Pro-arrhythmic effects

Anticholinergic drugs suppress parasympathetic control of heart rate which is associated with an increased incidence of tachyarrhythmias and myocardial ischaemia.8 Although tachyarrhythmias are often considered benign, they confer an increased risk of embolic strokes and sudden death in susceptible cardiac patients.9 At the time of regulatory approval, tiotropium was known to be associated with mild decreased heart rate variability,10 which is a known risk factor for cardiac mortality and sudden cardiac death.11 The Lung Health Study investigators reported a RR for hospitalisations due to supraventricular tachycardia of 4.5 (95% CI 0.97 to 20.8) associated with use of short-acting inhaled ipratropium versus placebo.12 The accurate prevalence of tachyarrhythmias in randomised controlled trials (RCTs) of tiotropium has not been reliably ascertained because 24 h Holter ECG monitoring was conducted in only a few small studies,13 ,14 and resting 12-lead ECGs performed during clinic visits rarely detect them. However, despite these limitations, an increased risk of arrhythmias associated with tiotropium has been consistently reported in many clinical trials of both formulations.3 ,15–17 Furthermore, the large RCTs of ipratropium and tiotropium have shown a trend towards an increased risk of arrhythmias in comparison with placebo and other bronchodilator drugs. For example, the Understanding Potential Long-Term Impact of Tiotropium on Lung Function Trial (UPLIFT), a large 4-year study of the efficacy of the Spiriva Handihaler, reported relative risks of tachyarrhythmias of 3.70 (95% CI 0.79 to 17.4) and atrial tachycardias of 7.39 (95% CI 0.92 to 59) reported as adverse events.17 In a recent analysis of three placebo-controlled RCTs from the indacaterol development programme of patients who completed 24 h Holter ECG monitoring those randomised to tiotropium Handihaler had a RR of newly occurring, non-sustained ventricular tachycardia of 1.84 (95% CI 0.89 to 3.43).18

Serious ischaemic cardiovascular effects

In 2002 the Lung Health Survey reported that ipratropium increased the risk of cardiovascular death compared with placebo (RR 2.57; 95% CI 1.08 to 6.15; p=0.03).12 In 2008 a meta-analysis of 17 RCTs of inhaled anticholinergics demonstrated a statistically significant increased risk (RR 1.60; 95% CI 1.22 to 2.10) for the combined endpoint of stroke, myocardial infarction (MI) and cardiovascular deaths (including sudden death) with inhaled anticholinergic inhalers; the RR for all-cause mortality was 1.29 (95% CI 1.00 to 1.65).19 In 2011, a meta-analysis of five RCTs reported a statistically significant increased risk of all-cause mortality (RR 1.52; 95% CI 1.06 to 2.16) and cardiovascular death (RR 2.05; 95% CI 1.06 to 3.99) with tiotropium Respimat inhaler compared with placebo.20 In this meta-analysis there was a dose–response effect on all-cause mortality, which was primarily due to an excess of cardiac deaths. It was estimated that one excess death would be expected for every 124 patients treated with 5 μg of tiotropium administered via the Respimat for 1 year.

In the largest Respimat study, patients with COPD in the Respimat group with known rhythm and cardiac disorders at baseline had an especially high risk for cardiac death (RR 8.6; 95% CI 1.1 to 67.2 in those with rhythm disorders at baseline; RR 4.03; 95% CI 1.15 to 14.13 in those with cardiac disorders at baseline). This risk was not statistically significant in patients with baseline use of cardiovascular medications (RR 1.62; 95% CI 0.53 to 4.95) or a history of coronary artery disease (RR 1.08; 95% CI 0.15 to 7.67).17 This observation is particularly important when the findings of no significant risk of mortality (HR 0.89; 95% CI 0.79 to 1.02) or MI (RR 0.73; 95% CI 0.53 to 1.00) with tiotropium in the UPLIFT trial are considered. The UPLIFT trial excluded patients with COPD with unstable arrhythmias, a history of heart failure, or a heart attack within the previous 6 months. Furthermore, the UPLIFT trial did not monitor adverse events for over 40% of participants who terminated treatment. Sudden deaths were categorised as ‘general deaths’ and not included as cardiovascular deaths.17 These design limitations reduced the ability of the trial to ascertain ischaemic and arrhythmic events. It is notable that the RR for serious adverse events for angina reported as a serious adverse event was 1.44 (95% CI 0.91 to 2.26) with tiotropium in UPLIFT.3

The other significant recent study is the Prevention Of Exacerbations with Tiotropium in COPD (POET-COPD) trial21 ,22 Patients with MI or congestive heart failure within a year, severe cardiovascular disease, those with cardiac arrhythmias or those with moderate to severe renal failure (creatinine clearance >50 ml) were excluded from the trial. Although the trial was not adequately powered for individual cardiovascular endpoints an excess of angina pectoris (9/3707 vs 5/3669), myocardial ischaemia (11/3707 vs 6/3669) and MI events (20/3707 vs 13/3669) was noted with tiotropium Handihaler compared with salmeterol. This risk occurred despite a significant reduction in severe exacerbations of COPD with tiotropium.

In contrast to the sponsor's previous analysis of 29 placebo-controlled trials which reported an excess of stroke with tiotropium,17 UPLIFT reported no significant difference in the composite endpoint of number of patients with stroke with tiotropium (82 vs 80).16 This composite endpoint of all-cause stroke may have masked an imbalance in the potential risk of ischaemic stroke 16 In another pooled analysis of three placebo-controlled RCTs, inhaled tiotropium Handihaler was the only COPD drug (compared with formoterol or indacaterol) associated with a statistically significant increased risk of cardiovascular and cerebrovascular adverse events compared with placebo (RR 1.67; 95% CI 1.02 to 2.73).18

The findings above should be viewed in the context of the overall results of arrhythmias and ischaemic cardiovascular events and mortality occurring in long-term trials of inhaled anticholinergics (>24 weeks), as shown in online table 1, identified through systematic reviews and reviews of regulatory documents.17 ,19 ,20 A significant increase in overall mortality with tiotropium Respimat is largely due to cardiovascular deaths. The tiotropium Handihaler and Respimat studies are consistently characterised by an increase in supraventricular arrhythmias and myocardial ischaemia manifest as angina in several large trials, including UPLIFT and the POET-COPD trial.

None of these trials were designed to measure cardiovascular outcomes and several long-term trials such as UPLIFT and POET-COPD excluded participants with cardiovascular risk. Incompleteness of reporting and incomplete ascertainment bias all the estimates towards the null. The absence of formal statistical significance in underpowered safety studies should not be construed as proof of safety.23