INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchial provocation tests (BPTs) are widely used both in the laboratory and in the field to identify persons with bronchial hyperresponsiveness (BHR) (1, 2). Histamine and methacholine, given by inhalation, are the most commonly used pharmacologic agents in such tests (3). Exercise and eucapnic hyperventilation tests are also used, particularly for identifying persons with exercise-induced bronchoconstriction (4, 5). BPT's are also used in patients with asthma to assess disease severity and response to treatment (4, 6, 7). In 1981 we reported that the airways of asthmatic patients narrowed in response to the inhalation of nonisotonic aerosols of sodium chloride (8). This finding was later confirmed and extended to show that increasing airway osmolarity with dextrose rather than sodium chloride also provoked airway narrowing in asthmatic individuals (9). Subsequently, at BPT done with hypertonic (4.5%) saline was developed and standardized for use in adults and children (10). Hypertonic saline has been used to assess the acute and chronic effect of drugs employed in treating asthma (11, 14).

The stimulus and mechanism whereby hypertonic saline causes airway narrowing is thought to respectively involve an increase in osmolarity and release of mediators from mast cells and sensory nerves (17). Antihistamines given both orally (18) and by aerosol (20) markedly reduce the airway response to hypertonic saline. In vitro, the human lung mast cell releases histamine in response to an increase in osmolarity (21). At the same osmolarity, mannitol is more potent than sodium chloride in causing this release of mast-cell histamine (21).

We therefore decided to investigate the effect of inhaling mannitol from a dry powder inhaler in asthmatic subjects known to be responsive to hypertonic saline delivered by an ultrasonic nebulizer. We measured the repeatability of the response to mannitol and the time course of recovery from mannitol challenge, and compared the responses to those with inhaled methacholine. The long-term aim of our study was the use of the dry powder preparation of mannitol for a BPT for diagnosing and assessing the severity of airway hyperresponsiveness, and for assessing the effects of medication used in the treatment of asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Forty-three subjects between 18 and 39 yr of age (32 females and 11 males) were recruited from the local community (Table 1). All subjects had been prescribed medication for their asthma. Twenty-four were taking inhaled corticosteroids on a daily basis, and 41 were taking ₂-agonists (Table 1). All subjects had at least one positive skin test (2 mm wheal or more) in response to a common aeroallergen (dust, grass, fungi, animal dander), and had a baseline FEV₁ of greater than 50%. All were required to have a 20% decrease in lung function (FEV₁) during challenge with 4.5% NaCl. The subjects were nonsmokers and had not had a chest infection in the previous 6 wk. They were asked to refrain from taking short-acting bronchodilators for 6 h and long-acting bronchodilators for 12 h prior to a study. No corticosteroids were taken by the subjects on the day of the study, and no antihistamines were taken for 3 d before the study day. The study was approved by the Central Sydney Area Health Service Ethics Committee (X93-0061), and all subjects signed a consent form prior to commencement of the study. The study was performed under the Clinical Trials Notification scheme Nos. 94-492 and 96-0809 of the Therapeutic Goods Administration of Australia.

Additionally, seven healthy subjects, aged 19 to 26 yr, were given an inhalational challenge with mannitol. They were selected on the basis of normal spirometry (flow rates and volume) and the lack of a history or symptoms of asthma or any other significant lung disease, a history of smoking, and a family history of asthma. Three members of this group had a positive skin test to a common aeroallergen but no symptoms of allergy.

Mannitol Capsule Preparation

The mannitol powder (Mallinckrodt AR grade and Mannitol BP; Rhône Poulenc Chemicals Pty., Ltd., Brookvale, NSW Australia) used in the study was prepared by spray drying (Buchii 190 Mini Spray Drier, Flawil, Switzerland) a solution containing mannitol at 15 mg/ml. The initial batch was prepared at Genentech (South San Francisco, CA), and the spray-dried powder was collected and used to fill glass vials in a laminar-flow hood, which were then shipped to Sydney. Bioburden analysis was done by Northview Pacific Laboratories, Inc. (Berkeley, CA). The second batch of mannitol powder was prepared at the University of Sydney School of Pharmacy. The batch was irradiated with approximately 4.7 kGy at Steritech (Wetherill Park, Australia). For this batch, a bioburden analysis was done at Stanford Laboratories (Rydalmere, Australia). The results for both yeast and mold showed a value less than 10 cfu/g, and no coliforms or other pathogens were detected.

The particle size of the powdered mannitol was measured with a multistage liquid impinger (Astra Pharmaceuticals; Lund, Sweden) and assayed by gravimetric analysis (Sartorius, Gottingen, Germany) or vapor-pressure osmometry (Knauer, Germany), which were always performed by the same investigator (K.C.). Gelatin capsules (Gallipot, St. Paul, MN) were hand-filled with 5, 10, 20, and 40 (± 0.2) mg of mannitol powder on an analytical balance (Sartorius BA11OS), as required under controlled conditions (relative humidity 40%, temperature 20 ± 1° C) in Sydney. The capsules were stored in a container with silica gel and kept in a cool and dry environment.

Delivery Device

Three devices were used to deliver the mannitol powder. All were single-dose devices, permitting different doses to be loaded during the challenge. Subjects 1 to 12 received the mannitol powder via a Halermatic (Fisons Pharmaceuticals Pty., Ltd, Loughborough, UK), Subjects 13 to 17 received the mannitol powder via an Inhalator (Boehringer Ingelheim Pty., Ltd., Ingelheim, Germany), and Subjects 18 to 43 received the mannitol via a Dinkihaler (Rhône-Poulenc Rorer, Collegeville, PA). The Inhalator and the Halermatic were chosen as they were available commercially and many of their characteristics are known (22). The Dinkihaler became available for evaluation with the second batch of mannitol.

Flow Measurement

For optimal airway deposition of mannitol with the Halermatic, the subjects were required to inhale the mannitol powder at a flow rate of 50 to 120 L/min. Because it was not possible to measure flow directly at each inhalation, a pressure transducer (Viggo-Spectromed DTX Disposable Pressure Transducer, Oxnard, CA) was used to approximate flow changes. To calibrate the pressure transducer, flows of 50 to 120 L/min were generated through a rotameter in line with the Halermatic device that was used to deliver the powder. Pressure changes were measured through a side port of the Halermatic at each flow rate (50 to 120 L/min in 10 L/min intervals), and the data were graphically related to provide estimates of flow measurements for each known pressure change. Pressure tracings were recorded on a chart recorder (Miniwriter Type WTR771A; Watanabe Instruments Corp., Tokyo, Japan) during the challenge to provide instantaneous readings. The traces were analyzed more accurately after testing.

For optimal airway deposition using the Inhalator, the subjects were required to inhale the mannitol powder at a minimum flow rate of 28 L/min. The inspired flow rate was checked by having the subject inhale maximally from the Inhalator while it was attached to an anemometer (AS 800; Minato Medical Science Co. Ltd., Osaka, Japan) prior to challenge on all of the test days. The best of three attempts was recorded. To calibrate the anemometer, flows of 25 to 95 L/min were generated through a rotameter (Series 2000; GEC-Elliott, Croydon, UK).

It was not possible to measure the flow rate through the Dinkihaler during the challenge with mannitol. All subjects had measurements made of their maximum peak inspiratory flow rate. Following this they practiced submaximal maneuvers in order to achieve an inspiratory flow rate of 80 to 120 L/min.

Lung-function Measurements

Spirometry was performed on an Autospiro AS-300 spirometer (Minato Medical Science Co. Ltd.), and the measurement of FEV₁ was used as an index of change in airway caliber. The predicted FEV₁ values used in the study were taken from Quanjer and colleagues (25). The spirometer was calibrated each morning, using a 2-L syringe.

Oxygen Saturation

Oxygen saturation (Sa_O2) was measured by oximetry as an index of safety during challenge with the dry mannitol (Ohmeda Biox 3700e; BOC Health Care, Louisville, KY).

Study Design

Subjects were asked to attend the laboratory on up to five occasions. The challenge with 4.5% NaCl (wet) was performed first, and thereafter the subject underwent either one or two challenge tests with dry mannitol, a challenge with methacholine chloride, or, in some cases, a further challenge with 4.5% NaCl.

4.5% NaCl (Wet) Challenge

Subjects were challenged with 4.5% NaCl generated from an ultrasonic nebulizer (MistO₂gen 143A; Timeter, Oregan Pike, PA) at a minimum rate of 1.2 ml/min according to the protocol of Smith and Anderson (10). The aerosol was inhaled via a two-way valve (No. 2700; Hans Rudolf, Inc., Kansas City, MO) for progressively increasing periods (i.e., 0.5. 1.0, 2.0, 4.0, and 8.0 min). FEV₁ was measured 60 s after each period of exposure. The test was terminated when the FEV₁ had fallen by 20% or more of the baseline FEV₁. All the subjects had a 20% decrease in FEV₁ after less than 23 ml of 4.5% NaCl (i.e., 1,035 mg of NaCl) had been delivered.

Mannitol Capsule Challenge

Baseline FEV₁ was measured on arrival of each subject at the laboratory, and was measured again 10 min later to establish its stability. Subjects underwent either one (n = 43) or two (n = 23) challenges with mannitol.

The dose protocol consisted of 0 (empty capsule acting as a placebo), 5, 10, 20, 40, 80, 160, 160, and 160 mg mannitol. The 40-, 80-, and 160-mg doses were given in multiples of either 20- or 40-mg capsules. Three FEV₁ maneuvers were performed 60 s after each dose, and the highest FEV₁ measurement was recorded. The FEV₁ value measured after the 0-mg capsule was used to calculate the percent decrease in FEV₁ in response to the mannitol challenge. If the subject had a decrease of > 10% then, the dose producing this was repeated for safety reasons. The challenge was stopped when a 20% decrease in FEV₁ was measured or a total cumulative dose of 635 mg had been given. The PD₁₅ and PD₂₀ for mannitol were calculated from the relationship between the percent decrease in FEV₁ and the cumulative dose of mannitol required to provoke this. To assess spontaneous recovery following completion of the first capsule challenge, the FEV₁ of 25 of the 43 subjects was measured at 5 min and then at 10 min intervals for at least 30 min and for a maximum of 60 min, until the FEV₁ had returned to within 5% of the baseline FEV₁ value. Following completion of the first or second challenge, 11 subjects (Subjects 13 to 23) received 0.5 mg terbutaline sulfate actuated into and inhaled from the Nebuhaler, and then underwent spirometry at 5- and 10-min intervals for 30 min or until the subject's FEV₁ had returned to within 5% of the baseline FEV₁ value.

For the healthy subjects the mannitol was given by the Dinkihaler and the challenge was stopped after 635 mg of mannitol had been inhaled.

Methacholine Challenge

Twenty-five subjects who inhaled mannitol from the Dinkihaler were also given an inhalation challenge with methacholine. The median time between challenges was 7 d (range: 1 to 34 d). DeVilbiss No. 40 (Somerset, PA) hand-held nebulizers were used to deliver the methacholine aerosol, using the protocol described by Yan and colleagues (26). A solution of methacholine chloride USP (ACI Chemicals, Brantford, ON, Canada) was prepared to a concentration of 5% wt/ vol in 0.9% saline. Solutions containing 0.3, 0.6, and 2.5% methacholine chloride were made by dilution of the 5% solution through the addition of 0.9% saline. The same nebulizer was used to deliver each concentration. Initially, the subject received one inhalation of methacholine, and the FEV₁ was measured 60 s later. The dose of methacholine was increased by giving one or more inhalations (from just below FRC to TLC) of each concentration (0.31, 0.63, 2.5, and 5%). The dose was increased if the decrease in FEV₁ was less than 10% of the baseline, and was repeated if the decrease in FEV₁ was 10% or greater. The output of each nebulizer was measured gravimetrically, using a solution of 0.9% saline. The dose of methacholine is expressed in micromoles. The dose increments were 0.1, 0.4, 0.8, 1.2, and 3.6 µmol, and the maximum cumulative dose did not exceed 7.8 µmol. Spontaneous recovery of FEV₁ for up to 60 min after methacholine and mannitol challenges were compared in 10 subjects.

Statistical Analysis

The baseline or prechallenge percent predicted FEV₁ and percent predicted FEV₁ following the 0 mg capsule were expressed as mean ± SD and compared, using Student's t test for paired values. The geometric mean ± 95% CI and range were calculated for the PD₁₅ (mg) and PD₂₀ (mg) values, and the log PD₁₅ (mg) and log PD₂₀ (mg) values were also compared again using Student's paired t test.

The Pearson correlation (r_p) and significance values were calculated for the relationship between the 4.5% NaCl and the mannitol challenge, and for the mannitol and methacholine challenges.

The repeatability of the two mannitol capsule challenges was expressed in terms of doubling doses according to the equation of Peat and associates (27), and was illustrated according to Bland and Altman (28).

The peak inspiratory flow rates (L/min) were calculated for the three dose delivery devices separately, and expressed as median and range values. The duration of the challenge and the number of capsules used were expressed as median values.

An analysis of variance (ANOVA) was used to compare the values for FEV₁ during spontaneous recovery and recovery with bronchodilator. In addition, Student's paired t test was used to compare the spontaneous recovery values at 30 and 60 min and the recovery values following bronchodilator at 5 and 60 min. For this analysis, all recovery values of FEV₁ exceeded the baseline FEV₁ were considered to be 0 (i.e., FEV₁ had returned to the baseline value).

For the healthy subjects, the response-dose ratio was calculated as the percent decrease in FEV₁ at the end of challenge divided by the cumulative dose of mannitol given to induce that decrease in FEV₁ in a manner similar to that previously described for histamine (29). The values were log transformed for analysis. The geometric mean value and 95% confidence intervals (CIs) for the healthy subjects were obtained and expressed as percentage decreases in FEV₁ per milligram of mannitol delivered to the subject.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Individual dose-response curves for the initial challenge with mannitol powder are illustrated in Figures 1 and 2. The geometric mean for the PD₁₅ for mannitol was 64 mg (CI: 45 to 91 mg) (n = 43) and that for the PD₂₀ 89 mg (CI: 65 to 120 mg) (n = 42). The FEV₁ of one subject did not reach a 20% decrease with a PD₁₅ 343 mg. Values for PD₁₅ and PD₂₀ for the different dose-delivery devices are given in Table 2.

View larger version (32K): [in this window] [in a new window]   Figure 1.   Individual dose-response curves relating the percentage decrease in FEV1 to the cumulative dose of a dry powder of mannitol in the 17 subjects who inhaled mannitol for the first time through the Halermatic (n = 12) (solid line) and the Inhalator (n = 5) (broken line).

View larger version (34K): [in this window] [in a new window]   Figure 2.   Individual dose-response curves relating the percentage decrease in FEV1 to the cumulative dose of a dry powder of mannitol in the 26 subjects who inhaled mannitol for the first time through the Dinkihaler.

The values for the PD₁₅ and PD₂₀ obtained with mannitol were significantly lower (p < 0.001) than those observed when the wet aerosol of saline was inhaled (PD₁₅ geometric mean [GM]: 155 mg [CI: 109 to 223 mg]); PD₂₀ GM: 220 mg [CI: 162 to 299 mg]. The relationship between the wet aerosol of NaCl and mannitol is illustrated in Figure 3 (r_p = 0.54 p < 0.001, n = 43).

View larger version (14K): [in this window] [in a new window]   Figure 3.   Values for PD15 for wet aerosol of NaCl in relation to the PD15 for mannitol (rp = 0.54; p < 0.001; n = 43).

All 25 subjects who inhaled methacholine had a positive response (i.e., PD₂₀ less than 7.8 µmol). The GM (95% CI) for the PD₂₀ of methacholine was 0.7 µmol (CI: 0.41 to 1.2 µmol), and all had a PD₁₅ of mannitol of less than 400 mg (Figure 4). Thirteen subjects had a PD₂₀ of methacholine of less than 1 µmol (i.e., in the moderately responsive range). For these subjects, the GM of the PD₁₅ of mannitol was 43 mg (CI: 21 to 87 mg), and 10 of 13 had a PD₁₅ < 75 mg. The 12 subjects with a PD₂₀ of methacholine exceeding 1 µmol (i.e., in the range of mild hyperresponsiveness) had a GM PD₁₅ of 119 mg (CI: 81 to 174 mg). There was a significant relationship between the PD₁₅ of mannitol and the PD₂₀ of methacholine (r_p = 0.50, p < 0.05). The mildest response to methacholine was a PD₂₀ of 7.38 µmol, and the subject for whom this occurred had a mild response to mannitol with a PD₁₅ of 344 mg.

View larger version (16K): [in this window] [in a new window]   Figure 4.   Individual values for the PD15 for mannitol in relation to the PD20 for methacholine. The axis are marked in micromoles, as used in Yan and colleagues technique (26), and the equivalent concentrations are marked in milligrams per milliliter as used in Cockcroft and coworkers technique (1). The values for mild, moderate, and severe are those commonly used to describe bronchial responsiveness to methacholine.

The FEV₁ percent predicted before the mannitol challenge was 82.9 ± 12.9%. There was a small but significant difference in the FEV₁ percent predicted values measured on the day on which the wet aerosol of 4.5% NaCl was given (85.5 ± 13.6%; p < 0.0025).

For the group of 23 subjects who underwent two challenges with mannitol, there was no significant difference for the baseline mean ± 1 SD FEV₁ values expressed as a percentage of the predicted FEV₁ between the two test days (first mannitol challenge: 83.4 ± 15.5%; second mannitol challenge: 82.7 ± 16.3%; p = 0.6). There was also no significant difference between the mean ± 1 SD FEV₁ expressed as a percentage of the predicted FEV₁ on the two mannitol test days, following the 0-mg capsule (first challenge 82.8 ± 15.6%; second challenge 81.1 ± 16.7%; p = 0.2).

All 23 subjects achieved a 15% decrease in FEV₁ on repeat challenge with dry mannitol powder. The repeatability of the PD₁₅ was independent of dose (r_p = 0.07, p = NS) (Figure 5). There was no significant difference between the PD₁₅ on both test days (GM [95% CI]: 59 mg [36 to 97 mg] versus 58 mg [35 to 94 mg]; p = 0.91; n = 23]. The median time (range) between challenges was 6 (2 to 36) d. The SD of the difference between the log₁₀ PD₁₅ for the two challenge tests was 0.34, with the mean difference being 0.008. The repeatability of the PD₁₅ for the dry mannitol challenges was ± 1.64 doubling doses (i.e., there was a 95% probability that a PD₁₅ FEV₁ measurement is either 1.64 doubling doses below or above the initial PD₁₅ FEV₁). The repeatability of the PD₁₅ for the three devices is given in Table 2 together with the inspiratory flow rates and the percentage of the particle mass in the aerosol cloud that was below 7 µm. From the in vitro analysis, the dose emitted, as a percentage of the loading dose (3 × 40 mg), was 79.3% for the Dinkihaler and 67.2% for the Inhalator. For the Dinkihaler, 47%, and for the Inhalator, 66.9% of the aerosol cloud was below 7 µm. Thus, for the Dinkihaler 37.4%, and for the Inhalator, 44.9% of the loaded dose had a particle size less than 7 µm.

View larger version (15K): [in this window] [in a new window]   Figure 5.   A Bland and Altman plot (29) relating the geometric mean for the PD15 for the first and second challenges with mannitol to the difference between the log10PD15 values for the 23 subjects who underwent repeated challenge. The dashed line illustrates the point of no difference between the first and second challenge. The repeatability was independent of the dose (rp = 0.07; p = NS) The difference in log10PD15 values for all but one subject was ± 0.5.

The mean ± SD for the maximum percent decrease in FEV₁ for the normal healthy subjects was 3.03 ± 1.49% after 635 mg had been inhaled. The geometric mean of the response-dose ratio was 0.0043 (95% CI: 0.0023 to 0.008).

Oxygen Saturation During Mannitol Challenge

The initial challenge value and the lowest Sa_O2 value measured were used to calculate the decrease in Sa_O2 during challenge. All subjects had an Sa_O2 measurement on at least one occasion during the challenge with mannitol. During the challenge with mannitol, the Sa_O2 fell by 2% or less in 37 of the 43 subjects. For the remaining six subjects the decrease in Sa_O2 was 3%. The lowest Sa_O2 measured during mannitol challenge was 93%.

Recovery Following Mannitol Capsule Challenge

Spontaneous recovery was measured in 25 of the subjects following the first mannitol challenge. The mean maximum decrease in FEV₁ expressed as a percentage of baseline was 26.4 ± 4.9%. At 30 min after challenge, the mean (± SD) percent change from the baseline FEV₁ was 11.9 ± 8.0%, and at 60 min this change was 5.4% ± 4.8. The FEV₁ had returned to 95% of the baseline value within 1 h in 18 of the 25 subjects.

Recovery following bronchodilator was compared with spontaneous recovery in 11 subjects (Figure 6). Five minutes after the administration of bronchodilator, the mean (± SD) percent change from the baseline FEV₁ was only 7 ± 4.3%, and was significantly different from the 22 ± 6.0% change observed when recovery was spontaneous (p = 0.001). On the day on which bronchodilator was given, all subject's FEV₁ values returned to baseline by 20 min, and this compared with a mean (± SD) percent reduction in the baseline FEV₁ of 18 ± 5.9% for the spontaneous recovery day (p = 0.004). At 60 min the mean (± SD) percent reduction in baseline FEV₁ was 5 ± 5.8% for the spontaneous recovery day, and this was not significantly different from the value on the day on which bronchodilator was given (p = 0.17).

View larger version (13K): [in this window] [in a new window]   Figure 6.   The mean ± SEM for FEV1 expressed as a percentage reduction from the baseline prechallenge value in the 11 subjects who spontaneously recovered after the first challenge with dry mannitol (solid line), and who were given 0.5 mg of terbutaline aerosol immediately after the second challenge (broken line). The value at time 0 was the maximum reduction in FEV1 recorded, and the time after bronchodilator or spontaneous recovery is shown. For eight of the 11 subjects, the FEV1 had returned to within 90% of baseline within 10 min after taking bronchodilator. For the remaining subjects, recovery took 20 to 30 min.

The mean ± SD percent reduction in FEV₁ from the baseline following methacholine was 26.5 ± 5.1%, and following mannitol challenge was 27.1 ± 5.2% (p = NS). There was no significant difference in the time required for spontaneous recovery from the two challenges (Figure 7).

View larger version (17K): [in this window] [in a new window]   Figure 7.   The mean ± SEM for FEV1 expressed as a percentage reduction from the baseline value in 10 subjects who spontaneously recovered after challenge with mannitol and after challenge with methacholine. There was no significant difference in the recovery between the two challenge tests.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results of this study clearly demonstrate that a dry powder preparation of mannitol, delivered from a capsule via a Halermatic, Inhalator, or Dinkihaler device, can provoke airway narrowing in asthmatic subjects who are sensitive to a wet aerosol preparation of 4.5% NaCl and methacholine. The subjects studied differed in the severity of their BHR, as demonstrated by the wide range of their PD₂₀ values for the wet NaCl aerosol and methacholine. They similarly differed in their sensitivity to mannitol, with the most sensitive subjects responding with a 15% decrease in FEV₁ at a dose of 2 mg and the least sensitive subject responding at a dose of 343 mg. All our asthmatic subjects had the expected response, and the healthy subjects did not have a decrease in FEV₁.

This study was designed to determine the feasibility of using a dry powder preparation of mannitol in place of a wet aerosol of hypertonic saline as an osmotic challenge, and to evaluate methacholine responsiveness in the same subjects. It was not designed to determine sensitivity and specificity of airway responses to mannitol in a random population, since greater numbers of subjects would be required for this. However, it is important that we did not document airway responsiveness in the healthy subjects that we did study. We studied only seven healthy subjects with mannitol, since we did not have sufficient powder available, and used what we had preferentially for investigating asthmatic subjects. The percent decrease in FEV₁ observed was similar to that observed for a group of healthy subjects who inhaled hypertonic (4.5%) NaCl, which was 4.6% (SD = 3.1) (13).

This study was not designed to compare different dose- delivery devices, but to see whether different devices could be used to achieve the same outcome. There were no qualitative differences between the devices, and any one of them could be used for the challenge. There was a small difference between the Dinkihaler and Inhalator for the percentage of the loaded dose that was of less than 7 µm. The percentage of the loaded dose remaining in the capsule for the Halermatic is not available, but is likely to have been similar to that for the Inhalator, since the same batch of mannitol was used for both devices. Although there are differences between devices with respect to the percentage of the loaded dose having particles in the respirable range, the outcome (i.e., a positive test in response to mannitol) was the same. An analogy can be made to the use of different nebulizers with different outputs to deliver methacholine or histamine (1).

The response to the dry mannitol powder had good repeatability, and compares well with the response to other provocation tests, including both pharmacologic and osmotic challenges (14, 30). The technique used here to analyze repeatability has been previously used for BPTs (27).

We do not know where in the respiratory tract the mannitol was deposited or what percentage of the inhaled dose was deposited in the lower respiratory tract. However, the repeatability of the response suggests that there was no major difference in amount or site of deposition of the mannitol on the two test days. The relatively small changes observed in Sa_O2 suggest that the site of deposition was more likely to have been the larger airways. Further studies with labeled mannitol are required to determine the ratio of peripheral to central deposition of the powder in the airways. There were no adverse events requiring medical intervention, and Sa_O2 remained at 93% or greater during challenge with mannitol. Spontaneous recovery occurred within 60 min after the first challenge with mannitol, but two subjects required bronchodilator during recovery after the second challenge. The recovery following the standard dose of a ₂-adrenoceptor agonist given from a pressurized metered dose inhaler (MDI) via a spacing device was rapid. The FEV₁ values of 60% of the subjects had recovered to within 5% of their baseline FEV₁ values within 10 min.

The mannitol was extremely well tolerated, and there was no difficulty encountered in inhaling the powder at the flow rates considered to be optimal for good deposition in the respiratory tract (22). The inhalation of mannitol did not provoke significant cough or any gagging. Cough did occur in some patients, but at a time at which deposition of the mannitol would have occurred. Cough and gag have been a problem for some subjects when inhaling dry powder preparations of NaCl (34). Subjects spontaneously remarked that they thought they were inhaling "icing sugar". There was an expressed preference for the mannitol over the wet NaCl aerosol.

The highest dose we gave for one inhalation of mannitol was 40 mg. The ability to use this dose without problems is important, as it will permit persons with mild BHR to be identified with less than 15 capsules of mannitol. Based on the methacholine challenge, our mildest patient required 343 mg of mannitol for a 15% decrease in FEV₁, and this required 12 capsules.

We used asthmatic subjects, with all but two taking medication regularly, with known hyperresponsiveness to hypertonic saline, a feature considered to be consistent with currently active, asthma (11). Thus, the subjects' bronchial responsiveness to hypertonic saline was not completely controlled with inhaled corticosteroids, as has been found for some asthmatic individuals taking similar doses (16). Asthmatic individuals responsive to hypertonic saline are likely to have a PD₂₀ to methacholine or histamine of less than 8 mg/ml or 4 µmol (13, 15, 34), and this was confirmed in the present study. We analyzed the PD₁₅ data because a 15% decrease in FEV₁ is greater than the mean plus 3 SD in healthy subjects in response to inhaling a wet aerosol of 4.5% NaCl (12, 13), and correspondingly, subjects we studied had a mean + 3 SD of 7.5% for their FEV₁ with dry mannitol. Using a 15% decrease in FEV₁ as a cutoff for abnormality would reduce the time of challenge for most patients to less than 12 min, reducing the number of capsules and the time for spontaneous recovery.

Mannitol (C₆H₁₄O₆; MW: 182) is a naturally occurring sugar alcohol found in most vegetables. It is not absorbed by the gastrointestinal tract and is not metabolized to any appreciable extent when injected. Mannitol is an antioxidant and is commonly used as an excipient for tablets, and up to 20 g may be ingested per day. It is also used intravenously in a dose of 13,000 mg (for a 65 kg subject) given over 3 to 5 min to treat cerebral edema. It is a stable substance and flows well when prepared as a spray-dried powder because the particles are close to being spherical. As a spray-dried powder, mannitol retains its crystalline structure and resists moisture resorption at high relative humidities. These characteristics make it an attractive substance to encapsulate for inhalation (35). This is the first long report of the airway narrowing effects of dry particles of mannitol in known asthmatic subjects. The challenge with mannitol would appear to be as safe and reproducible as any other challenge that we have experience with in the laboratory (i.e., methacholine, histamine, exercise, hyperventilation and hypertonic saline). There were no adverse events requiring medical intervention. The test was well tolerated by all subjects, and the airway narrowing was rapidly reversed after inhalation of a standard dose of bronchodilator. Spontaneous recovery compared well with that following methacholine challenge.

The advantage of dry powders over the existing wet aerosol challenge is the simplicity and inexpensive nature of the equipment used for their delivery. The capsule system for delivering dry mannitol, and the disposable nature of the device, could provide an end to the need and time for cleaning and maintaining equipment for bronchial provocation testing. Although initially such a challenge may be used only by respiratory specialists and in hospital outpatient departments, it has the potential to appeal to the wider health-care community. The potential advantage in using an osmotic challenge rather than a pharmacologic challenge with methacholine is likely to be in the greater specificity for identifying persons with BHR responsive to treatment with asthma drugs (11, 15, 16). Further studies are now required to define the specificity and sensitivity of the airway response to mannitol, and to compare the response to mannitol with that to other stimuli known to narrow the airways of asthmatics subjects, such as exercise.