Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Duloxetine for female stress urinary incontinence

Women in the trials
Outcomes
Trial design
Results
Efficacy
Harm
Comment

Urinary incontinence is a problem that may affect as many as a third of adult women. Most common is stress incontinence, involuntary leakage brought on by effort or exertion, or sneezing or coughing. Treatments include pelvic floor muscle training, behavioural therapies, and even surgery. Duloxetine is one of the first pharmacological treatment. There are now four published randomised trials of the 40 mg twice-daily dosing regime [1-4], which offer an opportunity not only to examine efficacy, but also to try and understand how trials in stress incontinence are conducted. Bandolier has performed a swift review, therefore.

Women in the trials

Women were 18 years or older with troublesome stress urinary incontinence lasting at least three months. They had to have seven or more episodes of stress urinary incontinence every week, where an episode was defined as an easily noticed leakage of urine that wet a pad or clothing, and that occurred with physical stress such as coughing, sneezing, or exercising. They had to have a diurnal frequency of fewer than nine per day, a nocturnal frequency of fewer than three per night, and have no urge incontinence.

Additionally, women underwent objective testing of normal bladder capacity and stress incontinence. The bladder was filled with saline. Women were excluded if unable to tolerate filling to 400 mL , with a first sensation of bladder filling at less than 100 mL, or with no sensation at any time during filling. After filling, both a positive cough test (visible leakage on coughing), and a positive stress pad test (leakage of more than 2 grams) were required.

Outcomes

The main outcome was incontinence episode frequency, which was recorded in real-time daily diaries for a week before each visit. Two diaries were completed before randomisation, and one at each of the last weeks of three monthly assessments. Other outcomes were quality of life questionnaires, adverse events, and compliance.

Trial design

Before randomisation there were three two-week periods, for screening for suitability for entry, and with drug free and placebo periods where diaries were completed. Randomisation was by remote central randomisation, stratified by incontinence frequency. Duloxetine and placebo tablets were identical and taken twice a day for 12 weeks.

Results

There were four trials [1-4] with 958 women on duloxetine 80 mg daily, and 955 on placebo. Women were aged from 22 to about 80 years, with an average age of about 50 years. The average number of incontinence episodes per week at baseline was 14-18 per week in the trials with an overall average of about 17 episodes per week, and over half of the women had more than 14 incontinence episodes a week.

Efficacy

Treatment with both duloxetine and placebo reduced the number of incontinence episodes over 12 weeks. The overall reduction was 52% with duloxetine and 33% with placebo, with consistency between trials (Figure 1). Not unsurprisingly there was a similar change in the proportion of women who had a 50% or greater reduction in the number of incontinence episodes (Table 1).



Figure 1: Percentage reduction in weekly incontinence frequency with duloxetine and placebo







Table 1: Benefits and harms in randomised trials of duloxetine in stress incontinence in women




Number of
Percent of


Outcome
Trials
Patients
Duloxetine
Placebo
Relative benefit or risk
(95% CI)
NNT/NNH
(95% CI)
Benefit
At least 50% reduction in incontinence episodes
3
1635
54
36
1.5 (1.3 to 1.7)
5.7 (4.5 to 7.8)
No incontinence episodes in last week
3
1432
11
8
1.4 (1.02 to 2.0)
32 (16 to 765)
Harm
Patients with at least one AE
4
1913
73
55
1.3 (1.2 to 1.4)
5.5 (4.5 to 7.2)
Serious AE
4
1913
2
1
1.8 (0.8 to 3.9)

Adverse event withdrawal
4
1913
18
4
5.1 (3.6 to 7.3)
6.8 (5.7 to 8.3)
Nausea
4
1913
23
4
6.4 (4.5 to 9.1)
5.2 (4.6 to 6.2)
Fatigue
4
1913
12
4
3.4 (2.3 to 4.9)
12 (9 to 16)



The number needed to treat with duloxetine 80 mg rather than placebo for 12 weeks for one additional woman to have a more than 50% reduction in incontinence episodes was 5.7 (4.5 to 7.8). It was 32 (16 to 765) for one woman to have no incontinence episodes (Table 1).

Quality of life measures used in the trials all showed significant improvements for duloxetine compared with placebo. These included avoidance and limiting behaviour, psychological impact, and social embarrassment, quality of life measures specific to urinary incontinence.

Harm

The majority of women reported having at least one adverse event over the 12 weeks (Table 1). About one patient in five on duloxetine withdrew because of adverse events, but few adverse events were described as serious, and these were not significantly higher for duloxetine over placebo (Table 1).

The most common adverse event was nausea (NNH 5), which was moderate or severe in about half the cases. Other common adverse events were fatigue (12%, NNH 12), dry mouth (13%), insomnia (13%), constipation (11%), headache (10%), dizziness (10%), somnolence (7%) and diarrhoea (5%).

Comment

Duloxetine pharmacology involves effects on serotonin and noradrenaline, so the adverse effects seen in the trials were much to be expected. The effect on stress incontinence was large, with over half the women given duloxetine experiencing a major reduction in stress incontinence events. Given the starting point of about 17 episodes a week, the median reduction was by about eight or nine episodes a week, or one a day.

The NNT of 5.7 was not startling, because there was also a large effect in women given placebo, where a third had a reduction in episodes of more than 50%. This should not be seen as the power of placebo to cause these effects, or mind over matter. It is more likely to be a reflection of the nature of the trials themselves, and the stringent entry requirements that may have maximised the size of the problem, creating conditions where doing nothing was beneficial.

What we have is four high-quality trials that describe a new drug treatment. They give us much information about effect and common adverse effects. There was some inconsistency of reporting, but that happens when information is compressed into the word limits demanded of scientific papers, so we have a more fuzzy picture than may have been.

The key will be to establish just which women seen in urology clinics or general practice would benefit, and which would find the adverse event profile bearable.

References:

  1. PA Norton et al. Duloxetine versus placebo in the treatment of stress urinary incontinence. American Journal of Obstetrics and Gynecology. 2002 187: 40-48.
  2. RR Dmochowski et al. Duloxetine versus placebo for the treatment of north American women with stress urinary incontinence. Journal of Urology 2003 170: 1259-1263.
  3. RJ Millard et al. Duloxetine vs placebo in the treatment of stress urinary incontinence: a four-continent randomised clinical trial. BJU International 2004 93: 311-318.
  4. P van Kerrebroeck et al. Duloxetine versus placebo in the treatment of European and Canadian women with stress urinary incontinence. British Journal of Obstetrics and Gynaecology 2004 111: 249-257.

previous or next story