HAY-SMITH - Which Anticholinergic Drug for Overactive Bladder Symptoms

8/6/2019 HAY-SMITH - Which Anticholinergic Drug for Overactive Bladder Symptoms

1/109

Which anticholinergic drug for overactive bladder symptoms

in adults (Review)

Hay-Smith J, Ellis G, Herbison GP

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library2009, Issue 1http://www.thecochranelibrary.com

Which anticholinergic drug for overactive bladder symptoms in adults (Review)

Copyright 2009 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
http://www.thecochranelibrary.com/http://www.thecochranelibrary.com/


2/109

T A B L E O F C O N T E N T S

1HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19 AUTHORS CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20 ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

26CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Analysis 1.1. Comparison 1 One anticholinergic versus another, Outcome 1 Cure/improvement. . . . . . . . 73

Analysis 1.2. Comparison 1 One anticholinergic versus another, Outcome 2 Leakage episodes in 24hrs. . . . . . 74

Analysis 1.3. Comparison 1 One anticholinergic versus another, Outcome 3 Change in leakage episodes in 24hrs. . 75

Analysis 1.4. Comparison 1 One anticholinergic versus another, Outcome 4 Micturitions in 24hrs. . . . . . . 76

Analysis 1.5. Comparison 1 One anticholinergic versus another, Outcome 5 Change in micturitions in 24 hrs. . . 77

Analysis 1.6. Comparison 1 One anticholinergic versus another, Outcome 6 Maximum cystometric capacity. . . . 78

Analysis 1.7. Comparison 1 One anticholinergic versus another, Outcome 7 Change in maximum cystometric capacity. 79

Analysis 1.9. Comparison 1 One anticholinergic versus another, Outcome 9 Change in volume at first contraction. . 80

Analysis 1.10. Comparison 1 One anticholinergic versus another, Outcome 10 Residual volume. . . . . . . . 81

Analysis 1.11. Comparison 1 One anticholinergic versus another, Outcome 11 Change in residual volume. . . . . 82

Analysis 1.12. Comparison 1 One anticholinergic versus another, Outcome 12 Withdrawals due to adverse events. . 83

Analysis 1.13. Comparison 1 One anticholinergic versus another, Outcome 13 Dry mouth. . . . . . . . . . 84 Analysis 2.1. Comparison 2 Different doses of tolterodine, Outcome 1 Cure/improvement. . . . . . . . . . 85

Analysis 2.3. Comparison 2 Different doses of tolterodine, Outcome 3 Change in leakage episodes per 24hrs. . . . 86

Analysis 2.5. Comparison 2 Different doses of tolterodine, Outcome 5 Change in number of micturitions per 24hrs. 87

Analysis 2.7. Comparison 2 Different doses of tolterodine, Outcome 7 Change in maximum cystometric capacity. . 88

Analysis 2.9. Comparison 2 Different doses of tolterodine, Outcome 9 Change in volume at first contraction. . . . 89

Analysis 2.11. Comparison 2 Different doses of tolterodine, Outcome 11 Change in residual volume. . . . . . . 90

Analysis 2.12. Comparison 2 Different doses of tolterodine, Outcome 12 Withdrawal due to adverse events. . . . 91

Analysis 2.13. Comparison 2 Different doses of tolterodine, Outcome 13 Dry mouth. . . . . . . . . . . . 92

Analysis 3.1. Comparison 3 Extended versus immediate release preparations, Outcome 1 Cure/improvement. . . . 93

Analysis 3.2. Comparison 3 Extended versus immediate release preparations, Outcome 2 Leakage episodes in 24hrs. 94

Analysis 3.3. Comparison 3 Extended versus immediate release preparations, Outcome 3 Change in leakage episodes per

24hrs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Analysis 3.4. Comparison 3 Extended versus immediate release preparations, Outcome 4 Micturitions in 24 hrs. . . 96Analysis 3.5. Comparison 3 Extended versus immediate release preparations, Outcome 5 Change in micturitions per

24hrs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Analysis 3.6. Comparison 3 Extended versus immediate release preparations, Outcome 6 Maximum cystometric capacity. 98

Analysis 3.8. Comparison 3 Extended versus immediate release preparations, Outcome 8 Volume at first contraction. 99

Analysis 3.11. Comparison 3 Extended versus immediate release preparations, Outcome 11 Change in residual volume. 100

Analysis 3.12. Comparison 3 Extended versus immediate release preparations, Outcome 12 Withdrawal due to adverse

events. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Analysis 3.13. Comparison 3 Extended versus immediate release preparations, Outcome 13 Dry mouth. . . . . . 102

Analysis 4.3. Comparison 4 One extended release preparation against another, Outcome 3 Change in leakage episodes in

24 hrs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Analysis 4.5. Comparison 4 One extended release preparation against another, Outcome 5 Change in micturitions in 24

hrs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

iWhich anticholinergic drug for overactive bladder symptoms in adults (Review)



3/109

Analysis 4.12. Comparison 4 One extended release preparation against another, Outcome 12 Withdrawals due to adverse

events. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Analysis 4.13. Comparison 4 One extended release preparation against another, Outcome 13 Dry mouth. . . . . 104

104WHATS NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

105SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

106INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iiWhich anticholinergic drug for overactive bladder symptoms in adults (Review)



4/109

[Intervention Review]

Which anticholinergic drug for overactive bladder symptomsin adults

Jean Hay-Smith1, Gaye Ellis2, G Peter Herbison3

1Rehabilitation Teaching and Research Unit, Department of Medicine, Wellington South, New Zealand. 2Womens and Childrens

Health, University of Otago, Dunedin, New Zealand. 3Department of Preventive & Social Medicine, Dunedin School of Medicine,

University of Otago, Dunedin, New Zealand

Contact address: JeanHay-Smith, Rehabilitation Teaching and Research Unit, Departmentof Medicine, WellingtonSchool of Medicine

and Health Sciences, University of Otago, PO Box 7343, Wellington South, Wellington, New Zealand. [email protected].

Editorial group: Cochrane Incontinence Group.

Publication status and date: Edited (no change to conclusions), published in Issue 1, 2009.

Review content assessed as up-to-date: 24 May 2005.

Citation: Hay-Smith J, Ellis G, Herbison GP. Which anticholinergic drug for overactive bladder symptoms in adults. CochraneDatabase of Systematic Reviews2005, Issue 3. Art. No.: CD005429. DOI: 10.1002/14651858.CD005429.


A B S T R A C T

Background

Around 16% to 45% of adults have overactive bladder symptoms (urgency with frequency and/or urge incontinence - overactive

bladder syndrome). Anticholinergic drugs are common treatments.

Objectives

To compare the effects of different anticholinergic drugs for overactive bladder symptoms.

Search strategy

We searched the Cochrane Incontinence Group specialised trials register (searched 17 January 2002) and reference lists of relevant

articles. A search for full publications of abstracts identified in January 2002 was completed in July 2003.

Selection criteria

Randomised trials in adults with overactive bladder symptoms or detrusor overactivity that compared one anticholinergic drug with

another, or two doses of the same drug.

Data collection and analysis

Two authors independently assessed eligibility, trial quality and extracted data. Data were processed as described in the Cochrane

Reviewers Handbook.

Main results

Forty nine trials, 39 parallel and 10 cross-over designs were included (11,332 adults). Most trials were described as double-blind, but

were variable in other aspects of quality. Crossover studies did not present data in a way that could be included in the meta-analysis.

Four trials collected quality of life data (the primary outcome measure) using validated measures; none reported useable data.

Oxybutynin versus tolterodine: There were no statistically significant differences for patient perceive improvement, leakage episodes or

voids in 24 hours, but fewer withdrawals due to adverse events (RR 0.57, 95% CI 0.43 to 0.75), and less risk of dry mouth (RR 0.60,

95% CI 0.54 to 0.66), with tolterodine.

1Which anticholinergic drug for overactive bladder symptoms in adults (Review)

mailto:[email protected]:[email protected]


5/109

Different doses tolterodine: The usual recommended starting dose (2 mg twice daily) was compared with two lower (0.5 mg and 1 mg

twice daily), and one higher dose (4 mg twice daily). The effect of 1 mg, 2 mg and 4 mg doses was similar for leakage episodes andmicturitions in 24 hours, with greater risk of dry mouth with 2 and 4 mg doses.

Extended versus immediate release preparations of oxybutynin and/or tolterodine: There were no statistically significant differences

for cure/improvement, leakage episodes or micturitions in 24 hours, or withdrawals due to adverse events, but there were few data.

Overall, extended release preparations had less risk of dry mouth.

Oneextendedrelease preparation versusanother:Therewas less risk ofdry mouth with oral extended release tolterodinethan oxybutynin

(RR 0.75, 95% CI 0.59 to 0.95), but no difference between transdermal oxybutynin and oral extended release tolterodine although

some people withdrew due to skin reaction at the trandermal patch site.

Authors conclusions

Where the prescribing choice is between oral immediate release oxybutynin or tolterodine, tolterodine might be preferred for reducedrisk of dry mouth. With tolterodine, 2 mg twice daily is the usual starting dose, but a 1 mg twice daily dose might be equally effective

with less risk of dry mouth. If extended release preparations of oxybutynin or tolterodine are available, these might be preferred to

immediate release preparations because there is less risk of dry mouth. There is l ittle or no evidence available about quality of life, costs,

or long-term outcome in these studies. There were insufficient data from trials of other anticholinergic drugs to draw any conclusions.

P L A I N L A N G U A G E S U M M A R Y

Which anticholinergic drug for overactive bladder symptoms in adults

Many adults have symptoms of overactive bladder. A person with overactive bladder syndrome feels a very strong urge to pass urine,

and they may not make it to the toilet before they leak urine. Other common problems are a feeling of needing to urinate often duringthe day and/or night. This problem seems to be caused by an overactive bladder muscle, and it becomes more common with age.

Treatments are conservative measures such as bladder training, or drugs. Anticholinergic drugs can reduce the overactivity of the bladder

muscle, and the feeling of urgency. The review found that there are several anticholinergic drugs prescribed for adults with overactive

bladder symptoms. The two most studied drugs are oxybutynin and tolterodine. These two drugs have similar effects, but on average

those taking oxybutynin were more likely to withdraw from the studies because of adverse effects, mainly dry mouth. However, both

drugs can give dry mouth, and this problem is less likely if an extended release formulation of either drug is used.

B A C K G R O U N D

Overactive bladder syndrome

People with overactive bladder syndrome report urgency (with

or without urge incontinence), usually in combination with fre-

quency and/or nocturia (Abrams 2002). To be called overactive

bladder syndrome these symptoms cannot be caused by metabolic

problems such as diabetes, problems with the urinary tract such

as urinary tract infection, or neurological diseases such as multi-

ple sclerosis. Overactive bladder syndrome may also be called urge

syndrome or urgency-frequency syndrome.

Urgency is the sudden and compelling desire to pass urine, which

isdifficult todefer (Abrams 2002). Sometimes thereis involuntary

leakage of urine with the feeling of urgency, and this is called urge

incontinence. Urgency and urge incontinence usually result from

an involuntary increase in bladder pressure due to detrusor (blad-

der smooth muscle) over-activity. If further investigation of ur-

gency/urge incontinence with urodynamics demonstrates sponta-

neous or provoked detrusor muscle contraction in the filling phase

of the test then detrusor overactivity is diagnosed. If there is no

defined cause for the overactivity this is called idiopathic detrusor

overactivity, but if there is a relevant neurological condition then

the term neurogenic detrusor overactivity (previously detrusor hy-




6/109

perreflexia) is used.

Frequency is the complaint of needing to void too often duringthe day, while nocturia is waking once or more per night to void (

Abrams 2002). In clinical practice, a person who voids more than

eight times during the day would be considered to have daytime

frequency; waking from sleep more than once at night to void

would be considered nocturia.

Overactive bladder symptoms are very common in community

dwelling adults, with recent large studies showing prevalences of

17% in Europe (Milsom 2001)andtheUSA(Stewart 2001), 31%

inKorea(Choo2001)and45%inAsianmen(Moothy 2001).The

Asian studies may have overestimated prevalence, because people

whose only symptom was frequency were counted as cases of

overactive bladder. Several large population studies have reported

that the prevalence of overactive bladder symptoms increases withageinmenandwomen( Brown 1999; Milsom 2001; Moller 2000;

Stewart 2001; Ueda 2000). Although the diagnosis of overactive

bladder excludes people with known causes of detrusor overactiv-

ity, e.g. neurological disorders, this combination of symptoms is

nevertheless common in such groups. In fact, urinary dysfunction

appears to be more common in the neurologically impaired than

unimpaired population; the most frequently reported problems

are urgency and/or frequency (Hennessey 1999).

Research on the amount of bother caused by overactive bladder

syndrome, and the effect on quality of life, is only just beginning.

However, it seems that frequency and/or urgency might be just

as bothersome as actual leakage (Milsom 2001), and overall the

effects of overactive bladder symptomson quality of lifearemarked(Jackson 1997). It also seems clear that many of thepeople affected

by overactive bladder symptoms do not seek help from health care

professionals (Milsom 2001; Ueda 2000).

Treatment of overactive bladder symptoms

The two main treatment options for overactive bladder syndrome

are bladder training and pharmacotherapy, i.e. drugs. A separate

Cochrane review on bladder training (Wallace 2004) is available,

and the scope of this review is confined to drug treatment.

While the pathophysiology of the overactive bladder remains to

be fully elucidated, the two most widely accepted explanationsare the myogenic and neurogenic theories; these are not mutally

exclusive (Hashim 2004). In the first, partial denervation of the

detrusor muscle is thought to increase excitability. Changes in cen-

tral nervous system pathways that inhibit bladder activity, and/or

over-sensitivity of the sensory nerve endings in the bladder, are the

basis of the second. In both theories, the outcome is overactivity

of the detrusor muscle. The motor nerve supply to the bladder is

via the parasympathetic nervous system (via sacral nerves S2,3,4)

(Abrams 1988; Ouslander 1982; Ouslander 1986), which affects

detrusor muscle contraction. This is mediated by acetylcholine

acting on muscarinic receptors at the level of the bladder. There

are currently five recognised subtypes of muscarinic receptor; the

M1, M2, and M3 subtypes are of interest in bladder activity. Mus-

carinic receptors are found in other parts of the body too, e.g.in the gut, salivary glands, tear ducts. Pharmacotherapy relies on

the use of drugs with anticholinergic properties. The rationale for

using anticholinergic drugs in the treatment of overactive bladder

syndrome is to block the parasympathetic acetylcholine pathway

and thus abolish or reduce the intensity of detrusor muscle con-

traction. Unfortunately noneof the anticholinergic drugsavailable

to date is specific to the muscarinic receptorsin the bladder, and as

a result the drugs can cause side effects by acting in other parts of

the body too, e.g. dry mouth or eyes, constipation or nausea. For

the purpose of this review the term anticholinergic medications

will refer to both pure antimuscarinic drugs and antimuscarinic

drugs with mixed actions given specifically for bladder symptoms.

Drugs with mixed actions will be included where their clinical ef-fect is thought to be antimuscarinic, rather than via direct action

on bladder muscle. Medications with secondary anticholinergic

effects, e.g. tricyclic depressants will be excluded.

The number of anticholinergic drugs available on the market

is increasing and various studies, both observational and ran-

domised controlled trials, have evaluated effectiveness. A previous

Cochrane review compared anticholinergic drugs with no treat-

ment or placebo treatments (Hay-Smith 2003). While statisti-

callysignificant, the differences between anticholinergic drugs and

placebo were small and of uncertain clinical significance, apart

from the rate of dry mouth which is a common side effect of an-

ticholinergic therapy.

Despite this, anticholinergics are commonly used in primary andsecondary caresettings for the treatment of overactive bladder syn-

drome, and this has considerable resource implications (Kobelt

1997). If anticholinergic therapy is prescribed, there is still un-

certainty about which anticholinergic drugs are most effective, at

which dose, and by which route of administration. There is also

uncertainty about the role of anticholinergic drugs in different

patient groups (e.g. the elderly, male and female).

There are many studies of the effects of anticholinergic drugs.Four

Cochrane reviews will consider them. The present review com-

pares anticholinergic drugs with each other, to see whether dif-

ferent anticholinergic drugs have different effects. A previous re-

viewcompared anticholinergic drugswith no treatmentor placebo

treatments (Hay-Smith 2003). Two further reviews will consider:1) whetheranticholinergic drugsare better than other active (non-

drug) therapies (Patrick 2004); and 2) whether anticholinergic

drugs are better than other drug treatments (Dublin 2004).

Two previous systematic reviews of anticholinergic drugs for urge

urinary incontinence were found (Haeusler 2002; Harvey 2001).

Haeusler et al (Haeusler 2002) included only placebo controlledtrials, and no comparisons of anticholinergic drugs. Harvey et al (

Harvey 2001) included trials that compared tolterodine and oxy-

butynin; the review, using Cochrane methods, is now outdated be-

cause other relevant trials have been published. In addition, Har-

veyetal(Harvey 2001) did not include all possible anticholinergic




7/109

drug comparisons. There is clearly a need for a regularly updated

and comprehensive systematic review of the effectiveness of anti-cholinergic drugs, versus each other, in the management of adults

with overactive bladder syndrome.

O B J E C T I V E S

To determine the effects of anticholinergic drugs in the treatment

of overactive bladder symptoms.

The following comparisons were made:

1. A particular anticholinergic drug versus another in the manage-

ment of overactive bladder symptoms.

2. Oneroute of anticholinergic drugadministration versus another

(e.g. oral, transdermal, rectal, intravesical)

3. Higher doses of anticholinergic drugs versus lower doses.

M E T H O D S

Criteria for considering studies for this review

Types of studies

All randomised or quasirandomisedcontrolled trials of anticholin-

ergic drugs for the treatment of overactive bladder symptoms or

detrusor overactivity.

Types of participants

All adult men and women with a symptomatic diagnosis of over-

active bladder syndrome, with or without a urodynamic diagnosis

of detrusor overactivity. Although people with neurological disor-

ders cannot, by definition, have overactive bladder syndrome they

often experience overactive bladder symptoms secondary to their

neurologic disease, and are offered anticholinergic drugs. There-fore, trials that recruited people with neurologic disorders com-

plaining of overactive bladder symptoms, and/or with a diagnosis

of neurogenic detrusor overactivity, were included.

Types of interventions

One arm of the study was allocated an anticholinergic drug and

at least one other arm used a different anticholinergic drug, an

anticholinergic drug given via a different route, or a different dose

of the same anticholinergic drug. To be included the drug had to

be an anticholinergic/muscarinic antagonist, and be given for the

purpose of decreasing symptoms of overactive bladder. Trials of

the following drugs were considered: emepronium bromide/car-

rageenate, darifenacin, dicyclomine chloride,oxybutynin, propan-theline bromide, propiverine, solifenacin (YM905), tolterodine,

and trospium chloride. Terodoline, an anticholinergic drug pre-

viously used in the treatment of overactive bladder, was excluded

because it was withdrawn from the market.

Other drugs with less direct anticholinergic effects were excluded,

e.g. smooth muscle relaxants (flavoxate hydrochloride, calcium

channel blockers, potassium channel openers, beta-adrenoceptor

agonists, alpha-adrenoceptor antagonists, and prostaglandin syn-

thetase inhibitors) and tricyclic antidepressants.

Types of outcome measures

The primary outcomes of interest were:1. Generic (e.g. Short Form 36) (Ware 1993) and condition spe-

cific quality of life (e.g. Incontinence Impact Questionnaire), for

example (Shumaker 1994), and psychosocial measures.

The secondary outcomes of interest were:

2. Patients observations e.g. symptom scores, perception of cure/

improvement, satisfaction with outcome.

3. Quantification of symptoms e.g. number of leakage episodes,

frequency and volume (urinary diary).

4. Clinicians measures e.g. urodynamic measures (e.g. maximum

cystometric capacity), and clinical findings.

5. Socioeconomics - direct and indirect costs of interventions (for

patients and providers), resource implications of differences in

outcome, formal economic analysis (e.g. cost effectiveness, costutility), desire or need for further treatment.

6. Other - e.g. adverse events, compliance measures, long term

follow up and any other outcome not pre-specified but judged

important when performing the review.

Search methods for identification of studies

This review has drawn on the search strategy developed for

the Incontinence Review Group. Relevant trials were identi-

fied from the Cochrane Incontinence Group Specialised Reg-

ister of controlled trials, which is described under the Groups

details in The Cochrane Library (For more details please see theSpecialized Register section of the

Groups module in The Cochrane Library). The register contains

trials identified from MEDLINE, CINAHL, the Cochrane Cen-

tral Register of Controlled Trials (CENTRAL) and handsearching

of journals and conference proceedings.

The date of the most recent search of the specialised register for

this review was: 17 January 2002.

The trials in the Incontinence Group Specialised Register are also

contained in the Cochrane Central Register of Controlled Trials

(CENTRAL).

The Incontinence Group Specialised Register was searched using

the Groups own keyword system, the search terms used were:


http://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.htmlhttp://www.mrw.interscience.wiley.com/cochrane/clabout/articles/INCONT/frame.html


8/109

topic.urine.incon*

AND({design.cct*} OR {design.rct*})

AND

({intvent.chem.drug.*})

Key: * = truncation symbol.

(All searches were in the keywords field of Reference Manager 9.5

N, ISI ResearchSoft).

The reference lists of relevant articles were searched for other pos-

sible relevant trials.

A further search (by first author, in MEDLINE and EMBASE)

was done in July 2003, to see if any of the abstracts identified in

2002 had reached full publication.

We did not impose any language or other restrictions on any of

these searches.

Data collection and analysis

Screening for eligibility

Trials underconsideration for inclusion in the review wereassessed

independently for their appropriateness by two authors without

prior consideration of their results. Any disagreements that were

not resolved by discussion were considered by a third person. Ex-

cluded studies are listed with reasons for their exclusion.

Assessment of methodological quality

The authors independently madean assessmentof methodological

quality using the Incontinence Groups quality assessment tool,

which includes evaluation of quality of random allocation and

concealment, description of dropouts and withdrawals, analysis by

intention to treat, and blinding during treatment and at outcome

assessment. Disagreements wereresolvedby discussion witha third

person.

Data extraction

Data were independently abstracted by at least two authors and

cross-checked. Where data were collected but were not reported,

or were reported in a form not suitable for inclusion in the formal

analysis, further clarification was sought from trialists.

Data analysis

Included trial data were processed as described in the Cochrane

Reviewers Handbook (Alderson 2004).

A priori subgroup analyses were planned to investigate the effects

of age, sex, severity of symptoms, and cause of overactive bladder

symptoms (i.e. idiopathic versus neurogenic).

Data plots were examined for evidence of heterogeneity (dissim-

ilarity), and a formal (statistical) test of heterogeneity was used.Where heterogeneity was observed (based on I-squared and the

test for heterogeneity) an explanation was sought and offered in

the text. Where three or more trials contributed to a single data

plot, the data were reanalysedafter removal of the trial(s) that were

the apparent cause of the dissimilarity. The secondary analysis is

reported in the text. Where clinically important (but not statisti-

cally significant) heterogeneity was observed, this is also noted in

the text.

R E S U L T S

Description of studies

See: Characteristics of includedstudies; Characteristicsof excluded

studies.

Included/excluded studies

Fifty-seven trial reports were identified. Eight were excluded (

Appell 1997; Gaudenz 1978; Larsson 1999; Mundy 2001; Rosario

1995; Tincello 2000; Wein 1999; Yoon 2001), and the reasons

are listed in the Characteristics of excluded studies table. Forty-

nine trials were included in the review; 39 parallel designs and 10crossover designs. One trial report contained data from two paral-

lel designs (Sussman 2002). The trials recruited a total of 11,332

adults, including 7,915 women and 2,190 men (NB: some trials

did not report sex data). Sample sizes ranged from 10 (Di Stasi

2001a) to 1529 (VanKerrebroeck 2001).

Trials made three types of comparison: comparisons of different

anticholinergic drugs, comparisons of different doses of the same

drug, and comparisons of immediate (IR) versus extended release

(ER) preparations. Some trials had more than two arms, and made

more than one comparison. Trials that compared an ER prepara-

tion of one drug with an IR preparation of another were included

in the comparison of different anticholinergic drugs, and the com-

parison of IR versus ER preparations. Trials that compared ERpreparations of different drugs were included in the comparison of

different anticholinergic drugs, and the comparison of ER prepa-

rations.

Two trials were published in German (Froehlich 1998; Wehnert

1992) and one in Flemish (Kramer 1987). Information and data

were abstracted from the original papers.

Sample characteristics

Parallel arm studies




9/109

Inclusion and/or exclusion criteria were not always well defined.

People with a presumed or urodynamic diagnosis of idiopathicdetrusor overactivity were included in five trials (Abrams 1998;

Chapple 2002; Froehlich 1998; Madersbacher 1999; Rentzhog

1998), and six trials restricted entry to those with a presumed or

urodynamic diagnosis of neurogenic detrusor overactivity(Abrams

1996; Gajewski 1986; Madersbacher 1995; Osca 1997; Stohrer

2002; VanKerrebroeck 1998). Thirteen trials included people

with urodynamic detrusor overactivity (both types, or type not

further specified) (Birns 2000; Chaliha 1998; Davila 2001b;

Drutz 1999; Jacquetin 2001; Jonas 1997; Junemann 2000; Le-

ung 2001; Mazur 1995; Millard 1999; Salvatore 1995; Thuroff

1991; VanKerrebroeck 1997). In the remaining trials partici-

pants had symptoms consistent with overactive bladder syndrome

(Anderson 1999; Appell 2001; Davila 2001a; Diokno 2003;Dmochowski 2003; Hofner 2000; Homma 2002; Junemann

1999; Lee2001; Malone-Lee 2001a; Malone-Lee2001b; Sussman

2002; VanKerrebroeck 2001; Versi 2000). Mean age of the partic-

ipants ranged from 31.3 years (no standard deviation) to 75 years

(no standard deviation). Two trials restricted entry to older people

(Malone-Lee 2001a) (65 years or more); (Malone-Lee 2001b) (50

years or more). Apart from contraindications to anticholinergics,

the most common exclusion criteria were evidence of voiding dys-

function or bladder outlet obstruction.

Crossover studies

Five studies included people with urgency or urge incontinence (Bagger 1985; Holmes 1989; Kramer1987;Wehnert1992; Zeegers

1987). These people usually had urodynamics as well but it was

not one of theselection criteria. The other five required a diagnosis

of detrusor overactivity (Burton 1994; Di Stasi 2001a; Di Stasi

2001b; Massey 1986; Nilsson 1997). Three studies did not report

age; in three other studies the mean age ranged from 42 to 54

years, one study had a range of ages of 17 to 55 years and another

16 to 78 years. There were a variety of exclusion criteria.

Interventions

Parallel arm studies

In many trials treatment was preceded by a washout period, or

treatment with co-medications was specifically excluded. This in-

formation was not given in 15 of the 39 parallel arm studies, but

given the short half life of the drugs this may not be important (

Appell 2001; Birns 2000; Chaliha 1998; Davila 2001a; Diokno

2003; Drutz1999; Gajewski 1986; Hofner2000;Junemann 1999;

Lee2001; Leung2001; Osca 1997; Salvatore 1995; Sussman 2002

(two studies)).

The following between drug comparisons were made:

(1) oxybutynin versus: (a) tolterodine (Abrams 1998;Appell 2001;

Dmochowski 2003; Diokno 2003; Drutz 1999; Homma 2002;

Lee 2001; Leung 2001; Malone-Lee 2001b; VanKerrebroeck

1997), (b)trospium (Froehlich1998; Hofner 2000; Madersbacher1995; Osca 1997), (c) propantheline (Gajewski 1986; Thuroff

1991), and (d) propiverine (Madersbacher 1999; Stohrer 2002);

(2) tolterodine versus: (a) oxybutynin (see above), (b) trospium (

Junemann 2000), and (c) solifenacin (Chapple 2002).

Trials also compared different doses of the same drug:

(1) tolterodine (Abrams 1996; Jacquetin 2001; Jonas 1997;

Malone-Lee2001a; Millard 1999; Rentzhog 1998; Sussman 2002;

VanKerrebroeck 1998);

(2) oxybutynin (Davila 2001a; Salvatore 1995; Sussman 2002);

(3) trospium (Chaliha 1998; Junemann 1999);

(4) propiverine (Mazur 1995);

(5) solifenacin (Chapple 2002).

Comparisons of IR and ER preparations included:(1) IR versus ER oxybutynin (Anderson 1999; Birns 2000; Davila

2001b; Versi 2000);

(2) IR versus ER tolterodine (VanKerrebroeck 2001);

(3) IR oxybutynin versus ER tolterodine (Homma 2002);

(4) IR tolterodine versus ER oxybutynin (Appell 2001).

There were also comparisons of different ER preparations:

(5) ER oxybutynin versus ER tolterodine (Diokno 2003;

Dmochowski 2003).

Treatment duration ranged from two weeks to three months in

nearly all studies. The exceptions were one trial that investigated

the effect of a single dose (Froehlich 1998), and two studies that

hadtreatment periods of one year ormore(Hofner 2000; Salvatore

1995).

Crossover studies

Some studies were preceded by a week-long placebo washout pe-

riod. Thelength oftreatment variedfromone dose (DiStasi2001a;

Di Stasi 2001b) to 60 days (Nilsson 1997) with a median length

of three weeks. For the one dose studies the washout period was

one week. Six of the 10 studies had no washout period. Three

washout periods were of one week and the other two weeks. Given

the short half life of the drugs used a washout period may not be

important. Four of the studies had twoarms, three had three arms,

one had four arms and two had six arms. Some arms werethe samedrug at different doses. Some studies had different combinations

of drugs so that they, for example, tested four drugs in a three

arm study (Kramer 1987) and some had even more complicated

arrangements with different doses (Massey 1986).

Two studies compared oxybutynin with emepronium (Kramer

1987; Zeegers 1987), one oxybutynin with propiverine (Wehnert

1992), one oxybutynin with propantheline (Holmes 1989), one

emepronium at different doses (Massey 1986), one immediate re-

lease with controlled release oxybutynin (Nilsson 1997), one oxy-

butynin three times a day with as needed (Burton 1994), and two

oral oxybutynin versus intravesical oxybutynin versus intravesical

oxybutynin with an electric current applied (Di Stasi 2001a; Di




10/109

Stasi 2001b).

Outcome measures

Overall there was a lack of consistency in the choice of outcome

measures by trialists, and a lack of consistency in the way data

werereported.Fourteen trials reported outcomesof interest but no

useable data were provided (Bagger 1985; Burton 1994; Chaliha

1998; Chapple 2002; Davila 2001a; Di Stasi 2001a; Di Stasi

2001b; Leung 2001; Massey 1986; Osca 1997; Salvatore 1995;

Stohrer 2002; Wehnert 1992; Zeegers 1987). Due to deficiencies

in datareporting (e.g. point estimate without measure of variation)

many trials contributed little or no data to the review. The lack of

similarity in measures limited the possibilities for combining datafrom individual trials.

The primary outcome of interest in the review was quality of

life. Validated incontinence specific quality of life measures were

only reported by four trials (Davila 2001a; Dmochowski 2003;

Leung 2001; VanKerrebroeck 2001). The Incontinence Impact

Questionnaire (IIQ) and the Kings Health Questionnaire (KHQ)

were each used by two trials, but no useable data were reported

(see results). Therefore, the authors chose the patients perception

of cure/improvement as an alternate primary outcome. About a

third of trials appeared to collect data on the patients perceptions

of symptomatic cure or improvement; but fewer reported these

findings.

About half the studies used micturition diaries to record num-ber of leakage episodes and number of micturitions over varying

lengths of time. In order to combine these data in the pooled anal-

ysis the number of leakage episodes and number of micturitions

in 24 hours was calculated. A wide range of urodynamic mea-

sures were reported. In view of the lack of correlation between

urodynamic measures and clinical outcome the formal compar-

isons were limited to maximum cystometric capacity, volume at

first contraction and residual volume. These were measures that

trialists most consistently reported, whichsuggests that researchers

and clinicians have thought these data were important in making

judgements about the effects of the drugs on overactive bladder.

Some authors reported post treatment urodynamic measures, and

others reported change in urodynamic measures, so both are in-cluded in the formal comparisons. Dry mouth was the adverse

event reported by the largest number of studies. No trial included

socioeconomic data.

In order to use the data from crossover studies in a meta-analysis

it must be presented as the mean and standard deviation of the

difference between two treatments for continuous data, or a two

by two table for binary data, as the correlation between measure-

ments on the same individual may be important. Only three of

the 10 crossover studies presented data in this way (Holmes 1989;

Kramer 1987; Nilsson 1997), Holmes for three of our predeter-

mined outcomes, and Kramer and Nilson for one each. These few

data have not been included in the pooled estimates.

Further characteristics of the trials are reported in the Character-

istics of Included Studies table.

Risk of bias in included studies

The method of group allocation was rarely described. Perhaps au-

thors of drug trials feel that a statement about single or double

blind design is sufficient to describe the randomisation process and

allocation concealment. While double blinding should adequately

conceal group allocation, this is not guaranteed. For the purposes

of the review, the 30 parallel arm studies that stated group alloca-

tion was double-blind were coded as having adequate allocation

concealment. In the other nine parallel armstudies it was not clear

if allocation was adequately concealed (Davila 2001a; Froehlich1998; Gajewski 1986; Homma 2002; Leung 2001; Mazur 1995;

Salvatore 1995; Sussman 2002). Only one trial specifically stated

that outcome assessors were blind to group allocation (Leung

2001). Some studies stated that the code was broken at the com-

pletion of the study, and in some it was specified that this was

after the analysis. This would imply that the final measurement

was done blinded.

Baseline comparability of the groups was not mentioned in 14

studies (Abrams 1996; Birns 2000; Chaliha 1998; Chapple 2002;

Davila 2001a; Davila 2001b; Hofner 2000; Junemann 1999;

Junemann 2000; Lee 2001; Leung 2001; Osca 1997; Rentzhog

1998; VanKerrebroeck 1997). The remaining trials stated that

the groups were comparable at baseline, although two studiesdid not provide supporting data (Homma 2002; Stohrer 2002).

In 17 trials the evaluation of treatment efficacy was conducted

on intention to treat principles (Abrams 1998; Appell 2001;

Birns 2000; Diokno 2003; Dmochowski 2003; Froehlich 1998;

Homma 2002; Jacquetin 2001; Junemann 1999; Madersbacher

1999; Malone-Lee 2001a; Malone-Lee 2001b; Millard 1999;

Sussman 2002; VanKerrebroeck 1997; VanKerrebroeck 2001).

Eight trials specifically stated that a per protocol analysis was used

to assess treatment efficacy (Abrams 1996; Anderson 1999; Drutz

1999; Hofner 2000; Junemann 2000; Rentzhog 1998; Stohrer

2002; VanKerrebroeck 1998). Baseline comparability is not an is-

sue of concern for crossover studies.

Thedescription ofwithdrawalsor dropouts wasnot adequate in 15trials (Abrams 1996; Chaliha 1998; Chapple 2002; Davila 2001a;

Davila 2001b; Dmochowski 2003; Homma 2002; Junemann

1999; Junemann 2000; Lee 2001; Leung 2001; Osca 1997;

Stohrer 2002; VanKerrebroeck 1998; VanKerrebroeck 2001).

There were no dropouts from the trials using single intravesi-

cal or oral doses of medication (Di Stasi 2001a; Di Stasi 2001b;

Froehlich 1998). In 16 trials the dropout rate was 10% or less (

Bagger 1985; Birns 2000; Chapple 2002; Davila 2001b; Homma

2002; Jacquetin 2001; Jonas 1997; Junemann 1999; Malone-Lee

2001a; Mazur 1995; Millard 1999; Rosario 1995; Thuroff 1991;

VanKerrebroeck 2001; Versi 2000; Wehnert 1992). In the re-

mainder, drop out rates ranged from 11% (Dmochowski 2003;




11/109

Madersbacher 1995) to 39% (Salvatore 1995) in parallel designs,

and 0% (Wehnert 1992) to 40% (Zeegers 1987) in the crossoverstudies. Very few parallel design trials included any follow up be-

yond the initial assessment of outcome. In those trials that did fol-

low up participants this was usually for very short periods of time,

i.e. one or two weeks. The stated purpose of these short follow up

periods was adverse events monitoring. One study followed par-

ticipants for two years(Salvatore 1995). Long term follow up does

not provide information on the differences in treatment effects for

crossover studies.

The primary, or only, reference for 15 trials was a conference ab-

stract (Abrams 1996; Burton 1994; Chaliha 1998; Chapple 2002;

Davila 2001a; Hofner 2000; Homma 2002; Junemann 1999;

Junemann 2000; Lee 2001; Leung 2001; Osca 1997; Salvatore

1995; Stohrer 2002; VanKerrebroeck1997). All ofthese trialswerecomplete at the time of reporting, and full publications were not

found with subsequent searching. All abstracts reported limited

details of method, and few results.

Some of the large multicentre/multinational trials were reported

in multiple publications. These publications usually presented

subsets of the main trial results (e.g. data from one country)

and subsequent publications rarely provided further methodolog-

ical detail. The most notable example is Van Kerrebroek 2001 (

VanKerrebroeck2001); 18 separate reports were identified. Where

there were multiple publications of the same trial the main trial

report was selected as the primary reference; the primary reference

is cited throughout the review for simplicity.

Effects of interventions

Very few of the crossover studies presented any data in a manner

that could be formally combined in a meta-analysis. Because of

this the results of the crossover studies are qualitatively assessed,

after the results for the parallel arm studies are presented.

One anticholinergic versus another

1. Oxybutynin versus tolterodine

Ten parallel arm studies were included ( Abrams 1998;

Appell 2001; Diokno 2003; Dmochowski 2003; Drutz 1999;

Homma 2002; Lee 2001; Leung 2001; Malone-Lee 2001b;

VanKerrebroeck 1997). The following comparisons were made by

one trial each: IR tolterodine versus ER oxybutynin, ER toltero-

dineversus IR oxybutynin, ER oraltolterodine versus ER transder-

mal oxybutynin, and ER oral oxybutynin versus ER oral toltero-

dine. The other six trials compared immediate release oral prepa-

rations of tolterodine and oxybutynin. Leung et al (Leung 2001),

reported as a conference abstract, gave no useable data.

Primary outcome measure (01.01)

Two trials collected quality of life data (Dmochowski 2003; Leung

2001), none of which was useable. Two trials reported the pa-

tients perception of change in symptoms (Abrams 1998; Malone-

Lee 2001b). There was no statistically significant difference be-

tween the groups in the proportion of people reporting cure/im-

provement (145/307, 47% cured/improved in tolterodine group

and 136/306, 44% cured/improved in oxybutynin group: RR for

cure/improvement 1.06, 95% CI 0.89 to 1.26). Patient reported

data were collected in three further trials, but were not included

in the pooled analysis because they were not dichotomous data

(i.e. were not either/or data such as cure/no cure), were not able to

be dichotomised by the authors, or were not reported (Lee 2001;

Leung 2001; VanKerrebroeck 1997).

Secondary outcome measures (01.02, 01.03, 01.04, 01.05,

01.12, 01.13)

One trial ( Appell 2001), reported data on number of leakage

episodes in 24 hours; there was no statistically significant dif-

ference between IR tolterodine and ER oxybutynin groups. Five

trials reported the change in number of leakage episodes in 24

hours (Abrams 1998; Dmochowski 2003; Drutz 1999; Lee 2001;

Malone-Lee 2001b). There was no statistically significant differ-

ence between the groups (WMD -0.15, 95% CI -0.47 to 0.16).

Of the five trials, only one (Lee 2001), found a statistically signif-

icant difference between the treatments and this appears to havecontributed to the statistically significant heterogeneity observed

in this comparison. Lee et al (Lee 2001) was reported as a confer-

ence abstract; the minimal detail available meant it was difficult

to tell why this trial might be different from the others.

The same five trials reported change in micturitions in 24 hours;

there was no statistically significant difference between the groups

(WMD -0.25, 95% CI -0.61 to 0.10). Again, Appell et al (Appell

2001) reported data on number of micturitions in 24 hours. There

was a statistically significant difference in favour of the ER oxybu-

tynin group.

There were no useable data available for any of the three clinicians

measures, maximum cystometric capacity, volume at first contrac-

tion, or residual volume.Six trials reported withdrawals due to adverse events (Abrams

1998; Diokno 2003; Dmochowski 2003; Drutz 1999; Lee 2001;

Malone-Lee 2001b). Therewere fewer withdrawals amongst those

taking tolterodine (72/1051,7% tolterodine group and 126/1046,

12% oxybutynin group, RR 0.57, 95% CI 0.43 to 0.75). Statis-

tically significant heterogeneity was observed in this comparison.

In Dmochowski et al Dmochowski 2003) some women withdrew

because of reactions at the transdermal application site. This was

in addition to those who withdrew because of anticholinergic side

effects. If data from Dmochowski et al (Dmochowski 2003) are

removed from the analysis there is still less risk of withdrawal due

to adverse events in the tolterodine group (RR 0.62, 95% CI 0.47




12/109

to 0.82), but there is no longer statistically significant heterogene-

ity.Dry mouth was the most frequently reported side effect; data were

available from nine studies (Abrams 1998; Appell 2001; Diokno

2003; Dmochowski 2003; Drutz 1999; Homma 2002; Lee 2001;

Malone-Lee 2001b; VanKerrebroeck 1997). Statistically signifi-

cant heterogeneity was observed in this comparison. If the data

from trials that compared IR with ER preparations (Appell 2001;

Homma 2002), in addition to comparing tolterodine with oxybu-

tynin, were removed from the analysis heterogeneity was reduced,

but was still statistically significant. If data comparing transdermal

and oral methods of delivery were also removed the heterogeneity

was no longer statistically significant. There was about half to a

third the risk of dry mouth for those taking tolterodine (RR for

dry mouth 0.60, 95% CI 0.54 to 0.66).

2. Oxybutynin versus trospium

Four parallel arm studies were included (Froehlich 1998; Hofner

2000; Madersbacher 1995; Osca 1997). Froehlich et al (Froehlich

1998) compared intravesical administration of single doses of tro-

spium and oxybutynin. The other three trials compared IR oral

preparations of trospium and oxybutynin. Osca et al (Osca 1997),

reported as a conference abstract, gave no useable data.

Primary outcome measure

None of the four trials collected data on the patients observations

of cure/improvement.

Secondary outcome measures (01.06, 01.10, 01.12, 01.13)

Only one trial collected data on leakage episodes and micturi-

tions in 24 hours (Hofner 2000), but data were presented without

a measure of variation. Two trials reported maximum cystomet-

ric capacity and residual volume (Froehlich 1998; Madersbacher

1995). Therewas no statistically significant difference between the

groups for either comparison. The confidence intervals for bothoutcomes in both trials were wide.

Madersbacher et al (Madersbacher 1995) found no statistically

significant difference between the groups for withdrawal due to

adverse events. Two trials reported data on dry mouth (Hofner

2000; Madersbacher 1995); those taking trospium had lower rates

of dry mouth (RR 0.74, 95% CI 0.59 to 0.93).

3. Oxybutynin versus propantheline

Two trials were included (Gajewski 1986; Thuroff 1991); both

compared IR oral preparations of propantheline and oxybutynin.

Primary outcome measure (01.01)

When pooled data from both trials were combined there was astatistically significant difference in proportions reporting cure/

improvement in favour of oxybutynin (RR 0.71, 95% CI 0.53 to

0.96).

Secondary outcome measures (01.06, 01.07, 01.08, 01.11,

01.12, 01.13)

Neither trial reported data on leakage episodes or micturitions

in 24 hours. One trial reported maximum cystometric capacity

post treatment (Gajewski 1986), the other change in maximum

cystometric capacity (Thuroff 1991). Neither showed a statisti-

cally significant difference between the treatments; the confidence

intervals were wide. Thuroff et al (Thuroff 1991) also reported

change in volume at first contraction and change in residual vol-

ume. There was no statistically significant difference between the

treatments for volume at first contraction, but residual volumes

were statistically significantly greater with oxybutynin.

There was no statistically significant difference in withdrawals due

to adverse events (RR 1.43, 95% CI 0.53 to 3.89). Thuroff et al (

Thuroff 1991) found no statistically significant difference between

the groups for dry mouth.

One crossover trial assessed this comparison (Holmes 1989). Al-

though the maximum cystometric capacity was greater during

treatment with oxybutynin than when on propantheline, there

were no significant differences in other outcomes of interest. This

study was small so the risk of a type II error must be high.

4. Oxybutynin versus propiverine

Two trials wereincluded(Madersbacher 1999; Stohrer2002);both

compared IR oral preparations of propiverine and oxybutynin.


Neither trial reported useable data on the patients observations of

cure/improvement.

Secondary outcome measures (01.06, 01.10, 01.12, 01.13)

Neither trial reported useable data on leakage episodes or number

of micturitions in 24 hours. Both reported maximum cystomet-

ric capacity post treatment; there was no statistically significant

difference between the groups (WMD -6.42, 95% CI -33.94 to

21.10). Madersbacher et al (Madersbacher 1999) did not find any

statistically significant difference in residual volume.

Based on data from a single trial (Madersbacher 1999) there was

no statistically significant difference between the groups for with-

drawals due to adverse events. Combined data from both trials

found statistically significantly fewer reports of dry mouth in those

taking propiverine (RR 0.78, 95% CI 0.66 to 0.92).




13/109

One crossover study assessed this comparison (Wehnert 1992).

None of the primary outcomes were statistically significantly dif-ferent. The maximum cystometric capacity was higher on oxybu-

tynin. This study was very small (n=10) so the risk of a type II

error was high.

5. Other

One trial compared trospium and tolterodine (Junemann 2000).

The trial was reported in a conference abstract. No useable data

were published. Another trial compared solifenacin (YM905) and

tolterodine (Chapple 2002). This trial was also reported as a con-

ference abstract, without any useable data.

Two crossover studies assessed the differences between oxybutynin

and emepronium (Kramer 1987; Zeegers 1987). Both found oxy-butynin to be better on most outcomes, but on a few they were

not statistically significantly different.

Different doses of tolterodine

It is worth noting that the three trials that contributed the bulk of

data in the following comparisons were all per protocol analyses

(Abrams 1996; Rentzhog 1998; VanKerrebroeck 1998). None of

the trials contributing to these comparisons collected quality of

life data, but as these were dose ranging (iePhase II) studies quality

of life data would not usually be collected.

1. 2 mg versus 0.5 mg

Three trials compared 0.5 mg and 2 mg IR oral tolterodine twice

daily (Abrams 1996; Rentzhog 1998; VanKerrebroeck 1998).


All three trials measured the patients perception of change in

symptoms, but the data could not be dichotomised to report cure/

improvement.

Secondary outcome measures(02.03, 02.05, 02.07, 02.09,02.11, 02.12, 02.13)

Based on data from two trials (Rentzhog 1998; VanKerrebroeck

1998) there was no statistically significant difference between the

groups for the change in number of leakage episodes (WMD 0.71,

95% CI -0.19 to 1.61), or change in number of micturitions per

24 hours (WMD 0.64, 95% CI -0.32 to 1.60).

All three trials reported the urodynamic measures. There was no

statistically significant difference between the two doses for changein maximum cystometric capacity (WMD -19.73, 95% CI -75.21

to 35.76), change in volume at first contraction (WMD -18.14,

95% CI -68.61 to 32.33), or change in residual volume (WMD -

23.71, 95% CI -50.88 to 3.45).

Only Rentzhog et al (Rentzhog 1998) reported withdrawals due

to adverse events; the single case was in the lower dose group. All

three trials published dry mouth data; there were fewer reports of

dry mouth in the lower dose group (RR 0.36, 95% CI 0.13 to

0.95).

2. 2 mg versus 1 mg

Seven trials compared 1 mg with 2 mg IR oral tolterodine twice

daily (Abrams 1996; Jacquetin 2001; Jonas 1997; Malone-Lee

2001a; Millard 1999; Rentzhog 1998; VanKerrebroeck 1998).

Malone-Lee et al (Malone-Lee 2001a) reported their data as me-

dians, with 95% CI, sothe data could not be enteredin the formal

comparisons (see Additional Tables) (Table 1; Table 2).

Table 1. Different doses. Change in leakage episodes in 24 hrs

Drug Study Dose/data Dose/data Difference

Tolterodine Malone-Lee (2001a) 1mg twice daily. Medianchange -0.3 (95% CI -0.8 to -

0.1) n=61

2mg twice daily. Medianchange -0.7 (95% CI -1.3 to -

0.2) n=73

Median difference -0.4 (95%CI not calculable)

Table 2. Different doses. Change in micturitions in 24 hrs





14/109

Table 2. Different doses. Change in micturitions in 24 hrs (Continued)

Tolterodine Malone-Lee (2001a) 1mg twice daily. Median

change -0.7 (95% CI -1.9 to

0.0) n=61

2mg twice daily. Median

change -0.7 (95% CI -1.1 to -

0.3) n=73

Median difference 0.0 (95%

CI not calculable)

Primary outcome measure(02.01)

Only Millard et al (Millard 1999) reported dichotomised cure/

improvement data. Those taking 2 mg twice daily were statistically

significantly more likely to report cure/improvement (RR 0.69;95% CI 0.53 to 0.89).


Based on data from four trials ( Jacquetin 2001; Millard 1999;

Rentzhog 1998; VanKerrebroeck 1998), there was no statistically

significant difference between the two doses for the change in

number of leakage episodes (WMD 0.22, 95% CI -0.21 to 0.64)

or change in number of micturitions in 24 hours (WMD 0.03,

95% CI -0.48 to 0.53).

Four trials reported the urodynamic measures(Abrams1996;Jonas1997; Rentzhog 1998; VanKerrebroeck 1998). There was no sta-

tistically significant difference for change in maximum cystomet-

ric capacity (WMD -16.90, 95% CI -44.73 to 10.93), change

in volume at first contraction (WMD -13.51, 95% CI -44.54 to

17.51), or change in residual volume (WMD -10.07, 95% CI -

24.49 to 4.34).

Pooled data from five trials (Jacquetin 2001 ; Jonas 1997; Malone-

Lee 2001a; Millard 1999; Rentzhog 1998) found no statistically

significant differencein the likelihoodof withdrawal dueto adverse

events (RR 0.70, 95% CI 0.36 to 1.40). All seven trials reported

dry mouth data; those taking the lower dose were less likely to

report dry mouth (RR 0.65, 95% CI 0.52 to 0.80).

3. 4 mg versus 2 mg

Four trialscompared 4 mgand 2 mgdosesof tolterodine. Three tri-

als used oral IR preparations twice daily (Abrams 1996; Rentzhog

1998; VanKerrebroeck 1998), and one used an oral ER prepara-

tion taken once daily (Sussman 2002).


Only Sussman et al (Sussman 2002) reported dichotomised cure/

improvement data. Those taking the higher dose were statistically

significantly more likely to report cure/improvement.


Based on data from two trials (Rentzhog 1998; VanKerrebroeck

1998) there was no statistically significant difference between the

twodoses for change in number of leakage episodes (WMD -0.25,

95% CI -1.32 to 0.82) or change in number of micturitions in 24

hours (WMD 0.18, 95% CI -1.03 to 1.40).

Three trials reported the urodynamic measures (Abrams 1996;

Rentzhog 1998; VanKerrebroeck1998). The change in maximum

cystometric capacity was statistically significantly greater with the

higher dose (WMD 73.83, 95% CI 18.05 to 129.59), but the

difference between the groups for change in volume at first con-

traction favouring the higher dose was not statistically significant(WMD 47.72, 95% CI -10.80 to 106.25). The change in residual

volume was statistically significantly greater with the higher dose

(WMD 92.98, 95% CI 25.56 to 159.40).

Two trials (Rentzhog 1998; Sussman 2002) reported withdrawals

due to adverse events; there was no statistically significant differ-

ence between the groups (RR 0.91, 95% CI 0.54 to 1.54). How-

ever in the trial by Rentzhog et al (Rentzhog 1998), there were

no withdrawals due to adverse events in the lower dose group;

it is not clear how stable the calculation of the point estimate

and confidence interval is in these circumstances. Three trials re-

ported useable data ondry mouth (Abrams 1996; Rentzhog 1998;

VanKerrebroeck 1998); although there were more reports in the

higher dose group, the difference was not statistically significantlydifferent (RR 1.67, 95% CI 0.91 to 3.08).

Different doses of other anticholinergics

1. Oxybutynin




15/109

Salvatore et al (Salvatore 1995) compared two dose regimens of

oxybutynin. In one group the starting dose was 2.5 mg twice dailyincreasing over six weeks to a maximum dose of 5 mg three times

a day. In the other group the starting dose was 5 mg at night,

increasing over six weeks to a maximum of 5 mg three times a

day. The trial was reported as a conference abstract, but did not

include any useable data.

Davila et al (Davila 2001a) compared three doses of transdermal

ER oxybutynin. The trial was reported in conference abstracts;

there were no useable data.

Sussman and Garely (Sussman 2002) compared 5 mg and 10 mg

doses of once daily oral ER oxybutynin. Useable data were pre-

sented for two outcomes, which showed statistically significantly

more reports of cure/improvement with the higher dose (see Addi-

tional Table) (Table 3), with no statistically significant differencein proportion of withdrawals due to adverse events (see Additional

Table) (Table 4).

Table 3. Different doses. Cure/improvement


Oxybutynin Sussman (2002) 10mg oxybutyninER. 146/244 5mg oxybutynin ER. 103/259 RRfor difference 1.50 (95% CI

1.25 to 1.80)

Table 4. Different doses. Withdrawals due to adverse events


Oxybutynin Sussman (2002) 5mg ER. 39/313. 10mg ER. 46/307. RR 1.20 (95% CI 0.81 to 1.79)

Propiverine Mazur (1995) 15mg. 1/45 30mg. 1/47 RR 1.04 (95% CI 0.07 to 16.20)

45mg.1/49 30mg. 1/47 RR 0.96 (95% CI 0.06 to 14.90)

60mg. 3/40 30mg. 1/47 RR 3.53 (95% CI 0.38 to 32.57)




16/109

In a crossoverstudyBurton(Burton 1994) assessed oxybutynin

three times a day versus oxybutynin taken when needed. Patientpreference favoured taken as needed, as did reports of dry mouth.

There was no statistically significant difference in frequency, leak-

age episodes or maximum cystometric capacity.

2. Trospium

Juneman et al (Junemann 1999) compared 40 mg once daily oral

trospium with 40 mg twice daily. The trial was reported in a con-

ference abstract. Useable data were reported for changes in maxi-

mum cystometric capacity, volume at first contraction and residual

volume; there was no statistically significant difference between

the groups for any of these outcomes (see Additional Tables) (

Table 5; Table 6; Table 7). Chaliha et al (Chaliha 1998) compared

twice daily 10 mg, 20 mg and 40 mg oral trospium. This trial wasalso reported as a conference abstract, but no data were included

in the publication.

Table 5. Different doses. Change in maximum cystometric capacity


Trospium chloride Junemann (1999) 40mg twice daily. Mean

change 86.11 (SD 86.06)

n=56

40mg once daily. Mean

change 69.98 (SD 83.66) n=

56

Mean difference 16.13(95%

CI -15.31 to 47.57)

Table 6. Different doses. Change in volume at first contraction



change 135.41 (SD 130.95)

n=56


change 113 (SD 155.23)

n=56

Mean difference 21.75(95%

CI -31.44 to 74.94)

Table 7. Different doses. Change in residual volume



change 16.00 (SD 44.78)

n=56


change 14.59 (SD 41.00)

n=56

Mean difference 1.41 (95%

CI -14.49 to 17.31)




17/109

3. Propiverine

Mazur et al (Mazur 1995) compared four daily doses of oral

propiverine hydrochloride, 15 mg, 30 mg, 45 mg and 60 mg.

Those in the15 mg and30 mg groupstookmedication twice daily;

for the higher doses this necessitated taking medication three and

four times daily respectively. The review compares 15 mg, 45 mg

and 60 mg with 30 mg propiverine. Useable data on number of

micturitions in 24 hours, volume at first contraction, withdrawals

due to adverse events and dry mouth were presented. Those tak-

ing 15 mg or 60 mg had statistically significantly more voids in

24 hours than those taking 30 mg propiverine (Table 8). There

were no statistically significant differences between the groups for

volume at first contraction or withdrawals due to adverse events(Table 9; Table 4). There were statistically significantly fewer re-

ports of dry mouth in the 15 mg compared to the 30 mg dose

group, with no statistically significant difference for the other two

dose comparisons (Table 10).

Table 8. Different doses. Micturitions in 24 hours


Propiverine Mazur (1995) 15mg. Mean 9.5 (SD 4.7) n=45 30mg. Mean 7.3 (SD 2.4) n=45 Mean difference 2.20 (95% CI

0.66 to 3.74)

45mg. Mean 7.5 (SD 2.8) n=48 30mg. Mean 7.3 (SD 2.4) n=45 Mean difference 0.20 (95% CI -

0.86 to 1.26)

60mg. Mean 8.7 (SD 3.8) n=39 30mg. Mean 7.3 (SD 2.4) n=45 Mean difference 1.40 (95% CI

0.02 to 2.78)

Table 9. Different doses. Volume at first contraction


Propiverine Mazur (1995) 15mg. Mean 166 (SD 86) n=45 30mg. Mean 186 (SD 83) n=46 Mean difference -20.00 (95%

CI -54.74 to 14.74)

45mg. Mean 192 (SD 85) n=48 30mg. Mean 186 (SD 83) n=46 Mean difference 6.00 (95% CI -

22.96 to 39.96)

60mg. Mean 189 (SD 102)

n=39

30mg. Mean 186 (SD 83) n=46 Mean difference 3.00 (95% CI -

37.00 to 43.00)




18/109

Table 10. Different doses. Dry mouth


Propiverine Mazur (1995) 15mg. 3/45 30mg. 11/47 RR 0.28 (95% CI 0.08 to 0.95)

45mg.11/49 30mg. 11/47 RR 0.96 (95% CI 0.46 to 2.00)

60mg. 11/40 30mg. 11/47 RR 1.18 (95% CI 0.57 to 2.42)

4. Solifenacin (YM905)

Chapple et al (Chapple 2002) compared four doses of once daily

oral solifenacin, 2.5 mg, 5 mg, 10 mg and 20 mg. The trial was

reported in a conference abstract that contained no useable data.

5. Emepronium

Two crossover trials compared different doses of emepronium car-

rageenate. Bagger (Bagger 1985) compared 500 mg or 1000 mg

per day and found no statistically significant differences in number

of micturitions in 72 hours, leakage episodes in 72 hours, or dry

mouth. Massey and Abrams (Massey 1986) as part of a compli-cated trial compared 1600 mg with 2000 mg and found no signif-

icant differences in micturitions in 24 hours, leakage episodes per

24 hours, volume at first contraction, and maximum cystometric

capacity.

Extended versus immediate release preparations

1. Extended (ER) versus immediate release (IR) oxybutynin

Four trials were included ( Anderson 1999; Birns 2000; Davila

2001b; Versi 2000). Three compared oral ER with oral IR oxy-butynin (Anderson 1999; Birns 2000; Versi 2000). Davila et al

(Davila 2001b) compared transdermal ER with oral IR oxybu-

tynin.


Only Birns et al (Birns 2000) reported the patients perception of

cure/improvement; therewas no statistically significant difference.

Secondary outcome measures(03.02, 03.06, 03.08, 03.11,

03.12, 03.13)

All four trials collecteddata on leakage episodes, but only Davila et

al (Davila 2001b) reported useable data. There was no statistically

significant difference between the groups in the number of leakage

episodes in 24 hours post treatment. None of the trials reported

useable dataon number of micturitionsin 24 hourspost treatment.

Davila et al (Davila 2001b) found a lower maximum cystomet-

ric capacity and larger volume at first contraction in the ER, al-

though only the latter was statistically significant. There was no

statistically significant difference between the groups for change in

residual volume measured using ultrasound (Anderson 1999) and

urodynamics (Davila 2001b). The confidence intervals in both

studies were wide.All four trials reported withdrawals due to adverse events and dry

mouth. There was no statistically significant difference in with-

drawals due to adverse events between IR and ER groups. There

were no withdrawals due to adverse eventsin one arm of two trials;

it is not clear how stable the calculation of the point estimate and

confidence interval is in these circumstances. There were fewer

reports of dry mouth from those using ER preparations (RR 0.77,

95% CI 0.66 to 0.91). Statistically significant heterogeneity was

observed in this comparison. Visual inspection suggested that the

findings of Birns et al(Birns 2000) differed from those of the other

three trials, but it was not clear why this trial might be clinically

heterogenous.

One crossover trial assessed this comparison (Nilsson 1997). Therewas no statistically significant difference either in the number of

voids per day, nor the numbers with dry mouth.

2. ER versus IR tolterodine

Onetrialwasincludedinthiscomparison(VanKerrebroeck 2001).




19/109


Quality of life data were collected, but were not useable. No data

were collected on patients perception of cure or improvement.

Secondary outcome measures(03.03, 03.05, 03.12, 03.13)

Van Kerrebroeck et al (VanKerrebroeck 2001) did not find any

statistically significant differences between ER and IR tolterodine

for change in leakage episodes or change in micturitions in 24

hours. No urodynamic data were reported. There was no statis-

tically significant difference between the groups for withdrawals

due to adverse events, but there were fewer reports of dry mouth

for those using the ER preparation.

3. ER oxybutynin versus IR tolterodine

One trial was included (Appell 2001).


No data were collected on patients perception of cure or improve-

ment.


Appellet al(Appell 2001) didnot finda statistically significant dif-ference between ER oxybutynin and IR tolterodine for the num-

ber of leakage episodes in 24 hours, but there was a statistically sig-

nificant difference in favour of oxybutynin for the number of mic-

turitions in 24 hours. No urodynamic data were collected. There

was no statistically significant difference between the groups for

number of withdrawals due to adverse events, or for dry mouth.

4. ER tolterodine versus IR oxybutynin

One trial was included (Homma 2002), and was reported in a

conference abstract that did not include any useable data, apart

from dry mouth.



ment.

Secondary outcome measures

Data on leakage episodes and micturitions were collected, but no

useable data were reported. No urodynamic data were collected.

Homma et al (Homma 2002) found that the risk of dry mouth

was less for those taking ER tolterodine (03.13).

One extended release preparation versus another

1. Tolterodine versus oxybutynin

Two trials wereincluded(Diokno 2003; Dmochowski 2003). Dio-

kno et al (Diokno 2003) compared oral ER preparations; Dmo-

chowski (Dmochowski 2003) compared oral ER tolterodine with

transdermal ER oxybutynin.



ment.


Dmochowski et al (Dmochowski 2003) did not find any statisti-

cally significant difference in change in leakage episodes or mic-

turitions in 24 hours. Neither trial collected urodynamic data.

Therewas no statistically significant difference between the groups

for withdrawals due to adverse events, but statistically significant

heterogeneity was observed in this comparison. Dmochowski et al

(Dmochowski 2003) recorded two reasons for withdrawal due to

adverse events, anticholinergic side effects and/or reaction to the

skin patch. It may not be appropriate to combine these data.

Therewas no statistically significant difference between the groups

for dry mouth. Visual inspection of the data suggests clinical het-

erogeneity. Diokno et al (Diokno 2003) found statistically signif-icantly fewer reports of dry mouth with oral ER tolterodine than

oral ER oxybutynin. Dmochowski et al (Dmochowski 2003) did

not find a difference in risk of dry mouth between oral ER toltero-

dine and transdermal ER oxybutynin, but therewere fewer reports

of dry mouth with transdermal delivery. It is possible that there are

differences in dry mouth depending on the method of delivery.

Quality of life, socioeconomics, long-term follow up, and

adherence

Although quality of life was the primary outcome of interest only

four trials measured this using a validated instrument, and none

of them reported useable data (Davila 2001a; Dmochowski 2003;Leung 2001; VanKerrebroeck 2001). The trials by Davila et al (

Davila 2001a) and Van Kerrebroeck et al (VanKerrebroeck 2001)

had fourand threearms respectively, includinga placebo arm.Both

trials reported quality of life data for one of the active treatment

arms versus placebo, rather than for comparisons of active treat-

ments (e.g. one drug dose versus another, one type of preparation

versus another). Dmochowski et al (Dmochowski 2003) reported

change from baseline for only one (travel) of the four (physical

activity, social relationships, travel, emotional health) domains of

the IIQ. Leung et al (Leung 2001) gave no data in their abstract.

Notrial reported socioeconomic data. Therewere no dataavailable

evaluating outcomebeyond the end of the treatmentperiod, which




20/109

was usually short (i.e. two weeks to three months). Some trials

used pill counts to assess adherence to taking medication. Eighttrials reported per protocol analyses, i.e. data from those who were

adherent. There were not sufficient data in any comparison to be

sure that size or direction of treatment effect was different in these

trials.

Subgroup analysis

The planned subgroup analyses were not appropriate; there were

insufficient data in any comparison.

D I S C U S S I O N

This review is the second of a series of reviews of anticholinergic

drug therapy for overactive bladder symptoms and it should be

viewed in that context. The first anticholinergic review consid-

ered whetheranticholinergic drugs were better than placebo (Hay-

Smith 2003). Two further reviews will consider: 1) whether anti-

cholinergic drugs are better than other active (non-drug) therapies

(Patrick 2004); and 2) whether anticholinergic drugs are better

than other drug treatments (Dublin 2004).

General Observations

Considering this review as a whole, it seemed that oxybutynin

and tolterodine demonstrated similar efficacy, but those taking

tolterodine had less risk of withdrawal due to adverse events and

less risk of dry mouth. However, those taking tolterodine did not

always have less risk of dry mouth, because this also depended on

the method of drug delivery. There was less risk of dry mouth with

extended release preparations regardless of whichdrugs were being

compared. There was some evidence that, clinically (i.e. leakage

and micturition data), the effects of 1 mg versus 2 mg and 2 mg

versus 4 mg oral immediate release tolterodine twice daily were

similar but confidence intervals were wide. However, there was

less risk of dry mouth with 1 mg. There were insufficient data in

any other comparison in the review to observe any clear patternsof effect.

In contrast to many othertreatments for urinary dysfunction there

are a relatively large number of trials comparing anticholinergic

drugs with each other. In general, the reported method of the

parallel arm trials was of moderate to high quality. However, the

methods of group allocation were rarely described in sufficient

detail to be sure that group allocation was adequately concealed.

Only one trial clearly used blinded outcome assessors. The report-

ing of group allocation of dropouts and/or reasons for withdrawal

was adequate in about two thirds of trials. Crossover trials were

usually poorly reported. In order to be included in a meta-analysis

the results from crossover studies for continuous variables must be

reported as a difference in means and either the standard devia-

tion, standard error or 95% confidence interval of that difference.This does coincide with what is of clinical interest, but can get

complex when there are more than two treatments. Dichotomous

outcomes (e.g. cure/not cured) should be reported so that it is

clear whether the events on each treatment occurred in the same

or different people. Only three of the ten crossover trials reported

outcomes in this manner, but only one outcome of interest each.

In view of the number of trials, it is disappointing that it was not

possible to combine more data. There was considerable variation

in chosen outcomes, and also variation in how the same outcome

was measured and reported. Relatively few trials sought the pa-

tients opinion on satisfaction with, and acceptability of, treat-

ment but these are important factors in the choice of manage-

ment. Only four trials addressed quality of life, and none reportedsocioeconomic outcomes, and these areas need to be addressed in

future research.In view of the lack of validated normal values and

the lack of correlation between urodynamic findings and clinical

outcome, the emphasis in some studies on urodynamic measures

needs to be re-evaluated. In the vast majority of included trials,

the primary endpoint was measured after 12 weeks or less of treat-

ment. Overactive bladder syndrome is a chronic condition, and

anticholinergic drugs are not curative; continued use of the drug is

likely to be needed if the benefits are to be maintained. The short

duration of most studies and the lack of long term follow up gives

little information about the long term effects and acceptability of

the different anticholinergic therapies.

None of the included trials reported results for those w

HAY-SMITH - Which Anticholinergic Drug for Overactive Bladder Symptoms

Documents

Transcript of HAY-SMITH - Which Anticholinergic Drug for Overactive Bladder Symptoms