Thorax 1995;50:511-516 511 Adenosine, methacholine, and exercise challenges in children with asthma or paediatric chronic obstructive pulmonary disease Avraham Avital, Chaim Springer, Ephraim Bar-Yishay, Simon Godfrey Institute of Pulmonology, Hadassah University Hospital, Kiryat Hadassah, P.O. Box 12000, Jerusalem 91120, Israel A Avital C Springer E Bar-Yishay S Godfrey Reprint requests to: Dr A Avital. Received 18 October 1994 Returned to authors 12 December 1994 Revised version received 12 January 1995 Accepted for publication 16 January 1995 Abstract Background - Bronchial hyperreactivity to methacholine is present in children with asthma and other types of paediatric chronic obstructive pulmonary disease (COPD), while hyperreactivity to exercise is more specific for asthma. Adenosine 5'monophosphate (AMP) is a potent bronchoconstrictor and, like exercise, may provoke asthma by activating mast cells. This study investigated the suitability of AMP as a specific challenge for asthma in children. Methods - Bronchial provocation challenges with methacholine and AMP were performed in a double blind fashion using tidal breathing in 51 children with asthma, 21 with paediatric COPD of various types, and in 19 control children. Each subject also underwent a standardised exercise challenge after inhalation challenges were completed. Sensitivity and specificity curves were constructed and the intersection point of sensitivity and specificity for each type of challenge was determined. Results - When the asthmatic patients were compared with the children with COPD, the intersection points for AMP, exercise and methacholine were 90%, 85%, and 50%, respectively. When compared with the controls the same intersection points were 98%, 84%, and 92%, and when children with paediatric COPD were compared with controls they were 55%, 50%, and 82%. Conclusions - Methacholine distinguishes both asthma and paediatric COPD from controls with a sensitivity of 82-92%, but does not distinguish between asthma and paediatric COPD; exercise and AMP distinguish asthma from controls with a sensitivity and specificity of 84-98% but they also distinguish asthma from paediatric COPD with a sensitivity and specificity of 85-90%. AMP inhalation is a practical aid for diagnosing asthma and distinguishing it from COPD in children of all ages. (Thorax 1995;50:51 1-516) Keywords: methacholine, adenosine, asthma, children. Bronchial hyperreactivity to methacholine or histamine is a characteristic feature of asthma.'-' Other chronic lung diseases, especially those associated with chronic infection, may also be associated with increased methacholine reactivity." However, in adult patients with chronic infections, bronchial hyperreactivity by cold air or isocapnic hyperventilation has been found in a minority of the patients.9'0 Moreover, we have recently shown" that children with asthma and those with chronic lung disease have hyperreactivity to methacholine, but only those with asthma are hyperreactive to physical exercise. The pathophysiology of exercise-induced asthma is still disputed but there is increasing evidence of mediator release being involved. Exercise is associated with significant increases in plasma histamine and neutrophil chemotactic activity'2"15 which have been considered markers for the release of mediators by mast cells.'6'7 Adenosine is a potent bronchoconstrictor in asthmatic patients, possibly by stimulating or enhancing the release of mediators from mast cells.' 819 Adenosine has been shown to potentiate the release of preformed but not newly generated mediators from mouse bone marrow-derived mast cells in tissue culture20 and has also been used for bronchial challenge in asthmatic patients.21-23 Thus, both exercise and adenosine may induce bronchoconstriction by releasing mediators from airway mast cells. Oosterhoff et a124 compared adults with chronic obstructive pulmonary disease (COPD) with adult asthmatic patients and found that non-smoking adults with COPD were less responsive to adenosine than to methacholine, while the asthmatic patients were similarly responsive to both. They did not study the response to exercise or hyperventilation. The differentiation of asthma from other types of COPD can be difficult in children since the clinical picture can be very similar. In the present study we have compared bronchial challenges by exercise, methacholine, and adenosine 5'-monophosphate (AMP) in children and young adults with bronchial asthma, in young patients with chronic obstructive lung diseases, and in healthy subjects of similar age. The object of the study was to determine whether AMP, like exercise, was a more specific stimulus than methacholine in differentiating asthma from chronic obstructive lung diseases in such patients. Methods SUBJECTS Ninety one children and young adults aged 6-25 years participated in the study; 51 met the Avital, Springer, Bar-Yishay, Godfrey 512 Table 1 Anthropometrtic and spirometric data of the subjects Group n Asthma Paediatric COPD Cystic fibrosis Bronchiolitis obliterans Primary ciliary dyskinesia Bronchiectasis 51 Control 19 7 7 4 3 Mean (range) age (years) Sex (MIF) Mean (SE) baseline FEV, (O/) 13 0 (6-25) 32/19 86-4 (1 6) (9-12) (7-15) (7-19) (9-13) 4/3 3/4 2/2 2/1 14 0 (8-23) 10/9 11-1 11-4 13 7 11-3 American Thoracic Society's diagnostic criteria for asthma25 and 21 had chronic obstructive lung disease (seven with cystic fibrosis, seven with bronchiolitis obliterans, four with primary ciliary dyskinesia, and three with bronchiectasis). We have termed this latter group paediatric chronic obstructive pulmonary disease (COPD). A further 19 patients were originally referred to the department for lung function and bronchial provocation tests because of subjective complaints of mild cough or breathlessness. Five children had anxietyrelated sighs, five mild post-viral cough, five mild intermittent cough, three were normal children with shortness of breath during exercise without asthma, and one had heartburn. In every case full investigation and follow up failed to reveal any objective evidence of lung disease and were used as a control group for the analyses. The exclusion of asthma or COPD was established before performing the challenges. All subjects were recruited from the paediatric clinic of the Institute of Pulmonology, Hadassah University Hospital, Jerusalem. Anthropometric and spirometric data of the subjects are presented in table 1. Of the asthmatic patients, 30 were using inhaled bronchodilators as needed, four used sodium cromoglycate, and 17 used inhaled beclomethasone dipropionate. There was no history of upper or lower respiratory tract infection 100-00 r 10-00 K A v A AA v gv E 0, 0 E I 1-00 h- A V *A Fresh solutions of methacholine and AMP were made up in phosphate buffer solution in a range of concentrations from 0-03 to 32 mg/ ml (0 15-163 5 mmol/1) for methacholine and from 0-39 to 400mg/ml (1-12-1152mmol/1) for AMP. Two sets of 11 syringes (2 0 ml) with doubling concentrations of methacholine and AMP were prepared and coded so that neither the patient nor the person performing the investigation knew whether methacholine or AMP was being used. Inhalation challenges were performed by the method described by 0 *0 *0 1 Asthma Ak A AA A COPD Control PC20 for methacholine (MCH) in children with asthma, paediatric COPD, and control subjects. STUDY DESIGN The three challenges were performed on the same day in 55 patients (86%), within seven days in 78 patients (86%), and within 15 days in 86 patients (95%). In the five remaining subjects the challenges were performed within 30 days. The inhalation challenges were always performed in random order on the same day. When the exercise challenge was also performed on the same day as the inhalation challenge it was always the last challenge of the day. The exercise test was performed on a separate day in 16 subjects and on the same day as one of the inhalation challenges in 20 subjects. The interval between challenges on any one day was 2-5-3 hours which allowed the FEV1 to return to within 10% of baseline value of the first challenge in every case before the second and third challenges were performed. Salbutamol inhalation was administered after the last challenge if the fall in FEV, was greater than 20%, if the patient was dyspnoeic, or if the FEV, did not return to within 10% of the baseline value. BRONCHIAL PROVOCATION AND MEASUREMENT OF PULMONARY FUNCTION A Figure during the four weeks before the studies in any of the subjects. At study entry all the control subjects and all but 10 patients (two with asthma, four with bronchiolitis obliterans, two with cystic fibrosis, one with bronchiectasis, and one with primary ciliary dyskinesia) had a forced expiratory volume in one second (FEVI) above 70% of the predicted value. All patients avoided bronchodilator therapy for at least 12 hours and sodium cromoglycate for at least 20 hours before the study. Inhaled corticosteroid therapy was continued unchanged. None of the 72 subjects was treated with theophyllinerelated compounds. The study was approved by the local ethics committee and written informed consent was obtained from the older subjects or from one parent. A 0-a- 0-01 95 0 (2 1) A *0 0-10 K (6 2) (3-8) (7 1) (9 9) A .0~ 0 78-9 69 6 80 8 80 3 Bronchial provocation in children with asthma and COPD 513 of FEVy at each interval was recorded. The concentrations of methacholine or AMP causing a 20% fall in FEV, (PC20) from the post60 f phosphate buffer value were derived from a plot of percentage fall in FEV, against log . 0 methacholine or log AMP concentration. If the fall in FEVy was less than 10% at the maximal 40 Fconcentration, the next doubling values - that w is, 64 mg/ml for methacholine and 800 mg/ml Ufor AMP - were arbitrarily taken as PC20. *. *X The exercise challenge was carried out using A *0 20K six minutes of treadmill running as described C) 0 previously.26 The treadmill was set at a slope 0L)a) *0. of 100 and the subjects ran continuously at a * 0 01) -A A speed of 5 km/h. This produced a heart rate of x wU 0 160-180/min and represented approximately VWVV two thirds of the maximal working capacity. FEV1 was measured before and 3, 5, 10, and 15 minutes after exercise. The challenge was performed in an air conditioned laboratory and -20 Control COPD Asthma the subjects breathed room air with a temperature of 22-260C and a relative humidity Figure 2 Percentage fall in FEV, after exercise in children with asthma, paedfiatric of 48-56%. The result of the exercise challenge COPD, and control subjects. was calculated as the greatest fall in FEV, after exercise expressed as a percentage of the preCockcroft et al.' The nebuliser char iber (Ac- exercise value. corn II, Marquest, Englewood, USA) was filled with 2 0 ml of test solution and an oairflow of 5 1/min resulted in a mean rate of net)ulisation STATISTICAL ANALYSIS of 0 39 ml/min. The nebuliser was c-onnected PC20 values were log transformed before statthrough a one-way valve system to:a mouth- istical analysis. Data are expressed as means piece through which the child breath ed norm- (SE). Statistical comparisons between groups ally while wearing a nose clip. The first and within groups were made by analysis of inhalation comprised phosphate biuffer so- variance (ANOVA) and paired Student t tests lution, and then methacholine or AAAP in in- as appropriate. All tests of significance were creasing concentrations. Each solution was two tailed. Differences were taken as significant inhaled for two minutes during tidal breathing when p was <0 05. with measurement of lung function ait 30, 90, and 180 seconds after each inhalationi until the FEV, had fallen by 20% or more from the post- Results buffer value, or until the maximium con- There was no significant difference between centration was reached. Lung func-tion was the mean age of the subjects in all three groups. measured in duplicate with a pneumo- The mean baseline FEV, (table 1) was sigtachograph based system (Vitalogralph Com- nificantly (p<001) lower in the paediatric pact, Buckingham, UK) and the higIiest value COPD (77 9 (3 4)%) and the asthmatic groups (86-4 (1-5)%) than in the control group (95-0 (2-1)%), and the difference between the asthmatic and paediatric COPD patients was also significant. There was no significant difference 1000 between the mean baseline FEV, before the _vvvvwv three types of challenge within any of the three VW groups. Individual results for the three challenges in the three groups are shown in figs 1-3. It can 0 be seen that in the control group 11 subjects (58%) failed to respond to the highest concentration of methacholine used and 16 subE CD jects (84%) failed to respond to the highest *0*** 10 r000 concentration of AMP. In the paediatric COPD all responded to methacholine but 15 group #0ft subjects (71%) failed to respond to AMP. In Clthe asthmatics all but one subject responded to methacholine and all but one (not the same subject) to AMP. By its nature the exercise s*0 challenge provided a result in every case even 000*0 if there was a small rise (negative fall) in FEV1 after exercise and the values were normally distributed in each group. The mean pern Asthma COPD Control centage fall in FEV, after exercise in the asthFigure 3 PC20 for AMP in children with asthma, paediatric COPD, and con,ztrol subjects. matic group (20-5 (14.8)%) was significantly *e AAAAA AA AA*AAAA A A A 100 A AL A 0 Avital, 514 100 0- 80 .2 0. a 60 _ 40 _ co Q) aI) 20 _ 0 *1 0- 100 10 1 1000 PC20AMP (mg/ml) Figure 4 Sensitivity (M) and specificity (*) of PC20 AMP in differentiating asthma from paediatric COPD at various AMP concentrations. 100 r 0- 80 Co ._ 60 CL C 40 'a 20 (a) en 0 - 0.1 1 100 10 PC20AMP (mg/ml) 1000 Figure 5 Sensitivity (A) and specificity (*) of PC20 AMP in differentiating asthma from controls at various AMP concentrations. different (p<00001) from that of the paediatric COPD (2'8 (541)%) and control groups (1P8 (2 1)%). There was no significant difference in the response to exercise between the paediatric COPD and control groups. One of the patients with paediatric COPD (primary ciliary dyskinesia) had a 21% fall in FEV, after exercise even though he had no personal or family history of atopic disease or asthma. Since all but one asthmatic and all the patients with paediatric COPD responded to methacholine it was possible to compare the responses in these groups. The geometric mean PC,0 for methacholine of the asthmatic group (028 mg/ ml, range 022-0O34 mg/ml) was not sigTable 2 Intersection points of sensitivity and specificity curves of the three challenges Sensitivity and specificity at intersection (%) Value of test at intersection 50 92 82 0-28 mg/ml 2-20 mg/ml 3-80 mg/ml controls Paediatric COPD v controls 85 84 50 5% fall in FEV, 4% fall in FEV, 3% fall in FEV, Adenosine: Asthma v paediatric COPD Asthma v controls Paediatric COPD v controls 90 98 55 25 mg/ml 210 mg/ml >400 mg/ml Methacholine: Asthma v paediatric COPD Asthma v controls Paediatric COPD v controls Exercise: Asthma Asthma v v paediatric COPD Springer, Bar-Yishay, Godfrey nificantly different from that of the paediatric COPD group (0.36 mg/ml, range 0-22058 mg/ml). With the reservation that a nominal (minimum) value for PC20 was arbitrarily assigned to non-responders to methacholine and AMP, the geometric mean PC,0 for AMP of the asthmatic group (4 43 mg/ml, range 3 57-5 50 mg/ ml) was significantly different (p<0O0001) from that of the paediatric COPD group (320-38 mg/ ml, range 294-3-550-7 mg/ml) which was not significantly different from that of the control group (800 mg/ml). Sensitivity and specificity in differentiating between asthma and paediatric COPD (which are not affected by arbitrary values assigned to non-responders) were calculated for every concentration of methacholine and AMP used and for every 5% fall in FEVy after exercise. Sensitivity of a challenge for asthma compared with paediatric COPD was defined as the ratio of the number of asthmatics with a positive test at the chosen concentration (or fall in FEVy) to the total number of asthmatics. Specificity for asthma was defined as the ratio of the number of patients with paediatric COPD with a negative test at the chosen concentration (or fall in FEVI) to the total number of patients with paediatric COPD. Curves of sensitivity and specificity for every step of the challenges were constructed as shown for the AMP challenge in fig 4. As sensitivity increases with increasingly strong stimuli, so specificity falls, and the best combination of sensitivity and specificity that can be obtained is at the point where they cross. These intersection points are given in table 2 from which it can be seen that the value for AMP was higher than that for exercise, and that both AMP and exercise yielded considerably greater sensitivity and specificity than methacholine in separating asthma from paediatric COPD. Similarly, sensitivity and specificity curves for differentiating between asthma and controls and between paediatric COPD and controls were calculated from the data of the asthmatic versus control and paediatric COPD versus control groups respectively. The intersection points of these analyses are also given in table 2 and the plot for AMP in distinguishing asthma from controls is shown in fig 5. This analysis shows that, to distinguish between asthma and controls, all these types of challenge yielded a similarly high level of sensitivity and specificity. Neither exercise nor AMP were particularly helpful in distinguishing paediatric COPD from controls (because the patients with paediatric COPD mostly failed to respond to these challenges), but methacholine did yield a reasonable specificity and sensitivity. Discussion Methacholine challenge has been widely used for the detection and quantitation of bronchial hyperreactivity in asthmatic patients. 13 Hargreave and coworkers27 suggested a cut-off limit of 8 mg/ml for the PC20 to methacholine since all their non-asthmatic subjects had a PC20 above this value while all their asthmatic sub- Bronchial provocation in children with asthma and COPD jects with recent symptoms had a lower PC20 value. In two previous studies" 28 performed in our laboratory on 52 and 182 asthmatic children the upper limits (least reactivity) of methacholine responsiveness (using log mean + 2 log SD) were 4-4 and 7-7 mg/ml, respectively. In the present study the upper limit of PC20 for methacholine (log mean + 2 log SD) in the asthmatic group was 5-6 mg/ml. Although a few of the control group had PC20 values below this range, the analysis of sensitivity and specificity (table 2) showed that, in concurrence with our previous study, " methacholine was useful in distinguishing the asthmatic subjects from the controls but not in differentiating between asthma and other types of chronic obstructive lung disease in children. Exercise-induced bronchoconstriction has also been regarded as an important hallmark of asthma, but the percentage fall in FEVy taken as the cut-off limit for asthma in the literature ranges from about 10% to 20%.29 M Burr and coauthors32 studied a group of 812 children aged 12 years who were healthy and unrelated to asthmatic subjects and found that 92% had a fall in peak expiratory flow rate after exercise of less than 10%. In our previous study" the upper limit of percentage fall in FEV, in control children (taken as mean + 2SD) was 8&2% and in the present study it was 6-0%. In the present study 40 of the 51 asthmatic subjects (78%) had a fall in FEV, of more than 6-0% while only one of the patients with paediatric COPD showed significant bronchoconstriction after exercise. This is very similar to the results of our previous study where none of the 22 children with paediatric COPD responded to exercise by bronchoconstriction." The inhalation challenges were performed randomly but the exercise test was performed at the end of the day in most cases. There is a slight possibility that the inhalation challenges might have influenced the exercise test. However, the protocol was identical for the different groups of patients and therefore, if there was any effect of one challenge on the response of the other, it would be the same in the three groups. From table 2 it can be seen that the intersection ofsensitivity and specificity for exercise was a little less than for methacholine in distinguishing the asthmatic from the control subjects, but it was much greater in distinguishing asthma from paediatric COPD. There is a problem in the evaluation of the mean PC20 of the paediatric COPD and control groups since we may have underestimated the true PC20 of the non-responders and the assumed values taken for statistical analysis were the least values for PC20. Given these restrictions, the significance of the very large differences in the responses of the various groups is almost certainly underestimated. The response of patients with paediatric COPD to AMP and exercise was poor or nonexistent up to the highest concentration used in the challenge. These subjects had lower basal lung function but, since they responded very well to methacholine, this could not have accounted for their lack of response to AMP and exercise. As a further check on the possible 515 effect of baseline lung function we selected nine children with asthma and nine with paediatric COPD who were matched for age (11 e 16 (3 2) and 11 7 (3 2) years, respectively) and for baseline FEV1 (89-3 (10-0)% and 88-2 (8-9)%, respectively). There was no significant difference in the geometric mean PC20 to methacholine (0-29mg/ml (range 0-14-0 6) for paediatric COPD and 0 47 mg/ml (range 0-31 to 0 72) for asthmatics), while there was a significant difference for mean percentage fall in FEV1 after exercise (1-7 (3 0)% in those with paediatric COPD and 20-6 (18-5)% in asthmatics, p<001) and for geometric mean PC20 to AMP (257-3mg/ml (range 148-8444 9) for patients with paediatric COPD and 6-35 mg/ml (range 5 06-7-97) for asthmatics, p<00001). These results closely reflect those of the whole groups so that baseline lung function was not the cause of the differences in response. In all but two asthmatic patients, whenever the exercise challenge was positive (>6.0% fall in FEVI), there was hyperreactivity to AMP at a concentration below the mean + 2SD of the asthmatic group (98-2mg/ml). On the other hand, 12 of the 49 asthmatic patients with a positive AMP challenge failed to respond to exercise. The mean PC20 of these 12 asthmatic children compared with that of the 39 children positive to exercise was not significantly different either for methacholine or AMP. These observations are reflected in the greater intersection point of AMP (98%) compared with exercise (84%) in distinguishing asthma from controls (table 2). The reason for this is unclear but could be related to the multiple factors known to influence the response of asthmatic patients to exercise at any one time.33 Adenosine challenge has clear advantages over exercise in that it can be performed in patients of all ages including the very young and older subjects in whom an exercise challenge would be undesirable. Moreover, the present study shows adenosine to be as specific as exercise in differentiating asthma from paediatric COPD, but more sensitive than exercise in revealing bronchial hyperreactivity in asthmatic subjects. Adenosine challenge should therefore be a very useful tool in the differential diagnosis of asthma and COPD in patients of all ages in whom the diagnosis is clinically uncertain. In conclusion, our control group was insensitive to AMP, exercise and low concentrations of methacholine. Asthmatic subjects were responsive to all three challenges while patients with paediatric COPD mainly responded to methacholine. These findings suggest a final common pathway of hyperreactivity to methacholine in all types of chronic lung disease in children, including asthma, while a more specific and possibly mast cell-related pathway may exist only in asthmatic subjects. Adenosine, like exercise, is of much more value than methacholine in the differentiation of asthma from COPD in children. These inhalation challenges, unlike exercise, are suitable for virtually all age groups and seem to be a useful tool in the differential Avital, Springer, Bar-Yishay, Godfrey 516 diagnosis of asthma in children and people. young The authors thank the pulmonary function laboratory technicians Elat Bardach, Leah Cohen, Miriam Bahir, and Ahuva Shina for their technical assistance and Sara Berger for her assistance in the preparation of the manuscript. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Cockcroft DW, Killian DN, Mellon JJA, Hargreave FE. Bronchial reactivity to inhaled histamine: a method and clinical survey. Clin Allergy 1977;7:235-43. Townley RG, Bewtra AK, Nair NM, Brodkey FD, Watt GD, Burkes BS. Methacholine inhalation challenge studies.] Allergy Clin Immunol 1979;64:569-74. Juniper EF, Frith PA, Dunnet C, Cockcroft DW, Hargreave FE. Reproducibility and comparison of responses to inhaled histamine and methacholine: relationship to minimum treatment to control symptoms of asthma. Thorax 1978;33:705-10. Price JF, Weller PH, Harper SA, Matthew DJ. Response to bronchial provocation and exercise in children with cystic fibrosis. Clin AllergD, 1979;9:563-70. Skorecki K, Levison H, Crozier DN. Bronchial lability in cystic fibrosis. Acta Paediatr Scand 1976;65:39-44. Avital A, Sanchez I, Chernick V. Efficacy of salbutamol and ipratopium bromide in decreasing bronchial hyperreactivity in children with cystic fibrosis. Pediatr Pul?noniol 1992;13:34-7. Ramsdell JW, Machtwey FJ, Moser KM. Bronchial hyperreactivity in chronic obstructive bronchitis. Anm Rev Respir Dis 1982;126:829-32. Bahous J, Cartier A, Ouimet G, Pineau L, Malo J-L. Nonallergic bronchial hyperexcitability in chronic bronchitis. Anm Rev Respir Dis 1984;129:216-20. Ramsdale EH, Morris MM, Roberts RS, Hargreave FE. Bronchial responsiveness to methacholine in chronic bronchitis: relationship to airflow obstruction and cold air responsiveness. Thorax 1984;39:912-8. Ramsdale EH, Roberts RS, Morris MM, Hargreave FE. Differences in responsiveness to hyperventilation and methacholine in asthma and chronic bronchitis. Thorax 1985; 40:422-6. Godfrey S, Springer C, Noviski N, Maayan Ch, Avital A. Exercise but not methacholine differentiates asthma from chronic lung disease in children. Thorax 1991;46:488-92. Lee TH, Nagy L, Nagakura T, Walpor MJ, Kay AB. Identification and partial characterization of an exercise-induced neutrophil chemotactic factor in bronchial asthma. .7 Clin Invest 1982;69:889-99. Anderson SD, Bye PTP, Schoeffel RE, Seale JP, Taylor KM, Ferris L. Arterial plasma histamine levels at rest, during and after exercise in patients with asthma: effects of terbutaline aerosol. Thorax 1981;36:259-67. Belcher NG, Murdoch RD, Dalton N, House FR, Clark TJH, Rees PJ, et al. A comparison of mediator and catecholamine release between exercise- and hypertonic saline-induced asthma. Am Rev Respir Dis 1988;137:102632. Venge P, Henriksen J, Dahl R, Hakansson L. Exerciseinduced asthma and the generation of neutrophil chemotactic activity. _7 Allergy Clin Inimunol 1990;85:498-504. 16 Nagy L, Lee TH, Kay AB. Neutrophil chemotactic activity in antigen induced late asthmatic reactions. N Engl ] Med 1982;306:497-501. 17 Metzger WJ, Richerson HB, Wasserman SI. Generation and partial chracterization of eosinophil chemotactic activity and neutrophil chemotactic activity during early- and latephase reactions. .fAllerg C/in Immunol 1986;78:282-90. 18 Cushley MJ, Holgate ST. Adenosine-induced bronchoconstriction in asthma: role of mast cell-mediator release. . Allergy, Clin Inimrunol 1985;75:272-8. 19 Driver AG, Kukoly CA, Metzger WJ, Mustafa SJ. Bronchial challenge with adenosine causes the release of serum neutrophil chemotactic factor in asthma. Amii Rev Respir Dis 1991;143:1002-7. 20 Marquardt DL, Walker LL, Wasserman SI. Adenosine receptors on mouse bone marrow-derived mast cells: functional significance and regulation by aminophylline. 7 Ininmunol 1984;133:932-7. 21 Cushley MJ, Tattersfield AE, Holgate ST. Inhaled adenosine and guanosine on airway resistance in normal and asthmatic subjects. Br y Clin Pharmacol 1983;15:161-5. 22 Cushley MJ, Tattersfield AE, Holgate ST. Adenosineinduced bronchoconstriction in asthma: antagonism bv inhaled theophylline. Anm Rev Respir Dis 1984;129:380 4. 23 Phillips GD, Ng WH, Church MK, Holgate ST. The response of plasma histamine to bronchoprovocation with methacholine, adenosine 5' monophosphate, and allergen in atopic nonasthmatic subjects. Ami Rev Respir Dis 1990; 141:9-13. 24 Oosterhoff Y, De Jong JW, Jansen AM, Koeter GH, Postma DS. Airway responsiveness to adenosine 5'-monophosphate in chronic obstructive pulmonary disease is determined by smoking. Amn Rezv Respir Dis 1993;147: 553-8. 25 Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma. Am Ret' Respir Dis 1987;136:225-44. 26 Anderson SD, Silverman M, Konig P, Godfrev S. Exerciseinduced asthma. Br3' Dis Chest 1975;69: 1-39. 27 Hargreave FE, Ryan G, Thomson NC, O'Byrne PM, Latimer K, Juniper EF. Bronchial responsiveness to histamine or methacholine in asthma: measurement and clinical significance. Eur 7 Respir Dis 1982;63:79-88. 28 Avital A, Noviski N, Bar-Yishay E, Springer C, Levy M, Godfrey S. Nonspecific bronchial reactivity in asthmatic children depends on severity but not on age. Am Rev Respir Dis 1991;144:36-8. 29 Chatham M, Bleecker ER, Smith PL, Rosenthal RR, Mason P, Norman PS. A comparison of histamine, methacholine, and exercise airway reactivity in normal and asthmatic subjects. Ani Rev Respir Dis 1982;126:235-40. 30 Clough JB, Hutchinson SA, Williams JD, Holgate ST. Airway response to exercise and methacholine in children with respiratory symptoms. Arch Dis Child 1991;66:579 83. 31 Cropp GJA. The exercise bronchoprovocation test: Standardization of procedures and evaluation of response. 7 Allergy Clini Inmnmunol 1979;64:627-33. 32 Burr ML, Eldridge BA and Borysiewicz LK. Peak inspiratory flow rates before and after exercise in school children. Arch Dis Child 1974;49:923-6. 33 Godfrev S, Bar-Yishay E. Exercise induced asthma revisited. Respir Med 1993;87:331-4.