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

Tiotropium for COPD

Bandolier looked at tiotropium for COPD (Bandolier 98) when the first two large trials were published. It reduced exacerbations and hospital admissions. Four years on we have a new systematic review [1], and it is interesting to see what has changed.


Systematic review

The review [1] used five search strategies, including Cochrane, and searching was up to the first quarter of 2006, making the review current. It also looked at bibliographies and requested studies from manufacturers. The review included studies in adults with stable COPD satisfying US and European criteria, where tiotropium was compared with placebo or long acting beta-agonists, that were randomised, and where the outcomes included exacerbations, hospital admission, and mortality. Only full publications were accepted.


Results

Thirteen studies with 6,078 patients were available, with 11 coming from the manufacturer (86% of patients). Most patients were men (80%), with FEV 1 between 34% and 43% of predicted. Almost all studies used tiotropium at an 18 μg dose inhaled once a day. Trials were as short as one week and as long as one year, and all but one had reasonably high reporting quality, and all were described as both randomised and double blind.

Eight trials compared tiotropium with placebo (Figure 1). Overall there were fewer exacerbations with tiotropium (relative risk 0.83; 95% confidence interval 0.76 to 0.91), and the number needed to treat to prevent one exacerbation was 21 (Table 1). The NNT was similar for three large trials lasting six months or longer, and five small trials lasting six weeks or less (Table 1), but event rates were much lower in the shorter trials. However, the average number of exacerbations per week with placebo was consistent between trials at about 1.3 (Figure 2).



Figure 1: Exacerbations in individual trials comparing tiotropium with placebo







Table 1: Results of trials comparing tiotropium with placebo, with outcomes of exacerbations, and of hospital admissions



Number of
Percent with outcome with
Outcome
Trials
Patients
Tiatropium
Placebo
NNT
(95% CI)
Exacerbations (all trials)
8
4279
27
31
21 (13 to 50)
Large trials ≥6 months
3
3552
31
36
21 (13 to 57)
Small trials ≤6 weeks
5
727
4.6
8.6
25 (13 to 270)
Hospital admissions
3
3552
7.6
13
20 (14 to 34)




Figure 2: The mean number of exacerbations per week of trials for placebo in individual trials. Symbol size is related to the number given placebo





Tiotropium also reduced hospital admissions in the three larger studies of at least six months duration (Table 1), with a similar NNT to prevent one hospital admission, of 20 (14 to 34). In trials of six months or longer, mortality was 1.7% for both tiotropium and placebo.

Most other outcomes were recorded in only a few trials, though spirometric testing in most trials consistently showed slight but significant improvement with tiotropium over placebo.

Comparison with long acting beta-agonists in two or three trials showed no significant difference for exacerbations, but lower hospital admissions with tiotropium, with an NNT to prevent one admission of 33 (17-1000), and significantly better results for spirometry.

Dry mouth was significantly more common with tiotropium than placebo or long acting beta agonist.


Comment

The estimates of efficacy from this review are broadly similar to those found in the original trials, despite there being five times more patients for the comparison with placebo. Tiotropium reduces exacerbations and hospital admissions compared with placebo, and probably against long acting beta-agonists. We could do with more comparisons with the latter to prove the point.

What was interesting was the comparison between larger, longer, trials, and smaller, shorter, trials. This could have been a source of clinical heterogeneity, or possible lack of validity, but actually the studies had a consistent rate of exacerbations with placebo, of about 1.3 per week, with only small studies varying much (Figure 2). This can provide a benchmark for knowing whether COPD candidates are like the patients in the studies.

For collectors of examples of how duplicate studies can creep into systematic reviews, this paper describes how subjects in six trials excluded from this review were included at least twice in a previous review.

Reference:

  1. GJ Rodrigo, LJ Mannini. Tiotropium for the treatment of stable chronic obstructive pulmonary disease: a systematic review with meta-analysis. Pulmonary Pharmacology & Therapeutics 2006 [epub ahead of print].

previous or next story