A placebo-controlled, crossover comparison of ... - ATS Journals

A Placebo-controlled, Crossover Comparison of Salmeterol and Salbutamol in Patients with Asthma PIERRE LEBLANC, ALLAN KNIGHT, HARVEY KREISMAN, CORNELIA M. BORKHOFF, and PATRICK R. JOHNSTON Centre de Pneumologie, H6pital laval et Universite laval, Ste-Foy, Quebec; Sunnybrook Health Science Centre, Toronto, Ontario; Dept. of Medicine, 5MBD Jewish General Hospital, Montreal, Quebec, Canada; Medical Sciences Division, Glaxo Canada Inc., Mississauga, Ontario, Canada

We compared the effects of salmeterol (Sm) (50 ~g twice daily) with that of salbutamol (Sb) (200 ~g four times daily) and placebo (P) in patients with mild-to-moderate asthma with asthma symptoms and related the effective-ness of these therapies between patients who used concurrent inhaled corticosteroids (ICS) and those who did not. The study was a 12-wk, multicenter, double-blind, placebocontrolled crossover trial with 367 adult asthmatics randomized to each trial medication for 4 wk. Inhaled Sb was provided as rescue medication to all patients throughout the trial. Only 80% of patients, albeit the majority, were receiving maintenance treatment with ICS throughout this trial; this reflects practice current in early 1990. Each study day, patients recorded their morning and evening peak expiratory flows (PEF), assessment of asthma symptoms, and use of rescue medication. Both morning and evening PEF were greater during treatment with Sm than with Sb (mean differences between the treatments of 29.8 and 14.3 l/min, respectively) or P (27.7 and 20.3 L/min, respectively) (p < 0.0001). Sm was also more effective than Sb or P in lowering diurnal variation in PEF and increasing the percentage of symptom-free days and rescue-free days and nights with no sleep disturbance (p ~ 0.0004). Sb was more effective than P in increasing evening PEF and the percentage of symptom-free days (p < 0.05) and rescue-free days (p < 0.0001). The same clinical superiority of Sm compared with Sb and P was observed in those patients using ICS (p < 0.001 for all treatment comparisons), and to a greater extent than in those patients not using ICS (i.e., Sm was more effective than Sb and P in just six of the 20 treatment comparisons; p < 0.05). In conclusion, Sm 50 ~g twice daily is effective in the management of mild-to-moderate asthma and it further improves asthma control in patients already using ICS. Leblanc P, Knight A, Kreisman H, Borkhoff eM, Johnston PRo A placebo-controlled, crossover comparison of salmeterol and salbutamol In patients with asthma. AM

Current asthma management guidelines recommend that inhaled corticosteroids be used as first-line therapy for all patients except those with the mildest asthma symptoms (I). For most patients then, management of asthma usually includes inhaled corticosteroid therapy to prevent inflammation and a short-acting l}2-agonist for the immediate relief of break-through symptoms only. Salmeterol, a long-acting l}2-agonist, provides protection against breakthrough symptoms, and it is recommended for asthmatics who continue to have symptoms despite maintenance treatment with anti-inflammatory drugs. l}2-Adrenoceptor agonists are effectivebronchodilating agents, protect against a wide variety of bronchoconstrictor stimuli, and are associated with a low incidence of side effects at clinically relevant doses (2-5). Salmeterol xinafoate is a long-acting highly

(Received in original form December 30, 1994 and in revised form December 26,

1995)

Supported by a grant from Gfaxo Canada Inc. Correspondence and requests for reprints should be addressed to Dr. Pierre leblanc, Centre de Pneumologie, H6pital laval, 2725 Chemin Ste-Foy, SteFoy, Quebec, G1V 4G5 Canada. Am

J Respir Crlt Care Med Vo11S4. pp 324-328, 1996

J RESPIR CRIT CARE MED

1996;15-4:J24-8.

selective l}2-adrenoceptor agonist with a slower onset of action than salbutamol and a duration of action of at least 12 h (6). The safety and efficacy of salmeterol 50 ug given twice daily for the management of mild-to-moderate asthma has been well established, and it has been shown to be of particular benefit in reducing symptoms of nocturnal asthma and dyspnea upon awakening (7-10). The purpose of our study was to assess the effects of regular, twice-daily therapy of inhaled salmeterol in patients with mildto-moderate asthma with asthma symptoms. This study provides a rare opportunity to examine the response to salmeterol treatment with and without concurrent inhaled steroid therapy. METHODS Patients Patients 18 to 70 yr of age were eligible for this study if they demonstrated both a FEY, of at least 600/0 of their predicted value (11) and a reversible component to their disease. defined as an increase in FEY I of at least 15% after inhalation of 200 I1g salbutamol. In addition. on 4 of the last 7 d of the prerandomization period. patients had to be either symptomatic or demonstrate a greater than 20% diurnal variation in peak expiratory flow (PEF). Patients unable to discontinue the use of methylxanthines, inhaled anticholinergics, and oral corticosteroids werenot enrolled into this study. Patients receiving inhaledcorticosteroids,

325

Leblanc, Knight, Kreisman, et 01.: Salmeterol's Role in Asthma Management inhaled sodium cromoglycate, and antihistamines could be enrolled provided that treatment had been initiated at least I mo before and that the dose remained constant throughout the study.

TABLE 1 PATIENT DEMOGRAPHY AND ASTHMA HISTORY (n

367) Patients

Study Design

This randomized, double-blind, placebo-controlled, three-way crossover 15-wk study was carried out at 15 respiratory clinics across Canada between July 1990and June 1992. The study wasconducted in accordance with the Declarationof Helsinki. Patients underwenta 2-wkprerandomization assessment. Eligiblepatients werethen randomized to receive one of the six possible sequences of the three treatments of salmeterol 50 Jlg twice daily, salbutamol 200 ug four times daily, and placebo. Each treatment lasted 4 wk. Double-blinding was ensured through the use of two matching inhalers, each containing either active drug or placebo. Patients self-administereddrug four times daily, taking two puffs from one inhaler in the morning and at bedtime and two puffs from the other inhaler at lunchtime and dinnertime. For those patients receiving salmeterol treatment, active drug was administered in the morning and at bedtime. For those patients receiving salbutamol treatment, active drug was administered at all four dosing times. Patients were provided with inhaled salbutamol to be used as rescuemedication. After completion of treatment, patients resumed their usual asthma therapy; after I wk, they returned to the clinic for a post-treatment assessment. PEF was monitored on a daily basis by each patient through the use of a mini-Wright peak flowmeter and was recorded on a daily record card. Morning PEF was measured prior to receiving the morning dose of study medication, approximately 9 h after the last dose of medication. Evening PEF was measured each day at bedtime prior to taking the last daily dose of medication. Pulmonary function (PEF, FEV I, and FVC),after havingwithheld rescue medication for 4 h, was also assessed in the clinic at the end of each of the three treatment periods. Each day, patients recorded a self-rated overall assessment of their asthma symptoms experienced during the night and in the early morning upon awakening and during the daytime at the end of each day. The use of rescue medication during the day and at night was also recorded. Statistical Methods

All 367randomizedpatients(the intent-to-treat population) were included in the analysis. All statistical tests were two-sided and used a 5070 significancelevel. All efficacyvariableswereanalyzedusing the SASilI GLM procedure with a three-treatment, three-period,crossover model containing within-subject effects for period, carryover,and center-by-treatment interaction (12). The period effect controls for the possibility that, for example, PEF increases over the three mo of the trial. The carryover effect controls for the possibleinfluence of the precedingtreatment. The primary measure of efficacy was morning PEF; all other measures of efficacy wereconsidered secondary. The estimated within-subject standard deviation for morning PEF was 26.6 Llmin. The study had a > 99070 power to detect treatment differences in PEF of 25 L/min at the 5070 level. For efficacy data recorded on the daily record card (e.g., morning PEF), treatments were compared using the mean of the last 14daily observations from each l-mo treatment period. Treatment subpopulations (patients using or not using concurrent inhaled corticosteroids) werecompared bycalculating the differencebetweenthe treatment differences. Compliance with study medication was measured by calculating the percentage of scheduled doses recordedas taken on the daily record card. Patients were considered compliant if they recorded more than 70070 of the scheduled study doses (i.e., at most 4 d of the last 14 daily observations from each I-mo treatment period could be recorded as anything other than four puffs of Inhaler A and four puffs of Inhaler B on any single day). RESULTS

Three hundred sixty-seven patients were randomized to treatment (93070 were compliant with study medication). Demographic data are presented in Table I. The majority (80070) of patients were using inhaled corticosteroids throughout the trial. Sixty-six patients (18070) were withdrawn; 38 (10.4070) experienced an exacerbation of asthma, 14 (3.8070) experienced major adverse events, four (1.1070) had not met the percent predicted FEV, inclusion criteria, four dropped out, two (0.5070) were noncompliant, two

(n)

Sex Male Female Age. yr

Smoking history, n (%) Smoker Nonsmoker Previous smoker Reversibility of FEV1, % FEV1, % pred Duration of asthma, yr 10 Concurrent asthma medications Inhaled corticosteroids Antihistamine/antiallergics Adrenomimetics

(%)

164 45 203 55 39 ± 13.9111 30 218 59 38 10 23.5 ± 7.877.1 ± 10.710 89 71 197

3 24 19 53

295 26 2

80 7 5

• Values are mean ± SD.

failed to return, and two used restricted medications. Fewer patients experienced exacerbations of asthma leading to withdrawal from the study during salmeterol treatment (eight patients) and salbutamol (12 patients) than during placebo treatment (18 patients) (p < 0.05). The proportion of patients who withdrew from the study was similar across treatment sequences. Treatment with salmeterol 50 ug twice daily resulted in a significantly greater morning PEF (433.2 Llmin) then either treatment with salbutamol200 ug four times daily (403.4 L/min) or placebo (405.5 Llmin) (Table 2). The mean difference between treatments are shown in Table 3. There was no significant period or carryover effect. Furthermore, there was no qualitative centerby-treatment interaction (13). The results favored salmeterol over salbutamol or placebo for all efficacy variables (p < 0.0001 for all comparisons except for nights with no sleep disturbance, with p = 0.0004 for the salmeterol-salbutamol comparison). In addition, salmeterol was found to be significantly more effective than placebo in increasing morning and evening PEF for those patients using inhaled corticosteroids than for those not using them (p < 0.05) (Table 4). The mean difference in diurnal variation in PEF was - 4.1 percentage points between salmeterol and salbutamol treatments, - 2.2 percentage points between salmeterol and placebo, and 1.8 percentage points between salbutamol and placebo (p < 0.00(1) (Table 3). Compared with salbutamol, diurnal variation in PEF was also significantly lower when receiving placebo among those using inhaled corticosteroids (mean difference of 1.5 percentage points; p < 0.001). Diurnal variation in PEF was lowest among the patients using inhaled corticosteroids while receiving salmeterol (Table 2). The superiority of salmeterol treatment observed from analysis of the daily PEF measures was supported by lung function data recorded at each clinic visit (clinic visit PEF, FEV.. and FVC). This improvement in lung function was also reflected in the decreased incidence of asthma symptoms during the day and night. The percentage of symptom-free mornings and days and nights with no sleep disturbance were all significantly greater when receiving salmeterol than when receiving salbutamol or placebo (p ~ 0.0004). Salmeterol resulted in a greater percentage of symptom-free days compared with placebo for patients using in-

326

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE TABLE 2

=

MEANS OF EFFICACY PARAMETERS FOR ALL PATIENTS (n 367), PATIENTS USING INHALED CORTICOSTEROIDS (n 295), AND PATIENTS 72) NOT USING INHALED CORTICOffiROIDS (n

=

=

Efficacy Parameters Patients Using ICS

All Patients Group Morning PH, L/min Evening PEF, Llmin Diurnal variation in PEF, %0 Clinic visit PH, Llmin Clinic visit FEV!, L Clinic visit FVC, L Symptom-tree mornings, % Symptom-tree days, % Nights with no sleep disturbance, % Rescue medication-tree days, %

Patients Not Using ICS

Sm

Sb

PI

Sm

Sb

PI

Sm

Sb

PI

433.2 440.3 1.0 460.0 2.73 3.82 28.5 22.6 79.7 62.0

403.4 426.0 5.1 443.1 2.62 3.71 18.1 14.7 74.1 48.9

405.5 420.0 3.3 430.1 2.58 3.70 19.4 11.2 72.4 33.6

434.7 440.4 0.8 458.2 2.68 3.78 31.4 24.3 82.9 64.6

404.5 425.0 4.6 442.0 2.58 3.67 20.3 16.0 77.1 50.3

405.0 418.8 3.1 426.4 2.54 3.65 20.6 11.5 76.6 34.1

427.7 442.0 2.4 466.4 3.04 4.12 20.7 16.3 57.8 52.4

403.2 430.7 5.9 453.8 2.86 4.03 12.9 11.5 54.0 45.9

414.2 434.5 4.3 445.8 2.81 4.03 16.5 11.5 50.4 29.9

Definition of abbreviations: ICS = inhaled corticosteroids; Sm • salmeterol; Sb • salbutamol; PI• placebo; PEF • peak expiratory flow. o Percent diurnal variation in PH = {previous evening PEF - next morning PEF)/previous evening PH x 100.

TABLE 3 DIFFERENCES BETWEEN TREATMENTS IN EFFICACY PARAMETERS FOR ALL PATIENTS (n 367)*

=

Efficacy Parameters

Group

SalmeterolSalbutamol

SalmeterolPlacebo

Morning PH, L/min Evening PEF, Llmin Diurnal variation in PH, % Clinic visit PH, Llmin Clinic visit FEV1, L Clinic visit FVC, L Symptom-tree mornings, % Symptom-tree days, % Nights with no sleep disturbance, % Rescue medication-tree days, %

29.8 %4.7t 14.3%4.1t -4.1 % 0.8 t 16.9 %8.6t 0.11 %0.05t 0.11 %0.05t 10.4% 3.3t 7.9 %3.4t 5.6 %3.1:1: 13.1 %4.8t

27.7 %4.7t 20.3% 4.1t -2.2 %0.8t 29.9 %8.6t 0.15 %0.05t 0.12 %0.05t 9.1 %3.3t 11.4 %3.4t 7.3 %3.1t 28.5 %4.8t

• Values are means

% 95%

SalbutamolPlacebo

-2.1 % 4.7 6.0 %4.1§ 1.8 %0.8t 13.0 %8.6§ 0.04%0.05 0.01 %0.05 -1.3 %3.3 3.5 % 3.4§ 1.8 %3.1 15.3 %4.8 t

confidence intervals.

t p < 0.0001. :I: p = 0.0004. §p

< 0.05.

TABLE 4 DIFFERENCES BETWEEN TREATMENTS IN EFFICACY PARAMmRS FOR PATIENTS USING INHALED CORTICOffiROIDS (n 295) AND FOR PATIENTS NOT USING INHALED CORTICOffiROIDS (n = 72)*

=

Efficacy Parameters Patients Using ICS Group Morning PH, Llmin Evening PH, Llmin Diurnal variation in PEF, % Clinic visit PEF, Llmin Clinic visit FEV1, L Clinic visit FVC, L Symptom-tree mornings, % Symptom-tree days, % Nights with no sleep disturbance, % Rescue medication-tree days, %

Sm-Sb

30.2 %5.1t 15.4 %4.5t -3.8 %0.8t 16.2 %9.3:1: 0.10 %0.06:1: 0.10 %0.06:1: 11.1 %3.9t . 8.3 %3.6 t 5.8 %3.3:1: 14.3 %5.5 t

Sm-Pl

29.7 %5.1 til 21.6 %4.5 tl -2.3 %0.8t 31.8 %9.3t 0.14 %O.06t 0.12 %O.06t 10.8 %3.9t 12.8 %3.6 tl 6.4 %3.3:1: 30.5 %5.5t

• Values are mean ± 95% confidence intervals. p < 0.0001. :I: p < 0.001. § P < 0.05. np < 0.05 for comparison between patients using and not using inhaled corticosteroids.

t

Patients Not Using ICS Sm-Sb

24.5 %15.4§ 11.3 %13.4 -3.5 %2.5§ 12.6 %27.1 0.18 %0.17§ 0.09 %0.17 7.7%7.1§ 4.8%5.8 3.9% 10.0 6.6 %13.7

Sm-PI

13.5 %15.3 7.5 %13.2 -1.9%2.5 20.6 %26.8 0.23 %0.17§ 0.09 %0.17 4.1 %7.0 4.8% 5.8 7.5 %9.9 22.5 %13.5§

VOL 154

1996

Leblanc, Knight, Kreisman, et aJ.: Salmeterol's Role in Asthma Management

haled corticosteroids than for patients not using inhaled steroids (p < 0.05) (Table 4). Salmeterol was also more effective than salbutamol or placebo in lowering the requirement for rescue medication. The'percentage of days free of rescue medication use increasedsignificantly in the salmeterol and the salbutamol treatment groups as compared with the placebo treatment group, and also when receiving salmeterol as compared with salbutamol (p < 0.0001). Use of rescue medication was greatest when receiving placebo (2.6 puffs per day), next greatest when receiving salbutamol (1.7puffs per day), and least when receiving salmeterol (1.2 puffs per day).

DISCUSSION This study provides further evidence that inhaled salmeterol 50 ug administered twice daily is-welltolerated and more effective than inhaled salbutamol 200 ug four times daily or placebo in the management of mild-to-moderate asthma. Treatment with salmeterol was statistically significantly better than treatment with either salbutamol or placebo for all efficacy measures. Most interestingly, our findings indicate that the addition of salmeterol is more effective in improving asthma control in patients already using inhaled corticosteroids. In examining the differences between treatments for all patients, treatment with salmeterol, compared with salbutamol or placebo, resulted in a clinically important improvement in morning PEF of 29.8 and 27.7 L/min, respectively. Because morning PEF was measured prior to the morning dose, approximately 9 h after the last dose, salbutamol (4 to 6-h duration of action) is rendered practically equivalent to placebo. As it happens, our finding that diurnal variation in PEF was significantly lower when receiving placebo than when receiving salbutamol (p < 0.0001) is likely due to both treatments producing a similar morning PEF upon awakening, but salbutamol treatment producing a higher evening PEP. Furthermore, treatment with salmeterol compared with treatment with salbutamol or placebo resulted in a highly statistically significant and clinically important increase in evening PEF, days free of rescue medication use, symptom-free mornings and days, and nights with no sleep disturbance. These observations reflect the long-acting nature of salmeterol. As salmeterol and salbutamol produce almost the same degree of maximal bronchodilation (6), this and other studies (8-10) indicate that it is salmeterol's long duration of action that provides asthmatics with the greatest clinical benefit during the night and in the morning. Current asthma management guidelines recommend that regular J3ragonists be used in conjunction with inhaled corticosteroids and not as monotherapy (1). Only 80070 of patients, albeit the majority, were using concurrent inhaled corticosteroids throughout this trial, reflecting practice current in 1990. As a result, the effectiveness of salmeterol in users and nonusers of inhaled corticosteroids was considered. Our findings extend those of other studies (14, 15), which showed that the addition of salmeterol to inhaled steroids had a greater effect on asthma control than an increase in inhaled steroids alone. In this study, the addition of salmeterol to inhaled steroids had a greater effect on asthma control than the addition of salmeterol alone. Additional treatment with salmeterol can be seen as complimentary to inhaled steroids. Salmeterol treatment showed significant benefit over salbutamol treatment and placebo among those patients using inhaled corticosteroids (p < 0.001 for all treatment comparisons), but the extent of salmeterol's benefit was not as great in those patients not using inhaled corticosteroids (i,e., four of the 10 salmeterol-salbutamol treatment differences and two of the 10 salmeterol-placebo treatment differences were significant at p < 0.05).

327

The patients in this study who used inhaled corticosteroids appear to have had more severe asthma (as judged by the lower morning PEF, eveningPEF, and clinic lung function while receiving placebo). Meanwhile, these same patients had a lower diurnal variation in PEF and a higher percentage of symptom-free mornings and days and nights with no sleep disturbance, indicating the inhaled steroid's positive impact on asthma control. By comparing the effectiveness of salmeterol in users versus nonusers of inhaled corticosteroids, we found salmeterol to be significantly more effective than placebo in increasing morning and evening PEF and symptom-free days for those patients using inhaled corticosteroids than for those not using them (p < 0.05). Despite the fact that only three of the total 20 treatment parameter differences (10 salmeterol-salbutamol and 10 salmeterolplacebo) were statistically significantly greater for patients receiving inhaled steroids, an additional 14 were greater, although not significantly so. This indicates that those patients already receiving inhaled steroids are likely to benefit most by being placed on salmeterol. Our results are clearly different from those of previous studies (8, 9) in which little or no difference in the effectiveness of salmeterol between users and nonusers of concurrent inhaled corticosteroids was found. A greater number of patients in this study (80%) used concurrent inhaled corticosteroids than in these previous salmeterol studies: 31% in a U.S. study (8) and 63% in a European study (9), perhaps reflective of prescribing practices of inhaled steroids throughout the world. In conclusion, our findings indicate that the addition of salmeterol to the therapy of patients already using inhaled corticosteroids further improved asthma control. This study supports the current recommendation that patients must be using inhaled corticosteroids before salmeterol is added to their asthma therapy. Regular maintenance therapy with salmeterol should be considered for asthmatics across the severity spectrum who have breakthrough symptoms despite maintenance treatment with antiinflammatory drugs. Acknowledgment: The writers thank the following respiratory clinics for conducting this study: Hopital Laval, Ste-Foy, Quebec (Drs. P. leblanc and l.-P. Boulet); 5MBD Jewish General Hospital, Montreal, Quebec (Drs. H. Kreisman and N. Wolkove); Hopital du Sacre-Coeur, Montreal (Drs. A. Cartier and J.-l. Malo); Sunnybrook Health Science Centre, Toronto, Ontario (Dr. A. Knight); Mount Sinai Hospital, Toronto (Drs. R. Grossman and A. Day); Vancouver General Hospital, Vancouver, B. C. (Drs. M. Yeung and S. Kwan); Toronto Hospital-Western Division, Toronto (Drs. S. Kesten and K. R.Chapman); Grey Nuns Hospital, Edmonton, Alberta (Dr. G. F. MacDonald); University of Alberta Hospital, Edmonton (Drs. S. F. P. Man and B. J. Sproule); Victoria Hospital, london, Ontario (Drs. J. Mazza and W. Moote); University of Manitoba Respiratory Hospital, Winnipeg (Dr. P. Warren); Foothills Hospital, Calgary, Alberta (Drs. R. l. Cowie and S. Field); Trois-Rlvieres, Quebec (Drs. M. Lapointe and F. Corbeil); Hotel-Dieu de Montreal (Dr. R. Amyot); Centre Hospitalier de Verdun, Verdun, Quebec (Dr. F. Plante). The writers are also grateful to Sandra Cooper and Tamara Rudnicki for their handling of the data.

References

I. National Heart, Lung, and Blood Institute, National Institutes of Health. 1992. International consensus report on diagnosis and treatment of asthma. Eur. Respir. J. 5:601-641. 2. Kerrebijn, K. F. 1991. Beta agonists. In M. A. Kaliner, P. J. Barnes, and C. G. A. Persson, editors. Lung Biology in Health and Disease, Vol. 39: Asthma: its pathology and treatment. Marcel Dekker, New York. 523-558. 3. Malo, J. L., H. Ghezzo, C. Trudeau, J. L'Archeveque, andA. Cartier. 1992.Salmeterol, a newinhaled beta-2 adrenergic agonist, has a longer blocking effect than albuterol on hyperventilation-induced bronchoconstriction. J. Allergy Clin. Immunol. 89:567-574. 4. Newnham, D. M., C. G. Ingram, J. Earnshaw, J. B. D. Palmer, and D. P. Dhillon. 1993.Salmeterol provides prolonged protection against exercise-induced bronchoconstriction in a majority of subjects with mild, stable asthma. Respir. Med. 87:439-444. 5. Twentyman, O. P., J. P. Finnerty, A. Harris, J. Palmer, and S. T. Holgate. 1990. Protection against allergen-induced asthma by salmeteroJ.

328

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

Lancet 336:1338-1342. 6. Ullman, A., and N. Svedmyr. 1988. Salmeterol, a new long acting inhaled /3, adrenoceptor agonist: comparison with salbutamol in adult asthmatic patients. Thorax 43:674-678. 7. Dahl, R., J. Earnshaw, and J. B. D. Palmer. 1991. Salmeterol: a four week study of a long-acting beta-adrenoceptor agonist for the treatment of reversible airways disease. Eur. Respir. J. 4:1178-1184. 8. Pearlman, D. S., P. Chervinsky, C. La Force, J. M. Seltzer, D. L. Southern, J. P. Kemp, R. J. Dockhorn, J. Grossman, R. F. Liddle, S. W. Yancey, D. M. Cocchetto, W. J. Alexander, and A. Van As. 1992. A comparison of salmeterol with albuterol in the treatment of mildto-moderate asthma. N. Engl. J. Med. 327:1420-1425. 9. Britton, M. G., J. S. Earnshaw, and J. B. D. Palmer. 1992. A twelve month comparison of salmeterol with salbutamol in asthmatic patients. Eur. Respir. J. 5:1062-1067. 10. Fitzpatrick, M. F., T. Mackay, H. Driver, and N. J. Douglas. 1990.

II.

12. 13. 14. 15.

VOL 154

1996

Salmeterol in nocturnal asthma: a double-blind, placebo-controlled trial of a long-acting inhaled /3, agonist. B.M.J. 301:1365-1368. Quanjer, P. H., editor. 1983.Report of Working Party, European Community for Coal and Steel. Standardized lung function testing: standardization of lung function tests. Bull. Eur. Physiopathol. Respir. 19(5uppl. 5):7-87. SAS Institute Inc. 1989. SAS/STAT@Users Guide. Version 6, 4th ed. SAS Institute Inc., Cary, NC. Gail, M., and R. Simon. 1985. Testing for qualitative interactions between treatment effects and patient subsets. Biometrics 41:361-372. Greening, A. P., P. W.lnd, M. Northfield, and G. Shaw. 1994. Added salmeterol versus higher-dose corticosteroid in asthma patients with symptoms on existing inhaled corticosteroid. Lancet 344:219-224. Woolcock, A. 1994. Effect of the addition of salmeterol versus doubling the dose of inhaled steroid in asthmatics. Eur. Respir. J. 7 (Suppl. 18):365S.