The Safety and Efficacy of Arformoterol and Formoterol in COPD

ABSTRACT

This study evaluated the safety and efficacy of arformoterol and formoterol over 6-months in subjects with COPD. In a multi-center, 6-month randomized, double-blind, double-dummy trial, subjects with COPD (mean FEV₁ 1.21 L, ∼41.0%% predicted) were randomized to receive either nebulized arformoterol (15μg BID [[n == 149]][[ARF 15]], 25μg BID [[n == 147]][[ARF 25]]), or racemic formoterol (12μg BID [[n == 147]][[FORM]]) delivered by DPI. The proportion of subjects with any post-treatment adverse event for ARF 15, ARF 25 μg, and FORM was 67.8%%, 76.2%% and 66.7%%, respectively, and those with at least one COPD exacerbation was 32.2%%, 30.6%%, and 22.4%%, respectively. Pulmonary function improved for all treatment groups and was maintained throughout the study. Mean change from baseline at 6-months for ARF 15, ARF 25 and FORM in peak FEV₁ was 0.30L, and 0.34L, and 0.26L, respectively, in 24-hour trough FEV₁ was, 0.10L, 0.14L, and 0.09L, and in inspiratory capacity was, 0.20L, 0.37L, and 0.23L. Dyspnea, (mean Transition Dypsnea Index (TDI) focal score) improved in all treatment arms (ARF 15: 1.4, ARF 25: 1.5, and FORM: 1.4) at 6 months, as did rescue short-acting β₂-agonists use (mean range: −1.1 to −1.3 actuations/day) and ipratropium bromide (mean range: −0.3 to −0.8 actuations/day). Health status, measured by St George's Respiratory Questionnaire, improved from baseline at 6-months in all treatment groups (mean change: −3.7 to −6.8). In this 6-month study, arformoterol and formoterol were well-tolerated, and their use was associated with improvement in pulmonary function and health status in subjects with COPD with no apparent development of tolerance.

Keywords: :

• Arformoterol
• Formoterol
• COPD
• Long-acting β₂-agonist
• Bronchodilator
• Safety
• Efficacy

INTRODUCTION

Chronic obstructive pulmonary disease (COPD) is a multi-component disease characterized by airflow obstruction that is poorly reversible (1, 2). It affects approximately 24 million people in the United States, and is currently the fourth-leading cause of death and is expected to be the third-leading cause of death by 2020 (3,4). Evidence-based guidelines recommend the use of inhaled long-acting bronchodilators including long-acting β₂-agonists (LABAs) for the initial maintenance treatment in patients with COPD (2,3,5). Nonetheless, multiple studies have found that physicians under-prescribe long-acting bronchodilators for maintenance treatment of patients with moderate to very severe COPD (6–11). Although not recommended by the guidelines, (5) many physician prescribe short-acting β-agonists (SABAs) as regular treatment for this patient population (6–9).

Several factors may influence the prescription of such medications to patients with COPD. Recent concerns have been raised about the safety of prolonged use of LABAs in patients with asthma (12). Furthermore, some studies suggest that the long-term use of β₂-agonists can be associated with diminished efficacy over time (e.g., tolerance) in patients with COPD (13–20). Few studies have investigated the long-term safety (at least 6-months) and efficacy of racemic formoterol (19–22). Until recently, long-acting bronchodilators were administered primarily by metered dose inhalers (MDI) or dry powdered inhalers (DPI).

However, some patients with COPD, in particular those who cannot manipulate hand-held devices and those who cannot achieve a sufficient inspiratory flow for adequate drug delivery may have difficulty using these devices (23). A nebulized delivery in such situations may prove superior to these more common delivery methods (23). Arformoterol, the (R,R)-isomer of racemic formoterol, has recently become available as a nebulized formulation for the maintenance treatment of COPD (15). Previous trials have shown that arformoterol significantly improved airway function at the end of the dosing interval (trough FEV₁) and significantly improved post-dose FEV₁ when compared with placebo and salmeterol (13,15,16).

Only one previously published study investigated the long-term safety and efficacy of arformoterol using a dose (50 μg) that was higher than the approved and marketed dose (15 μg) (13,24). In this study, we examined the safety and efficacy of nebulized arformoterol 15 μg BID, 25 μg BID, and racemic formoterol over 6 months in subjects with COPD.

METHODS

This was a 6-month double-blind, double-dummy, multicenter (62 sites), randomized, active-controlled, parallel group study designed to examine the long-term safety and efficacy of twice daily arformoterol 15 μg and 25 μg and racemic formoterol 12 μg. This study was conducted according to the Declaration of Helsinki (25) and approved by appropriate IRBs. A written informed consent was obtained from all subjects. The study (## NCT00250679) is posted in clinicaltrials.gov.

Study subjects

Subjects were 35 years of age or older, had a ≥ 15 pack-year smoking history, and had a physician-diagnosis of COPD (including emphysema and/or chronic bronchitis) with baseline breathlessness severity grade (Modified Medical Research Council [[MMRC (26)]] Dyspnea Scale Score) of ≥ 2. Eligible subjects were also required to have pre-dose baseline FEV₁ ≤ 65%% predicted, pre-dose FEV₁ > 0.70L, and a pre-dose FEV₁/FVC ratio of ≤ 70%%. Subjects were excluded if they had history of asthma, life-threatening or unstable respiratory illness, upper or lower respiratory tract infection within 30 days prior to screening, or malignancy. Subjects were also excluded if they had pre-existing cardiovascular conditions or other co-morbidities that may have precluded them from completing the study.

Study design

The primary endpoint for this study was the frequency of adverse events, including COPD exacerbations over 6 months. Secondary endpoints included change in pulmonary function (FEV₁, inspiratory capacity) and the extent of use of rescue short-acting inhaled bronchodilators. Other secondary endpoints included dyspnea measured by Transition Dypsnea Index (TDI) (27), exercise tolerance assessed by the 6-minute walk test (28), and health status measured by the St George's Respiratory Questionnaire (SGRQ)(29).

Baseline values were obtained prior to the first dose of study medication. Medical event calendars were completed daily. Approximately 7 to 14 days after screening, subjects were randomized in a 1:1:1 ratio to 26 weeks of one of three treatment regiments: (1) nebulized arformoterol (Brovana, Sepracor, Inc., Marlborough, Massachusetts, USA) 15 μg BID plus placebo DPI BID (ARF 15), (2) nebulized arformoterol 25 μg BID plus placebo DPI BID (ARF 25), or (3) nebulized placebo plus racemic formoterol (Foradil® Aerolizer^TM; Schering, Kenilworth, New Jersey, USA) DPI BID (FORM). To facilitate an assessment of potential treatment differences by race, the randomization schedule was stratified by Caucasian, Black or Other. Subjects self-administered nebulized arformoterol or the nebulized placebo using the PARI LC PLUS^TM nebulizer and Trek^TM compressor (PARI Respiratory Inc., Richmond, Virginia, USA) followed by formoterol or placebo using the Aerolizer^TM Inhaler DPI, respectively. The nebulized medication/placebo was administered first followed by DPI administration. Subjects administered medication twice (approximately 12 hours apart) during the day at about 8 AM (± 1 hour) and 8 PM (± 1 hour).

During the study period, the use of long-acting bronchodilators other than the study medications was prohibited. Inhaled and oral corticosteroids (≤ 10 mg/day dose of prednisone or equivalent), short-acting xanthines and leukotriene inhibitors were allowed if subjects were on a stable regimen for ≥ 14 days prior to study entry. During the trial, up to two 14-day courses of oral corticosteroids separated by at least 6 weeks were permitted for treatment of COPD exacerbations. Short-acting β₂-agonists (SABAs) [[racemic albuterol MDI (VENTOLIN®, GlaxoSmithKline, Research Triangle Park, North Carolina, USA) and levalbuterol (Xopenex®, Sepracor Inc., Marlborough, Massachusetts, USA)]] were supplied to be used as rescue medication, and ipratropium bromide MDI (ATROVENT®, Boehringer Ingelheim, Ridgefield, CT) as supplemental medication to be used as needed during the study. Study medication and rescue and supplemental medication were required to be withheld for ≥ 8 hours and ≥ 6 hours, respectively, prior to clinic visits.

To evaluate compliance, subjects completed daily diaries recording study medication use. In addition, investigators maintained a log of all study drug that was dispensed, used and returned, and compliance was calculated separately for the UDV and the DPI. Subjects were contacted by telephone on weeks 1, 2, 6, 9, 16, 19, and 22 to discuss study drug adherence, adverse events, concomitant medication and recording of information in the medical events calendar. Clinic visits were scheduled at randomization and at weeks 3, 13, and 26 (end of treatment). At each clinic visit COPD questionnaires and medical event calendars were collected and vital signs, heart rate, MMRC dyspnea scale, and spirometry were evaluated. Blood samples for laboratory evaluation were collected at screening, at randomization, and week 26.

Spirometry was performed in accordance to the current American Thoracic Society Guidelines (30) prior to administering the study medication, immediately post-dose, and 30 minutes, 1, 2, 4, 6, 23, and 24 hours post-dose. All reported FEV₁ values were the highest of three acceptable maneuvers. Inspiratory capacity was measured pre-dose and 2-hours post-dose at randomization and at weeks 3, 13, and 26.

Baseline Dyspnea Index (BDI) was assessed at randomization and TDI was evaluated before administering the study medication at weeks 13 and 26 (27). The 6-minute walk test (28) and SGRQ (29) were evaluated at randomization and at weeks 13, and 26 prior to administering the study medication. The six minute walk test was also performed 3 to 4 hours post-dose.

Statistical analysis

Safety and efficacy summaries were based on the Intent-to-Treat (ITT) population, defined as randomized subjects who received at least one dose of study medication. In accordance with the ICH E9 Guidance for Industry document (Statistical Principles for Clinical Trials), safety and tolerability data from this study were analyzed descriptively (31). Significance testing was not performed on any safety comparisons, except for the key safety endpoint of COPD exacerbation. The sample size was outside of statistical considerations and was not powered to detect specific efficacy differences between treatment groups.

Continuous efficacy outcomes were summarized using descriptive statistics. For categorical variables, statistical summaries included counts of subjects and percentages. Confidence intervals for means were based on the t-distribution. In general, missing observations were treated as missing at random, and no imputation was performed. Last observation carried-forward was used for the calculation of FEV₁ AUC_0–6 parameters to impute the 6-hour FEV₁ value if missing.

Subgroup analyses were prospectively defined by race (Caucasian and Black) and baseline disease severity defined as baseline percent predicted FEV₁ values: < 30%%, 30%% to < 50%%, and ≥50%%.

Treatment-emergent adverse events were summarized by MedDRA system organ class and preferred term (32). In addition, certain prospectively defined composite adverse events were assessed including any adverse event categorized as follows: An event-based COPD exacerbation was defined as an increase in symptoms that necessitated any change in baseline medications other than bronchodilators or required subjects to have additional medical attention (Emergency Department, urgent care physician visit or hospitalization). Symptom-based COPD exacerbation included event-based COPD exacerbations in addition to reported adverse events associated with a worsening or new respiratory symptom. Cardiovascular adverse events were classified based on MedDRA terms into arrhythmic and ischemic events.

RESULTS

Of the 1014 subjects screened, 444 were randomized and 443 received at least one dose of medication (ITT population) (Figure 1). Baseline characteristics and demographics were similar among the three treatment groups (Table 1). Baseline characteristics were also similar among treatment subgroups when stratified by disease severity (FEV₁< 30%% predicted, 30%% to < 50%% predicted, and 50%% to 65%% predicted) and by race (Supplemental Tables 1 and 2). Of the subjects in the ITT population 311 (70%%) completed the 26 weeks study, with similar frequency of discontinuation among the three treatment groups (Figure 1). The most common reasons for discontinuation were adverse events (9.7%%) and voluntary withdrawal (10.8%%). Both the nebulized and DPI active treatment groups reported good average compliance (95%%–100%%) with the protocol-specified treatments. Blacks discontinued from the study due to adverse events at a slightly higher frequency than Caucasians (Supplemental Table 4). Prior to screening, inhaled or oral corticosteroids were used by 35.4 to 40.1%% of subjects in the three treatment arms, and the long acting inhaled anticholinergic tiotropium by 11.6 to 20.1%% of subjects.

Figure 1.  Consort diagram indicating the number of subjects enrolled (top), randomized to each treatment arm (middle) and either completed or withdrew from the study (bottom). Reasons for withdrawing from the study are indicated (bottom).

Table 1.  Demographics and baseline characteristics

	Formoterol 12 μg n == 147	Arformoterol 15 μg n == 149	Arformoterol 25 μg n == 147
Mean age, years (SD)	63.9 (7.8)	65.4 (9.0)	63.8 (8.6)
Female, n (%%)	64 (43.5)	52 (34.9)	56 (38.1)
Race, n (%%)
Caucasian	127 (86.4)	126 (84.6)	122 (83.0)
Black	19 (12.9)	21 (14.1)	23 (15.6)
Other	1 (0.7)	2 (1.3)	2 (1.3)
Current smoker, n (%%)	75 (51)	60 (40)	65 (44)
Pack-years smoked, n (%%)
≥15–< 30 years	18 (12.2)	23 (15.4)	24 (16.3)
≥ 30 years	129 (87.8)	126 (84.6)	123 (83.7)
Oral or Inhaled Corticosteroid use, n (%%)	52 (35.4)	59 (39.6)	59 (40.1)
Mean FEV1, L (SD)	1.20 (0.42)	1.22 (0.47)	1.20 (0.42)
Mean FEV1%% predicted (SD)	41.2 (11.5)	40.9 (13.6)	40.2 (12.2)
Mean inspiratory capacity, L (SD)	1.74 (0.63)	1.86 (0.55)	1.76 (0.58)
Reversibility**
Median percent change in FEV1 (25%%, 75%% percentiles)	12.1 (0.0, 0.23)	12.7 (6.1, 20.8)	13.4 (7.5, 23.0)

**Reversibility was measured following 2 actuations of 45 μg of levalbuterol HFA MDI or 90 μg racemic albuterol HFA MDI administered approximately 15–30 minutes prior to pulmonary function testing.

Safety outcomes

Adverse events

The overall frequency of adverse events was similar among the three treatment groups. The majority of subjects reported at least one adverse event (Table 2) and about one-third reported a respiratory event (34.2%% for ARF 15, 34.0%% for ARF 25, and 27.2%% for FORM). Less than 10%% of the subjects in each treatment group reported a cardiovascular adverse event. The frequencies of serious adverse events were low among the three treatment groups (6.8 to 13.4%%; Table 2). One subject treated with ARF 25 died of myocardial infarction, although this was not considered by the investigator as related to the study medication.

Table 2.  Summary of adverse events

	Formoterol 12 μg n == 147 n (%%)	Arformoterol 15 μg n == 149 n (%%)	Arformoterol 25 μg n == 147 n (%%)
All adverse events (AEs)	98 (66.7)	101 (67.8)	112 (76.2)
Serious adverse events (SAE’s)	14 (9.5)	20 (13.4)	10 (6.8)
Discontinued due to adverse events	12 (8.2)	15 (10.1)	18 (12.2)
COPD adverse event	7 (4.8)	9 (6.0)	7 (4.8)
Cardiovascular adverse event	0	2 (1.3)	3 (2.0)
Most common adverse events (≥ 5%%)
Nasopharyngitis	17 (11.6)	8 (5.4)	15 (10.2)
Bronchitis	9 (6.1)	8 (5.4)	10 (6.8)
Event-based COPD exacerbations**	33 (22.4)	48 (32.2)	45 (30.6)
Comparison between formoterol and arformoterol treatments, treatment difference (95%% C.I.)		9.8 (1.3, 18.2)	8.2 (−0.3, 16.6)
Symptom-based COPD exacerbations^##	41 (27.9)	53 (35.6)	56 (38.1)
Comparison between formoterol and arformoterol treatments, treatment difference (95%% C.I.)		7.7 (−1.2, 16.5)	10.2 (1.2, 19.2)
Overall cardiovascular adverse events	9 (6.1)	9 (6.0)	9 (6.1)
Arrhythmic adverse events	5 (3.4)	5 (3.4)	2 (1.4)
Ischemic adverse events	2 (1.4)	1 (0.7)	8 (5.4)

†COPD adverse events refers to an adverse event with a MedDRA preferred term of COPD. Cardiac disorders included acute coronary syndrome, atrial flutter, bundle branch block left, congestive cardiac failure and myocardial infarction.

**Event-based COPD exacerbation was protocol defined as an increase in symptoms that necessitated any change in baseline medications other than bronchodilators or required subjects to have additional medical attention (Emergency Department visit, hospitalization, etc.).

^##Symptom-based COPD exacerbation included event-based COPD exacerbations in addition to AEs associated with a COPD exacerbation. These AEs were pre-defined, and included events coded to the MedDRA (32) System Organ Class of Respiratory, Thoracic, and Mediastinal disorders and to the System Organ Class of Infections and Infestations.

COPD exacerbations

Over the 6-month treatment period event-based COPD exacerbations were observed to be less frequent in subjects treated with FORM (22.4%%) compared with either ARF 15 μg (32.2%%) or 25 μg (30.7%%) (Table 2). Using the more inclusive symptom-based definition of exacerbation, the proportion with exacerbations was also numerically less frequent in subjects treated with FORM (27.9%%) versus ARF 15 μg (35.6%%), and 25 μg (38.1%%). The frequency of both event-based and symptom-based COPD exacerbations appeared to be lower in the last 13 weeks of the study than in the first 13 weeks (Figure 2).

Figure 2.  Percent of subjects with event-based and/or symptom-based COPD exacerbations over time. Bars represent the proportion of subjects treated with FORM (solid), ARF 15 (stripe) and ARF 25 (speckled) with event-based (grey) and/or symptom-based (black) COPD exacerbations. The 95% CI for the proportions are indicated.

β-adrenergic mediated adverse effects

The occurrence of β-adrenergic mediated treatment effects, including dizziness, tremor, nervousness, insomnia, and paresthesia was low (ARF 15; 10.7%%, ARF 25; 15.6%%, and FORM; 17.0%%). There was no meaningful difference in mean pre-dose glucose and potassium serum concentrations over the 26-week study period. Mean change [[SD]] in ventricular heart rate 2-hours post-dose from baseline and at week 26 was similar among the three treatment groups (ARF 15: −2.8 [[10.3]] bpm, ARF 25: −1.5 [[10.0]] bpm, and FORM : −1.6 [[9.8]] bpm). The frequency of adverse events, COPD exacerbations, and β-adrenergic mediated effects were similar in the different disease severity groups regardless of treatment (Supplemental Table 3).

Rescue/supplemental medication use

During the pre-randomization period, 75.2%% of subjects used ipratropium bromide at least once as supplemental medication with a mean use of 2.4 to 2.7 actuations per day. The use of ipratropium bromide declined over the treatment period for all treatment groups by a mean of −0.83 to −1.09 actuations per day. Mean reductions were observed by week 3 and were maintained during the remainder of the treatment period. During the pre-randomization period, 79.5%% of subjects used a SABA at least once and baseline SABA use averaged 2.9 actuations per day. The use of rescue SABA medication declined from baseline over the treatment period for all treatment groups by a mean of −1.12 to −1.32 actuations per day.

Pulmonary function outcomes

Mean change in trough FEV₁ at the 24-hour time point following the morning dose improved similarly for all three treatment groups from baseline (Table 3). This improvement was seen following the first dose of the study medication (mean of 0.16 L to 0.17 L), with stable improvement (mean of 0.09 L to 0.114 L) maintained over weeks 3 to 26.

Table 3.  Change in pulmonary function outcomes from study baseline

	Formoterol 12 μg n == 147	Arformoterol 15 μg n == 149	Arformoterol 25 μg n == 147
Mean change in Trough FEV1 (24 hour time point from previous morning dose) L (95%% C.I.)
Randomization	0.16 (0.13, 0.20)	0.17 (0.14, 0.21)	0.16 (0.12, 0.19)
Week 3	0.10 (0.06, 0.13)	0.12 (0.09, 0.16)	0.13 (0.10, 0.17)
Week 13	0.09 (0.06, 0.13)	0.12 (0.08, 0.15)	0.10 (0.06, 0.14)
Week 26	0.09 (0.05, 0.13)	0.10 (0.05, 0.15)	0.14 (0.09, 0.18)
Mean peak change in FEV1, L (95%% C.I)
Randomization	0.29 (0.26, 0.33)	0.31 (0.28, 0.35)	0.36 (0.32, 0.39)**
Week 3	0.28 (0.24, 0.32)	0.31 (0.27, 0.34)	0.33 (0.29, 0.37)
Week 13	0.25 (0.21, 0.29)	0.32 (0.27, 0.36)**	0.32 (0.28, 0.37)**
Week 26	0.26 (0.22, 0.31)	0.30 (0.25, 0.35)	0.34 (0.30, 0.39)**
Mean change in predose (trough) inspiratory capacity, L (95%% C.I.)
Week 3	0.06 (−0.03, 0.15)	0.03 (−0.06, 0.13)	0.12 (0.04, 0.21)
Week 13	0.11 (0.02, 0.21)	0.01 (−0.11, 0.12)	0.12 (0.01, 0.24)
Week 26	0.04 (−0.05, 0.13)	0.06 (−0.03, 0.14)	0.15 (0.05, 0.26)
Mean change in 2 hour post-dose inspiratory capacity, L (SD)
Randomization	0.24 (0.16, 0.32)	0.30 (0.23, 0.36)	0.33 (0.25, 0.41)
Week 3	0.21 (0.11, 0.30)	0.24 (0.15, 0.33)	0.33 (0.22, 0.44)
Week 13	0.30 (0.17, 0.41)	0.23 (0.14, 0.33)	0.28 (0.17, 0.40)
Week 26	0.23 (0.12, 0.35)	0.20 (0.10, 0.31)	0.37 (0.25, 0.49)

**Indicates difference between arformoterol and formoterol treatments was significant (95%% C.I. of the treatment differences in means did not include zero).

Mean change in time-normalized FEV₁AUC_0–6 post-dose improved from baseline for all treatment groups at each study visit (Figure 3). These increases in FEV₁AUC_0–6 were sustained throughout the 26-week treatment period (Figure 3). Following the first dose, mean peak improvement in FEV₁ over 6 hours post-dose ranged between 0.29 L and 0.36 L for the three treatment groups (Table 3). The extent of this improvement from study baseline FEV₁ level was maintained for all treatment groups over 26-weeks (Table 3). Mean trough FEV₁, FEV₁ AUC_0–6, and peak FEV₁, improved from baseline for all disease severity groups independent of treatment with the single exception of trough FEV₁ in the 50 to 65%% predicted group (Figure 4A-C). Improvement from baseline continued to be evident after 26 weeks for all treatment groups, although the magnitude of improvement declined to some extent at 26 weeks for subjects with less severe baseline FEV₁ compromise (i.e., the 30 to <50%% predicted and 50 to 65%% predicted groups) except for those with severe (30 to < 50%%) COPD treated with ARF 25 μg (Figure 4A-C). This was especially true for mean trough FEV₁ values.

Figure 3.  Mean change in time normalized FEV₁AUC_0–6 from baseline. Bars represent the change in FEV₁ AUC_0–6 for FORM (white), ARF 15 (grey) and ARF 25 (black). Standard errors are indicated. **denotes that the mean difference between an arformoterol treatment group and FORM were significant (95%% CI for the difference between treatment groups mean did not include zero). ## indicates the mean difference between ARF 25 and ARF 15 was significant (95%% CI for the difference between treatment groups means did not include zero).

Figure 4.  Pulmonary function outcomes stratified by baseline FEV₁%% predicted (< 30%%, 30%% to < 50%%, and 50%% to 65%%). (A) Mean change in FEV₁AUC_0–6 (B) 6-hour peak change in FEV₁, (C) 24-hour trough FEV₁, and (D) pre-dose (trough) inspiratory capacity for patients with < 30%% (left), 30%% to < 50%% (middle), and 50%% to 65%% (right) percent predicted FEV₁ at baseline for FORM (white), ARF 15 (grey), and ARF 25 (black). Standard error bars are indicated. **denotes that the mean difference between an arformoterol group and FORM was significant (95%% CI for the difference between treatment group means did not include zero). ## indicates the mean difference between ARF 25 and ARF 15 was significant (95%% CI for the difference between treatment groups means did not include zero).

Mean change in pre-dose (trough) inspiratory capacity from baseline values over the 26-week treatment period consistently improved only for ARF 25 μg treatment group (Table 3). Mean change in inspiratory capacity 2 hours post-dosing improved similarly for all treatment groups following the first dose (Table 3). The extent of this improvement was sustained throughout the study. Trough inspiratory capacity improved from baseline only for subjects with very severe COPD and for those with severe (30 to < 50%% predicted) COPD treated with ARF 25 μg (Figure 4D).

Forced expiratory (FEV₁) values improved to a similar extent in both Caucasian and Black COPD subjects, (Supplemental Table 5) although Blacks had less improvement in inspiratory capacity than Caucasians.

SYMPTOMS AND FUNCTIONAL OUTCOMES

Dyspnea, assessed by TDI focal scores, improved for all three treatment groups at weeks 13 and 26 (Table 4). At both weeks 13 and 26 approximately 50%% of the subjects in each treatment group had at least 1 unit improvement (the minimal clinically important difference [[MCID]]) in the TDI focal score.

Table 4.  Change in dyspnea, 6 minute walk, and health status from baseline

	Formoterol 12 μg n == 147	Arformoterol 15 μg n == 149	Arformoterol 25 μg n == 147
Baseline dyspnea index, mean (SD)	5.3 (1.7)	5.3 (2.2)	5.1 (2.4)
Transition dyspnea index focal score
Week 13, mean (95%% C.I)	1.5 (1.1, 1.9)	1.5 (1.0, 1.9)	1.6 (1.1, 2.1)
Subjects with change of ≥ 1 unit at week 13, n (%%)	59 (51.3)	59 (52.2)	61 (53.5)
Week 26, mean (95%% C.I)	1.4 (0.9, 2.0)	1.4 (0.9, 2.0)	1.5 (0.9, 2.0)
Subjects with change of ≥ 1 unit at week 26, n (%%)	56 (51.9)	49 (48.0)	54 (52.4)
6-minute walk test, (meters)
Baseline (pre-dose), median (25%%, 75%% percentiles)	335.3 (251.5, 426.7)	311.7 (231.0, 386.3)	320.3 (231.0, 387.3)
Change in pre-dose 6-minute walk test from baseline, median (25%%, 75%% percentiles)
Week 13	4.9 (−30.4, 53.1)	2.6 (−30.0, 53.4)	22.0 (−11.0, 53.4)
Week 26	18.3 (−15.2, 64.0)	8.2 (−22.9, 61.9)	26.8 (−16.1, 65.5)
Change in postdose 6-minute walk test from visit pre-dose, median (25%%, 75%% percentiles)
Randomization	15.5 (−10.7, 37.2)	11.6 (−9.6, 35.2)	15.2 (−7.9, 40.2)
Week 13	7.8 (−9.1, 30.5)	3.1 (−15.2, 25.2)	4.6 (−24.4, 26.8)
Week 26	0.0 (−11.0, 22.0)	0.0 (−7.0, 11.9)	7.3 (−11.6, 24.4)
SGRQ total score
Baseline, mean (SD)	48.9 (14.9)	49.4 (17.4)	52.4 (17.7)
Change in SGRQ from baseline
Week 13, mean (95%% C.I)	−4.2 (−6.1, −2.2)	−3.4 (−5.8, −1.1)	−1.9 (−.3.8, −0.0)
Subjects with ≥ 4 unit improvement at week 13, n (%%)	49 (46.7)	47 (46.1)	39 (37.1)
Week 26, mean (95%% C.I)	−6.8 (−8.9, −4.7)	−3.7 (−6.4, −1.0)	−3.7 (−6.1, −1.3)
Subjects with ≥ 4 unit improvement at week 26, n (%%)	55 (56.7)	43 (45.7)	42 (43.3)

All treatment groups had small improvements from baseline in median 6-minute walk distance (Table 4). These improvements were modest and unlikely to be clinically meaningful, and they did not change substantially between weeks 13 and 26. The median post-dose 6-minute walk distance was also only slightly greater than pre-dose median distance for all treatment groups, and again unlikely to be clinically meaningful. Mean SGRQ total scores improved from baseline for all treatment groups at weeks 13 and 26 (Table 4). FORM group had the greatest numerical mean improvement at both week 13 (−4.2 units versus −3.4 and −1.9 for ARF 15 and ARF 25 respectively) and at week 26 (−6.8 versus −3.7 for both ARF 15 and ARF 25). The proportion of subjects who had ≥ 4 unit (the MCID) improvement in SGRQ at week 26 was, 45.7%% for ARF15, 43.3%% for ARF25, and 56.7%% for FORM (Table 4).

DISCUSSION

This is the first study to compare the safety and efficacy of extended (up to 6 months) use of nebulized arformoterol compared with racemic formoterol. Overall, both doses of arformoterol, 15 μg and 25 μg, and racemic formoterol were well tolerated with pulmonary function improvement being maintained over the 6-month treatment period. The overall frequency of adverse events and serious adverse events were similar among treatment groups. The proportion of subjects with respiratory or cardiac adverse events was similar among the three treatment groups. Over the 6 month treatment period, COPD exacerbations resulting in a change in therapy or urgent visit (event-based definition) were observed less frequently in the FORM treatment group (33 of 147 subjects; 22.4%%) than in either the ARF 15 (48 of 149; 32.2%%) or in ARF 25 (45 of 147; 30.7%%) treated subjects. Whether this difference is real and related to pharmacologic properties inherent in the different therapies, due to unmeasured confounding factors not controlled at randomization, or chance is not clear. Nonetheless, the absence of differences for other safety outcomes for these treatment arms, and the lack of dose-response (higher exacerbation rate for ARF 15 versus ARF 25 μg) does not likely suggest a clear safety/tolerability difference for the 3 treatments.

Furthermore, the proportion of subjects with β₂-adrenergic adverse effects such as tremor, insomnia, dizziness etc. was low and similar among the three treatment groups. During this 6-month study, the frequency of serious adverse events in all three treatment groups was low and was similar to prior reports (15,16,13,21). Pulmonary function improvement following treatment was maintained over the 6-month treatment period in the 3 treatment groups.

The use of LABAs for maintenance treatment in COPD is currently recommended by all evidence based guidelines (2,3). However, recent studies in asthma have raised questions regarding the long-term safety of these medications in asthma (12,33). Less is known regarding the safety of regular LABAs use in COPD. Data from the recently published TORCH trial, revealed that the three year use of the LABA, salmeterol, either as monotherapy or in combination with fluticasone, was not associated with an increased risk of death or exacerbation frequency (17). The current study supports the safety of formoterol and arformoterol when used for maintenance therapy in COPD.

Some prior studies investigating the regular LABA use (13–17,34) have demonstrated a small loss in the improvement in pre-dose (trough) FEV₁ from baseline after a few weeks following the initiation of medications, while other studies failed to show such changes (35,36). Clinically meaningful tolerance, measured by loss of bronchodilator response and disease control, was not observed in this study in any of the treatment groups. However, the extent of improvement observed in trough FEV₁ on the first day diminished by week 3, but remained improved from baseline and did not decline further following weeks 3 in all three treatment groups. Of interest, there was little or no reduction in the magnitude of mean peak improvement in FEV₁ or FEV₁ AUC _0–6 from the first day of dosing to week 26. The fact that a decrease in trough FEV₁ but not other pulmonary function measures is often detected with regular LABA therapy (13–17, 34) suggests that this outcome may be a particularly sensitive indicator of tolerance.

The frequency of COPD exacerbations, a key measure of disease stability, did not increase during the course of the study; in fact, rate estimates decreased between the first and second half of the treatment period for all treatment groups (Figure 2). This phenomenon was also observed in the prior 12 week (15,16) and 52 week (13) arformoterol treatment trials. Higher exacerbation rates early in this and previous trials could have been contributed to by subjects with more severe disease discontinuing participation in trial after experiencing an exacerbation. This would result in the profile of subjects who completed the trial being less exacerbation-prone, with corresponding decrease in the observed rate late in the trial. Similarly, a possible beneficial impact of treatments on exacerbation risk that accrued over a long duration of treatment could also have contributed to this finding. Moreover, the sustained decrease in the use of rescue medication and supplemental medication for all three treatment groups further indicates that there was no worsening in the course of the disease.

Dyspnea (BDI/TDI) improved by more than 1 unit for approximately 50%% of subjects, a finding consistent with results seen in prior arformoterol (11,13) and formoterol studies (13–15,37,38). Health status, a measure that may be a better predictor of patient response to treatment, hospitalization (39) and risk of death than FEV₁ (40) improved in all treatment arms, as was the case in prior studies (14,15,19,34). The use of FORM was associated with a mean change in the total SGRQ score of −6.8 which exceeds the metric (4 unit improvement) that has been proposed as the minimum clinical important difference (MCID) for this outcome (41). Improvement for both arformoterol doses also approached this cutoff (3.7 units). For all treatments the SGRQ symptom domain score had the greatest improvement after 6-months of treatment (mean change from baseline: FORM μg: −8.4, ARF 15: −6.1, and ARF 25 −7.0) compared with the other SGRQ sub-scores as has been noted in previous formoterol (14,42) and arformoterol (15) reports.

Multiple genetic, behavioral, socioeconomic, and underlying disease factors, may impact how a patient with COPD responds to therapy. In general, subjects with very severe COPD at the start of the study had a greater absolute and percent improvement in airway function (particularly notable for trough FEV₁ and inspiratory capacity) measures following 6-months of therapy than COPD subjects with less severe disease. Subjects with severe COPD treated with ARF 25 also had the greatest sustained improvements in pulmonary function, suggesting a possible dose-response effect. This differential and greater improvement for subjects with most severe disease is supportive of findings from a prior study that investigated the use of arformoterol and salmeterol in COPD (16).

Several limitations must be taken into account when considering our study results. This study did not include a placebo arm. Hence, safety and efficacy outcomes of regular LABA therapy relative to as-needed bronchodilator therapy were not evaluated. However, two large prior 12-week efficacy and safety trials (13, 14) were conducted using the two BID arformoterol dose regimens included in this trial and an additional 50 μg QD regimen, along with a placebo treatment group. The safety profile in those trials for subjects in each of the two BID arformoterol treatment groups did not differ from those in the placebo group. In contrast, the primary purpose of this trial was to make a comparative assessment of longer term safety and efficacy among the LABA treatments. As long-acting bronchodilator treatments are widely endorsed by independent guidelines (5) as effective for use in COPD patients whose disease severity is consistent with the entry criteria for this trial, no placebo group was included.

Nonetheless, safety signals among the 3 treatment groups in this trial did not occur with a greater frequency over 6 months than reported in prior clinical trials (13,15,16,22,21). A longer study enriched for exacerbation-prone subjects would have increased the ability of the study to quantify possible exacerbation-risk differences among the treatments. As stated above, a substantial number of subjects (about 30%% for each treatment group) discontinued prior to completion of the 6-month trial. Subjects who withdrew early from the trial may have differed from those who completed in such a way as to impact outcomes measured at weeks 13 and 26, such as early versus late exacerbation rates.

In summary, this randomized trial supports the long-term safety and efficacy of the arformoterol and racemic formoterol in COPD patients (13,19,21,37,38,40). These results add to a growing body of evidence that support the importance of long-acting bronchodilators in controlling symptoms and optimizing functional status of patients with COPD.

Declaration of interest

Dr. Hanania has received research grant support from Sepracor, GlaxoSmithKline, Boehringer Ingelheim, Novartis, and Dey. Dr. Hanania has also served as a consultant to Sepracor, GlaxoSmithKline, Novartis, & Dey. Dr. Donohue has received research grants from Sepracor and Novartis and served on advisory boards and as a consultant to both. Arformoterol is manufactured by Sepracor and formoterol by Novartis. Dr. Nelson has consultant arrangements with Genentech/Novartis, Abbott Laboratories, MediciNova, AstraZeneca, Amgen, GlaxoSmith-Kline, Schering-Plough, Dyson, Sepracor and Nycomed. He has also received grant/research support from Shering-Plough, Novartis, Genentech, Ception, and AstraZeneca. Lastly, he is a member of the Speakers' Bureau for GlaxoSmithKline. Mr. Sciarappa is a full-time employee of Sepracor Inc. which markets arformoterol (brand name: Brovana). Dr. Goodwin is a full-time employee of Sepracor Inc. which markets arformoterol (brand name: Brovana). Dr. Baumgartner is a previous employee of Sepracor Inc. Dr. Hanrahan is a former employee of Sepracor Inc. Dr. Hanrahan was the Senior Medical Director responsible for the design and conduct of this trial. Drs N.A. Hanania, J.F. Donohue and H. Nelson participated as investigators in this study. K. Sciarappa PhD and E. Goodwin PhD are Sepracor employees. Dr. J. P. Hanrahan and Dr. R.A. Baumgartner are former Sepracor employees. Support for this study provided by Sepracor Inc. Data were compiled and statistical analyses were performed by Sepracor Inc. and were available to all authors. All authors had access to the data and were involved in the manuscript preparation.