Eletrocardiographic Monitoring in COPD Patients Receiving Tiotropium

Abstract

Tiotropium is a once‐daily, inhaled anticholinergic for the treatment of chronic obstructive pulmonary disease that acts as a prolonged antagonist of the M3‐receptor. To ascertain whether electrophysiologic effects can be detected following tiotropium treatment in patients with chronic obstructive pulmonary disease, serial electrocardiograms were incorporated into multiple placebo‐controlled clinical trials including long‐term (6 and 12‐month) trials with tiotropium 18 mcg daily (n = 2,128) and a 4‐week dose‐ranging study with tiotropium up to 36 mcg daily (n = 169). In addition, 24‐hour electrocardiographic (Holter) monitoring was performed as part of a 6‐week, placebo‐controlled trial with tiotropium 18 mcg daily (n = 121). Electrocardiograms were performed before and up to 6 times during treatment in the 12‐month trials, and before and at the end of treatment in the 6‐month trials. For both the 12 and 6‐month trials, electrocardiograms were recorded as adverse events if significant changes occurred, and were retrospectively sent for centralized analysis. During the 6‐week trial, Holter monitoring was performed prior to the first dose and following 6 weeks of treatment. In all of these trials, no significant differences were observed in any of the electrocardiogram or Holter outcome parameters compared to placebo. Specifically, there was no clinically relevant difference in heart rate, atrio‐ventricular conduction or the occurrence of ventricular or supraventricular arrhythmias. In conclusion, tiotropium was not associated with any signs of cardiac safety concerns as defined by electrocardiographic evaluations in placebo‐controlled clinical trials.

• COPD
• Tiotropium
• Electrocardiograms
• Holter monitoring

Introduction

Inhaled bronchodilators are the recommended first‐line treatment for patients experiencing symptoms from chronic obstructive pulmonary disease (COPD) [[1]][[2]]. The advantage of inhaled products includes the ability to deliver medication directly to the target organ in small but clinically effective doses that also limit systemic exposure and hence adverse effects. Despite the selective delivery to the lungs of inhaled airway medications, there is systemic bioavailability of the minute doses generally used. As a result, systemic effects, albeit infrequent, can be observed. For example, inhaled beta‐agonist administration may result in tremor, hypokalemia and tachyarrhythmias [[3]]. Inhaled corticosteroids have also been associated with systemic effects [[4]]. Hence, it is reasonable to critically evaluate the potential systemic effects of virtually all inhaled medications.

Inhaled anticholinergics such as ipratropium (Boehringer Ingelheim, Ridgefield, CT, USA) are bronchodilators commonly recommended for the treatment of COPD [[1]][[2]]. Systemically administered anticholinergic medications can result in pharmacologically predictable effects on the central nervous system, gastrointestinal tract, genitourinary system, eyes and the cardiovascular system including tachycardia and a decrease in the ventricular effective refractory period [[5]][[6]][[7]]. The development of an anticholinergic with an N‐quaternary structure such as ipratropium bromide results in the ability to deliver an antimuscarinic effect on bronchial smooth muscle (i.e., relaxation) while limiting systemic effects due to minimal gastrointestinal absorption and failure to cross the blood‐brain barrier [[8]]. Nevertheless, anticholinergic effects such as dry mouth can occur [[9]]. In addition, it is possible that tachyarrhythmias might occur although there has been a long history of widespread use with ipratropium with minimal reports of such events [[10]][[11]][[12]].

Tiotropium (Boehringer Ingelheim, Ridgefield, CT, USA) is a new, inhaled anticholinergic with a 24‐hour duration of action that has its effect through prolonged M3‐receptor antagonism [[13]]. Tiotropium was delivered as a dry powder capsule with the HandiHaler® device in the clinical trials mentioned below [[14]]. Long‐term trials over 12 and 6 months with tiotropium have demonstrated its efficacy as a once‐daily maintenance therapy [[15]][[16]][[17]][[18]]. Tiotropium has been shown to have greater clinical efficacy compared to ipratropium in clinical outcomes such as bronchodilation, dyspnea, health status and exacerbations [[16]]. Pharmacologically, tiotropium differs from ipratropium with regard to potency and the dissociation half‐life from muscarinic receptors. It is more potent and the dissociation half‐life from M3‐receptors is approximately 35 hours for tiotropium and 0.25 hours for ipratropium [[13]].

The clinical development program for tiotropium included electrocardiographic evaluations in long‐term studies of 6 months and 12 months duration, and in a 4‐week, multiple‐dose, dose‐ranging study. In addition, the 24‐hour duration of effect was confirmed in a 6‐week trial during which patients were required to remain in the clinic over a 24‐hour period for spirometric evaluation [[19]]. This study design provided the opportunity to evaluate the effects of tiotropium on heart rate and rhythm using 24‐hour Holter monitoring. The large database of electrocardiographic (ECG) evaluations from these various studies assists in the evaluation of potential effects that tiotropium may have on cardiac electrophysiologic parameters. The present report summarizes this clinical ECG database to look for signs of cardiac risk associated with tiotropium in patients with COPD.

Materials and Methods

ECGs were part of safety monitoring throughout the clinical development program of tiotropium. The methods for the trials included in this report are summarized below. In all COPD clinical trials, patients were required to have a diagnosis of COPD according to the American Thoracic Society definition [[2]]. Patients were required to be at least 40 years of age, have at least a 10‐pack year history of smoking, an FEV₁ ≤ 65% predicted and to have stable disease. Patients were excluded if they had a history of asthma, atopy, allergic rhinitis or a recent respiratory tract infection. The only cardiac exclusions were patients with a recent history (i.e., one year or less) of myocardial infarction, a history of any cardiac arrhythmia requiring drug therapy or patients who have been hospitalized for heart failure within the past 3 years.

Twelve‐Month and Six‐Month Trials

Two 12‐month, multicenter, randomized, double‐blind, placebo‐controlled, parallel group trials evaluated tiotropium 18 mcg daily [[15]]. The 2 protocols had identical inclusion and exclusion criteria as well as study designs. A standard 12‐lead ECG was performed on all patients at baseline and repeated at 3 month intervals throughout the study period.

Two 6‐month trials were conducted to compare tiotropium 18 mcg daily to salmeterol 50 mcg bid metered dose inhaler (GlaxoSmithKline, Research Triangle Park, NC, USA) and matching placebos in patients with COPD. The trials were multinational and used a double‐blind, double‐dummy, parallel group design [[17]][[18]]. These trials were nearly identical with the only difference between the protocols being the duration of the pulmonary function testing (3 vs. 12 hours) on clinic test days (baseline and at 2, 8, 16 and 24 weeks of treatment). ECGs were performed upon entry and at completion of the trial.

For the 12‐month and 6‐month trials, the initial interpretations of the ECGs were performed by the investigators or their designees. The investigators reported ECG abnormalities as adverse events if a change from the baseline ECG was observed that was: considered clinically significant in the medical opinion of the investigator, required treatment, or lead to discontinuation of the study drug.

Retrospectively, ECGs from the tiotropium and placebo arms of the 12‐month and 6‐month studies were sent to a central laboratory (eResearchTechnology, Inc., Philadelphia, PA, USA) for a high‐resolution digital manual measurement of the cardiac intervals and morphological changes. ECG measurements were performed by experienced technicians using digitization software with magnification of the ECG and point‐to‐point determination on the digitizing pad experienced technicians and by a centralized cardiologist who was blinded to the tracings.

The analyses of the ECG intervals of heart rate, PR, QRS, QT, QTcB and QTcF included change from baseline to the average of on‐treatment values for a given patient and change from baseline to the maximum of on‐treatment values for a given patient. All of the listed parameters were determined from individual time points. For all parameters, each patient was tallied as having at least one notable change or no notable changes. Outlier analyses for heart rate, PR and QRS used the following criteria:

• Heart Rate (HR) changes reflecting a 25% decrease from baseline to a HR < 50 bpm (bradycardic event) or a 25% increase from baseline reflecting a HR > 100 bpm (tachycardic event)

• PR change from baseline: ≥ 25% when PR > 200 msec vs. ≤ 200 msec

• QRS change from baseline: ≥ 25% when QRS > 100 msec vs. ≤ 100 msec

For QT parameters, change from baseline to an average of on‐treatment values for a given patient and a determination of those patients who attain QT values > 500 msec when not present at baseline (new onset) were assessed. The calculated QTc interval was determined using both Bazett's (QTcB) and Fridericia's formula (QTcF) [[20]][[21]]. QTc changes were categorized as follows: < 30 msec, 30–60 msec and > 60 msec in QTc from baseline (described as the proportion of patients meeting these criteria for each treatment group). The reproducibility of this system was 1.4 ± 2.2 milliseconds (msec).

Multidose Dose‐Ranging Trial

In a 4‐week, placebo‐controlled, parallel group, multidose, dose‐ranging trial, a 12‐lead ECG and a 2‐minute rhythm strip were performed prior to drug administration: at 1, 3 and 5 hours post drug administration, and after 1, 2 and 4 weeks of treatment [[22]]. These ECGs were electronically transmitted to a central monitoring laboratory (MERX Medical Diagnostics, Reno, NV, USA) and all were read and evaluated by cardiologists. The 15 ECG tracings and rhythm strips performed following study drug administration were compared to the baseline tracings and rhythm strips performed immediately prior to the first administration of the study drug. Any changes seen were characterized using the following terms: transient (seen once), intermittent (seen occasionally), stable (seen consistently post treatment), and progressive (worsening over the course of the trial).

Holter Monitoring

A 6‐week, double‐blind, parallel‐group, placebo‐controlled study was conducted to determine whether the time of dosing (morning [9 a.m.] or evening [9 p.m.]) at steady state influenced the bronchodilator response of tiotropium 18 mcg daily in patients with COPD [[19]]. As part of the protocol, in addition to the standard 12‐lead ECG performed on all patients at baseline and repeated at the end of the study, 24‐hour ECG (Holter) monitoring was conducted on all patients at Day 0 and after 6 weeks on medication. The initial Holter monitoring included 2 hours of monitoring following the first dose. However, this post‐dose period was excluded from the baseline (Day 0) analysis presented in this report. Holter monitoring was repeated after 6 weeks of drug treatment. On study days when Holter monitoring was performed, patients stayed in the facility for the 24‐hour monitoring period.

Results

In total, there were 2,418 patients in the trials. From the total population, 1,167 patients were randomized to receive tiotropium, 405 were randomized to receive salmeterol and 846 were randomized to receive placebo. Overall, there was a predominance of men with the population having a mean age of approximately 65 years. On average, FEV₁ was approximately 40% predicted.

Twelve‐Month and Six‐Month Placebo Controlled Trials

As previously indicated, abnormalities in the ECGs were recorded as adverse events only if they were considered clinically significant, required treatment or led to discontinuation of the study drug. In addition, ECGs were retrospectively obtained for centralized reading. The demographics of the tiotropium (n = 550) and placebo (n = 371) populations from the 12‐month trials and for the tiotropium (n = 402), salmeterol (n = 405) and placebo (n = 400) groups for the 6‐month trials were balanced across treatment arms within trials. Patients on average were approximately 65 years of age with approximately 65 to 75% being men. Mean FEV₁ was 1.02 L (38% predicted) in the 12‐month trials and was 1.09 L (39% predicted) in the 6‐month trials.

ECG Abnormalities Recorded as Adverse Events

In the 12‐month, placebo‐controlled trials, 14 patients developed ECG changes: 5 (0.9%) patients receiving tiotropium, and 9 (2.4%) patients receiving placebo. There were no statistically significant differences noted with regard to the type of ECG abnormalities between the tiotropium and the placebo group (Table 1). Supraventricular tachyarrhythmias were noted in one patient receiving tiotropium and 2 patients receiving placebo in the placebo‐controlled trials. There were no ventricular tachyarrhythmias. In the 6‐month studies, 19 patients developed clinically significant ECG changes: 8 (2%) tiotropium patients, 8 (2%) salmeterol patients and 3 (0.7%) placebo patients. There were no statistically significant differences noted with regard to the type of ECG abnormalities between the tiotropium and control groups (Table 1).

Table 1.  ECG Changes Reported as Adverse Events in the 12‐Month and the Six‐Month Studies

	12‐Month studies				6‐Month studies
	Tiotropium		Placebo		Tiotropium		Salmeterol		Placebo
	n	%	n	%	n	%	n	%	n	%
Total treated	550	(100.0)	371	(100.0)	402	(100.0)	405	(100.0)	400	(100.0)
Total with any ECG adverse event	5	(0.9)	9	(2.4)	8	(2.0)	8	(2.0)	3	(0.7)
Extrasystoles	0	(0.0)	0	(0.0)	2	(0.5)	2	(0.5)	2	(0.5)
Atrial fibrillation	0	(0.0)	2	(0.5)	2	(0.5)	1	(0.2)	0	(0.0)
Multifocal atrial tachycardia	1	(0.2)	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)
Sinus tachycardia	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)	0	(0.0)
Myocardial infarction	0	(0.0)	2	(0.5)	2	(0.5)	2	(0.5)	0	(0.0)
Other changes	4	(0.7)	5	(1.5)	2	(0.5)	3	(0.7)	1	(0.2)

Centralized Reading of ECGs

In the 12‐month trials, there were 2,359 ECGs completed by tiotropium patients and 1,505 completed by the placebo patients. ECGs were obtained from clinical sites from approximately 94% of the participants in each treatment group. In the 6‐month trials, there were 390 ECGs completed by tiotropium patients, 379 ECGs completed by salmeterol patients and 364 completed by the placebo patients. ECGs were obtained from 92% of tiotropium, 90% of the salmeterol and 86% of placebo treated patients. The sample sizes for different parameters varied somewhat due to the ability or inability to adequately characterize the variable from the ECGs.

There was no indication of any relevant effects of tiotropium on heart rate, PR, QRS or QT intervals in the 12‐month or 6‐month trials (Tables 2 and 3). The only statistically significant difference was a minor difference in QRS duration in the 12‐month trials which was not seen in the 6‐month trials. The QTc interval by Bazett [[20]] and Fridericia [[21]] corrections gave comparable results. In the 12‐month trials, the incidence of an abnormal rhythm was 9% in the tiotropium group and 10.1% in the placebo group. A change in conduction was observed in 2.9% of the tiotropium group and in 4.2% of the placebo group. In the 6‐month trials, the incidence of an abnormal rhythm was 4.4%, 4.1% and 4.7% in the tiotropium, salmeterol and placebo groups respectively. A change in conduction was observed in 2.7%, 1.4% and 2.2% of the respective groups.

Table 2.  ECG Monitoring in the 12‐Month Placebo Studies

	Tiotropium	Placebo
Patients (n)	518	343
Heart rate
Mean change (bpm)	0.1	− 0.1
Maximum change (bpm)	6.7	6.2
% with bradycardia events	0.8	1.2
% with tachycardia events	2.5	2.0
PR
Mean change (msec)	1.2	1.0
% outliers	0.2	0.9
QRS
Mean change (msec)	1.0	0.1
% outliers	1.2	0.6
QT
Mean change (msec)	0.4	–0.1
% New QT > 500 msec	0	0
QTc by Bazett
Mean change (msec)	1.0	–0.2
Maximum change (msec)	14.3	13.1
% New QTcB > 500 msec	0.8	0.3
% Change of 30–60 msec	21.3	22.2
% Change > 60 msec	2.7	1.5
QTc by Fridericia
Mean change (msec)	0.6	− 0.3
Maximum change (msec)	11.8	11.1
% New QTcB > 500 msec	0.2	0.3
% Change of 30–60 msec	14.7	15.2
% Change > 60 msec	1.4	0.3

*Varies according to ability to characterize each parameter.

Table 3.  ECG Monitoring in the 6‐Month Salmeterol and Placebo‐Controlled Trials

	Tiotropium	Salmeterol	Placebo
Patients (n)	346	338	315
Heart rate
Mean change (bpm)	2.6	1.3	1.7
% with bradycardia events	0.9	0.3	0
% with tachycardia events	1.4	1.2	1.6
PR
Mean change (msec)	1.6	− 0.3	0.8
% Outliers	0.6	0	0.3
QRS
Mean change (msec)	1.4	0.5	0.3
% Outliers	1.4	0.6	0.3
QT
Mean change (msec)	− 4.5	− 1.7	− 4.0
% New QT > 500 msec	0	0	0
QTc by Bazett
Mean change (msec)	2.0	2.1	0.4
% New QTcB > 500 msec	0	0	0
% Change of 30–60 msec	9.0	9.8	9.2
% Change > 60 msec	1.4	2.4	1.0
QTc by Fridericia
Mean change (msec)	− 0.3	0.8	− 1.2
% New QTcF > 500 msec	0	0	0
% Change of 30–60 msec	6.4	6.8	5.1
% Change > 60 msec	0.3	0.9	0.3

*Varies according to ability to characterize each parameter.

Multidose Dose‐Ranging Trial

The demographics of the tiotropium and placebo populations were similar. The mean age was 65 years with 57% of patients being men. The mean FEV₁ was 1.08 L (42% predicted). All changes in heart rate, rhythm or conduction observed in the multidose, dose‐ranging study are recorded in Table 4. The numbers indicate the number of patients who had the associated abnormality on any ECG while on treatment. There was no statistically significant effect on heart rate, rhythm or conduction and, as expected for an anticholinergic agent, there were no effects on prolongation of QT interval. The changes seen were predominantly premature atrial and ventricular contractions. Most changes were either transient or intermittent. Only one patient exhibited progressive changes (placebo patient with borderline changes in the PR interval). One case of each of the following tachyarrhythmias was observed in the tiotropium treatment groups: sinus tachycardia (18 mcg dose), atrial fibrillation (9 mcg dose) and ventricular tachycardia (4.5 mcg dose). One case of sinus tachycardia was observed in the placebo group. All tachyarrhythmias were transient in nature.

Table 4.  Changes in Heart Rate, Rhythm or Conduction Observed in the Multidose Dose‐Ranging Study.

	Tiotropium bromide				Placebo
	4.5 mcg	9 mcg	18 mcg	36 mcg
Total patients treated (n)	34	33	33	34	35
Total patients with changes (n)	24	22	23	16	20
Borderline PR interval	0	1	0	0	1
First degree A‐V block	1	0	1	0	1
Atrial fibrillation	0	1	0	0	0
Sinus arrhythmia	0	0	0	1	0
Sinus tachycardia	0	0	1	0	1
Consistent with ventricular tachycardia	1	0	0	0	0
Repetitive premature systoles	2	0	2	0	1
Multiform premature systoles	4	2	1	2	0
Premature systoles, ventricular	11	7	10	8	9
Premature systoles, atrial	10	10	10	6	10
Sinus arrest or sino‐atrial block	2	2	0	0	0
Ectopic atrial rhythm	0	2	1	0	2
Borderline QT interval	0	0	1	0	1

*The numbers indicate the number of patients who had the associated abnormality on any ECG while on treatment.

Holter Monitoring

The populations in each group had similar demographics. The mean age was 67 years with 65% of the patients being men. The mean FEV₁ was 1.04 L (39% predicted). There was no evidence of an effect of tiotropium on heart rate or rhythm. Mean heart rates were similar before and after treatment in all groups (Table 5). For all 3 groups, there were no incidences of atrial flutter, atrial fibrillation, sustained ventricular tachycardia or ventricular fibrillation during the Holter monitoring. A small number of patients in all groups experienced episodes of supraventricular tachycardia and non‐sustained ventricular tachycardia (Table 5). Overall, there were no marked shifts in atrial or ventricular ectopic beat counts noted in any of the three treatment groups. There was one episode of second degree atrio‐ventricular block post‐treatment in the placebo group. There were no episodes of third degree atrio‐ventricular block.

Table 5.  Holter Monitoring Summary for Heart Rate and Number of Patients with Supraventricular Tachycardia (SVT) and Ventricular Tachycardia (V Tach)

		Tiotropium AM (n = 37)		Tiotropium PM (n = 35)		Placebo (n = 31)
		Baseline	6 Weeks	Baseline	6 Weeks	Baseline	6 Weeks
Heart rate (beats/min)	Mean	76	78	73	74	77	78
	Min/max	61/108	61/109	57/106	58/104	62/111	60/110
SVT	# Patients	7	7	5	5	3	7
V tach	# Patients	1	2	1	2	2	1
	Longest run	20	6	12	6	6	5

Discussion

Tiotropium is an inhaled anticholinergic medication for the treatment of COPD. Tiotropium has a unique pharmacokinetic profile with kinetic receptor subtype selectivity for the three muscarinic receptors that have been identified in the human lung (M1, M2, and M3). The dissociation half lives for each are 14.6, 3.6 and 34.7 hours for the M1, M2 and M3 receptors respectively [[13]]. The prolonged M3‐receptor antagonism is responsible for the clinical outcome of continuous bronchodilation throughout the day and night with once‐daily dosing [[19]]. The predominant muscarinic receptor in the heart is the M2 receptor which can decrease heart rate directly through stimulation of receptors in the atria and indirectly decrease contractility through stimulation of receptors in the ventricle [[7]]. The existence of other muscarinic receptor subtypes in the human heart is unclear. Any effect of an inhaled anticholinergic would therefore be predominantly mediated through antagonism of M2 receptors. As previously mentioned, the dissociation half‐life for tiotropium is shortest for the M2 receptor [[13]]. While the dose of tiotropium is only 18 mcg once daily, the clinical trial program incorporated extensive ECG monitoring to understand whether electrophysiologic effects could be observed. Centralized analysis of ECGs from over 2,000 patients participating in clinical trials of tiotropium 18 mcg once daily did not detect any effect of tiotropium on ECG intervals, heart rhythm or morphology. A 4‐week multiple‐dose, dose‐ranging study with serial ECGs did not reveal any association of abnormalities with doses of tiotropium up to 36 mcg daily. Furthermore, 24‐hour Holter data collected in a 6‐week trial did not find adverse effects on cardiac rhythm with tiotropium.

There is extensive documentation that inhaled medications are systemically bioavailable and even small quantities can have measurable systemic effects. Perhaps the most widely reported and most readily quantifiable are those with inhaled corticosteroids. Hypothalamic‐pituitary‐adrenal suppression has been observed in patients with airway disease with moderate to high doses of inhaled steroids [[23]][[24]]. In addition, changes in bone metabolism, easy bruising, skin thinning and possible ocular associations have been the subject of reviews [[4]]. Inhaled beta‐agonists can lead to cardiovascular effects that can be considered serious such as ventricular tachyarrhythmias and prolongation of QTc intervals. This may result directly from activation of beta‐receptors on the myocardium or indirectly through mechanisms such as hypokalemia, peripheral vasodilation and decreases in oxygenation [[25]][[26]][[27]]. This needs to be considered in the context of the intended population. Patients with COPD are at an increased risk of cardiovascular disease for multiple reasons including exposure to tobacco smoke [[28]]. Data on potential adverse effects should therefore be based both on the pharmacologic mechanisms and the population that will receive the drug. An interaction is suggested by the study by Cazzola et al. who observed adverse cardiac effects associated with long‐acting beta‐agonists in a population of COPD patients suffering from pre‐existing cardiac arrhythmias and hypoxemia [[29]]. Given the frequency of cardiovascular disease in COPD patients, objective data on cardiac effects of novel agents for COPD are warranted.

Anticholinergic agents antagonize muscarinic receptors that can be found in multiple organ systems in the body. Antagonism of muscarinic receptors have therapeutic value when considering effects such as mydriasis, cycloplegia, tachycardia and decreases in smooth muscle contraction in the bladder, gastrointestinal tract and airway. However, many of these same effects can also be considered unwanted and lead to morbidity. The N‐quarternary structure of ipratropium bromide has permitted administration of antimuscarinic agents to the airways leading to effective bronchodilation while limiting systemic bioavailability and the associated undesirable effects.

There has been over 25 years of experience with administration of inhaled N‐quarternary anticholinergic agents in man. Although the sample sizes are relatively small, studies of ipratropium which have included ECG rhythm strips and Holter monitoring have reported no serious electrocardiographic events or development of arrhythmias [[30]][[31]][[32]][[33]]. Studies in congestive heart failure did not detect a higher risk of arrhythmias associated with ipratropium [[34]][[35]][[36]].

The mechanism of action of tiotropium is similar to ipratropium in that they both antagonize muscarinic receptors [[13]]. In this regard, the types of systemic activities, including cardiac effects, with tiotropium would be expected to be similar to ipratropium. For example, dry mouth is the most commonly reported adverse anticholinergic effect observed with tiotropium and is also reported with ipratropium [[9]][[15]][[16]]. However, the differences in molecular structure are likely responsible for differing pharmacokinetic and pharmacodynamic profiles. Tiotropium has a higher potency compared with ipratropium, along with a more prolonged antagonism of muscarinic receptors which manifests in the clinical correlate of sustained bronchodilation of at least 24 hour duration [[13]][[16]][[19]]. Therefore, it is relevant to conduct independent evaluations of tiotropium rather than extrapolating all conclusions from ipratropium observations.

The ECG data generated from the tiotropium clinical trials, including thousands of ECGs, have provided reassuring information regarding the cardiac safety of tiotropium that is consistent with the experience with ipratropium. It is acknowledged that within the context of clinical trials, all clinical scenarios are not fully evaluated. There may be specific subpopulations with yet unidentified factors that, under unique conditions, may sensitize to anticholinergic mechanisms. While risk factors for cardiac disease were not exclusionary criteria and a retrospective review of the populations indicated that approximately 25% to 35% of patients had known cardiac disease at the time of randomization, potentially unstable cardiac patients (i.e. recent myocardiac infarction, recent hospitalization for cardiac failure) were excluded.

As with many chronic diseases, the treatment of symptomatic COPD involves the prescription of medications on a regular basis. The administration of medications requires an understanding of the pharmacologic effects that result in both the efficacy and the safety profile of the product in the intended population. It is well recognized that inhaled airway medications are, in general, extremely well tolerated. The present summary of the electrocardiographic findings in multiple clinical trials has shown an absence of effects on cardiac rhythm and conduction of tiotropium 18 mcg in patients with stable COPD. While this does not exclude the possibility of cardiac events secondary to anticholinergic mechanisms, the ECG monitoring results support the favorable benefit‐risk assessment of tiotropium in patients with COPD.