Acute Bronchodilator Responses Decline Progressively Over 4 Years in Patients With Moderate to Very Severe COPD

Donald P Tashkin; Ning Li; Eric C Kleerup; David Halpin; Bartolome Celli; Marc Decramer; Robert Elashoff

Respiratory Research. 2014;15(102)

Abstract and Introduction

Abstract

Background: We previously reported a progressive decline in absolute responses of FEV₁ and FVC to a near-maximal dose of 2 different short-acting bronchodilators over 4 years. Since varying host factors and the method of expressing the response may impact the time trend of acute bronchodilator responses, we now examined the potential influence of salient host characteristics on changes in bronchodilator responses over time expressed in different ways.

Methods: As part of the 4-year, placebo-controlled Understanding Potential Long-term Impacts on Function with Tiotropium (UPLIFT) trial, pre- and post-bronchodilator spirometry was performed at baseline and 1 month and every 6 months thereafter. Post-bronchodilator values for FEV₁ and FVC were analyzed for subjects completing at least the 1 year visit (Placebo – N = 2463; Tiotropium – N = 2579), stratified by GOLD stage, age, gender and smoking status and expressed as absolute, relative (%) and % predicted changes from pre-bronchodilator values. Annual changes in bronchodilator response were estimated using linear mixed effects models.

Results: For all subjects analyzed, FEV₁ and FVC bronchodilator responses showed progressive and highly significant (p < 0.0001) declines over 4 years. Declines were generally larger in patients with severe/very severe than mild/moderate airflow obstruction, in older patients (≥65 yrs) and in former than continuing smokers.

Conclusion: Acute FEV₁ and FVC responses to bronchodilators decline significantly over time in COPD patients, whether expressed as absolute, relative or % predicted changes, potentially impacting on the clinical responses to bronchodilator therapy as well as on the annual rate of decline in post-bronchodilator lung function.

Introduction

Bronchodilator responsiveness is a well-described feature of both asthma and COPD. While the response to a bronchodilator in COPD is never complete, nonetheless it often fulfills the currently accepted criteria for a significant response,^[1–3] although the degree of response (and the attainment of a significant response) is highly variable between testing sessions.^[4,5] Since COPD is usually a progressive disease characterized by a variably accelerated annual rate of decline in lung function, as determined from measurements of both pre- and post-bronchodilator forced expired volume in 1 sec (FEV1),^[6] it is possible that the response to a bronchodilator might also change over an extended period of time; however, few studies have examined the long-term course of responses to a bronchodilator in COPD with varying results.^[7–9]

We recently compared the annual rates of change in the pre- versus post-bronchodilator FEV1 and FVC over 4 years in 5041 COPD UPLIFT trial participants^[10] and observed that, on average, the absolute FEV1 and forced vital capacity (FVC) responses to a bronchodilator decreased progressively and significantly over the 4-year course of the trial, in contrast to findings from previous studies of 1–11 years duration in which either no change, small average changes of varying significance or substantial increases in responses were observed.^[7–9] These differences could be due to several factors, including differences in the study populations, especially regarding the severity of airflow obstruction, as well as differences in the methods used to measure the bronchodilator response. Such methods included a standard therapeutic dose of a beta-agonist followed only 10 minutes later by repeat spirometry in the IPPB trial^[7] and the Lung Health Study^[8] and 400 mcg salbutamol with repeat spirometry after only 15 minutes in the Evaluation of COPD Longitudinally to Identify Predictive Surrograte Endpoints (ECLIPSE) study,^[9] compared to double doses of both a beta-agonist and a muscarinic antagonist and performance of the post-bronchodilator spirometry at the expected time of peak or near-peak action of each of the two classes of bronchodilators in the UPLIFT trial.^[11] It is not unlikely, therefore, that the responses to a bronchodilator were sub-maximal in the earlier trials and near-maximal in the UPLIFT study.

Because of these differences in methodology for measuring the response to a bronchodilator and the possibility that varying host factors (including gender, age, severity of airflow obstruction, smoking status and use of inhaled corticosteroids) may impact the bronchodilator response over time, we extended our analysis of the time trend of bronchodilator responses (for both FEV1 and FVC) over the 4 years of the UPLIFT trial to examine the potential influence of these host factors on the changes over time in bronchodilator responses expressed in three different ways: absolute change in ml, percent change from baseline and change in percent predicted.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Baseline characteristics of subjects in the placebo and tiotropium arms of the UPLIFT trial included in the analysis of time trends in bronchodilator responses over the course of the trial
Characteristic	Placebo (N = 2463)	Tiotropium (N = 2579)
Age, yrs, mean (SD)	64.2 (8.39)	64.3 (8.39)
>65 yrs, N (%)	1158 (47.0%)	1223 (47.4%)
≤65 yrs, N (%)	1305 (53.0%)	1356 (52.6%)
Gender, male, N (%)	1852 (75.2%)	1965 (76.2%)
Pre-bronchodilator FEV₁, L, mean (SD)	1.12 (0.40)	1.11 (0.40)
Pre-bronchodilator FEV₁, % predicted, mean (SD)	39.9 (11. 8)	39.8 (11.9)
Post-bronchodilator FEV₁, L, mean (SD)	1.35 (0.44)	1.34 (0.43)
Post-bronchodilator FEV₁, % predicted, mean (SD)	48.2 (12.4)	48.1 (12.5)
Pre-bronchodilator FVC, L, mean (SD)	2.66 (0.83)	2.64 (0.80)
Pre-bronchodilator FVC, % predicted, mean (SD)	75.4 (18.0)	74.8 (17.9)
Post-bronchodilator FVC, L, mean (SD)	3.13 (0.90)	3.11 (0.86)
Post-bronchodilator FVC, % predicted, mean (SD)	88.7 (18.7)	88.2 (18.5)
Pre-bronchodilator FEV1/FVC ratio, %, mean (SD)	42.5 (10.3)	42.7 (10.4)
Post-bronchodilator FEV1/FVC ratio, %, mean (SD)	43.8 (10.5)	43.9 (10.7)
Smoking status
Sustained ex-smoker, N (%)	1471 (59.7%)	1502 (58.2%)
Intermittent smoker, N (%)	679 (27.6%)	761 (29.5%)
Continuing smoker, N (%)	313 (12.7%)	316 (12.3%)
Pack-yrs smoking, mean (SD)	48.0 (27.9)	49.0 (28.0)
GOLD grade of airflow obstruction
Grade I/II, N (%)	1179 (48.6%)	1226 (48.3%)
Grade III, N (%)	1059 (43.6%)	1118 (44.0%)
Grade IV, N (%)	189 (7.79%)	197 (7.75%)
Use of Inhaled Corticosteroids at baseline
Yes	1506 (61.1%)	1581 (61.3%)
No	957 (38.9%)	998 (38.7%)
SGRQ total scores, mean (SD)	45.2 (17.2)	45.0 (17.0)

Table 2. Estimated average change per year over 4 years in absolute bronchodilator response (Δ, in ml) (±SE) for FEV ₁ and FVC in the placebo arm of the UPLIFT trial by GOLD grading for airflow obstruction (I&II, III, IV), age (≤50 yrs, >50 yrs), gender, and smoking status (sustained ex-smoker, intermittent smoker, continuing smoker)
Group	FEV₁		FVC
Group	Estimated change in Δ (SE) per yr	p value	Estimated change in Δ (SE) per yr	p value
All	-10.9 (0.76)	<0.0001	-20.6 (1.87)	<0.0001
GOLD Stage
I & II	-11.1 (1.10)	<0.0001	-14.4 (2.42)¹	<0.0001
III	-11.2 (1.10)	<0.0001	-25.2 (3.08)	<0.0001
IV	-13.1 (2.35)	<0.0001	-40.6 (8.21)	<0.0001
Age, yrs
≤65 yrs	-11.6 (1.05)	<0.0001	-15.5 (2.64)²	<0.001
>65 yrs	-10.1 (1.09)	<0.0001	-2.66 (2.60)	<0.0001
Gender
Male	-10.7 (0.92)	<0.0001	-21.4 (2.27)	<0.0001
Female	-11.5 (1.21)	<0.0001	-18.4 (2.98)	<0.0001
Smoking status
Sustained ex-smoker	-11.9 (0.93) ³	<0.0001	-24.6 (2.37)³	<0.0001
Intermittent smoker	-10.5 (1.48)	<0.0001	-17.3 (3.57)	<0.0001
Continuing smoker	-6.77 (2.42)	0.0053	-9.17 (5.70)	0.11
Inhaled steroids (baseline)
No	-8.41 (1.24)⁴	<0.0001	-11.3 (3.01)⁴	0.0002
Yes	-12.5 (0.95)	<0.0001	-26.6 (2.38)	<0.0001

¹Significantly different from GOLD III (p = 0.0059) and GOLD IV (p = 0.0006).
²Significantly different from age >65 yrs (pp = 0.0022).
³Signficiantly different from continuing smokers (p = 0.0264 for FEV1 and p = 0.0088 for FVC).
⁴Significantly different from those with baseline inhaled steroids (p = 0.0081 for FEV1 and p < 0.0001 for FVC)).

Table 3. Estimated average change per year over 4 years in relative bronchodilator response (percent change Δ = post- minus pre-bronchodilator FEV1 and FVC/pre-bronchodilator value X 100) (±SE) in the placebo arm by GOLD grading for airflow obstruction (I&II, III, IV), age (≤50 yrs, >50 yrs), gender, and smoking status (sustained ex-smoker, intermittent smoker, continuing smoker)
Group	FEV₁		FVC
Group	Estimated change in Δ (SE) per yr	p value	Estimated change in Δ (SE) per yr	p value
All	-0.68 (0.08)	<0.0001	-0.66 (0.09)	<0.0001
GOLD Stage
I & II	-0.40 (0.11)¹	0.0001	-0.28 (0.10)¹	0.0085
III	-0.86 (0.14)²	<0.0001	-0.89 (0.15)²	<0.0001
IV	-2.08 (0.38)	<0.0001	-2.11 (0.47)	<0.0001
Age, yrs
≤65 yrs	-0.61 (0.11)	<0.0001	-0.34 (0.12)³	0.0058
>65 yrs	-0.77 (0.12)	<0.0001	-1.04 (0.13)	<0.0001
Gender
Male	-0.61 (0.10)	<0.0001	-0.64 (0.10)	<0.0001
Female	-0.90 (0.17)	<0.0001	-0.74 (0.19)	<0.0001
Smoking status
Sustained ex-smoker	-0.91 (0.11)⁴	<0.0001	-0.88 (0.11)³	<0.0001
Intermittent smoker	-0.54 (0.16)	0.0007	-0.55 (0.17)	0.0014
Continuing smoker	0.12 (0.25)	0.64	0.14 (0.26)	0.59
Inhaled steroids (baseline)
No	-0.30 (0.13)⁵	0.0273	-0.13 (0.14)⁵	0.35
Yes	-0.93 (0.11)	<0.0001	1.01 (0.12)	<0.0001

¹Significantly different from GOLD III (p = 0.0072 for FEV1 and p = 0.0008 for FVC) and GOLD IV (p < 0.0001 for FEV1 and p < 0.0001 for FVC).
²Significantly different from GOLD IV (p = 0.0018 for FEV1 and p = 0.0058 for FVC).
³Significantly different from age >65 (p < 0.0001).
⁴Signficiantly different from continuing smokers (p < 0.0001 for FEV1 and p = 0.0002 for FVC).
⁵Significantly different from those with baseline inhaled steroids (p = 0.0002 for FEV1 and p < 0.0001 for FVC).

Table 4. Estimated average change per year over 4 years in % predicted bronchodilator response (Δ = % predicted post-bronchodilator minus % predicted pre-bronchodilator FEV1 or FVC) (±SE) in the placebo arm by GOLD grading for airflow obstruction (I&II, III, IV), age (≤50 yrs, >50 yrs), gender, and smoking status (sustained ex-smoker, intermittent smoker, continuing smoker)
Group	FEV₁		FVC
Group	Estimated change in Δ (SE) per yr	p value	Estimated change in Δ (SE) per yr	p value
All	-0.33 (0.03)	<0.0001	-0.53 (0.05)	<0.0001
GOLD Stage
I & II	-0.33 (0.04)	<0.0001	-0.35 (0.07)¹	<0.0001
III	-0.35 (0.04)	<0.0001	-0.68 (0.09)	<0.0001
IV	-0.41 (0.08)	<0.0001	-1.05 (0.24)	<0.0001
Age, yrs
≤65 yrs	-0.33 (0.04)	<0.0001	-0.39 (0.07)²	<0.0001
>65 yrs	-0.32 (0.04)	<0.0001	-0.69 (0.08)	<0.0001
Gender
Male	-0.30 (0.03)³	<0.0001	-0.51 (0.06)	<0.0001
Female	-0.43 (0.06)	<0.0001	-0.61 (0.11)	<0.0001
Smoking status
Sustained ex-smoker	-0.37 (0.03)⁴	<0.0001	-0.65 (0.07)³	<0.0001
Intermittent smoker	-0.31 (0.05)	<0.0001	-0.46 (0.10)	<0.0001
Continuing smoker	-0.18 (0.08)	0.0322	-0.16 (0.15)	0.29
Inhaled steroids (baseline)
No	-0.23 (0.04)⁵	0.0273	-0.27 (0.09)⁶	0.0024
Yes	-0.39 (0.03)	<0.0001	-0.70 ((0.07)	<0.0001

¹Significantly different from GOLD III (p = 0.0023) and GOLD IV (p = 0.0013).
²Significantly different from no baseline inhaled steroids (p = 0.0040).
³Significantly different from women (p = 0.0428).
⁴Signficiantly different from continuing smokers (FEV1 p = 0.0267; FVC p = 0.0038).
⁵Significantly different from no baseline inhaled steroids (p = 0.0045).
⁶Significantly different from no baseline inhaled steroids (p < 0.0001).

Table 5. Between- and within-subjects variability in the annual change in bronchodilator response expressed as absolute, relative and percent predicted changes ¹
	Response measured by absolute change in FEV (ml)		Response measured by absolute change in FVC (ml)
	Tiotropium	Placebo	Tiotropium	Placebo
Between-subject variability in annual change (square-root of variance)	15.2	15.3	30.2	44.5
Within-subject variability (square-root of variance)	109.2	107.9	244.7	252.0
	Response measured by % change in FEV (ml)		Response measured by % change in FVC (ml)
	Tiotropium	Placebo	Tiotropium	Placebo
Between-subject variability in annual change (square-root of variance)	1.44	1.65	1.43	1.98
Within-subject variability (square-root of variance)	11.3	12.0	11.0	12.4
	Response measured by absolute change in FEV (% predicted)		Response measured by absolute change in FVC (% predicted)
	Tiotropium	Placebo	Tiotropium	Placebo
Between-subject variability in annual change (square-root of variance)	0.55	0.54	0.89	1.26
Within-subject variability (square-root of variance)	3.93	3.88	7.04	7.23

¹The variations were estimated from the linear mixed effects models.

Table 6. Long-term trends in the proportion of subjects in each treatment arm who achieved a significant bronchodilator response for FEV ₁, FVC and either FEV ₁ or FVC
Time	FEV				FVC				FEV or FVC
	Tiotropium		Placebo		Tiotropium		Placebo		Tiotropium		Placebo
	N	Proportion	N	Proportion	N	Proportion	N	Proportion	N	Proportion	N	Proportion
1 m	2535	0.42	2416	0.55	2535	0.47	2416	0.62	2535	0.59	2416	0.73
6 m	2515	0.42	2391	0.53	2515	0.44	2391	0.63	2515	0.57	2391	0.72
1 yr	2503	0.39	2382	0.53	2503	0.43	2382	0.63	2503	0.55	2382	0.72
1.5 yr	2380	0.40	2200	0.51	2380	0.43	2200	0.62	2380	0.56	2200	0.71
2 yr	2236	0.39	2040	0.49	2236	0.43	2040	0.60	2236	0.56	2040	0.69
2.5 yr	2132	0.39	1906	0.50	2132	0.42	1906	0.61	2132	0.55	1906	0.70
3 yr	2012	0.36	1792	0.47	2012	0.39	1792	0.59	2012	0.52	1792	0.69
3.5 yr	1894	0.36	1674	0.47	1894	0.41	1674	0.59	1894	0.52	1674	0.68
4 yr	1804	0.35	1599	0.47	1804	0.39	1599	0.56	1804	0.51	1599	0.66
Estimated OR (95% CI) for change in proportion per year*	0.92 (0.90, 0.94)		0.91 (0.89, 0.94)		0.93 (0.91, 0.95)		0.94 (0.92, 0.96)		0.92 (0.90, 0.94)		0.92 (0.90, 0.95)
p-value	<0.0001		<0.0001		<0.0001		<0.0001		<0.0001		<0.0001

*The OR (odds ratio) and its 95% CI were estimated with GEE methods using data starting from 1 m.
FEV: Tiotropium vs. placebo p = 0.79.
FVC: Tiotropium vs. placebo p = 0.60.
FEV or FVC: Tiotropium vs. placebo p = 0.84.