Artificial urinary sphincter implantation in the irradiated patient: safety, efficacy and satisfaction



Objective To compare the long-term outcome of artificial urinary sphincter (AUS) implantation in patients after prostatectomy, with and with no history of previous irradiation.

Patients and methods The study included 98 men (mean age 68 years) with urinary incontinence after prostatectomy for prostate cancer (85 radical, 13 transurethral resection) who had an AUS implanted. Twenty-two of the patients had received adjuvant external beam irradiation before AUS implantation. Over a mean (range) follow-up of 46 (5–118) months, the complication and surgical revision rates were recorded and compared between irradiated and unirradiated patients. The two groups were also compared for the resolution of incontinence and satisfaction, assessed using a questionnaire.

Results Overall, surgical revision was equally common in irradiated (36%) and unirradiated (24%) patients. After activating the AUS, urethral atrophy, infection and erosion requiring surgical revision were more common in irradiated patients (41% vs 11%; P < 0.05); 70% of patients reported a significant improvement in continence, regardless of previous irradiation. Patient satisfaction remained high, with > 80% of patients stating that they would undergo surgery again and/or recommend it to others, despite previous irradiation and/or the need for surgical revision.

Conclusions Despite higher complication and surgical revision rates in patients who have an AUS implanted and have a history of previous irradiation, the long-term continence and patient satisfaction appear not to be adversely affected.


Incontinence after radical prostatectomy presents significant social and hygienic difficulties, with considerable implications for healthcare resources and costs, in addition to the obvious detrimental effect on the patients' quality of life [1]. Reported incontinence rates after prostatic extirpative surgery alone are 3–87% [2]; this wide variation reflects differing definitions of incontinence, as well as true variations in surgical technique and patient characteristics. Whilst the addition of adjuvant or neoadjuvant radiation therapy may effectively improve cure rates in high-risk cancers, the associated adverse effect on healthy tissue-repair mechanisms accounts for the higher incidence of incontinence amongst men who undergo a combination of radical prostatectomy and radiotherapy compared with surgery alone [3].

In treating incontinence after prostatectomy, available options include the use of various urinary collecting devices and catheters, or injection with periurethral bulking agents [4]. The current ‘gold standard’ of all interventions remains the implantation of an AUS. Patient satisfaction with the AUS after prostatectomy remains uniformly high, with > 90% of patients reporting a significant improvement in their quality of life [5]. However, concerns have arisen about the safety and efficacy of the AUS in patients who have received radiation therapy in addition to prostatectomy [6]. For such patients, the aetiology of their incontinence is more complex; tissue repair capabilities are impaired and surgical dissection may be more difficult, reflected by the higher rates of AUS explantation and revision reported in the irradiated population [3]. The purpose of the present study was to prospectively document and compare the long-term outcomes of AUS placement for incontinence after prostatectomy in both irradiated and unirradiated patients.

Patients and methods

The study included 98 men (mean age 68 years, range 48–78) with a history of prostatic cancer and severe urinary incontinence after prostatectomy, who had an AUS implanted; 85 had undergone radical retropubic prostatectomy and 13 transurethral resection. Twenty-two patients had additionally received postoperative EBRT because of high Gleason tumour grade and/or positive surgical margins. Of the irradiated patients, 19 (87%) had previously undergone radical prostatectomy and three (14%) had undergone transurethral resection. Of the patients who had not received radiation, 66 (87%) had undergone open surgery and 10 (13%) had undergone endoscopic resection. The time between cancer surgery and AUS implantation is shown in Table 1 both overall and for the unirradiated and irradiated groups. Relevant comorbidity, defined as coexistent cardiovascular disease or diabetes, was equally prevalent in both groups.

Table 1.  Patient characteristics, causes of incontinence, AUS component features, surgical complications, revision rates and resolution of incontinence
  • *

    P < 0.05;

  • P < 0.01.

  • In three of these five patients, infection/erosion was caused by inappropriate urethral catheterization.

No. of patients987622
Mean age (years)686967
Radical prostatectomy, n (%)85 (87)66 (87)19 (86)
Transurethral prostatectomy, n (%)13 (13)10 (13)3 (14)
Mean months since cancer surgery14.113.814.3
Comorbidity, n (%)23 (23)16 (21)5 (23)
AUS cuff size, cm
  4.06 (6)5 (7)1 (5)
  4.590 (92)70 (92)20 (91)
  5.02 (2)1 (1)1 (5)
AUS reservoir pressure, cmH2O
  51–602 (2)1 (1)1 (5)
  61–7096 (98)75 (99)21 (95)
Surgical complication and revision rates
Overall26 (25)18 (24)8 (36)
Before activation9 (9)8 (11)1 (5)
Pump migration6 (6)5 (7)1 (5)
Haematoma2 (2)2 (3)0
Mechanical complication1 (1)1 (1)0
After activation17 (17)8 (11)9 (41)
Urethral atrophy9 (9)6 (8)3 (14)
Infection/erosion6 (6)1 (1)5 (23)
Mechanical complication 2 (2.0)1 (1)1 (5)0
Resolution of incontinence after sphincter implantation
No. of patients917318
Dry17 (19)15 (21)2 (11)*
Significant improvement64 (70)51 (70)13 (72)
Little/no improvement10 (11)7 (10)3 (17)*

Before AUS implantation all patients underwent a full history and physical examination. All were categorized as having severe stress urinary incontinence, as defined by the use of at least two diapers or four incontinence pads per day. Multichannel urodynamic studies were used to evaluate lower urinary tract storage and voiding capabilities. Intrinsic sphincter deficiency was confirmed by abdominal leak-point pressures of < 65 cmH2O in all patients. Cysto-urethroscopy excluded coexistent pathology such as bladder neck contracture or urethral/anastomotic stricture.

The AUS (AMS 800, American Medical Systems, Minnetonka, USA) was implanted using the same operative technique by the same surgeon in all cases (A.R.S.). The AUS component characteristics were similar in both patient groups (Table 1). Via a midline perineal incision, each AUS cuff was positioned around the distal urethral bulb, within the bulbospongiosus muscle; the cuff sizes are given in Table 1. Through a separate midline suprapubic incision, each AUS reservoir was placed in a pouch constructed below the rectus sheath, with reservoir pressures as shown in Table 1; 20 mL isotonic contrast medium was used for filling the fluid system. All sphincters were then deactivated and a 16 F Silastic urethral catheter inserted at the end of the procedure. Antibiotic prophylaxis was provided by intravenous ampicillin and gentamicin perioperatively, and irrigation of the surgical field and AUS components during implantation with 120 mL bacitracin/kanamycin solution, and oral cephalosporin therapy for 1 week after surgery. After removing the catheter on the first morning after surgery all patients were discharged, and returned for follow-up and sphincter activation 6 weeks thereafter.

The mean (range) follow-up was 46 (5–118) months, with a mean of 44 and 47 months in the unirradiated and irradiated patients, respectively. Surgical complication and revision rates were recorded and compared between groups, both before and after sphincter activation.

The patients' satisfaction with the resolution of their incontinence was assessed after surgery by an impartial reviewer using a previously reported written or telephone questionnaire [5]; Table 2 lists the questions used to assess the outcome variables before and after surgery, and each patient's perception of symptoms after surgery. Continence was then ranked as being either: (i) completely resolved (no leakage episodes and no requirement to use protective devices); (ii) significantly improved (frequency of leakage and protective device use reduced by at least half); or (iii) unchanged (including patients who reported some improvement in continence, but a reduction of less than half in leakage episodes or protective device usage). The patients' responses about their level of satisfaction with the results of surgery were also recorded, each patient being asked if they would: (i) undergo AUS implantation again and (ii) recommend the procedure to a friend. The pad scores, surgical complication and revision rates were assessed using univariate analysis and the unpaired Student's t-test. The responses to the patient questionnaire were analysed using contingency tables and compared using the chi-squared test; in all tests statistical significance was assumed for P < 0.05.

Table 2.  The outcome variables assessed before and after AUS implantation
VariablePossible responses
Frequency of leakageNone; Few times/month; Few times/week; Once/day; More than once/day; Continuous
LeakageNone; Few drops; <1 teaspoon; >1 teaspoon; >1 cup
Use of protective devicesNone; Pads (number); Diapers (number); Penile clamp; Catheter
Patient satisfactionVery satisfied; Satisfied; Undecided; Unsatisfied
Recommend AUS to friend?Yes; Undecided; No
Willingness to undergo procedure againYes; Undecided; No


The surgical complication and revision rates are shown in Table 1; overall, 26 patients required a revision, for the reasons given in Table 1. Whilst revisions tended to be more frequent in irradiated than in unirradiated patients the differences were not statistically significant (P = 0.06). The differences in revisions before sphincter activation were not statistically significant (Table 1) although only one irradiated patient required pump repositioning before sphincter activation; after sphincter activation urethral atrophy was more frequent in irradiated than in unirradiated patients (P = 0.01). Infection and erosion were also more common in irradiated patients (P = 0.006), although in three of the five irradiated patients with this complication it was secondary to inappropriate urethral catheterization. Mechanical complications (e.g. fluid leak or pump failure) occurred in one patient in each group.

Questionnaires were completed by 91 (93%) of the patients (73 unirradiated and 18 irradiated patients); 57 (63%) were contacted by telephone and 34 (37%) completed the written format. Sixty-four (70%) patients reported that their incontinence was significantly improved and there was no difference in the frequency with which this was reported between the groups (Table 1). Of the remaining 27 (30%) patients, complete resolution of incontinence was more common in the unirradiated patients (Table 1; P = 0.02), whilst incontinence remained unchanged more often in the irradiated group (P = 0.03).

Sixty-seven (92%) of the unirradiated patients reported that they were either very satisfied or somewhat satisfied with their AUS surgery, compared with 16 (89%) of the irradiated patients; the other responses are shown in Table 3; none of the differences were statistically significant. All of the 26 patients who required surgical AUS revision responded to questionnaire; their responses are also shown in Table 3, with the responses of the 65 responding patients who had not undergone revision. Whilst there was an overall trend for irradiated or surgically revised patients to be less satisfied with the results of their surgery than their counterparts who had no such history, these differences were not statistically significant.

Table 3.  The degree of satisfaction in patients treated or untreated with radiation and requiring surgical revision or not
Satisfaction (%)UnirradiatedIrradiatedNo surgical revisionSurgical revision
Very satisfied/ satisfied92899589
Would undergo AUS again90898685
Would recommend  AUS to a friend89838585


Since first introduced in 1973 by Scott et al.[7] the AUS is now accepted as the most effective intervention for incontinence after prostatectomy, with continence and patient satisfaction rates reported to be > 90% [8,9]. Some patients also develop incontinence after endoscopic prostatic resection for malignant disease and the AUS is similarly effective for these patients [2].

Radical prostatectomy is currently considered to provide the best opportunity for cure in men with localized prostate cancer. Several studies have reported lower local recurrence rates after surgery by giving adjuvant radiotherapy in selected patients [10]. Urinary incontinence after radical prostatectomy alone is reported in 3–87% of patients, largely depending upon whether the data are derived from objective observations of the physician or subjective patient perception [11]. Urinary incontinence after prostatic extirpative surgery has a significant adverse effect on quality of life [12]. Both the incidence of incontinence and its severity increase when radiotherapy is added to surgical intervention [6], attributed to the combination of radiation-induced ischaemia and late-effect radiotoxicity, which causes detrusor and interstitial fibrosis, thus reducing bladder compliance. Similar changes may also be expected to occur in the urethra, which may effectively reduce tissue repair capabilities and increase the potential for cuff erosion [13].

Despite these disadvantages, several authors have reported encouraging results with the AUS in patients with a previous history of radiation treatment [3,6]. The present study was aimed to further determine the immediate and long-term outcomes of AUS implantation in irradiated and unirradiated patients incontinent after prostatectomy, particularly the surgical complication and revision rates, resolution of incontinence and overall patient satisfaction.

The present results indicate that there was an overall trend to a higher surgical revision rate in irradiated patients, but this was only statistically significant if early revision surgery (before sphincter activation) was excluded from the analysis. In the initial period, before sphincter activation, revision surgery was more common in unirradiated patients, with only one patient in the irradiated group requiring reoperation. As early revision surgery was largely undertaken for indications such as pump migration or haematoma formation, it is likely that these early complications reflect surgical technique and the failure of patients to comply with directions to regularly exert traction on their pumps (to prevent proximal migration) rather than true biological or pathological differences between the groups.

After sphincter activation the longer term follow-up showed that recognized complications, e.g. urethral atrophy, infection and erosion, were more common in irradiated patients. This probably reflects the adverse long-term biological effects of radiation. Beyond induced cellular toxicity, the known sequelae of tissue ischaemia, fibrosis and necrosis can only be expected to mitigate against tissue repair. Radiation-induced small-vessel occlusion secondary to endarteritis obliterans appears to be the common pathophysiological pathway accounting for these adverse macroscopic changes. However, in this series, infection and erosion occurred in three of five irradiated patients as a direct result of inappropriate urethral catheterization, rather than as a result of previous irradiation per se. This may have been avoided by providing appropriate education for patients and medical staff outside the primary surgical team.

Complete resolution of incontinence was more frequent in unirradiated patients and ‘little or no improvement’ was reported more often by irradiated patients. Patients reporting complete resolution of incontinence (‘cure rate’) or ‘little or no improvement’ (‘failure rate’) represented only 30% of the patients overall. From a more optimistic perspective therefore, 70% of patients overall reported a significant improvement in continence, regardless of a history of previous irradiation.

Despite the apparent increase in complication rates in the irradiated patients, there were no significant differences in reported levels of satisfaction between the patient groups (Table 3). The high level of satisfaction reported by all the present patients is comparable with that in previous studies [5,14]. Amongst patients who underwent surgical revision, the vast majority stated that they would undergo AUS implantation again and would recommend the procedure to a friend. This is an important result and should be explained to patients when discussing treatment options, and one which physicians should consider when faced with such a decision. Adjuvant radiotherapy should not be withheld from patients in an effort to preserve urinary continence, nor should irradiated patients be discouraged from selecting the AUS for fear of surgical failure. The present results suggest that whilst previous irradiation tends to increase complication rates and decrease rates for complete resolution of incontinence, overall patient satisfaction is unaffected.

The present patients received localized radiotherapy to the prostate bed; thus, these findings may not be applicable to patients receiving more widespread irradiation, and implanting the AUS in such patients may be associated with more complications.

AUS implantation is successful in the irradiated patient but at a price; patients with significant incontinence after prostatectomy and who are candidates for adjuvant radiation therapy before having an AUS should be counselled on the increased complication and surgical revision rates, whilst being reassured that long-term continence and overall satisfaction should not be expected to be significantly affected.


I.K. Walsh, MD, FRCS(Urol), Visiting Assistant Professor.

S.G. Williams, MD, Chief Resident.

V. Mahendra, MBBS, MS(Surg), MCh(Urol), DNB (Urol), FRCS(Urol), Specialist Registrar in Urology.

T. Nambirajan, MBBS, MS(Surg), MCh(Urol), FRCS(Glasg), FRCS(Edin), Specialist Registrar in Urology.

A.R. Stone, MD, FRCS(Ed), Professor and Vice Chairman.

I.K. Walsh, Consultant Urologist, Belfast City Hospital, Lisburn Road, Belfast, BT9 7AB, UK.