Effect of Inhaled Preservatives on Asthmatic Subjects II. Benzalkonium Chloride 1 •2
Y. G. ZHANG, W. J. WRIGHT, W. K. TAM, T. H. NGUYEN-DANG, C. M. SALOME, and A. J. WOOLCOCK
Introduction
Recently reported cases of paradoxical bronchoconstriction with nebulized isotonic ipratropium bromide solution (1) and beclomethasone dipropionate (2) have led to the discovery that an additive in both solutions, namely benzalkonium chloride, causes bronchoconstriction in some asthmatic subjects. Benzalkonium chloride is a mixture of quaternary benzyldimethyl alkylammonium chlorides that has antibacterial properties. The mechanism by which benzalkonium causes bronchoconstriction is not known. Previous studies have suggested that an effect via mast cell mediator release is likely (1) and Miszkiel and colleagues have shown that histamine is responsible for at least part of the response (3). However, an effect on neural pathways has not been excluded and may account for the nonhistamine component of the response. In this report, we document a number of characteristics ofthe response to benzalkonium. Wehave examined the shape and position of the dose-response curves to benzalkonium in 28 asthmatic subjects and compared responsesto benzalkonium and histamine. The time course of the response and the effects of a response to benzalkonium on subsequent responses to histamine or benzalkonium have been described. Finally, the protective effect of 8 mg cromolyn sodium (SC) (Fisons plc, Leicestershire, England) and 160 ug ipratropium bromide (lB) (Atrovent't; Boehringer Ingelheim, FHERGMBH, West Germany) on challenges with benzalkonium was examined. Methods Subjects Twenty-eight subjects with stable asthma were studied. Subjects were patients from the Allergy Clinic of Royal Prince Alfred Hospital and volunteers from the University of Sydney. All subjects were free of other respiratory illnesses at the time of testing. All gave in-
SUMMARY Benzalkonium chloride has been used as a preservative in some antiasthma respirator solutions and is known to cause bronchoconstrictlon In asthmatic subjects. ToIncrease understanding of hoWIt causes bronchoconstrictlon, the characteristics of airway response in 28 asthmatic SUbjects were documented. SUbjects inhaled histamine, in doses ranging from 0.03 to 7.8 umel, or benzalkonium In doses ranging from 0.04 to 5.33 umol on separate days. The dose of histamine or benzalkonium that caused a 20% fall in the 1-sforced expiratory volume (PD20FEV 1) was measured. All subjects responded to histamine, with PD20FEV 1 values In the range of 0.14to 7.8 umol and 17 responded to benzalkonium, with PD20FEV1 values In the range 0.35 to 5.55 umet, Subjects who responded to benzalkonium were more sensitive to histamine (mean PD20FEV 1 0.44 j.1mol) than subjects who did not respond (mean PD20FEV1 1.84 j.1mol) and, among the benzalkonium responders, there was a significant correlation between PD2 0FEV1 values for histamine and benzalkonium (r == 0.5, P < 0.05). Inhalation of benzalkonium enhanced subsequent responses to histamine, causing a decrease In mean PD20FEV 1 from 0.51to 0.18 j.1mol histamine (p < 0.001), but did not alter subsequent responses to benzalkonium. The response to benzalkonium reached a maximum 1 min after inhalation and was slow to recover, taking up to 60 min to return to baseline values. Response to benzalkonium was Inhibited by 8 mg cromolyn sodium but not by 160 j.1g ipratropium bromide. The characteristics of the response to benzalkonium suggest a mechanism of action via release of mediators. AM REV RESPIR DIS 1990; 141:1405-1408
formed consent and the study was approved by the Hospital Ethics Committee.
Study Design Baseline studies included skin tests with 13 common allergens, measurement of spirometric function, a histamine inhalation test and, on a second day, a benzalkonium inhalation test. Eight subjects (Nos. 1 to 8) were studied on two additional days within ten days of the initial study. On Day 3 these subjects were challenged with benzalkonium, but were allowed to recover spontaneously without albuterol for at least 60 min. When the FEV 1 had returned to within 90010 of the baseline value subjects were challenged with histamine. On Day 4, a single dose of benzalkonium was administered. This dose was determined on the basis of the two previous benzalkonium challenges as the dose which, when given cumulatively, caused a 15to 25010 fall in FEV rSpirometric function was measured before challenge and then at 1, 2, 5, 10, 15, 30, 45, and 60 min after challenge and then at 15min intervals until the FEV 1 recovered to within 90010 of the baseline value. After 60 min, or when the FEV 1 had recovered, the single dose of benzalkonium was readministered. Spirometric function was measured at 1, 2, and 5 min later. Albuterol was administered
and spirometric function was measured again 10 min later. In addition, three subjects were retested on a fifth day using benzalkonium solution that was 1 month old to determine the shelf life of the solution. Two subjects (Nos. 2 and 14) were rechallenged with histamine and with benzalkonium 3 months after their first test after a period of treatment with aerosol steroids. Another five subjects (Nos. 9 to 13) were enrolled in a double-blind study to determine the effects of pretreatment with 8 rng SC or 160 J.1g IB on the response to benzalkonium. This study was carried out at the same time of day on four separate days within ten days. The drugs, which did not contain benzalkonium, were delivered by metered-dose aerosols. On each day baseline spirometric func-
(Received in original form July 25, 1989 and in revised form November 17, 1989) 1 From the Department of Medicine, University of Sydney, Sydney, and the Institute of Respiratory Medicine, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia. 2 Correspondence and requests for reprints should be addressed to Ms. C. M. Salome, Department of Medicine, University of Sydney, Sydney, NSW 2006, Australia.
1405
1406
ZHANG, WRIGHT, TAM, NGUYEN-DANG, SALOME, AND WOOLCOCK
TABLE 1 BENZALKONIUM CHLORIDE CHALLENGEDOSE SCHEDULE
Dose Number 1 2 3 4
5 6 7 8
Concentration (mglm/)
Inhalations*
3.1 3.1 6.2 6.2 12.5 12.5 25.0 25.0
1 1 1 2 2 4 4 8
Cumulative Dose t (pmo/) 0.044 0.088 0.18 0.35 0.71 1.41 2.82
5.64
• Nebulizer output: 0.005 ml (x 0.0004 SO) per inhalation. Molecular weight used in estimations: 354.
t
tion was measured. The subjects then inhaled eight puffs of SC (8 mg) or placebo and eight puffs of IB (160ug) or placebo. The benzalkonium inhalation tests were performed 10min later.
Lung Function Tests FEV 1 and FEC weremeasured using a Vitalograph dry spirometer (Vitalograph Ltd., Buckingham, England), with the subject standing. Forced expiratory maneuvers were repeated until a FEV 1, reproducible to within 100ml, was obtained. Values for FEV 1 and FVC were expressed as a percentage of the predicted values of Morris and coworkers (4). If the FEV I/FVC was less than 60070 or the FEV 1 was lessthan 60% predicted, the subjects were not challenged. Inhalation Tests Inhalation tests with histamine and with benzalkonium were carried out according to the method described by Yan and colleagues (5). Briefly, histamine, in concentrations of 3.1 mg/ml, 6.2 mg/ml, 25 mg/ml, and 50 mg/ml in normal saline, or benzalkonium, in concentrations of 3.1, 6.2, 12.5, and 25 mg/ml in normal saline, was administered using DeVilbiss No. 40 (DeVilbiss Co., Somerset, PAl hand-held glassnebulizers. Nebulizer output was checked prior to the study and it was found that nebulizers containing benzalkonium had a mean output of 0.005 ml (± 0.0004 SD) compared with a mean output of 0.003 ml (± 0.0003 SD) from the same nebulizers containing histamine. This difference was attributed to an effect of benzalkonium on the surface tension of the solution. Doses of histamine ranged from 0.03 to 7.8 umol and doses of benzalkonium from 0.042to 5.33 umol. The dose schedule for benzalkonium is shown in table 1. Three inhalations of normal saline wereadministered as a control, prior to the challenge. The subject inhaled over 1 to 2 s towards total lung capacity. At the beginning of the inspiration, the operator gave the bulb of the nebulizer one firm squeeze. After a slight breath-hold (3 s), the subject exhaled. Spirometric function was measured 1 min after the dose and the next dose given
within 3 min of the previous dose. The challenge was stopped when the FEV 1 had fallen 20% or more from the postsaline value, or when the highest dose had been given. To aid recovery, 200 J.1g albuterol wasgiven and FEV 1 was measured 10min later. Subjects withheld aerosol bronchodilators for 6 h or oral bronchodilators for 12 h prior to testing.
Data Analysis On the basis of the measured volume delivered by the nebulizer and the known concentration of the solution, the dose of histamine
and benzalkonium delivered to the mouth of the subject was calculated in urnoles. Because benzalkonium is a mixture of molecules of varying composition, an estimated average molecular weight of 354 was used to calculate doses. Response was measured as percentage change in FEV 1 from the post saline value. This was plotted on a linear scale against log dose of histamine or benzalkonium and the dose that caused a 20% fall in FEV 1 (PDzoFEV1) was read from the graph. The PDzoFEV1 values were logarithmically transformed and all subsequent analysis was performed on the log values. The geometric mean and 95% confidence intervals (ci) of PDzoFEV1 values from each study were calculated. Paired t tests were used to examine the significance of differences. Correlation and regression coefficients were calculated to determine the degree of association between values. Significance was taken at the 5% level. Results
Details of the 28 subjects are shown in table 2. Subjects were 19to 71 yr of age; all subjects had two or more positive responses to skin tests; and all were receiving treatment for asthma. All had baseline FEV 1 and FVC values greater than 60010 and 70070 of the predicted values,
TABLE 2 DETAILS OF SUBJECTS PDzo FEV1 SUbject
Agel Sex
Current Treatment*
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
231M 241M 291M 231M 351M 201M 191M 231M 45/M 311F 421F 221M 211M 55/M 211M 201M 30/F 36/F 201M 71/F 221M 231M 211M 40/F 281M 431M 56/F 48/M
A A, B,C A,B A A,C A, B A A, B A A, B, T A,B A, S A, B A, T A A,S A, B A,S A,C A, B, T A, B A,C A A, B A,B A,B A, B, I, T A, B, T
= ipratropium
bromide; T
61 73
77 98 93 60 74
79 91 114 65 89 63 55 86 88 71 79 80 86 81 84 103 82 89 80 76 66
FVC (% pred)
73 85 86 84 97 82 88 94 97 109 88 108
77 77 84 92 87 70 76 100 80 91 97 112 90 91 83 86
= albuterol; C = cromolyn sodium; B = theophylline.
Definition of abbreviations: A
I
FEV 1 (% pred)
Histamine 0.60 0.70 1.10 0.68 0.86 0.14 0.26 0.50 0.34 0.21 0.11 0.45 1.2 0.24 1.1 1.0 0.17 7.8 1.4 1.3 4.6 2.4 3.7 1.2 5.0 1.8 0.44 0.29
Benzalkonium chloride 2.2 5.3 1.4 3.0 2.0 1.0 2.5 1.2 0.9 2.4 3.3 3.6 3.2 0.4 5.5 5.5 0.35
= beclomethasone dipropionate;
1407
EFFECT OF BENZALKONIUM CHLORIDE ON ASTHMATIC SUBJECTS
50 40 ~
30
f
ZO
1:J~~~~ 0.04
0.10
1.00
10.00
DOSE BNZ (I'mol)
Fig. 1. Dose-response curves to inhaled benzalkonium chloride (BNZ) in 28 asthmatic subjects.
respectively. There were no significant differences between baseline FEV 1 measurements on different study days. Seventeen subjects (Nos. 1 to 17) responded to the benzalkonium inhalation test with PD20FEV 1 values between 0.35 and 5.5 umol (table 2). Dose-response curves to benzalkonium for all 28 subjects are shown in figure 1. Subjects who responded to benzalkonium were more sensitive to histamine (mean PD 20FEV 1 0.44 umol histamine, ci: 0.31, 0.64) than those who did not respond (1.84 umol histamine, ci: 1.0, 3.4) (p < 0.001). Among the benzalkonium responders, PD20FEV 1 values for histamine and benzalkonium were significantly correlated (r = 0.50, p < 0.05). Twosubjects with PD 20FEV 1 values of 0.41 and 5.1lJmol to benzalkonium were retested after 3 months of treatment with aerosol steroids. Their responsiveness to histamine had decreased from initial PD20FEV 1 values of 0.24 and 0.7 umol to final values 5.0 and 7.8 umol, and they were no longer responsive to benzalkonium. Responsesto benzalkonium werereproducible to within ± 1.8doubling doses. Geometric mean values for PD20FEV 1 on two study days for eight subjects were 2.03 umol (ci:1.4,2.9) and 2.17 umol (ci: 1.2,3.9) and werenot significantly different. Mean difference between values was 0.03 log units (ci: -0.16, 0.22). The time course of the changes in FEV 1 over 60 min after a single dose of benzalkonium is shown in figure 2. The fall in FEV 1 was immediate and reached a maximum after 5 min, with a mean fall of 280/0. Recovery was slow and 60 min after challenge, the mean FEV 1 remained 100/0 below the initial value. In a separate study, response to benzalkonium was easily reversed with 200 IJgalbuterol. After challenge, the mean FEV 1 fell by 30.60/0 from initial and, 10min after 200 IJg albuterol, it increased to within 1% of the initial value (p < 0.001).
The doses of benzalkonium acted cumulatively. When the total cumulative dose, which caused a 20% fall in FEV 1 during the progressive challenge, was given as a single dose there was no significant difference between the percentage falls in FEV 1. The mean percentage fall in FEV 1 in eight subjects after a single dose was 23.10/0, and after the equivalent dose given progressively it was 190/0 on Day 2 and 20% on Day 3. Challenge with benzalkonium enhanced responses to histamine inhaled 1 h later. Geometric mean values for PD 20FEV 1 in eight subjects decreased from 0.51 umol (ci: 0.32, 0.81) for histamine alone to 0.18 umol (ci: O. 09, 0.33) for histamine challenge 1 h after benzalkonium (p < 0.001). Mean initial FEV 1 was 760/0 predicted prior to challenge with histamine alone and 700/0 prior to challenge with histamine after benzalkonium. These values were not significantly different. There did not appear to be a refractory period after benzalkonium inhalation. Mean fall in FEV 1 in six subjects, after a singledose of benzalkonium was 30.3% (ci: 22.6, 38.0) and after the same dose given 1 h later was 23.5% (ci: 19.4,27.6). These values for the group were not significantly different. However, in four of the six subjects the response to the second dose of benzalkonium was inhibited by 22, 30, 34, and 3'8 %, suggesting that a larger sample may show some significant refractoriness. Of the two subjects who did not complete this stage of the study, one suffered an acute exacerbation of his asthma and was not studied on Day 4, and the second did not have sufficient improvement in FEV 1 2.5 h after the first challenge to allow a second challenge. Response to benzalkonium was inhibited by 8 mg SC but not by 160 IJg
100
0
:J
90
]
:!; '0
80
~
i" to.
70
Discussion
The present study showsthat in asthmatic subjects with severe or moderate bronchial hyperresponsiveness,benzalkonium chloride causes reproducible bronchoconstriction that is rapid in onset, is slow to recover,and causes enhanced responses to histamine. The response is inhibited by SC but not by lB. There were no complications or serious side-effects of the test, although all subjects experienced cough and a burning sensation within moments of inhalation. The bronchoconstrictive effects of benzalkonium chloride were originally described after observations of patients who experienced marked reductions in lung function after inhalation of IB nebulizer solution (l). Atrovenr" nebulizer solution now contains no benzalkonium, but previously it contained 250 ug/ml benzalkonium chloride. At this concentration, patients inhaling 1 ml of this solution during tidal breathing might have been expected to receive approximately 0.35 to 0.7 umol benzalkonium. In the present study PD 20FEV 1 values were in the range of 0.35 to 5.0 umol and only
No PDZO
Q)
g
lB. Figure 3 shows PD 20FEV 1 values in six subjects for challenge with benzalkonium after placebo, SC, and lB. There was no significant difference between mean PD 20FEV 1 values after placebo (1.96 umol, ci: 0.97, 3.98) and after IB (3.17IJmol, ci: 1.63,6.15). PD 20FEV 1 values weregreater than the maximum dose of 5.64 umol in five subjects after SC. The solutions of benzalkonium had a shelf life of at least 1 month. Three subjects were challenged with f-month-old benzalkonium solutions, and the mean PD 20FEV 1 value of 3.61 umol, was not significantly different from the mean PD 20FEV 1 value of 2.75 umol obtained using fresh solutions.
1
1
iLLL//f~r~1
R
100j
z
CD
"0
!
f
1.0
0 N
1111
0
0 e,
60 +---,-~--..----..--.------.--, -10 10 40 30 50 60
zo
TIME AFTER CHALLENGE (mins)
x~:
0.1
IPB~~~F~~m
Placebo
(1601'9)
Fig. 2. (left) Time course of response to benzalkonium in eight asthmatic subjects. Values are mean and standard deviation. Fig. 3. (right) Effect on PD 2oFE\t, values for benzalkonium challenge after placebo, ipratropium bromide (160 I1g) or sodium cromoglycate (8 mg).
1408
two subjects had responses in the 0.35 to 0.7 umol range. Beasley and coworkers (1) found a mean PC 2oFEV. value of 0.3 giL or approximately 1.7/lmol (range 0.7 to 11.3 umol) and Miszkiel and colleagues (3) found a mean PC 2 oFEV. of 3.98 giL or approximately 1.96 umol (range 0.45 to 11.8 umol), The results of these studies suggest that only a few very sensitive patients were likely to respond to the benzalkonium chloride in a normal dose of Atrovent" nebulizer solution, and accounts for the relative rarity of cases of paradoxical bronchoconstriction with this drug. The mechanism by which benzalkonium chloride causes bronchoconstriction is not known. However, an effect of benzalkonium on mediator-releasing cells is suggested by the results of this and previous studies. In vitro studies of rat serosal mast cells have shown that benzalkonium inhibits histamine release, but at higher concentrations (10ug/ml) induces release, probably by a lytic effect (6). Furthermore, benzalkonium enhances IgE-dependent release of 5-hydroxytryptamine (7). In human subjects, benzalkonium may also cause histamine release. Miszkiel and coworkers (3) have recently shown that the selective Hi-receptor antagonists, terfenadine and astemizole, shift the benzalkonium concentration-response curve to the right. However, the protection afforded by these agents was only partial and was confined to the early phase of the response. The authors suggest that histamine release represents a small component of the overall effect of the chemical. In the present study 11 subjects failed to respond to benzalkonium despite being responsive to inhaled histamine. These 11 subjects were less sensitive to histamine than the benzalkonium responders, suggesting that if benzalkonium does cause histamine release then the amount released was insufficient to cause a response in these subjects. Mediators other than histamines may be involved in the response. If benzalkonium causes mast cell cytolysis (6) then it is possible that lipid-derived mediators, such as prostaglandins and leukotrienes, are generated at the same time. Prostaglandin generation by immunologically activated human mast cells occurs between 10 and 30 min after challenge (8). Benzalkonium reduces FEV 1 for at least 10 to 30 min and recovery is slow. Pros-
ZHANG, WRIGHT, TAM, NGUYEN-DANG, SALOME, AND WOOLCOCK
taglandin D 2 and leukotriene E 4 are potent bronchoconstrictors that enhance responses to histamine (9, 10). We have shown that inhalation of benzalkonium caused an enhanced response to histamine 1 h later. This is unlikely to be due to any residual effect of the inspiratory and expiratory maneuvers involved in the benzalkonium challenge. We have shown previously that histamine inhalation, using the same inhalation technique as used for benzalkonium, gives reproducible responses when performed at hourly intervals (11). The enhanced response to histamine is more likely to be due either to increased contractility of the smooth muscle or to changes in airway geometry due to persistent increases in airway wall thickness. Although differences between baseline FEV 1 values prior to the two histamine challenges were not statistically significant, a change in mean FEV1 values from 760/0 predicted prior to the first test to 700/0 predicted prior to the second test suggests there may have been some residual airway narrowing at the second test. James and coworkers (12) have suggested that very small changes in airway wall thickness, insufficient to cause changes in baseline lung function, can account for quite large changes in BHR. The most likely causes of a small increase in airway wall thickness are edema or an inflammatory infiltrate. On the other hand, an effect of benzalkonium on nerves cannot be excluded. The failure of ipratropium to protect against the response suggests that benzalkonium has no effect on the vagal reflex. However, SC has been shown to block afferent C-fibers (13)and this may explain its protective effect against benzalkonium in this study. An axon reflex acting on blood vessels, leading to persisting edema, could explain the slow recovery period and enhancement of the response to histamine. We have shown that responses to benzalkonium, although rapid in onset, are prolonged and continue to affect the airways after lung function has returned to baseline. In contrast, responses to metabisulfite occur rapidly and are short lived, with little residual effect on the airways (14). These differences suggest that these two preservatives affect the airways of asthmatic subjects by different mechanisms. It is likely that benzalkonium
acts in asthmatic airways primarily by a non-lgE-dependent release of mediators whereas metabisulfite has its effect primarily via neural pathways. Continued study of these two provoking agents may help to elucidate the relative contributions of non-IgE-dependent mediator release and neural mechanisms to bronchial hyperresponsiveness in asthma. References 1. Beasley CRW, Rafferty P, Holgate ST. Bronchoconstrictor properties of preservatives in ipratropium bromide (Atrovent) nebuliser solution. Br Med J 1987; 294:1197-8. 2. Clark RJ. Exacerbation of asthma after nebulised beclomethasone diproprionate. Lancet 1986; 574. 3. Miszkiel KA, Beasley R, Rafferty P, Holgate ST. The contribution of histamine release to bronchoconstriction provoked by inhaled benzalkonium chloride in asthma. Br J Clin Pharmac 1988; 25:157-63. 4. Morris JF, Koski A, Johnson LC. Spirometric standards for healthy nonsmoking adults. Am Rev Respir Dis 1971; 103:57-67. 5. YanK, SalomeCM, WoolcockAJ. Rapid method for measurement of bronchial responsiveness. Thorax 1983; 38:760-5. 6. Read GW, Kiefer EF. Benzalkonium chloride: selective inhibitor of histamine release induced by compound 48/80 and other polyamines. J Pharmacol Exp Ther 1979; 211:711-5. 7. Coleman JW, Holgate ST, Church MK, Godfrey RC. Immunoglobulin E decapeptide-induced 5-hydroxytryptamine release from rat peritoneal mast cells. Biochem J 1981; 198:615-9. 8. Holgate ST, Burns GB, Robinson C, Church MK. Anaphylactic and calcium-dependent generation of prostaglandin Dz (PGD z), thromboxane B, and other cyclooxygenase products of arachidonic acid by dispersed human lung cells and relationship to histamine release.J Immunol1984; 133: 2138-44. 9. Fuller RW, Dixon CMS, Dollery CT, Barnes PJ. Inflammatory mediator interactions: PGD z potentiates histamine and cholinergic responsiveness. Am Rev Respir Dis 1985; 131:A29. 10. O'Hickey SP, Arm JP, ReesPJ, Spurr BW,Lee TH. The relative responsiveness to inhaled leukotriene E 4 , methacholine and histamine in normal and asthmatic subjects. Eur Resp J 1988; 1:913-7. 11. Salome CM, Wright WJ, Sedgwick CJ, Woolcock AJ. Acute effects of fenoterol (Berotec) and ipratropium bromide (Atrovent) alone and in combination on bronchial hyperresponsivenessin asthmatic subjects. Prog Clin BioIRes 1988; 263:405-19. 12. James AL, Pare PD, Hogg JC. The mechanics of airway narrowing in asthma. Am Rev Respir Dis 1989; 139:242-6. 13. Dixon M, Jackson DM, Richards 1M. The action of sodium cromoglycate on "C" fibre endings in the dog lung. Br J Pharmacol 1980; 70:11-3. 14. Wright W, Zhang YG,Salome CM, Woolcock AJ. Effect of inhaled preservatives on asthmatic subjects. I. Sodium metabisulfite. Am Rev Respir Dis 1990; 141:1400-4.
Y. G. ZHANG, W. J. WRIGHT, W. K. TAM, T. H. NGUYEN-DANG, C. M. SALOME, and A. J. WOOLCOCK
Introduction
Recently reported cases of paradoxical bronchoconstriction with nebulized isotonic ipratropium bromide solution (1) and beclomethasone dipropionate (2) have led to the discovery that an additive in both solutions, namely benzalkonium chloride, causes bronchoconstriction in some asthmatic subjects. Benzalkonium chloride is a mixture of quaternary benzyldimethyl alkylammonium chlorides that has antibacterial properties. The mechanism by which benzalkonium causes bronchoconstriction is not known. Previous studies have suggested that an effect via mast cell mediator release is likely (1) and Miszkiel and colleagues have shown that histamine is responsible for at least part of the response (3). However, an effect on neural pathways has not been excluded and may account for the nonhistamine component of the response. In this report, we document a number of characteristics ofthe response to benzalkonium. Wehave examined the shape and position of the dose-response curves to benzalkonium in 28 asthmatic subjects and compared responsesto benzalkonium and histamine. The time course of the response and the effects of a response to benzalkonium on subsequent responses to histamine or benzalkonium have been described. Finally, the protective effect of 8 mg cromolyn sodium (SC) (Fisons plc, Leicestershire, England) and 160 ug ipratropium bromide (lB) (Atrovent't; Boehringer Ingelheim, FHERGMBH, West Germany) on challenges with benzalkonium was examined. Methods Subjects Twenty-eight subjects with stable asthma were studied. Subjects were patients from the Allergy Clinic of Royal Prince Alfred Hospital and volunteers from the University of Sydney. All subjects were free of other respiratory illnesses at the time of testing. All gave in-
SUMMARY Benzalkonium chloride has been used as a preservative in some antiasthma respirator solutions and is known to cause bronchoconstrictlon In asthmatic subjects. ToIncrease understanding of hoWIt causes bronchoconstrictlon, the characteristics of airway response in 28 asthmatic SUbjects were documented. SUbjects inhaled histamine, in doses ranging from 0.03 to 7.8 umel, or benzalkonium In doses ranging from 0.04 to 5.33 umol on separate days. The dose of histamine or benzalkonium that caused a 20% fall in the 1-sforced expiratory volume (PD20FEV 1) was measured. All subjects responded to histamine, with PD20FEV 1 values In the range of 0.14to 7.8 umol and 17 responded to benzalkonium, with PD20FEV1 values In the range 0.35 to 5.55 umet, Subjects who responded to benzalkonium were more sensitive to histamine (mean PD20FEV 1 0.44 j.1mol) than subjects who did not respond (mean PD20FEV1 1.84 j.1mol) and, among the benzalkonium responders, there was a significant correlation between PD2 0FEV1 values for histamine and benzalkonium (r == 0.5, P < 0.05). Inhalation of benzalkonium enhanced subsequent responses to histamine, causing a decrease In mean PD20FEV 1 from 0.51to 0.18 j.1mol histamine (p < 0.001), but did not alter subsequent responses to benzalkonium. The response to benzalkonium reached a maximum 1 min after inhalation and was slow to recover, taking up to 60 min to return to baseline values. Response to benzalkonium was Inhibited by 8 mg cromolyn sodium but not by 160 j.1g ipratropium bromide. The characteristics of the response to benzalkonium suggest a mechanism of action via release of mediators. AM REV RESPIR DIS 1990; 141:1405-1408
formed consent and the study was approved by the Hospital Ethics Committee.
Study Design Baseline studies included skin tests with 13 common allergens, measurement of spirometric function, a histamine inhalation test and, on a second day, a benzalkonium inhalation test. Eight subjects (Nos. 1 to 8) were studied on two additional days within ten days of the initial study. On Day 3 these subjects were challenged with benzalkonium, but were allowed to recover spontaneously without albuterol for at least 60 min. When the FEV 1 had returned to within 90010 of the baseline value subjects were challenged with histamine. On Day 4, a single dose of benzalkonium was administered. This dose was determined on the basis of the two previous benzalkonium challenges as the dose which, when given cumulatively, caused a 15to 25010 fall in FEV rSpirometric function was measured before challenge and then at 1, 2, 5, 10, 15, 30, 45, and 60 min after challenge and then at 15min intervals until the FEV 1 recovered to within 90010 of the baseline value. After 60 min, or when the FEV 1 had recovered, the single dose of benzalkonium was readministered. Spirometric function was measured at 1, 2, and 5 min later. Albuterol was administered
and spirometric function was measured again 10 min later. In addition, three subjects were retested on a fifth day using benzalkonium solution that was 1 month old to determine the shelf life of the solution. Two subjects (Nos. 2 and 14) were rechallenged with histamine and with benzalkonium 3 months after their first test after a period of treatment with aerosol steroids. Another five subjects (Nos. 9 to 13) were enrolled in a double-blind study to determine the effects of pretreatment with 8 rng SC or 160 J.1g IB on the response to benzalkonium. This study was carried out at the same time of day on four separate days within ten days. The drugs, which did not contain benzalkonium, were delivered by metered-dose aerosols. On each day baseline spirometric func-
(Received in original form July 25, 1989 and in revised form November 17, 1989) 1 From the Department of Medicine, University of Sydney, Sydney, and the Institute of Respiratory Medicine, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia. 2 Correspondence and requests for reprints should be addressed to Ms. C. M. Salome, Department of Medicine, University of Sydney, Sydney, NSW 2006, Australia.
1405
1406
ZHANG, WRIGHT, TAM, NGUYEN-DANG, SALOME, AND WOOLCOCK
TABLE 1 BENZALKONIUM CHLORIDE CHALLENGEDOSE SCHEDULE
Dose Number 1 2 3 4
5 6 7 8
Concentration (mglm/)
Inhalations*
3.1 3.1 6.2 6.2 12.5 12.5 25.0 25.0
1 1 1 2 2 4 4 8
Cumulative Dose t (pmo/) 0.044 0.088 0.18 0.35 0.71 1.41 2.82
5.64
• Nebulizer output: 0.005 ml (x 0.0004 SO) per inhalation. Molecular weight used in estimations: 354.
t
tion was measured. The subjects then inhaled eight puffs of SC (8 mg) or placebo and eight puffs of IB (160ug) or placebo. The benzalkonium inhalation tests were performed 10min later.
Lung Function Tests FEV 1 and FEC weremeasured using a Vitalograph dry spirometer (Vitalograph Ltd., Buckingham, England), with the subject standing. Forced expiratory maneuvers were repeated until a FEV 1, reproducible to within 100ml, was obtained. Values for FEV 1 and FVC were expressed as a percentage of the predicted values of Morris and coworkers (4). If the FEV I/FVC was less than 60070 or the FEV 1 was lessthan 60% predicted, the subjects were not challenged. Inhalation Tests Inhalation tests with histamine and with benzalkonium were carried out according to the method described by Yan and colleagues (5). Briefly, histamine, in concentrations of 3.1 mg/ml, 6.2 mg/ml, 25 mg/ml, and 50 mg/ml in normal saline, or benzalkonium, in concentrations of 3.1, 6.2, 12.5, and 25 mg/ml in normal saline, was administered using DeVilbiss No. 40 (DeVilbiss Co., Somerset, PAl hand-held glassnebulizers. Nebulizer output was checked prior to the study and it was found that nebulizers containing benzalkonium had a mean output of 0.005 ml (± 0.0004 SD) compared with a mean output of 0.003 ml (± 0.0003 SD) from the same nebulizers containing histamine. This difference was attributed to an effect of benzalkonium on the surface tension of the solution. Doses of histamine ranged from 0.03 to 7.8 umol and doses of benzalkonium from 0.042to 5.33 umol. The dose schedule for benzalkonium is shown in table 1. Three inhalations of normal saline wereadministered as a control, prior to the challenge. The subject inhaled over 1 to 2 s towards total lung capacity. At the beginning of the inspiration, the operator gave the bulb of the nebulizer one firm squeeze. After a slight breath-hold (3 s), the subject exhaled. Spirometric function was measured 1 min after the dose and the next dose given
within 3 min of the previous dose. The challenge was stopped when the FEV 1 had fallen 20% or more from the postsaline value, or when the highest dose had been given. To aid recovery, 200 J.1g albuterol wasgiven and FEV 1 was measured 10min later. Subjects withheld aerosol bronchodilators for 6 h or oral bronchodilators for 12 h prior to testing.
Data Analysis On the basis of the measured volume delivered by the nebulizer and the known concentration of the solution, the dose of histamine
and benzalkonium delivered to the mouth of the subject was calculated in urnoles. Because benzalkonium is a mixture of molecules of varying composition, an estimated average molecular weight of 354 was used to calculate doses. Response was measured as percentage change in FEV 1 from the post saline value. This was plotted on a linear scale against log dose of histamine or benzalkonium and the dose that caused a 20% fall in FEV 1 (PDzoFEV1) was read from the graph. The PDzoFEV1 values were logarithmically transformed and all subsequent analysis was performed on the log values. The geometric mean and 95% confidence intervals (ci) of PDzoFEV1 values from each study were calculated. Paired t tests were used to examine the significance of differences. Correlation and regression coefficients were calculated to determine the degree of association between values. Significance was taken at the 5% level. Results
Details of the 28 subjects are shown in table 2. Subjects were 19to 71 yr of age; all subjects had two or more positive responses to skin tests; and all were receiving treatment for asthma. All had baseline FEV 1 and FVC values greater than 60010 and 70070 of the predicted values,
TABLE 2 DETAILS OF SUBJECTS PDzo FEV1 SUbject
Agel Sex
Current Treatment*
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
231M 241M 291M 231M 351M 201M 191M 231M 45/M 311F 421F 221M 211M 55/M 211M 201M 30/F 36/F 201M 71/F 221M 231M 211M 40/F 281M 431M 56/F 48/M
A A, B,C A,B A A,C A, B A A, B A A, B, T A,B A, S A, B A, T A A,S A, B A,S A,C A, B, T A, B A,C A A, B A,B A,B A, B, I, T A, B, T
= ipratropium
bromide; T
61 73
77 98 93 60 74
79 91 114 65 89 63 55 86 88 71 79 80 86 81 84 103 82 89 80 76 66
FVC (% pred)
73 85 86 84 97 82 88 94 97 109 88 108
77 77 84 92 87 70 76 100 80 91 97 112 90 91 83 86
= albuterol; C = cromolyn sodium; B = theophylline.
Definition of abbreviations: A
I
FEV 1 (% pred)
Histamine 0.60 0.70 1.10 0.68 0.86 0.14 0.26 0.50 0.34 0.21 0.11 0.45 1.2 0.24 1.1 1.0 0.17 7.8 1.4 1.3 4.6 2.4 3.7 1.2 5.0 1.8 0.44 0.29
Benzalkonium chloride 2.2 5.3 1.4 3.0 2.0 1.0 2.5 1.2 0.9 2.4 3.3 3.6 3.2 0.4 5.5 5.5 0.35
= beclomethasone dipropionate;
1407
EFFECT OF BENZALKONIUM CHLORIDE ON ASTHMATIC SUBJECTS
50 40 ~
30
f
ZO
1:J~~~~ 0.04
0.10
1.00
10.00
DOSE BNZ (I'mol)
Fig. 1. Dose-response curves to inhaled benzalkonium chloride (BNZ) in 28 asthmatic subjects.
respectively. There were no significant differences between baseline FEV 1 measurements on different study days. Seventeen subjects (Nos. 1 to 17) responded to the benzalkonium inhalation test with PD20FEV 1 values between 0.35 and 5.5 umol (table 2). Dose-response curves to benzalkonium for all 28 subjects are shown in figure 1. Subjects who responded to benzalkonium were more sensitive to histamine (mean PD 20FEV 1 0.44 umol histamine, ci: 0.31, 0.64) than those who did not respond (1.84 umol histamine, ci: 1.0, 3.4) (p < 0.001). Among the benzalkonium responders, PD20FEV 1 values for histamine and benzalkonium were significantly correlated (r = 0.50, p < 0.05). Twosubjects with PD 20FEV 1 values of 0.41 and 5.1lJmol to benzalkonium were retested after 3 months of treatment with aerosol steroids. Their responsiveness to histamine had decreased from initial PD20FEV 1 values of 0.24 and 0.7 umol to final values 5.0 and 7.8 umol, and they were no longer responsive to benzalkonium. Responsesto benzalkonium werereproducible to within ± 1.8doubling doses. Geometric mean values for PD20FEV 1 on two study days for eight subjects were 2.03 umol (ci:1.4,2.9) and 2.17 umol (ci: 1.2,3.9) and werenot significantly different. Mean difference between values was 0.03 log units (ci: -0.16, 0.22). The time course of the changes in FEV 1 over 60 min after a single dose of benzalkonium is shown in figure 2. The fall in FEV 1 was immediate and reached a maximum after 5 min, with a mean fall of 280/0. Recovery was slow and 60 min after challenge, the mean FEV 1 remained 100/0 below the initial value. In a separate study, response to benzalkonium was easily reversed with 200 IJgalbuterol. After challenge, the mean FEV 1 fell by 30.60/0 from initial and, 10min after 200 IJg albuterol, it increased to within 1% of the initial value (p < 0.001).
The doses of benzalkonium acted cumulatively. When the total cumulative dose, which caused a 20% fall in FEV 1 during the progressive challenge, was given as a single dose there was no significant difference between the percentage falls in FEV 1. The mean percentage fall in FEV 1 in eight subjects after a single dose was 23.10/0, and after the equivalent dose given progressively it was 190/0 on Day 2 and 20% on Day 3. Challenge with benzalkonium enhanced responses to histamine inhaled 1 h later. Geometric mean values for PD 20FEV 1 in eight subjects decreased from 0.51 umol (ci: 0.32, 0.81) for histamine alone to 0.18 umol (ci: O. 09, 0.33) for histamine challenge 1 h after benzalkonium (p < 0.001). Mean initial FEV 1 was 760/0 predicted prior to challenge with histamine alone and 700/0 prior to challenge with histamine after benzalkonium. These values were not significantly different. There did not appear to be a refractory period after benzalkonium inhalation. Mean fall in FEV 1 in six subjects, after a singledose of benzalkonium was 30.3% (ci: 22.6, 38.0) and after the same dose given 1 h later was 23.5% (ci: 19.4,27.6). These values for the group were not significantly different. However, in four of the six subjects the response to the second dose of benzalkonium was inhibited by 22, 30, 34, and 3'8 %, suggesting that a larger sample may show some significant refractoriness. Of the two subjects who did not complete this stage of the study, one suffered an acute exacerbation of his asthma and was not studied on Day 4, and the second did not have sufficient improvement in FEV 1 2.5 h after the first challenge to allow a second challenge. Response to benzalkonium was inhibited by 8 mg SC but not by 160 IJg
100
0
:J
90
]
:!; '0
80
~
i" to.
70
Discussion
The present study showsthat in asthmatic subjects with severe or moderate bronchial hyperresponsiveness,benzalkonium chloride causes reproducible bronchoconstriction that is rapid in onset, is slow to recover,and causes enhanced responses to histamine. The response is inhibited by SC but not by lB. There were no complications or serious side-effects of the test, although all subjects experienced cough and a burning sensation within moments of inhalation. The bronchoconstrictive effects of benzalkonium chloride were originally described after observations of patients who experienced marked reductions in lung function after inhalation of IB nebulizer solution (l). Atrovenr" nebulizer solution now contains no benzalkonium, but previously it contained 250 ug/ml benzalkonium chloride. At this concentration, patients inhaling 1 ml of this solution during tidal breathing might have been expected to receive approximately 0.35 to 0.7 umol benzalkonium. In the present study PD 20FEV 1 values were in the range of 0.35 to 5.0 umol and only
No PDZO
Q)
g
lB. Figure 3 shows PD 20FEV 1 values in six subjects for challenge with benzalkonium after placebo, SC, and lB. There was no significant difference between mean PD 20FEV 1 values after placebo (1.96 umol, ci: 0.97, 3.98) and after IB (3.17IJmol, ci: 1.63,6.15). PD 20FEV 1 values weregreater than the maximum dose of 5.64 umol in five subjects after SC. The solutions of benzalkonium had a shelf life of at least 1 month. Three subjects were challenged with f-month-old benzalkonium solutions, and the mean PD 20FEV 1 value of 3.61 umol, was not significantly different from the mean PD 20FEV 1 value of 2.75 umol obtained using fresh solutions.
1
1
iLLL//f~r~1
R
100j
z
CD
"0
!
f
1.0
0 N
1111
0
0 e,
60 +---,-~--..----..--.------.--, -10 10 40 30 50 60
zo
TIME AFTER CHALLENGE (mins)
x~:
0.1
IPB~~~F~~m
Placebo
(1601'9)
Fig. 2. (left) Time course of response to benzalkonium in eight asthmatic subjects. Values are mean and standard deviation. Fig. 3. (right) Effect on PD 2oFE\t, values for benzalkonium challenge after placebo, ipratropium bromide (160 I1g) or sodium cromoglycate (8 mg).
1408
two subjects had responses in the 0.35 to 0.7 umol range. Beasley and coworkers (1) found a mean PC 2oFEV. value of 0.3 giL or approximately 1.7/lmol (range 0.7 to 11.3 umol) and Miszkiel and colleagues (3) found a mean PC 2 oFEV. of 3.98 giL or approximately 1.96 umol (range 0.45 to 11.8 umol), The results of these studies suggest that only a few very sensitive patients were likely to respond to the benzalkonium chloride in a normal dose of Atrovent" nebulizer solution, and accounts for the relative rarity of cases of paradoxical bronchoconstriction with this drug. The mechanism by which benzalkonium chloride causes bronchoconstriction is not known. However, an effect of benzalkonium on mediator-releasing cells is suggested by the results of this and previous studies. In vitro studies of rat serosal mast cells have shown that benzalkonium inhibits histamine release, but at higher concentrations (10ug/ml) induces release, probably by a lytic effect (6). Furthermore, benzalkonium enhances IgE-dependent release of 5-hydroxytryptamine (7). In human subjects, benzalkonium may also cause histamine release. Miszkiel and coworkers (3) have recently shown that the selective Hi-receptor antagonists, terfenadine and astemizole, shift the benzalkonium concentration-response curve to the right. However, the protection afforded by these agents was only partial and was confined to the early phase of the response. The authors suggest that histamine release represents a small component of the overall effect of the chemical. In the present study 11 subjects failed to respond to benzalkonium despite being responsive to inhaled histamine. These 11 subjects were less sensitive to histamine than the benzalkonium responders, suggesting that if benzalkonium does cause histamine release then the amount released was insufficient to cause a response in these subjects. Mediators other than histamines may be involved in the response. If benzalkonium causes mast cell cytolysis (6) then it is possible that lipid-derived mediators, such as prostaglandins and leukotrienes, are generated at the same time. Prostaglandin generation by immunologically activated human mast cells occurs between 10 and 30 min after challenge (8). Benzalkonium reduces FEV 1 for at least 10 to 30 min and recovery is slow. Pros-
ZHANG, WRIGHT, TAM, NGUYEN-DANG, SALOME, AND WOOLCOCK
taglandin D 2 and leukotriene E 4 are potent bronchoconstrictors that enhance responses to histamine (9, 10). We have shown that inhalation of benzalkonium caused an enhanced response to histamine 1 h later. This is unlikely to be due to any residual effect of the inspiratory and expiratory maneuvers involved in the benzalkonium challenge. We have shown previously that histamine inhalation, using the same inhalation technique as used for benzalkonium, gives reproducible responses when performed at hourly intervals (11). The enhanced response to histamine is more likely to be due either to increased contractility of the smooth muscle or to changes in airway geometry due to persistent increases in airway wall thickness. Although differences between baseline FEV 1 values prior to the two histamine challenges were not statistically significant, a change in mean FEV1 values from 760/0 predicted prior to the first test to 700/0 predicted prior to the second test suggests there may have been some residual airway narrowing at the second test. James and coworkers (12) have suggested that very small changes in airway wall thickness, insufficient to cause changes in baseline lung function, can account for quite large changes in BHR. The most likely causes of a small increase in airway wall thickness are edema or an inflammatory infiltrate. On the other hand, an effect of benzalkonium on nerves cannot be excluded. The failure of ipratropium to protect against the response suggests that benzalkonium has no effect on the vagal reflex. However, SC has been shown to block afferent C-fibers (13)and this may explain its protective effect against benzalkonium in this study. An axon reflex acting on blood vessels, leading to persisting edema, could explain the slow recovery period and enhancement of the response to histamine. We have shown that responses to benzalkonium, although rapid in onset, are prolonged and continue to affect the airways after lung function has returned to baseline. In contrast, responses to metabisulfite occur rapidly and are short lived, with little residual effect on the airways (14). These differences suggest that these two preservatives affect the airways of asthmatic subjects by different mechanisms. It is likely that benzalkonium
acts in asthmatic airways primarily by a non-lgE-dependent release of mediators whereas metabisulfite has its effect primarily via neural pathways. Continued study of these two provoking agents may help to elucidate the relative contributions of non-IgE-dependent mediator release and neural mechanisms to bronchial hyperresponsiveness in asthma. References 1. Beasley CRW, Rafferty P, Holgate ST. Bronchoconstrictor properties of preservatives in ipratropium bromide (Atrovent) nebuliser solution. Br Med J 1987; 294:1197-8. 2. Clark RJ. Exacerbation of asthma after nebulised beclomethasone diproprionate. Lancet 1986; 574. 3. Miszkiel KA, Beasley R, Rafferty P, Holgate ST. The contribution of histamine release to bronchoconstriction provoked by inhaled benzalkonium chloride in asthma. Br J Clin Pharmac 1988; 25:157-63. 4. Morris JF, Koski A, Johnson LC. Spirometric standards for healthy nonsmoking adults. Am Rev Respir Dis 1971; 103:57-67. 5. YanK, SalomeCM, WoolcockAJ. Rapid method for measurement of bronchial responsiveness. Thorax 1983; 38:760-5. 6. Read GW, Kiefer EF. Benzalkonium chloride: selective inhibitor of histamine release induced by compound 48/80 and other polyamines. J Pharmacol Exp Ther 1979; 211:711-5. 7. Coleman JW, Holgate ST, Church MK, Godfrey RC. Immunoglobulin E decapeptide-induced 5-hydroxytryptamine release from rat peritoneal mast cells. Biochem J 1981; 198:615-9. 8. Holgate ST, Burns GB, Robinson C, Church MK. Anaphylactic and calcium-dependent generation of prostaglandin Dz (PGD z), thromboxane B, and other cyclooxygenase products of arachidonic acid by dispersed human lung cells and relationship to histamine release.J Immunol1984; 133: 2138-44. 9. Fuller RW, Dixon CMS, Dollery CT, Barnes PJ. Inflammatory mediator interactions: PGD z potentiates histamine and cholinergic responsiveness. Am Rev Respir Dis 1985; 131:A29. 10. O'Hickey SP, Arm JP, ReesPJ, Spurr BW,Lee TH. The relative responsiveness to inhaled leukotriene E 4 , methacholine and histamine in normal and asthmatic subjects. Eur Resp J 1988; 1:913-7. 11. Salome CM, Wright WJ, Sedgwick CJ, Woolcock AJ. Acute effects of fenoterol (Berotec) and ipratropium bromide (Atrovent) alone and in combination on bronchial hyperresponsivenessin asthmatic subjects. Prog Clin BioIRes 1988; 263:405-19. 12. James AL, Pare PD, Hogg JC. The mechanics of airway narrowing in asthma. Am Rev Respir Dis 1989; 139:242-6. 13. Dixon M, Jackson DM, Richards 1M. The action of sodium cromoglycate on "C" fibre endings in the dog lung. Br J Pharmacol 1980; 70:11-3. 14. Wright W, Zhang YG,Salome CM, Woolcock AJ. Effect of inhaled preservatives on asthmatic subjects. I. Sodium metabisulfite. Am Rev Respir Dis 1990; 141:1400-4.
