Hyperosmolarity as the stimulus to asthma induced by hyperventilation?

I-lyperosmolarity as the stimulus to asthma induced by hyperventilation? Christine Camper-down,

M. Smith,

B.Sc., Grad. Dip. Phty., and Sandra

D. Anderson,

Ph.D.

Australia

tiyperosmolarity of the epithelial jluid of the large airways caused by evaporative water loss (wloss) has been proposed as the stimulus to exercise-induced asthma. The aim of this study was to compare the wloss during hyperpnea with a theoretical wloss from a known hypertonic stimulus in order to determine whether comparable volumes of wloss will induce the same response. Since wloss also occurs during isocapnic hyperventilation (ISH), we decided to compare the airway response to ISH with the response obtained after inhaling 4.5% NaCl aerosol. Changes in FEV, were measured in I7 subjects with asthma in response to increasing rates of ventilation (ISH) and increasing doses of 4.5% NaCl aerosol. For ISH, wloss was calculated at 29 mglL of expired air and for 4.5% NaCl, at 4.0 ml11 ml of aerosol inhaled, as this is the volume of water that will bring the periciliary fluid to normal tonicity. Two doseresponse curves were drawn for each subject. These curves were similar both in position (PD,,) and in shape (i.e., the slope of the curve as estimated by the ratio of wloss for maximum recorded percent fall in FEV, [PD,,,J to PD,,). There was no signi$cant difference in the PD,, (.ISH, IO.3 ml, 95% conjidence limits 7.5 and 13.9; 4.5% NaCl, 12.3 ml, 95% confidence lairnits 8.9 and 17.1) or between the ratio of log PD,,,: log PD,, (ISH, I.19 -+ I SD, 0.14; 4.5% ElaCl, 1 .I 7 ? I SD, I .I 7; p = not significant). These Jindings support the concept that airway hyperosmolarity may be the mechanism for ISH and exercise-induced asthma. (.I ALL.ERGY CLIN IMMUNOLOGY 77:729-36, 1986.)

The severity of asthma induced by exercise is inversely related to the concentration of water in the air inspired during exercise” ’ and directly related to the amount of water lost from the airways by humidifying this air.’ The mechanism whereby respiratory wloss induces asthma remains unknown. Cooling of the intrathor,acic airways’. 5 and an increase in osmolar&y of the fluid lining the respiratory tract3. 6. ’ have both been proposed as the stimulus by which respiratory wloss induces asthma by either exercise or by ISH. It is likely that both the cooling and the hypertonic effects of respiratory wloss affect the asthmatic response in EIA and ISH.8 However, there are two reasons why the hypertonic stimulus may be more potent than airway cooling for inducing asthma. First, airway cooling may be reduced by increasing the inspired air temperature, and yet there is no attenuation of the From the Department of Thoracic Medicine, Royal Prince Alfred Hospital, Camperdown, Australia. Supported in part by a grant from the National Health and Medical Research Council of Australia. Received for puiblication April 11, 1985. Accepted for publication Nov. 12, 1985. Reprint requests: Sandra Anderson, Ph.D., Department of Thoracic Medicine, Rayal Prince Alfred Hospital, Camperdown, NSW 2050, Australia.

Abbreviations EIA: ISH: MVV: “Wloss”:

Wloss: PD?,: PD,,,:

used Exercise-induced asthma Isocapnic hyperventilation Maximum voluntary ventilation Amount of water required from mucosa to bring periciliary fluid back to normal tonicity Water loss Wloss for 20% fall in FEV, Wloss for maximum recorded percent fall in FEV,

asthmatic response after exercise or ISH.4. ‘-‘O The second reason is that patients with asthma have a reduction in FEV, after inhaling very small doses of aerosols of hypertonic solutions. ” The present study was undertaken to compare the amount of respiratory wloss during ISH with the theoretical amount of wloss from a known hypertonic stimulus when each challenge has induced a similar change in FEV,. An aerosol of 4.5% NaCl was administered as the hypertonic stimulus. In order to use wloss as the stimulus common to both challenges, it was necessary to convert the dose of saline inhaled to a theoretical “wloss” from the airways. This “wloss” was calculated as the amount of water that would be 729

730

Smith

and

J. ALLERGY CLIN. IMMUNOL. MAY 1986

Anderson

TABLE I. Morphometric asthma

who were

data, predicted values for FEV,, and medications challenged with ISH and 4.5% NaCl

Subject

Sex

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

F M F M

S, salbutamol; *As needed

Height

Predicted FEV, (liters)

(cm)

13 25 20 20 42 30 37 29 26 32 20 23 28

158 181 166 184 171 164 182 168 167 183 158 156 169

14

162

M

26 31

185 186

M

15

179

M M M M F M F

F M M M

F, fenoterol;

Age (vr)

B, beclomethasone;

T, theophylline;

SCG,

required from the mucosa to bring the periciliary fluid back to its normal tonicity. The wloss from the airways during ISH was estimated by taking the value of 29 mg of expired water, measured per liter in normal subjects, and multiplying this amount by the volume ventilated by each subject with asthma. By assessing both the challenges in terms of wloss, it was possible to make a direct comparison between the two doseresponse curves. We reasoned that if the two doseresponse curves to wloss were similar both in position and in shape, then the proposal that respiratory wloss during exercise or ISH is a hypertonic stimulus would be supported. SUBJECTS AND METHODS We studied 17 subjects with asthma, 12 male and five female, ranging in age from 13 to 42 years. The protocol

was approved by the hospital Ethics Review Committee, and informed consent was obtained from all subjects.Morphometric data ate listed in Table I, together with predicted values for FEV, and current medication. All medications were withheld for at least 4 hours before challenge with the exception of oral theophylline that was withheld for at least 12 hours. No subject was taking oral corticosteroidsat the time of the study. ISH The technique used to induce wloss by ISH was similar to that described by O’Byme et al.‘? to induce heat loss.

for the 17 patients

sodium

2.68 4.34 3.09 4.61 3.43 3.36 4.11 3.59 3.05 4.27 2.75 2.67 3.66 2.82 4.52 4.45 3.87

with

Medications

S, B, T S, T S, B, ‘IS* F, B S, B S, B, T, SCG S, B S, B, SCG S S* S* S* S, SCG NIL S*

S, T, SCG

cromoglycate.

The challenge was modified in two ways. First, the inspired air was not cooled but was kept at room temperature, and second, dry gas was inhaled from a tank of compressedair or compressedair containing 5% CO,. This concentration of CO, produces near normal end tidal CO, at ventilation rates of 30 to 105 L/min.” Initially, subjects inhaled room air through a Hans-Rudolph valve (No. 2700, KansasCity, MO.), which had been divided in order to reduce its dead space. After 3 minutes, measurementsof FEV, were made in triplicate by use of a hot-wire anemometer (Minato AS 800, Osaka, Japan), and the highest value was taken as the baseline FEV,. The first 3 minutes of the challenge were also performed at resting ventilation, but the subjects inhaled air through a demand valve from a tank of compressedmedical air. The FEV, was measuredat 0.5, 1.5, and 3.0 minutes and then at 2-minute intervals until it reached a plateau. For each subsequent3minute challenge period, the subjectswere askedto breathe from the tank containing 5% CO, at 30 to 40 L/min, then at 60 to 80 L/mm, and finally at their maximum voluntary ventilation. The change in FEV, after each period of ventilation was expressedas a percentage of the baselinevalue (i.e., as a percent fall in FEV,). If the reduction in FEV, was still <20%, a final challenge at maximum voluntary ventilation was performed for 6 minutes. To obtain the required rate of ventilation, the subject was instructed to keep a target balloon filled to a constant volume. The rate of gas entering the baloon was monitored by a rotameter (GEC-Elliot, Croydon, England). The expired gas was collected in a 350 L Tissot gasometer(W. E. Collins, Braintree, Mass.), and the ventilation was recorded

Hyperosmolarity

VOLUME 77 NUMBER 5

(Devices M 19, Herts, U. K.). From this trace both the rate of ventilation and total ventilation were measured. To ensure that the subjects remained isocapnic throughout the challenge, the end tidal COZ was monitored continuously by use of an infrared gas analyser (Godart, Bilthoven, Holland). The heart rate was also monitored (Pulseminder No. 77 19H, Computer Instruments Corp., Hempstead, N. Y.) in 10 subjects to document the extent to which heart rate increases during ISH. The dose-response curve for the cumulative wloss from ISH was obtained by summing the volume of ventilation in liters expired during each period of ventilation and multiplying this by 29 mg. A curve was constructed that related the percent fal.! in FEV, to the cumulative wloss from the respiratory tract during each period of ISH. The value of 29 mg was obtained by measuring the actual amount of water lost from the airways in six normal subjects who performed the challenge with ISH under similar inspired air conditions. (Inspired air temperature ranged from 22” to 26” C, ;and the air was dry.) For these subjects the expired air was passed through two stainless steel coils that were immersed in dry ice. Wloss was measured as the change in weight (Sartorius 1216MP, Gbttingen, Germany) in the coils before and after challenge and expressed in terms of a concentration. All the water was collected in the first coil. The wloss over a range of ventilation of 50 to 65 Limin was 28.9 -C I. 1 mg/L for the six subjects, and their expired temperature was 3 1.2 5 1.6” C.

Ultrasonicallly

nebulized

4.5% NaCl

For the hypertonic challenge, an ultrasonic nebulizer (Mistogen 143A, Mistogen Equipment Co, Oakland, Calif.) was used to generate an aerosol of 4.5% NaCl that was delivered by tubing to a two-way valve (Hans-Rudolph No. 2700).” For this challenge the valve used was not divided, because doing this would reduce the output of the aerosol from the valve to the patient. The volume of expired air was measured as for the ISH challenge. Subjects breathed 30 L of room air through the valve with a noseclip in situ. The FEV, measured after this procedure was taken as the baseline value. The subjects then inhaled the aerosol of 4.5% NaCl. After each dose of aerosol (10 L, 20 L, 40 L, 80 L, and 100 L), FEV, was measured at 0.5, 1.5, and 3 minutes and at 2-minute intervals until the values reached a plateau or began to rise. The challenge was terminated when the percent fall in FEV, was >20% or when a total of 250 L was exhaled. The total dose of 4.5% NaCl solution delivered to the inspiratory port of the valve was measured by weighing the nebulizer cannister and the tubing before and after the challenge (Sartorius 1216 MP). The dose of aerosol inhaled by each subject was calculated in milligrams per liter by dividing the total dose of 4.5% NaCl solution delivered (milligrams) to the subject by the total volume of expired air (liters) by the subject. The “wloss” caused by inhaling 4.5% NaCl was calculated as the amount of water required to bring the volume of saline delivered to isotonicity. Thus, for each milliliter of 4.5% saline inhaled into the airways, 4.0 ml of water would be required from the mucosa to bring the periciliary

as stimulus

to asthma

731

fluid back to its usual tonicity. The calculation of a theoretical “wloss” allowed a comparison to be made between the two challenges for the amount of water lost from the airways and the reduction in FEV,. To construct a doseresponse curve, the cumulative expired volume (liter) after each challenge period was multiplied by the output of the nebulizer (milligrams per liter). This value was then multiplied by 4.0 to ascertain the “wloss.”

Analysis

of data

A semilog scale was used to plot the percent fall FEV, in response to the cumulative wloss caused by ISH and in response to the cumulative “wloss” caused by inhalation of 4.5% NaCl. The PDZ, was determined by linear interpolation. A PD,,, was also determined. This represented the wloss required to induce the maximum response in FEV, that was common to both challenges. A paired t test was used to compare the PD2,, for ISH with that for 4.5% NaCl. Log PD,, values were used to calculate the t statistic. The mean PD,,, values were expressed in milliliters by taking antilogs, and 95% confidence limits were used as an index of the variation about the mean. The PD,, value was used to record the position of the doseresponse curve. The Spearman’s correlation coefficient was calculated by comparing the responses from least responsive to most responsive for the two challenges. The ratio of log PD,,, to log PD,, was used as an indication of the slope of the dose-response curve. To determine whether the two curves for each subject were parallel (i.e., that the curves were the same shape), this ratio was compared by use of a paired t test. Because the curves were constructed with a log scale for wloss, it was necessary to calculate the ratio with the log-dose values. Two linear regressions of log PD,,, on log PD,,, were compared by use of an analysis of covariance to determine whether they were the same in terms of residual variances, slope, and elevation. The statistical methods used are described in Snedecor and Co&ran.‘* The time course of the response was also examined. A paired t test was used to compare the time taken to reach the maximum reduction in FEV, after each challenge.

RESULTS In response to the wloss induced by ISH, the mean -t 1 SD, percent fall in FEV, was 39.8 ? 15.9, and all 17 subjects registered a percent fall in FEV, of 320%. Similarly, in response to the “wloss” induced by the inhalation of 4.5% NaCl, the mean 4 1 SD percent fall in FEV, for the 17 subjects was 37.7% + 15.5. Of the 17 subjects, 15 recorded a percent fall in FEV, >20% in response to 4.5% NaCl. The remaining two subjects recorded a 15% fall in FEV, after the 250 L of aerosol saline had been inhaled. The baseline values for FEV, expressed as a percentage of the predicted value were not significantly different for the two challenges, and the values

732

J. ALLERGY CLIN. IMMUNOL. MAY 1996

Smith and Anderson

TABLE II. Baseline FEV, (expressed as percent predicted), maximum percent fall in FEV, after each challenge, the provoking dose of water loss required to induce a 20% fall in FEV,, and the dose of water loss at the maximum fall in FEV, that was common to both challenges in the 17 patients with asthma who were challenged with ISH and 4.5% NaCl Baseline FEV, % predicted Subject 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Maximum % fall in FEV,

PDm “Wloss” (ml)

PL.* Wloss (ml)

NaCl

ISH

NaCl

ISH

NaCl

ISH

NaCl

ISH

67 74 98 76 74 60 86 84 91 87 103 89 118 87 109 99 88

67 68 104 80 76 61 86 78 84 85 93 93 114 85 112 96 94

73 45 38 32 35 37 64 27 24 46 55 33 38 44 27 15 15

80 34 46 37 50 39 53 50 33 35 22 35 50 20 22 30 26

t 6.8 5.6 12.6 7.8 t 11.6 12.0 25.0 11.7 12.8 29.5 13.0 7.0 26.0 $ $

1.3 16.0 6.8 10.4 5.3 6.4 14.4 4.8 10.2 9.2 9.6 15.1 10.6 7.4 33.0 11.5 17.5

8.0 8.2 9.0 29.0 15.7 4.0 29.0 16.2 30.0 16.5 13.5 51.2 32.0 7.0 29.0 t f

7.4 20.4 8.8 20.0 8.2 11.0 24.0 6.2 14.0 13.8 11.5 24.2 18.3 7.4 41.5 15.0 21.0

*Common to both dose-response curves. TSubject very sensitive; first percent fall in FEV, >20%. *Subject did not register a percent fall in FEV, >20%.

for each pair did not differ by more than 11% in any subject (mean ? 1 SD; 4.5% NaCl, 87.6 +- 15.0; ISH, 87.5 + 15.3; p = not significant) (Table II.) The dose-response curves obtained for the two challenges in three subjects are illustrated in Fig. 1. There was no significant difference in the position of the curves. For the 13 patients who recorded a 20% fall in FEV, for both tests, there was no significant difference in the wloss required to induce this reduction in lung function. The mean PD,, for ISH was 10.29 ml (95% confidence limits, 7.5 and 13.9), and the mean PD,, for 4.5% NaCl was 12.3 ml (95% confidence limits, 8.9 and 17.1) (p = not significant). The sensitivity to the loss of water was also similar for both challenge tests (Spearman’s r = 0.73, p < 0.001) (Fig. 2). To calculate the Spearman’s rank correlation, subjects Nos. 1 and 6, who registered a reduction in FEV, of >20% after the challenge with 4.5% saline, were ranked equal first. Subjects Nos. 16 and 17, who did not register a 20% fall in FEV, after 4.5% NaCl, were ranked equal last. The sensitivity of the patients to ISH and 4.5% NaCl were compared by a Spearman’s rank correlation for two reasons; first, because it was difficult to measure the

precise dose of hypertonic saline delivered to the airways, and second, because we were able to include all subjects in the statistical analysis. For each subject the two curves were the same shape. There was no significant difference in the ratio of the log PD,,, to the log PD,, between the two challenges, indicating that the curves were parallel (mean t 1 SD; ISH, 1.19 2 0.14; 4.5% NaCl, 1.17 -+ 0.97). The relationship of log PD,,, to log PD2,, was also similar for the two challenges (Fig. 3). An analysis of covariance between the regression lines demonstrated no significant difference in terms of variance (F = 2.6, p = not significant), slope (F = 0.001, p = not significant), or elevation (F = 0.01, p = not significant). The time taken to reach a maximum reduction in FEV, after 4.5% NaCl (mean f 1 SD, 2.0 -+ 1.7 minutes) was significantly shorter (t = 4.1, p < 0.001) compared with the time observed for the maximum response after ISH (mean ? 1 SD, 3.7 + 1.8 minutes). Thirteen of the 17 subjects responded more rapidly to 4.5% NaCl than to ISH, two responded within the same time interval, and only two subjects responded more rapidly to ISH than to 4.5% NaCl.

Hyperosmolarity

VOLUME 77 NUMBER 5

50

40

1

0 ISH lmll

!

as stimulus

to asthma

733

.

1

100

FIG. and and the lized

1. The relationship between the percent fall in FEV, the cumulative wloss for three subjects (Nos. 3, 4, 12). The broken, solid, and interrupted lines represent individual responses to ISH and ultrasonically nebu4.5% NaCI.

The heart rate increased significantly (p < 0.001) during the ISH challenge. At rest the mean heart rate + 1 SD was 74 -+ 11 hpm. The mean maximum heart rate & 1 SD during the ISH challenge was 97 + 12 bpm. DlSCUSSlClN The results of this study demonstrate that the doseresponse curves obtained after challenge with ISH or hypertonic saline aerosol are similar both in position and shape when the dose is considered in terms of respiratory wloss. Because the response to the wloss from ISH was similar compared with the response to “wloss” calculated from a known hypertonic stimulus, we believe that these findings support the proposal that respiratory wloss has a drying effect that causes the periciliary fluid to become hypertonic. The PD,, was used as an indicator of position of the dose-response curve because a 20% reduction in FEV, is the index most commonly used in the laboratory to document bronchial hyperresponsiveness. The ratio of log PD,,, to log PD,, was used to determine whether the curves were parallel (i.e., had the same shape). This technique of analysis is similar to that used by O’Byme et al.‘* who compared the PDZOto the PD,,, for respiratory heat exchange and methacholine in order to establish that the dose-response curves were parallel. Because we were interested in comparing the shape of the curve beyond a 20% fall in FEV,, we modified this technique and compared PI),,, common to both challenges to PD,,. This simple relationship between the PD,,, and PD,, was also used because the patients did not plateau in their response to the two challenges, and thus, insufficient numbers of data points were available to calculate a true slope. Our findings confirm that the airways of patients

2

4

6

8 10 12 14 16 18

RANK ORDER PD20 NaCl FIG. 2. The relationship PDB after the inhalation NaCI. The correlation Spearman’s rank test.

between PD2, after ISH and the of ultrasonically nebulized 4.5% coefficient was calculated with The line of identity is presented.

with asthma are very sensitive to the inhalation of very small doses of hypertonic aerosols. ” We have also demonstrated, as others have done,” that subfreezing air is not necessary to induce asthma, and ISH with dry air inspired at normal laboratory temperatures is a potent stimulus for reducing FEV,. This article appears to be the first comparing the doseresponse curves to these two challenge tests in the same group of patients with asthma. The significant Spear-man’s correlation may simply be a reflection of the fact that the airways of our patients were similarly sensitive to all nonspecific stimuli. Significant correlations have also been reported between the PD,, to methacholine and the PD,o to hyperventilation when cold air was inhaled,‘* between PD2,, for methacholine, histamine, and ISH,16 and between the PDZOfor ultrasonically nebulized distilled water and ISH. ” These findings by other workers suggest that the airways of subjects with asthma are similarly sensitive to all nonspecific stimuli. However, in contrast to their studies, we compared our challenges in terms of the common stimulus of wloss. For this reason we were able to extend our findings on sensitivity and compare directly the position and shape of the dose-response curve. We did not measure the actual amount of respiratory wloss by our patients. However, there is no reported difference in the expired air conditions between subjects with asthma and normal subjects ventilating under the same inspired air conditions.“. I9 For this reason we were able to use the direct measurements of wloss made in six normal subjects hyperventilating under the same inspired air conditions as the subjects

734

Smith

and

J. ALLERGY CLIN. IMMUNOL. MAY 1986

Anderson

1.6.

1.4 2 g P

:: J

13

1.0 -I

0.8 1

V-1

7 0.4

0.6

0.8

1.0

LOG

FIG. 3. The relationship between to both challenges in 12 subjects inhaled ultrasonically nebulized

1.2

1.4

1.6

PD20

PDm and PD,,, common who performed ISH and (UN) 4.5% NaCI.

with asthma. By weighing the actual amount of wloss from the airways during challenge with ISH, we were able to overcome the problems, also encountered by others, with the measurement of expired air temperature at high frequencies of breathing9 and the assumption that all the expired air is fully saturated with water vapor4 A value of 33 mg/L has previously been reported during exercise in subjects with asthma and normal subjects. “3 I9 The lower value of 29 mg/L that we measured in this study probably related to the lower temperature and water content of the compressed air inspired from a cylinder. The value of 29 mg/L we observed was similar to the value reported by Eschenbacher and Sheppard9 for the same expired air temperature with the same technique used to measure water loss during ISH. As the expired air temperature in our normal subjects remained between 29.0” C and 32.5” C, any changes in water concentration with changes in rate of ventilation would have been small. Either exercise or ISH could have been used for this study. We used ISH in preference to exercise to induce wloss for a number of reasons. First, the mechanism by which these challenges induce asthma is believed to be the same.*, ” Second, the ability to construct a cumulative dose-response curve over a single challenge allows comparison of both sensitivity and reactivity with that obtained from a known hypertonic stimulus. “. ‘* Third, bronchodilation does not occur in response to ISH, as it does with exercise, so that the reduction in lung function is progressive and comparable to the inhalation challenge with NaCl. Finally, although there is a small increase in heart rate, the cardiac output probably remains relatively

unchanged during ISH. This means that the bronchial circulation is unlikely to have been much different for the two challenges, particularly because ambient air, rather than cold air, was used for ISH. The assumption that theoretical “wloss” was that which was required to bring the delivered dose of saline to isotonicity is probably valid. Aerosol droplets entering the airways undergo rapaid equilibration with their environment in terms of temperature and osmolarity. Thus, droplets of 4.5% NaCl will grow rapidly as water is absorbed from the airways to make these particles isotonic. Providing that the droplets do not impact in the mouth and orophamyx, the process of evaporative water loss from droplet growth is probably similar to that which occurs from humidifying the inspired air.*‘, *’ The main limitation in estimating the “wloss” from the airways is in determining how much aerosol is delivered to the airways, as opposed to the apparatus, mouth, and orophatynx. As with other inhalational challenges, it is not possible to know the precise dose of aerosol inhaled. Our technique may have overestimated the dose of saline delivered to the airway. The documentation of similar values for PDzo wloss for the two challenges may thus have been a chance finding caused by the selection of that particular concentration of saline. However, the mean PD,, for the challenge with 4.5% NaCl was approximately 2 ml greater than that for ISH, and this value is probably a reflection of the overestimation. Until we have more information on the amount and site of deposition of 4.5% NaCl aerosol, we will be unable to assess the precise dose of hypertonic aerosol delivered to the airways. We believe that our results provide indirect evidence that respiratory wloss during exercise or ISH leads to hypertonicity of the periciliary fluid. Unfortunately, there are no direct measurements of changes in osmolarity of the respiratory tract fluid during either exercise or ISH. However, the human airway can and does maintain considerable osmotic gradients in response to drying of the mucosa. It is known that in patients with laryngectomies the respiratory secretions are hyperosmolar presumably because the upper airways are bypassed during respiration.23. 24Thus wloss from the airways can induce a hyperosmolar state in the human respiratory tract at resting ventilations. There are also several reports that demontrate that the osmolarity of the airway secretions in dogs increase at resting ventilations merely by changing from nose to mouth breathing.25. *’ Thus, it is not unreasonable to propose that an increase in osmolarity may have occurred at the high rates of ventilation performed by our patients during ISH with dry air.

VOLUME 77 NUMBER 5

Hyperosmolarity

From our data it appears that the rate of respiratory wloss during ISH is sufficient to increase the osmolarity of the periciliary fluid of the upper respiratory tract. The rate of respiratory wloss that induced a 20% reduction in FEV, was between 0.5 and 3.5 mlimin. It has been reported that when air is hyperventilated at room temperature most of the conditioning of the inspired air to body temperature occurs in the first six to seven generations of the airways.” If we assume that the air is humidified at the same time as it is heated,** then most of the respiratory wloss will occur in these six to seven generations of airways using the formula preslented by Weibe129for calculating the surface area of airways and a depth of periciliary fluid of 7 to 12 pm,3o then the volume of fluid lining the surface of the first six generations is between 0.3 and 0.4 ml. To maintain a constant osmolarity, the surface water would have to be replaced every 7 seconds at the high rate of wloss observed in some of our subjects. If the surface area of the first six to seven generations is around 325 cm*, then the rate of return of water would need to be up to 100 ~1 (0.1 ml)/cm’. It is unclear what effect an increase in osmolarity of the periciliary fluid would have on the rate of return of water to the airway or the production of mucus. As water moves along an osmotic gradient, it would be expected that a hypertonic state of the airways would lead to an increase in the release of fluid into the airway lumen.*‘. 3’. 3z The mechanism whereby hypertonicity of the periciliary fluid leads to asthma is unknown, but there is evidence that mediators may be released in response to a hyperosmolar stimulus. It is now clear that mast cells from human lung will release histamine in response to a hyperosmolar stimulus and that this response is enhanced in the presence of anti-IgE.33 The fact that sodium cromoglycate inhibits the response to hypertonic aerosols” is indirect evidence that lung mast cells may be involved. Hyperosmolarity does not have a direct contractile effect on isolated airway smooth muscle from humans.34 Furthermore, the contractile response to histamine is not enhanced in a hypertonic environment when human airway smooth muscle is challenged in vitro.34 We conclude that the position and shape of the doseresponse curves to ISH and 4.5% NaCl were the same when wloss was used as the common stimulus. These findings support the proposal that, during exercise or ISH, the increased rate of respiratory wloss acts as a hypertonic stimulus to induce asthma. REFERENCES 1. Chen WY. Horton and exercise-induced

DJ: Heat and water loss from the airways asthma. Respiration 34:305. 1977

as stimulus

to asthma

735

2. Strauss RH, McFadden ER, Ingram RH, Deal EC, Jaeger JJ, Stearns D: Influence of heat and humidity on the airway obstruction induced by exercise in asthma. J Clin Invest 61:433, 1978 3. Anderson SD, Schoeffel RE, Follet R, Perry CP, Daviskas E, Kendall M: Sensitivity to heat and water loss at rest and during exercise in asthmatic patients. Eur J Respir Dis 63:459, 1982 4. Deal EC, McFadden ER, Ingram RH, Strauss RH, Jaeger JJ: Role of respiratory heat exchange in production of exerciseinduced asthma. J Appl Physiol 46:467, 1979 5. McFadden ER, Ingram RH: Exercise-induced asthma: observations on the initiating stimulus. N Engl J Med 301:763, 1979 6. Anderson SD: Is there a unifying hypothesis for exercise-induced asthma? J ALLERGY CLIN IMMUNOL73:660, 1984 7. Hahn A, Anderson SD, Morton AR, Black JL, Fitch KD: A reinterpretation of the effect of temperature and water content of the inspired air in exercise-induced asthma. Am Rev Respir Dis 130:575, 1984 8. Anderson SD: Issues in exercise-induced asthma. J ALLERGY

CLIN IMMUNOL~~:~~~, 1985 9. Eschenbacher WL, Sheppard D: Respiratory heat loss is not the sole stimulus for bronchoconstriction induced by isocapnic hyperpnea with dry air. Am Rev Respir Dis 131:894, 1985 IO. Aitken ML, Marini JJ: Effect of heat delivery and extraction on airway conductance in normal and asthmatic subjects. Am Rev Respir Dis 131:357, 1985 Il. Anderson SD, Schoeffel RE, Finney M: Evaluation of ultrasonically nebulised solutions as a provocation in patients with asthma. Thorax 38:284, 1983 12. O’Byme PM, Ryan Cl, Morris M, McCormack D, Jones NL, Morse JLC, Hargreave FE: Asthma induced by cold air and its relation to nonspecific bronchial responsiveness to methacholine. Am Rev Respir Dis 125:281, 1982 13. Phillips YY, Jaeger JJ, Laube BL, Rosenthal RR: Eucapnic voluntary hyperventilation of compressed gas mixture: a simple system for bronchial challenge by respiratory heat loss. Am Rev Respir Dis 131:31, 1985 14. Snedecor GW, Cochran WG: Statistical methods. Ames, Iowa, 1967, The Iowa State University Press, pp 432-6 15. Kilham H, Tooley M, Silverman M: Running, walking, and hyperventilation causing asthma in children. Thorax 34:582, I979 16. Aquilina AT: Comparison of airway reactivity induced by histamine, methacholine, and isocapnic hyperventilation in normal and asthmatic subjects. Thorax 38:766, 1983 17. Fabbri LM, Mapp CE, Hendrick DJ: Comparison of ultrasonically nebulized distilled water and hyperventilation with cold air in asthma. Ann Allergy 53: 172, 1984 18. Anderson SD, Schoeffel RE, Black JL, Daviskas E: Airway cooling as the stimulus to exercise-induced asthma: a reevaluation. Eur J Respir Dis 67:20, 1985 19. Mitchell JW, Nadel ER, Stolwijk JAJ: Respiratory weight losses during exercise. J Appl Physiol 32:474, 1972 20. Deal EC Jr, McFadden ER Jr, Ingram RH Jr, Jaeger JJ: Hy perpnea and heat flux: initial reaction sequence in exerciseinduced asthma. J Appl Physiol 46:476, 1979 21. Martonen TD, Bell KA, Phallen RF, Wilson AF, Ho A: Growth rate measurements and deposition modelling of hygroscopic aerosols in human tracheobronchial models. Ann Ckcup Hyg 26:93, 1982 22. Ferron G: The size of soluble particles as a function of humidity of the air: applications to the human respiratory tract. J Aerosol Sci 8:251, 1977 23. Richardson P.S. Phipps RJ: The anatomy, physiology, pharmacology, and pathology of tracheobronchial mucus secretion

Smith

24. 25.

26.

‘27.

28. 29.

and

J. ALLERGY CLIN. IMMUNOL. MAY 1986

Anderson

and the use of expectorant drugs in human disease. Pharmacol Ther 3441, 1978 Potter JL, LeRoy WM, Spector S, Lemm J: Studies on pulmonary secretions. Am Rev Respir Dis 96:83, 1967 Man SFP, Adams GK, Proctor DF: Effects of temperature, relative humidity, and mode of breathing on canine airway secretions. J Appl Physiol 46:205, 1979 Boucher RC, Stutts MJ, Bromberg PA, Gatzy JT: Regional differences in airway surface liquid composition, J Appl Physiol 50:613, 1981 McFadden ER, Denison DM, Wailer JF, Assoufi B, Peacock A, Sopwith T: Direct recordings of the temperatures in the tracheobronchial tree in normal man. J Clin Invest 69:700, 1982 Ingelstadt S: Studies on the conditioning of air in the respiratory tract. Acta Otolaryngol 13 I(supp1): I, 1956 Weibel ER: Morphometrics of the lung. In Penn WO, Rahn H, editors: Handbook of physiology, respiration section 3.

30. 31.

32.

33.

34.

Washington, D. C., 1964, American Physiological Society, vol I, pp 285307 Kilbum KH: A hypothesis for pulmonary clearance and its implications. Am Rev Respir Dis 98:449, 1968 Marin M, Davis B, Nadel JA: Effect of acetylcholine on Cl and Na’ fluxes across dog tracheal epithelium in vitro. Am J Physiol 231: 1546, 1976 Lopata M, Barton AD, Lourenco RV: Biochemical characteristics of bronchial secretions on chronic obstructive pulmonary disease. Am Rev Respir Dis 110:730, 1974 Eggleston PA, Kagey-Sobotka A, Schleimer RP, Lichtenstein LM: Interaction between hyperosmolar and IgE-mediated histamine release from basophils and mast cells. Am Rev Respir Dis 130:86, 1984 Finney MJB, Black JL, Anderson SD: Effect of a hypertonic stimulus on human bronchial smooth muscle (BSM) contractility. Clin Exp Pharmacol Physiol (in press)

Occupational allergy to honeybee-body in a honey-processing plant N. K. Ostrom, M.D., M. C. Swanson, B.A., M. K. Agarwal, J. W. Yunginger, M.D. Rochester, Minn.

dust

Ph.D.*, and

We studied a honey-plant employee who developed severe asthma coincident with the seasonal honey-packing process. Symptoms correlated with duration of exposure inside the plant during the honey pack and improved in other environments during that season. The patient was asymptomatic inside the plant at other times of the year. Skin tests to seasonal outdoor aeroallergens were negative, as were inhalation challenges with two insecticides used inside the building during the honey pack. Skin test, RAST, and bronchial provocation test with honeybee whole body extract were positive. We used high-volume air samplers to collect ambient airborne particles on filter sheets outside the patient’s home and inside the honey plant, both during and after the honey pack. Skin tests, RASTs, and bronchial provocation tests with eluates from these filters were positive only to eluate from the filter exposed inside the plant during the honey pack. The patient’s positive honeybee whole body RAST could be inhibited by preincubation of her serum with this filter eluate but not by preincubation with eluate from a filter exposed outside the patient’s home. Collectively, these data support an IgE-mediated, seasonal, occupational sensitivity to honeybee-body dust. (J ALLERGY CLIN IMMUNOL 77:736-40, 1986.)

From the Departments of Pediatrics and Internal Medicine (Allergy), Mayo Graduate School of Medicine, and the Allergic Diseases Research Laboratory, Mayo Clinic and Foundation, Rochester, Minn. Supported in part by United States Public Health Service Grant AI21398 and by Mayo Foundation. Received for publication July 8, 1985. Accepted for publication Nov. 20, 1985. Reprint requests: J. W. Yunginger, M.D., Allergic Diseases Research Laboratory, 406 Guggenheim Building, Mayo Clinic, Rochester, MN 55905 *Present address: V.P. Chest Institute, University of Delhi, Delhi, India.

736

There have been previous articles of allergy caused by inhalation of airborne insect body antigens.le3 Immunotherapy with honeybee-whole body extract had been standard therapy for honeybee-venom hypersensitivity before the introduction of venom extracts.4.5 However, few articles exist of documented inhalant allergy to honeybee-whole body antigens.6-9 We report a patient who developed IgE-mediated occupational asthma at a honey-processing plant. The case illustrates use of an immunochemical method of defining specific causes in occupational allergy to unknown or amorphous airborne antigens.