Assessment of different methods of inhalation from salbutamol metered dose inhalers by urinary drug excretion and methacholine challenge

Abstract

Aims

Methods to determine the lung delivery of inhaled bronchodilators from metered dose inhalers include urinary drug excretion 30 min post inhalation and methacholine challenge (PD₂₀). We have compared these two methods to differentiate lung delivery of salbutamol from metered dose inhalers using different inhalation methods.

Methods

In phase 1 of the study, on randomized study days, 12 mild asthmatics inhaled placebo, one and two 100 µg salbutamol doses from a breath actuated metered dose inhaler, in randomized fashion on different days. In phase 2, they inhaled one 100 µg salbutamol dose from a metered dose inhaler using a SLOW (20 l min⁻¹) and a FAST (60 l min⁻¹) inhalation technique and a slow inhalation delayed until after they had inhaled for 5 s (LATE). Urinary excretion of salbutamol (0–30 min) and PD₂₀ were measured after each dose.

Results

Following placebo, one and two 100 µg salbutamol doses, the geometric mean for PD₂₀ was 0.10, 0.41 and 0.86 mg respectively and the mean (SD) urinary drug excretion after one and two doses was 2.25 (0.65) and 5.37 (1.36) µg, respectively. After SLOW, FAST and LATE inhalations the geometric mean for PD₂₀ was 0.50, 0.40 and 0.42 mg, respectively, and mean (SD) salbutamol excretion was 2.67 (0.84), 1.90 (0.70) and 2.72 (0.67) µg, respectively. Only the amount of drug excreted during the FAST compared with the SLOW and LATE inhalations showed a statistical difference (95% confidence interval on the difference 0.12, 1.54 and 0.06,1.59 µg, respectively).

Conclusions

Urinary salbutamol excretion but not PD₂₀ showed differences between the inhalation methods used. When using a metered dose inhaler slow inhalation is better and co-ordination is not essential if the patient is inhaling when they actuate a dose of the drug.

Introduction

Various methods have been proposed to characterize the lung deposition of a drug after it has been inhaled [1]. These include clinical endpoints such as spirometry and bronchoprovocation, pharmacokinetic methods and gamma scintigraphy. The relative bioavailability of salbutamol to the lungs is one of the proposed pharmacokinetic approaches. The amount of drug excreted in the urine during the first 30 min after the start of an inhalation is measured [2]. The method is based on the principle that following inhalation, there is a delay for the swallowed fraction of the inhaled dose being absorbed and appearing in the urine, whereas the salbutamol deposited into the lungs is immediately delivered to the systemic circulation and excreted in the urine. This method has been used to compare drug delivery to the lungs by different inhalers and inhalation methods [1]. A linear dose–response relationship for up to five doses has been demonstrated [3] and the method is reproducible [2, 3].

Clinical endpoints are regarded as the gold standard by the Regulatory Authorities to demonstrate the bioequivalence of inhaled products [4]. Standard bronchodilator methods using spirometry have a shallow dose–response relationship [5] and thus guidance on the use of a bioassay using bronchoprovocation by methacholine (PD₂₀) has been issued [4]. Preliminary indications are that a dose–response relationship exists [6] and that the method is reproducible [7].

The influence of the inhalation method in a bioequivalence study comparing inhaled products has not been determined. It has recently been shown that patients inhale at a fast rate when using their metered dose inhalers [8]. There is also confusion over the importance of co-ordination between dose release and the start of an inhalation. Studies that have recommended split second co-ordination have focused on releasing the dose before inhaling [9, 10] whereas it has been shown that if a dose is actuated during an inhalation, lung deposition is only slightly decreased [11].

The effect of inhalation technique on the relative lung deposition of salbutamol has not been determined using two methods simultaneously. The primary aim of this study was to determine the potential of urinary salbutamol excretion data and the methacholine challenge bioassay to describe the effect of inhalation methods on the lung deposition of salbutamol using a metered dose inhaler. Any difference between the inhalation methods will have practical implications on how patients should use their metered dose inhalers.

It has been recommended that when using bronchoprovocation end points a two-fold difference in response should be demonstrated for a doubling of the dose [6]. Therefore the study was divided into two phases. The first one involved different doses of salbutamol inhaled by the same method. The second phase was the comparison between the inhalation methods using the patients who completed phase 1. In addition, owing to variability in dosage emission from DeVilbis nebulizers [11, 12], in vitro characterization of the methacholine solutions was carried out.

Methods

Dosage characteristics of the methacholine emitted from each nebulizer

The bronchoprovocation method of Yan et al. [14] using DeVilbis no. 40 (DeVilbis, Co, USA) hand operated nebulizers was followed, each device containing either 0, 1.56, 3.15, 6.25, 12.5, 25.0, 50.0 or 100 mg ml⁻¹ methacholine in 0.9% sodium chloride. The pH and osmolarity of the solutions were measured and aerodynamic dose emission characteristics from each nebulizer were determined using a Malvern 2600 laser particle sizer (Malvern Instruments, UK).

Subjects

Approval was obtained from the Bradford Local Research Ethics Committee and all subjects gave written informed consent. All patients were mild asthmatics, were over 18 years and had a % predicted FEV₁ >80% and >15% reversibility to inhaled salbutamol. Those with a known hypersensitivity to inhaled salbutamol, and those who had taken theophylline, an inhaled long acting β-adrenoceptor agonist, inhaled anticholinergic therapy or oral prednisolone in the past 3 months, or who had a known history of sudden severe asthma were excluded. Their prescribed reliever was switched to terbutaline for the duration of the study. On the study days subjects had to demonstrate a less than 5% change in their FEV₁ following inhalation of 0.9, 1.8, 2.7 and 4.5% sodium chloride from a DeVilbis nebulizer. Baseline FEV₁ had to be within 10% of prestudy values. During these 4 prestudy day visits patients were also taught how to inhale salbutamol from the Easi-Breathe (IVAX, UK) metered dose inhaler and were told that they would need to provide a urine sample 30 min post inhalation. Patients were instructed to maintain their normal fluid intake. The inhaled doses were administered after the last FEV₁ measurement.

Study design

There were two phases to the study with each consisting of three visits. All appointments were scheduled for the same time each morning after a 12 h bronchodilator free period. Subjects then inhaled four doses of 0.9% sodium chloride from a DeVilbis Nebulizer to demonstrate a less than 10% change in FEV₁.

In phase I subjects inhaled from an active and placebo Easi-Breathe metered dose inhaler randomized as two doses from the placebo inhaler, or one dose from each inhaler, or two active doses. Each administration was separated by more than 7 days. Subjects voided their urine 15 min predosing and then provided a urine sample 30 min after the start of the first inhalation. FEV₁ was then measured and the bronchoprovocation challenge, using methacholine, was carried out according to the procedure described by Yan et al. [14] until a fall of at least 20% in the FEV₁ was detected. The PD₂₀ was estimated as the amount of methacholine that reduced the FEV₁ by 20%.

In phase II subjects were randomized to inhale one dose from the Easi-Breathe metered dose inhaler using

1. the recommended method of slow inhalation (the Easi-Breathe metered dose inhaler activates at an inhalation flow rate of 20 l min⁻¹) [SLOW].

2. Same technique as (1) except that dose release was delayed until each subject inhaled slowly for 5 s [LATE] and

3. a technique where the inhaler was modified to actuate at 60 l min⁻¹ and patients were trained to inhale at a rapid rate [FAST].

Urine collection and measurement of PD₂₀ were the same as phase I.

Data analysis

Salbutamol urine concentrations were measured by high performance liquid chromatography as described previously [2]. The limit of detection was 25 µg l⁻¹. The intra- and inter day coefficient of variation over the range of 25–1000 µg l⁻¹ was 4.0–2.4 and 7.5–3.2%, respectively. The mean (SD) measured concentration of urine standards containing nominal concentrations of 75 and 250 µg l⁻¹ were 74.0 (3.7) and 252.8 (3.3) µg l⁻¹, respectively. These measured values corresponded to 98.6 and 101.1% of the nominal values.

Log transformed PD₂₀ values were expressed in terms of doubling doses using data between the active and placebo doses [15]. Differences in FEV₁, doubling doses and urinary salbutamol excretion between the different inhalation techniques were examined using a one way anova with the application of a Bonferroni correction.

Results

In vitro emitted dosage characteristics for the methacholine from each of the DeVilbis nebulizers are shown in Table 1. The methacholine dose delivered from each nebulizer was determined from these data and each nominal PD₂₀ dose was adjusted to the emitted fine particle dose. The 0.9, 1.8, 2.7 and 4.5% solutions of sodium chloride had a pH of 5.5 and osmolarity ranging from 291 to 1437 mOsmol.

Table 1. Characteristics of the methacholine dose emitted from each nebulizer.

Nebulizer	Nominal methacholine concentration (mg ml−1)	Mean osmolarity (mOsmol) (n = 3)	pH	% coefficient of variation for the emitted dose dose	Mass median diameter (µm)	Fine particle dose (%)
1	0	291	5.5	2.7	10.1	31.8
2	1.56	309	5.5	4.1	11.4	29.1
3	3.13	322	5.3	5.6	11.7	27.0
4	6.25	350	5.3	3.3	12.4	26.8
5	12.5	409	5.0	3.7	8.9	36.1
6	25.0	522	5.0	4.0	11.4	28.5
7	50.0	753	4.7	3.1	7.7	39.1
8	100.0	1283	4.4	4.6	11.4	28.0

Fine particle dose = % of the nominal amount that contains particles < 5 µm.

Thirty-nine patients were initially recruited but only 12 completed the screening process and agreed to take part in the rest of the study. Their mean (SD) age was 21.1(2.0) years, weight 70.2 (17.2) kg and height 1.69 (0.09) m. The mean (SD) FEV₁ was 3.15 (0.53) l corresponding to a percentage predicted value of 87.1 (5.33)%. Statistical analysis of baseline FEV₁, prior to the inhalation of placebo, and the two doses of salbutamol did not reveal any period, sequence or treatment effects, and they did not differ from prestudy baseline values. Similarly FEV₁ values 30 min post inhalation (immediately before the methacholine challenge) did not change.

The mean (SD) PD₂₀ values following inhalation of 0, one and two 100 µg of salbutamol were 0.21 (0.24), 0.74 (0.70) and 1.69 (1.60) mg with geometric means of 0.10, 0.41 and 0.87 mg. The mean (SD) doubling doses values after the one and two 100 µg doses were 2.07 (0.94) and 3.14 (1.04), respectively. Individual values were all above 1 and 2 doubling doses, respectively. The mean (SD) amounts of salbutamol excreted in the urine following the one and two 100 µg doses were 2.25 (0.65) and 5.37 (1.36) µg, respectively. No salbutamol was detected in all the urine samples after placebo inhalation. Differences in the doubling dose and urinary salbutamol data between administrations of one and two 100 µg doses of salbutamol were each highly significant (P < 0.001). The mean difference (95% confidence interval) between the one and two 100 µg doses was 1.07 (0.88, 1.26) doubling doses and 3.12 (2.33, 3.91) µg of salbutamol excreted in the urine in the first 30 min post inhalation.

Baseline FEV₁ prior to the different inhalation methods (SLOW, FAST, LATE) was similar and there was no difference in each of these values compared with the prestudy baseline FEV₁ measurements. Statistical analysis highlighted no difference in the FEV₁ following each of the three different inhalation methods.

Individually log transformed PD₂₀ values following the different inhalation methods and the placebo dose are shown in Figure 1 and Table 2. The differences between the doubling doses were not statistically significant. Similar analysis of the non- and natural log transformed PD₂₀ data also showed no differences, as did analysis using the uncorrected PD₂₀ data.

Figure 1.

Individual PD₂₀ data following placebo and one 100 µg salbutamol dose using the SLOW, FAST and LATE inhalation methods

Table 2. Methacholine challenge and urinary salbutamol excretion data from phase 2 of the study.

	Inhalation method (mean (SD) unless stated)
	SLOW	FAST	LATE
PD20 (mg)	1.04 (1.28)	0.65 (0.55)	0.81 (0.88)
PD20 (geometric mean)	0.50	0.40	0.42
Doubling doses	2.25 (1.01)	1.95 (1.13)	2.15 (1.02)
Urinary salbutamol excretion (µg)	2.67 (0.54)	1.90 (0.70)	2.72 (0.67)
Mean difference (95% confidence interval) between the inhalation methods
	SLOW vs FAST	SLOW vs LATE	LATE vs FAST
Doubling doses	0.30 (−0.78, 1.39)	0.10 (−0.98, 1.18)	0.20 (−1.29, 0.88)
Urinary salbutamol excretion (µg)	0.78 (0.12, 1.54)*	−0.05 (−081, 0.71)	0.83 (0.06, 1.59)*

^*P < 0.05 otherwise not significant.

The amounts of salbutamol excreted in the 30 min after dosing for each inhalation method are shown in Figure 2 and Table 2. Significantly more salbutamol was excreted in urine using the SLOW and LATE methods compared with the FAST method.

Figure 2.

The amount of salbutamol excreted in the urine during the first 30 min after the inhalation of one 100 µg salbutamol dose using the SLOW, FAST and LATE inhalation methods

Urinary salbutamol excretion following inhalation of one dose in Phase I and SLOW inhalation in Phase II (identical inhalation methods) did not differ (95% confidence interval on the difference, −0.22, 1.06). Furthermore, no difference was found for the methacholine doubling dose values between these two study doses (95% confidence interval on the difference −0.12, 0.50).

Discussion

The urinary salbutamol excretion and PD₂₀ dose–response relationships are consistent with previous studies [3, 6]. Although only 12 asthmatic subjects completed the protocol, measurements of urinary salbutamol excretion were able to demonstrate a significant difference between the inhalation methods, in contrast to the methacholine challenge method. The results suggest that patients should be encouraged to inhale slowly and that it is acceptable to actuate a dose after they have started to inhale.

No difference in methacholine challenge results following inhalation from a metered dose inhaler and when it is attached to a spacer have been reported [16]. In contrast, differences have been observed for salbutamol urinary excretion [17] and plasma drug concentrations [18] and also using gamma scintigraphy [19, 20]. These data are consistent with our results and confirm limitations in the use of some clinical end-points to determine lung deposition.

Inter-subject variability in urinary salbutamol excretion was low and similar to that reported in previous studies [2, 3, 9]. The larger intersubject variability in the response to inhaled methacholine indicates that larger numbers need to be used in this type of bioequivalence study [4, 6]. In the present work with methacholine, to detect a 20% difference in the doubling doses measurement (80% power at the 5% significance level) would have required 80 subjects.

The results of the in vitro characterization of the methacholine solutions indicate that variability due to dose emission and bronchoconstriction (caused by changes in pH and osmolarity) was minimized. The significance of pH or osmolarity with respect to inhaled methacholine solutions or the quality of the nebulized methacholine dose have never been determined during challenge studies. Nebulizer solutions should be isotonic with a pH greater than 5 [21], but the two most concentrated methacholine solutions had a pH of less than 5. Whether or not these exaggerated the subsequent bronchoconstriction is not known. No significant changes in FEV₁ have been reported after the inhalation of solutions with osmolarities between 150 and 549 mOsmol [22]. Although the two most concentrated methacholine solutions had an osmolarity above this range, they were not affected by the nebulized 4.5% sodium chloride solution which had an osmolarity of 1437 mOsmol. These results and the variablility in dosage emission confirm the need to identify such factors during challenge studies of bioequivalence.

A recent study [8] of almost 400 patients, of all ages and severity of airways obstruction, revealed inhalation rates of greater than 100 l min⁻¹ through a metered dose inhaler. Our results show that emphasis should be placed on advising patients to slow down their inhalation rate when using a metered dose inhaler. We achieved the latter by training patients to inhale over 5 s. The similar amounts of salbutamol excreted in the urine after slow and late inhalations suggest the patient should be inhaling when a dose is actuated rather than be focusing on split second co-ordination of dose release at the start of their inhalation.

In conclusion the results show that urinary salbutamol excretion measurements have considerable potential for the determination of the relative lung deposition of inhaled salbutamol products. The differences between the inhalation methods confirm the importance of training of the inhalation technique during bioequivalence studies. In clinical practice patients should be encouraged to use a slow inhalation technique even if the dose is actuated after they have started to inhale.