sphinc ter may be more appropriate if one considers our results with these other operations when performed in conjunction with the sphincter. CONCLUSION
The artificial sphincter remains a good modality for provid ing continence in selected instances and is equally effective for
LONG-TERM RESULTS OF ARTIFICIAL SPHINCTERS IN CHILDREN
boys and girls so long as the latter have not had a prior bladder neck or urethral operation. Approximately 70% of the devices will be intact and functional at 5 years and nearly two-thirds of these will not need any revision. High balloon pressures, greater than 80 cm. water, are associated with more complica tions and a higher failure rate than lower pressure regulating balloons, and, thus, should be avoided. Although a third of the individuals with an intact sphincter will need revision at some time within the first 5 years after implantation, the need for revisions has decreased in recent years as the design for the sphincter has improved and the surgical technique is refined. REFERENCES
1. Berry, J. L.: A new procedure for correction of urinary inconti nence: preliminary report. J. Urol., 85: 771, 1961. 2. Kaufman, J. J.: Treatment of post-prostatectomy urinary inconti nence using a silicone gel prosthesis. Brit. J. Urol., 45: 646, 1973. 3. Scott, F. B., Bradley, W. E. and Timm, G. W.: Treatment of urinary incontinence by implantable prosthetic sphincter. Urol ogy, 1: 252, 1973. 4. Light, J. K., Hawila, M. and Scott, F. B.: Treatment of urinary incontinence in children: the artificial sphincter versus other methods. J. Urol., 130: 518, 1983. 5. Scott, F. B.: The artificial urinary sphincter: experience in adults. Urol. Clin. N. Amer., 16: 105, 1989.
399
6. Light, J. K. and Scott, F. B.: Complications of the artificial urinary sphincter in pediatric patients. Urol. Clin. N. Amer., 10: 551, 1983. 7. Barrett, D. M. and Parulkar, B. G.: The artificial sphincter (AS800). Experience in children and young adults. Urol. Clin. N. Amer., 16: 119, 1989. 8. Sidi, A. A., Reinberg, Y. and Gonzalez, R.: Comparison of artificial sphincter implantation and bladder neck reconstruction in pa tients with neurologic urinary incontinence. J. Urol., 138: 1120, 1987. 9. Gonzalez, R. and Sheldon, C. A.: Artificial sphincters in children with neurogenic bladders: long-term results. J. Urol., 128: 1270, 1982. 10. Roth, D. R., Vyas, P. R., Kroovand, R. L. and Perlmutter, A. D.: Urinary tract deterioration associated with the artificial urinary sphincter. J. Urol., 135: 528, 1986. 11. Bauer, S. B., Reda, E. F., Colodny, A.H. and Retik, A. B.: Detrusor instability: a delayed complication in association with the arti ficial sphincter. J. Urol., 135: 1212, 1986. 12. McGuire, E. J., Woodside, J. R. and Borden, T. A.: Upper urinary tract deterioration in patients with myelodysplasia and detrusor hypertonia: a followup study. J. Urol., 129: 823, 1983. 13. Light, J. K. and Pietro, T.: Alteration in detrusor behavior and the effect on renal function following insertion of the artificial urinary sphincter. J. Urol., 136: 632, 1986. 14. Woodside, J. R. and McGuire, E. J.: Technique for detection of detrusor hypertonia in the presence of urethral sphincteric in competence. J. Urol., 127: 740, 1982.