Effects of rac-albuterol on arterial blood gases in patients with stable hypercapnic chronic obstructive pulmonary disease

Abstract

Aims

Many patients with chronic obstructive pulmonary disease (COPD) are treated with high dose β₂-adrenoceptor agonists, which can increase ventilation/perfusion mismatching, and tremor and cardiac output, thereby increasing oxygen uptake and carbon dioxide output (VCO₂). Patients with severe COPD and hypercapnia may be unable to increase ventilation in response to increased VCO₂, in which case arterial carbon dioxide tension (P_aCO₂) may rise. Our aim was to determine whether high dose nebulized rac-albuterol could increase P_aCO₂ in patients with COPD, limited bronchodilator reversibilty and hypercapnia.

Methods

We compared 10 mg and 400 µg rac-albuterol, given in two doses 1 h apart on nonconsecutive days, in a double-blind randomized crossover study in 14 patients with severe COPD. P_aCO₂, arterial oxygen tension (P_aO₂) and heart rate were measured over 120 min and change from baseline was plotted against time to obtain an area under the curve.

Results

Mean P_aCO₂ fell slightly over 120 min, with no difference between treatments (0.03 kPa h⁻¹ (95% confidence interval 0.02, 0.04)) and only three subjects had an increase in P_aCO₂ after high dose rac-albuterol. High dose rac-albuterol caused a greater fall in P_aO₂ [0.1 kPa h⁻¹ (95% confidence interval 0, 0.2)] and increase in heart rate than the low dose, although the differences were small.

Conclusions

Under stable conditions most subjects with severe COPD and hypercapnia will have a fall in P_aCO₂ and P_aO₂ following 10 mg rac-albuterol, suggesting that they maintain capacity to respond to any increase in VCO₂ and prevent a rise in P_aCO₂.

Introduction

Acute exacerbations of chronic obstructive pulmonary disease (COPD) are a major cause of hospital admissions [1] and patients with hypercapnia have particularly high mortality rates [2]. The response of these patients to treatment has been little studied and guidelines have been unable to provide clear dose recommendations for bronchodilator treatment. Patients with severe COPD have a limited ability to bronchodilate and the adverse effects following high doses of β₂-adrenoceptor agonists can outweigh the benefit [3].

Potential problems with high dose β-adrenoceptor agonists in patients with severe COPD include an increase in ventilation/perfusion mismatching and an increase in carbon dioxide output (VCO₂). Carbon dioxide output may rise by over 25% [4–6] and has been attributed to an increase in CO₂ production as tremor increases, and a reduction in CO₂ stores as cardiac output increases [6–8]. The normal response to an increase in CO₂ flux to the lungs would be to increase ventilation, but patients with severe COPD and limited reversibility may not have the capacity to do this, particularly during an acute exacerbation, in which case the P_aCO₂ would rise and may worsen acidosis.

Due to the difficulties of studying patients in a controlled manner during an acute exacerbation, we have initially examined the effect of high dose nebulized rac-albuterol in patients with severe stable COPD and chronic hypercapnia. We chose to give a low dose of rac-albuterol as the control rather than normal saline to reduce the possibility that patients would require treatment to relieve breathlessness during the study and to avoid bronchoconstriction from normal saline [9].

We report the changes in arterial blood gas tensions from a crossover study in which 14 patients with stable COPD, documented hypercapnia and limited reversibility received 10 mg or 400 µg nebulized rac-albuterol on separate days, both given in two doses 1 h apart.

Methods

Subjects

Patients aged 40–85 years with a clinical diagnosis of COPD were recruited if they had a forced expiratory volume in 1 s (FEV₁) below 50% predicted and a FEV₁/forced vital capacity (FVC) ratio of less than 70%. Ten minutes after inhaling 200 µg rac-albuterol, all subjects had an increase in FEV₁ of not more than 200 ml and 15%. Subjects were required to have had an arterial carbon dioxide tension (P_aCO₂) of greater than 6 kPa on two occasions when clinically stable or have developed hypercapnia (P_aCO₂ > 6 kPa) during assessment for long-term oxygen.

Exclusion criteria were other causes of ventilatory failure, an additional unstable medical condition, an acute exacerbation of COPD requiring oral corticosteroids within the last 4 weeks and regular medication with a β₂-adrenoceptor antagonist or anticoagulant. Nottingham City Hospital Research Ethics Committee approved the study and written informed consent was obtained at least 24 h before the first study day.

Measurements

P_aCO₂, P_aO₂ and pH were measured using standard electrodes on a blood gas analyser (Rapidlab 840, Bayer, Leverkusen, Germany) that was calibrated daily. Heart rate, oxygen saturation and blood pressure were recorded on a central monitor (Dynascope DS-5100E, Fukuda Denshi, Tokyo, Japan), attached to a three-lead electrocardiograph, an oximeter sensor (Nellcor, Pleasanton, USA) and a diaphragm transducer and amplifier to measure blood pressure.

Study design and methods

This was a randomized, double-blind, crossover study performed on two nonconsecutive days, not more than 5 days apart. Subjects were required to withhold short-acting bronchodilators for at least 10 h, and theophylline, tiotropium and long-acting β₂-adrenoceptor agonists for at least 24 h before each study day; inhaled and oral corticosteroids and nonpulmonary medications were taken as usual. Subjects using home oxygen discontinued this 30 min before leaving home. After an early breakfast without caffeine-containing beverages, subjects came to the hospital by taxi, were escorted to the laboratory by wheelchair and rested in a reclining chair for 10 min. Three chest leads and the fingertip pulse oximeter were attached and Alan’s test performed on the radial and ulnar arteries to confirm the adequacy of local circulation. A cannula (FloSwitch, Becton Dickinson, Swindon, UK) was inserted into the radial artery under local anaesthetic and attached to the transducer. Two 1 ml blood samples were taken sequentially from the arterial cannula to measure pH, P_aO₂ and P_aCO₂ (mean value recorded) and repeated at 10-min intervals until consecutive mean P_aO₂ and P_aCO₂ readings were within 0.5 kPa. Consecutive heart rate readings also had to be within five beats min⁻¹, without an increase in rate. Baseline values for all variables were taken as the mean of the last two results.

On each day the patient was given rac-albuterol sulphate (Steri-Neb, Baker Norton, London, UK), either two doses of 5 mg in 2.5 ml or two doses of 200 µg diluted to 2.5 ml with normal saline, with the order determined by computer-generated random sequence. The first dose was delivered over 6 min via a nebulizer (IEC 601–1, Medic-Aid, Pagham, UK) driven by air at 5 l min⁻¹. Duplicate 1 ml samples of arterial blood were taken 10, 15, 20, 30, 45 and 60 min after starting drug delivery to determine P_aCO₂ and P_aO₂ and followed at each time point by heart rate, blood pressure and oxygen saturation measurements in that order. Sixty minutes after starting the first nebulization the second identical dose of rac-albuterol was delivered and the measurements repeated over the next hour. Subjects received the alternative dose of rac-albuterol on the second study day, using an identical protocol, except that the cannula was placed in the contralateral radial artery.

The primary endpoint was the difference in change in P_aCO₂ between two doses of rac-albuterol. Fourteen patients provided >95% power to find a mean difference of 0.05 kPa h⁻¹, assuming a standard deviation of 0.02 kPa h⁻¹. Secondary outcome measures were differences between doses for change in P_aO₂ and heart rate.

Analysis

Change in mean P_aCO₂, P_aO₂ and heart rate for each dose of rac-albuterol was plotted against time and area under the time-response curves (AUC for 0–120 min) calculated by trapezoid integration. t-Tests were conducted to exclude carry-over and period effects between baseline readings [10] and paired t-tests to determine differences between doses. Within subject change in P_aCO₂ (AUC) following high dose rac-albuterol was related to change in P_aO₂ (AUC) and heart rate (AUC) and baseline measures using Pearson’s correlation. Analyses were carried out using the statistical software program, SPSS version 11 (SPSS inc, Chicago, USA), with statistical significance accepted at P < 0.05.

Results

Of 15 subjects who agreed to participate, one was withdrawn after the first study day following a fall at home. Mean demographic data for the 14 patients (three women) who completed the study is found in Table 1. Twelve patients had had a P_aCO₂ above 6 kPa on two occasions and two had developed hypercapnia in response to oxygen (2 l min⁻¹) during assessment for long-term domiciliary treatment.

Table 1.

Mean (SD) baseline data for the 14 subjects

Age (years)	66 (7)
FEV1 (l)	0.71 (0.3)
BMI (kg m−2)	28 (6)
Cigarette consumption (pack years)	53 (34)
Reversibility (ml)	109 (57)
PaCO2 (kPa)	7.9 (1.0)
PaO2 (kPa)	7.2 (1.3)
Number of subjects taking respiratory medication
Short-acting β2-adrenoceptor agonists	13
Long-acting β2-adrenoceptor agonists	9
Inhaled corticosteroids	9
Short-acting antimuscarinics	9
Long-acting antimuscarinic	0
Oral theophylline	2
Home oxygen	8

Baseline P_aO₂ showed a positive correlation with FEV₁ (r= 0.6, P = 0.02) and a negative correlation with baseline P_aCO₂ (r = –0.6, P = 0.02).

There was no carry-over or period effect for the two study treatments.

Mean changes with rac-albuterol

Mean P_aCO₂ for the 14 patients fell gradually over the study period following both doses of rac-albuterol, but with no difference between doses for P_aCO₂ AUC (Table 2). The mean fall after 120 min was 0.32 and 0.29 kPa for the high and low doses of rac-albuterol, respectively (Figure 1).

Table 2.

Between dose comparisons of mean AUC for P_aCO₂, P_aO₂ and heart rate over 2 h

PaCO2 AUC
Difference (high vs. low)	0.03 kPa h−1 (95% CI 0.02, 0.04), P = 0.57
PaO2 AUC
Difference (high vs low)	0.1 kPa h−1 (95% CI 0, 0.2), P = 0.04
Heart rate AUC
Difference (high vs. low)	1.1 beats min−1 h (95% CI 0.4, 1.8), P = 0.005

Figure 1.

Mean change in heart rate, arterial carbon dioxide (P_aCO₂) and oxygen (P_aO₂) tension over 120 min after high (▴) and low dose (▪) nebulized rac-albuterol (n = 14)

There was a progressive fall in mean P_aO₂ following both doses of rac-albuterol, with a significant difference between doses (mean difference in P_aO₂ AUC 0.1 kPa h⁻¹; 95% confidence interval (CI) 0, 0.2, P = 0.04). The reduction at 120 min was 0.57 and 0.24 kPa for the high and low doses, respectively (Figure 1). There was a greater rise in mean heart rate with the high dose compared with the low dose (mean difference in heart rate AUC 1.1 beats min⁻¹ h; 95% CI 0.4, 1.8, P = 0.005); the heart rate increase at 120 min in the two groups was 4.9 and 1.1 beats min⁻¹ (Figure 1).

Within subject changes with high dose rac-albuterol

The change from baseline P_aCO₂ 120 min after high dose rac-albuterol ranged from +0.8 kPa to −0.82 kPa (Figure 2), with 11 of the 14 subjects showing a fall in P_aCO₂. Patients showing a rise in P_aCO₂ were more likely to have a lower baseline P_aCO₂ and a smaller rise in heart rate with rac-albuterol, i.e. P_aCO₂ AUC showed a negative correlation with baseline P_aCO₂ (r = –0.6, P = 0.04) and heart rate AUC (r = –0.6, P = 0.02). There was no significant correlation between P_aCO₂ AUC and P_aO₂ AUC (r= 0.2, P = 0.44) or other baseline measures (P_aO₂r = 0, FEV₁r = 0.4, reversibility r = 0).

Figure 2.

The effect of high dose rac-albuterol on P_aCO₂ and P_aO₂ for each patient (n = 14), comparing baseline reading with the reading at 120 min postdose

Patients with the largest falls in PaO₂ had higher baseline FEV₁ and P_aO₂ values and a greater rise in heart rate, i.e. P_aO₂ AUC showed a negative correlation with FEV₁ (r = –0.6, P = 0.04), baseline P_aO₂ (r = –0.7, P = 0.01) and heart rate AUC (r = –0.53, P = 0.05).

Discussion

We examined the effects of high and low dose nebulized rac-albuterol on arterial blood gas tensions in patients with stable COPD who had limited bronchodilator reversibility and either resting hypercapnia or a raised P_aCO₂ in response to oxygen therapy. The higher dose of rac-albuterol produced a greater fall in mean P_aO₂ and increase in heart rate. However, the differences between doses were small. Neither dose of rac-albuterol caused a rise in mean P_aCO₂, although three subjects showed a small increase in P_aCO₂ with high dose rac-albuterol.

β-adrenoceptor agonists have reduced P_aO₂ in previous studies in patients with less severe COPD and the changes were attributed to an increase in ventilation/perfusion mismatching [11–13]. Although these studies have not shown a rise in P_aCO₂, other studies have documented an acute increase in CO₂ output of over 25% in response to modest rac-albuterol doses in normal subjects and patients with asthma [5, 6]. This increase in VCO₂ has been attributed to increased CO₂ flux to the lungs, due to increased cardiac output, and increased CO₂ production as metabolic rate and skeletal muscle tremor increase. We hypothesized that P_aCO₂ might rise following higher doses of rac-albuterol in patients with severe COPD, hypercapnia and limited bronchodilator reversibility due to an inability to increase ventilation in response to an increase in VCO₂. We initially chose to test the hypothesis on patients with severe COPD in a stable condition in whom hypercapnia may suggest limited ability to further increase ventilation.

The mean P_aCO₂ in our patients fell after high dose rac-albuterol, suggesting that either VCO₂ did not rise for most subjects or that ventilation increased appropriately in response to an increase in VCO₂ despite severe airflow obstruction and hypercapnia. We did not measure VCO₂ as it was important that patients were as relaxed as possible. All but one of our subjects was taking regular β₂-adrenoceptor agonists and they may therefore have developed tolerance, thus reducing the increase in tremor and cardiac output following rac-albuterol. Tolerance has been shown for the increase in CO₂ production with β-adrenoceptor agonists in normal subjects [14] and it would also explain the relatively small increase in heart rate following rac-albuterol in our study.

There was some variation in change in P_aCO₂ in response to the high dose of rac-albuterol and a rise in P_aCO₂ was seen in three of the 14 subjects. Contrary to expectations, patients showing a rise in P_aCO₂ were more likely to have had low baseline P_aCO₂ values and a smaller increase in heart rate. This could reflect relaxation during the study although we tried to ensure that subjects were relaxed and baseline measurements stable before giving rac-albuterol. Nevertheless patients who were less relaxed would have a lower P_aCO₂ and higher heart rate prior to treatment, and these would tend to increase and decrease, respectively, during the study. We also considered whether rac-albuterol might potentiate the effects of hypoxaemia on chemoreceptor stimulation and ventilation, as there is putative evidence of β-adrenoceptors on central chemoreceptors [15, 16]. However, change in P_aCO₂ did not correlate with either baseline P_aO₂ or change in P_aO₂, which may reflect differences between patients in their ability to increase ventilation, in the extent to which tolerance to β₂-adrenoceptor agonists had developed and/or the presence of different β₂-adrenoceptor polymorphisms within the group.

Whatever the mechanism behind these changes, our findings confirm that 10 mg nebulized rac-albuterol reduces P_aO₂ and suggest that an increase in P_aCO₂ following a high dose of rac-albuterol is uncommon in patients with stable severe COPD who take β-adrenoceptor agonists regularly. Whether an increase would occur more often during an acute exacerbation of COPD when patients are less able to increase ventilation is uncertain.