Volume 115

Number 4

15 August 1991

Annals of Internal Medicine ARTICLES

Aminophylline Therapy for Acute Bronchospastic Disease in the Emergency Room Keith Wrenn, MD; Corey M. Slovis, MD; Frances Murphy, RN; and Raymond S. Greenberg, MD, PhD

• Objective: To assess the role of aminophylline in the treatment of acute exacerbations of bronchospastic disease when used in addition to inhaled beta-agonists and intravenous corticosteroids. • Design: Randomized, double-blind, placebocontrolled intervention study. • Patients: One hundred thirty-three adult patients with either asthma or chronic obstructive pulmonary disease who presented to the emergency department with asthma or wheezing. • Interventions: All patients received therapy with both aerosolized metaproterenol and intravenous methylprednisolone. Patients were randomly assigned to receive either a loading dose of aminophylline followed by a routine infusion (n = 65) or an equal volume of placebo as a loading dose and infusion (n = 68). • Measurements and Main Results: At discharge from the emergency department, the median serum theophylline concentration for the aminophylline group was 54 jimol/L (9.7 mg/L). The two groups showed no differences (P > 0.2) in measurements of forced expiratory volume at 1 second (FEVJ, forced vital capacity (FVC), or peak expiratory flow rate (PEFR) at baseline or at 60 or 120 minutes after aminophylline administration. Neither patient satisfaction nor physician assessment of response to therapy differed between the two groups. There was no difference (P > 0.2) in the frequency of side effects, except for a trend toward a higher frequency of nausea (P = 0.13) in the aminophylline group. There was, however, a threefold decrease in the hospital admission rate for patients treated with aminophylline (6%) compared with placebo recipients (21%) ( P = 0.016). • Conclusion: Aminophylline, in doses producing levels just below the commonly accepted therapeutic range, appears to decrease hospital admissions in patients with acute exacerbation of asthma or chronic obstructive pulmonary disease. This finding, if confirmed in larger studies, may represent a substantial cost savings. Annals of Internal Medicine. 1991;115:241-247. From Emory University School of Medicine and Emory University School of Public Health, Atlanta, Georgia; and University of Rochester School of Medicine, Rochester, New York. For current author addresses, see end of text.

O n c e a mainstay in the treatment of acute bronchospastic disease, aminophylline has recently been relegated to a second- or third-line role. The declining use of aminophylline can be attributed to the lack of convincing, well-designed trials that demonstrate aminophylline's efficacy (1) and to the drug's narrow therapeutic index and relatively low potency as a bronchodilator (2-5). Exacerbations of asthma and chronic obstructive pulmonary disease are common presenting problems in emergency departments and frequently cause hospital admission (6, 7). Inhaled beta-agonists and corticosteroids have assumed increasing importance in the treatment of acute exacerbations of bronchospastic disease because of their greater potency and wider therapeutic index. Our double-blind, placebo-controlled trial was designed to test the null hypothesis that intravenous aminophylline would add no important benefit to treatment with beta-agonists and corticosteroids in patients with acute bronchospastic disease presenting to an emergency department. Methods All adult patients (more than 16 years of age) presenting to the emergency department of Grady Memorial Hospital with asthma exacerbation or wheezing were eligible for inclusion in the study. Exclusion criteria included the use of a theophylline-containing product within the preceding 24 hours; a past history of an adverse reaction to theophylline; a contraindication to the use of corticosteroids or beta-agonists; insulindependent diabetes; possible myocardial ischemia; or pulmonary edema. Once enrolled in the study, patients were excluded from re-enrollment on a subsequent visit. The study was approved by the Emory University Human Investigations Committee. Patients who agreed to participate in the study gave informed consent. Baseline measurements of forced expiratory volume at 1 second (FEV,), forced vital capacity (FVC), and peak expiratory flow rate (PEFR) were done in all patients with a Respiradyne spirometer (Model 57905, CheeseboroughPonds, Greenwich, Connecticut). Patients then received metaproterenol sulfate (0.3 mL of a 5% solution in 2.2 mL saline) by hand-held nebulizer at intervals of 15 to 20 minutes for three back-to-back treatments. They also received methylprednisolone sodium succinate, 80 mg intravenously. During therapy with inhaled metaproterenol and intravenous corticosteroids, patients were randomly assigned in a double-blind fashion to receive a loading dose of either aminophylline, 5.6 mg/kg body weight over 20 minutes, or an equivalent volume ©1991 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

241

of placebo intravenously, followed by a constant infusion of aminophylline, 0.9 mg/kg per hour, or an equivalent volume of placebo. Patients continued to receive aerosolized metaproterenol at intervals of 30 to 60 minutes during the rest of their emergency department stay as deemed necessary by a house officer assigned to cover the 4 'asthma room" of the emergency department. The house officer had no connection with the study and no knowledge of whether the patient was receiving aminophylline or placebo. Spirometric measurements were repeated 60 and 120 minutes after administration of the study drug. Patients were observed closely for adverse side effects, including nausea, vomiting, anxiety, tremor, palpitations, dysrhythmias, seizures, and altered mental status. Data on the following variables were collected at baseline: age, sex, duration of disease (years), duration of the current attack (hours), severity of the attack ("mild," "moderate," or "severe") as judged by the patient, severity of the attack as judged clinically by the investigator, frequency of attacks of bronchospasm, smoking history ("current," "former," or "never"), amount of smoking (pack-years), use of caffeinecontaining beverages (cups per day of coffee or tea, cans per day of soft drinks), duration without medications, medications and doses, vital signs, and the presence of wheezes, rales, diaphoresis, or accessory muscle use. Baseline serum theophylline levels were not obtained because none of the study patients had taken theophylline within 24 hours. Chest radiographs and arterial blood gas assessments were done only if clinically indicated. At discharge (either to hospital admission or to home), patients were asked to assess their satisfaction with the regimen ("better," "same," or "worse"). An investigator also rendered an opinion of the response to the regimen on the same nominal scale. Time in the emergency department before a decision to admit or discharge was noted. Theophylline levels at discharge were obtained, but results were not available to the investigators before the final disposition. The decision to admit to the hospital or to discharge was not made by the investigators. The house officer assigned to the asthma room of the emergency department made the decision based on an independent clinical assessment of the patient's response to treatment. Preexisting guidelines for admission had been established for the emergency department. These included a history of recent relapse (previous emergency department treatment within 24 hours); the failure to clear after 6 hours of treatment; the self-reported (patient) inability to attain pre-exacerbation status; the presence of respiratory failure; and the inability of the patient to walk 100 feet without exacerbation of symptoms or signs. The study drug (aminophylline or placebo) was prepared and randomized in advance by pharmacy personnel who were not involved in the care of any of the patients. All spirometric measurements were obtained and medications were administered by a single investigator who was blinded to experimental drug treatment. During the study period, the Respiradyne spirometer was calibrated according to manufacturer's standards on a daily basis. Spirometric measurements with this spirometer were compared on a weekly basis with the results obtained with an Eagle II spirometer (Warren Collins, Inc., Braintree, Massachusetts) in randomly selected patients. We categorized patients according to the primary process responsible for wheezing: asthma or chronic obstructive pulmonary disease. Patients were classified as having asthma if they were under 45 years of age, had smoked for less than 20 pack-years, and had a duration of disease of less than 20 years or had onset of asthma in childhood. Patients were classified as having chronic obstructive pulmonary disease if they were over 45 years of age, had a smoking history of 20 pack-years, had a duration of disease of more than 20 years, or had onset of disease after the age of 45 years. All categorizations were reviewed by an investigator who did not have contact with the patient and final assignment to the asthma or chronic obstructive pulmonary disease category was made before unblinding. Statistical Methods Univariate comparisons of experimental and control subjects were done in aggregate and within disease groups using the 242

Pearson chi-square test for categorical variables, the Student Mest for normally distributed continuous variables, and the Wilcoxon two-sample rank-sum test for non-normal continuous variables. The primary response and outcome variables were FEV,, FVC, PEFR, patient satisfaction, physician assessment of response, time in the emergency department, and hospital admission status. Differences between the experimental and control groups in spirometric values were assessed with analysis of variance, whereas the categorical responses were compared using unconditional logistic regression models. All comparisons of responses were carried out for the total sample and separately for each disease group, with adjustment for any observed differences in baseline characteristics. Two-sided P values are reported for all statistical tests performed. Ninetyfive percent confidence intervals (CI) are given when appropriate. Results During the study period, 143 patients were randomly assigned to receive aminophylline or placebo and 133 patients were included in final statistical analysis. The 10 patients excluded from the analysis included 5 patients for whom the code for the study drug was lost; 3 patients who received the study drug but for whom no spirometric values were obtained; 1 patient in whom an intravenous line could not be successfully established; and 1 patient in whom infusion of the study drug was discontinued when a fingerstick blood glucose measurement was found to be elevated and the electrocardiogram showed new T-wave inversions. The median age of the patients with asthma who received aminophylline was 31 years (interquartile range, 20 to 39 years), whereas the median age of the patients with asthma who received placebo was 36 years (interquartile range, 31 to 43 years) (P = 0.007). The patients with asthma who received placebo had a history of heavier smoking (difference in means of 5 pack-years) than did the patients treated with aminophylline (P = 0.03). The prevalence of soft drink use was higher in both the patients with asthma who were receiving aminophylline (P = 0.04) and the patients with chronic obstructive pulmonary disease who were receiving placebo (P = 0.03). The difference between the aminophylline and placebo groups in the mean amount of soft drink use per day, however, was approximately 1/2 can per day. There were no other statistically significant differences between the groups (Table 1). The spirometric values at 60 and 120 minutes after aminophylline infusion are shown in Figure 1. There was no difference (P > 0.2) in the improvement in spirometric values by 60 and 120 minutes between the aminophylline and placebo groups. At discharge from the emergency department, the median theophylline level in patients receiving aminophylline was 54 /nmol/L (9.7 mg/L) with an interquartile range of 40 to 72 /xmol/L (7.2 mg/L to 13 mg/L), whereas the median level in the placebo group was 13 //,mol/L (2.4 mg/L) with an interquartile range of 0 to 19 /x,mol/L (0 to 3.4 mg/L) (P < 0.0001). There was no difference (P > 0.2) in the incidence of side effects between the aminophylline and placebo groups (Table 2), although there was evidence of more frequent nausea in patients receiving aminophylline (P = 0.13). Tremor was the most common side effect. Patient satisfaction, the investigator's assessment of re-

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

Table 1. Baseline Characteristics of Patients Variable

Patients with Asthma Placebo

Male, % Median duration of disease!, y Median frequency of attackst, nlmo Median duration of attackt, h Mean duration off medication^, h Current smoker, % Previous intubation,

%

Mean temperature^, °C Mean pulsed, beats/ min Mean systolic blood pressured, mm Hg Mean diastolic blood pressured, mm Hg Median respiratory ratet, breathslmin Wheezes, % Rales, % Diaphoresis, % Accessory muscle use, % Median FEV.t, L Median FVCt, L Median PEFRt, Llmin

AminophyUine

Total Patients

Patients with COPD P Value*

Placebo

AminophyUine

P Value*

Placebo

AminophyUine

P Value*

38

>0.2

19 (3-35)

16 (5-29)

>0.2

>0.2

1 (0-1)

1 d-2)

0.08

24 (13-66)

>0.2

24 (18-94)

24 (12-72)

0.15

68 ± 136 39

>0.2 >0.2

226 ± 1150 37

239 ± 1074 29

>0.2 >0.2

11

>0.2

13

17

>0.2

36.6 ± 0.6

36.4 ± 0.7

>0.2

36.7 ± 0.7

36.0 ± 0.46

> 0.2

>0.2

95 ± 14

93 ± 23

>0.2

91 ± 15

92 ± 19

>0.2

123 ± 18

> 0.2

131 ± 21

122 ± 40

>0.2

124 ± 18

123 ± 29

>0.2

78 ± 12

77 ± 13

>0.2

85 ± 14

75 ± 26

76 ± 19

>0.2

20 (20-24) 94 2 2

22 (20-27) 95 5 3

>0.2 >0.2 >0.2 >0.2

24 (20-28) 95 5 0

24 (20-26) 89 0 0

24 (20-26) 92 3 2

>0.2 >0.2 >0.2 >0.2

20 1.0(0.7-1.7) 1.8(1.4-2.3)

22 1.5(0.8-1.8) 2.1 (1.5-2.5)

> 0.2 >0.2 >0.2

>0.2 42 39 0.7(0.4-1.1) 0.7 (0.5-0.9) > 0 . 2 1.6(0.9-2.2) 1.4(0.9-1.8) > 0 . 2

>0.2

98 (77-160)

35

32

>0.2

74

46

17 (3-34)

18 (9-30)

>0.2

21 (4-51)

6 (4-24)

>0.2

1 (0-1)

1 (0-1)

0.14 1 d-2)

2 (1-2)

28 (20-96)

21 (12-72)

0.14 24 (12-86)

342 ± 1350 36

374 ± 1426 22

10

22

36.7 ± 0.7

35.6 ± 0 . 6

> 0.2

89 ± 15

91 ± 15

121 ± 17

> 0 . 2 34 ± 41 0.19 42 0.14 22

151 (115-233) 178(117-237)

108 (69-155)

0.06 46

0.11 80 ± 13 >0.2 >0.2 >0.2 >0.2

>0.2

22 (20-24) 94 3 2

29 27 0.9(0.6-1.6) 1.1 (0.6-1.6) 1.8(1.2-2.2) 1.7(1.2-2.2)

>0.2 >0.2 >0.2

142 (96-219) 138 (88-213) > 0 . 2

* Chi-square for categorical variables; Student Mest for normally distributed continuous variables; Wilcoxon rank-sum test for variables with non-normal distributions. t Interquartile range (25% to 75%) is given in parentheses. X Given ± SD.

sponse to therapy, and time in the emergency department were not affected by whether the patient received aminophyUine. Despite the absence of a difference in other response or outcome measures, there was a reduction in the hospital admission rate when aminophyUine was used. Six percent of the patients in the aminophyUine group were admitted, whereas 21% of the placebo group were admitted (difference, 15%; CI, 3% to 27%; P = 0.016). Using a logistic regression model controlling simultaneously for the effects of age, sex, duration of disease, duration of episode, and type of disease, the adjusted odds ratio for the effect of aminophyUine on admission rate was 0.22 (CI, 0.12 to 0.42). When analyzed by disease process, 6% of asthmatic patients treated with aminophyUine were admitted and 19% of asthmatic patients who received placebo were admitted (P = 0.08). Among the patients with chronic obstructive pulmonary disease, 7% of the patients on aminophyUine and 26% of patients on placebo were admitted to the hospital (P = 0.07). Among the admitted patients, the most common reason for admission was failure to clear after 6 hours (77% of patients on placebo, 100% of patients on aminophyUine) (Fisher exact test, P > 0.2). In the admitted patients who were receiving placebo, other reasons for admission included respiratory failure, chest pain, and

worsening of symptoms during treatment in one patient each. None of the admitted patients in either group could walk 100 feet without recurrence of symptoms. The mean (± SD) time in clinic for patients on placebo who were admitted because of failure to clear was 428 ± 91 min, whereas the mean time in clinic for the aminophyUine group who failed to clear was 351 ± 53 min (P = 0.08). The median theophylline level was 82 /xmol/L (14.8 mg/L) for admitted patients who were receiving aminophyUine and 53 /xmol/L (9.6 mg/L) for discharged patients on aminophyUine. Furthermore, among patients receiving aminophyUine, 45% of discharged patients had theophylline levels of more than 55 ju,mol/L (10 mg/L) and 50% of the admitted patients had theophylline levels of more than 55 /xmol/L (10 mg/L) (Fisher exact test, P > 0.2). When the admitted patients were compared with the discharged patients in the aminophyUine and placebo groups, no differences (P > 0.2) were found in frequency of attack, percent of patients with previous intubation, or the percent of patients rated as having a "severe" attack by the blinded investigator (Table 3). There was a difference in the mean duration of attack between admitted and discharged patients in the aminophyUine group (P = 0.0005) but not in the placebo group (P > 0.2). Examination of billing records after the study re-

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

243

Discussion

Figure 1. Spirometric values at baseline and at 60 and 120 minutes after administration of aminophylline or placebo. Top. Forced expiratory volume in 1 second (FEVi). Middle. Forced vital capacity (FVC). Bottom. Peak expiratory flow rate (PEFR). vealed that three patients discharged from the emergency department returned within 72 hours for further treatment (one patient in the aminophylline group and two patients in the placebo group). The average cost of treatment without admission was $350, including physician fees. The average cost of treatment when admitted to the hospital was $1750, excluding physician fees. 244

Despite many years of experience with the use of theophylline, its role in the treatment of acute exacerbations of asthma and chronic obstructive pulmonary disease is still controversial (1, 5, 8-11). It is generally accepted that theophylline is a useful adjunct to the outpatient management of stable asthma and chronic obstructive pulmonary disease (12-16). Using meta-analysis to study 13 controlled trials of aminophylline therapy in patients with severe, acute asthma, Littenberg (1) found no evidence to support or reject its use. In most of the studies analyzed, there were important methodologic problems, the most significant being an inadequate sample size (1). This randomized, double-blind, placebo-controlled trial was designed to assess whether the addition of intravenous aminophylline in standard doses would improve any outcome variables when added to a regimen of intravenous methylprednisolone and intensive aerosolized metaproterenol for moderate to severe acute exacerbations of asthma or chronic obstructive pulmonary disease. Intravenous aminophylline was used instead of oral theophylline to eliminate concerns about absorption. We used intravenous methylprednisolone to eliminate questions of poor absorption if aminophylline produced no favorable effects. As in previous studies (5, 8, 11), we found that adding aminophylline to beta-agonists had little effect on pulmonary function (Figure 1). The dosing regimen used (5.6 mg/kg as a loading dose, followed by a continuous infusion of 0.9 mg/kg per hour) achieved a median serum theophylline concentration of 54 /i,mol/L (9.7 mg/ L). We had hoped to see theophylline levels well within the accepted therapeutic range of 55 /rniol/L (10 mg/L) to 110 /xmol/L (20 mg/L), but the standard dosing schedule used was designed to produce a level of around 55 ju,mol/L (10 mg/L) in 95% of patients (17). A level of 55 /nmol/L (10 mg/L) has long been considered the lower end of the therapeutic range needed to relieve bronchoconstriction (18). Although a few studies have shown that higher therapeutic levels in the range of 94 to 110 /Ltmol/L (17 to 20 mg/L) may increase the effects of theophylline in bronchodilation (10, 19), there is evidence that the dose-response curve of theophylline is not linear but begins to plateau at levels of approximately 71 /Amol/L (12.8 mg/L) (20). It is estimated that 80% of theophylline's bronchodilator effects are provided at levels of 55 /xmol/L (10 mg/L) (21). There is considerable interindividual variation in the response to theophylline (13), and the magnitude of the response may depend on the degree of baseline airway reactivity (22). The incidence of side effects is known to increase as the serum level of theophylline increases (13). Our study failed to show any significant benefit of aminophylline in terms of patient satisfaction, physician assessment of the response to treatment, or time in the emergency department. At the relatively low serum levels achieved in the treatment group, there was no significant difference in the incidence of side effects between the aminophylline and placebo groups, but there was a trend toward a higher frequency of nausea with aminophylline.

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

Table 2. Side Effects, Patient Satisfaction, and Disposition Variable Nausea, % Anxiety, % Tremor, % Seizure, % Palpitations or arrhythmia, % Patients satisfied with treatment, % Patients showing improvement, % Median duration in the emergency department, min* Admitted to hospital, %

Placebo

Aminophylline P Value

3 8 18 1.5

9 9 23 0

0.13 >0.2 >0.2 >0.2

5

6

>0.2

77

78

>0.2

84

88

>0.2

220(180-310) i 205(160-275) 21 6

0.11 0.016

* Interquartile range is given in parentheses.

An unexpected finding was the marked difference in hospital admission rates between the aminophylline and placebo groups. Considering the entire group of 133 patients, only 6% of the patients receiving aminophylline were admitted as opposed to 21% of patients on placebo (P = 0.016). Because we assessed four primary categorical outcomes (patient satisfaction, patient response, minutes in the emergency department, and admission to the hospital), the statistical problem of multiple comparisons arises. One method to gauge the effects of multiple comparisons is the Bonferroni correction (23, 24). If the Bonferroni correction (a/number of outcome variables) is carried out, statistical significance is achieved only at P < .0125. Although the statistical significance of the decrease in admission rate becomes marginal after this correction, we feel that the prospective nature of the trial with predetermined end points makes the chance of a type I error small, especially given the magnitude of the difference in admission rate as well as the sample size. It has been argued that the Bonferroni correction may be overly conservative, particularly in the setting of a prospective study (23). Admissions for asthma account for almost a half million hospitalizations each year (25). We found that aminophylline treatment reduced the likelihood of hospital admission by 71% in patients presenting to the emergency department with an acute exacerbation of bronchospastic disease. If the effectiveness in reducing admission rates and the difference in outpatient and inpatient treatment costs that we found were reproducible and applied to the entire United States, universal adoption of aminophylline for treatment of acute exac-

erbations of asthma alone could correspond to a national cost savings of a half billion dollars yearly. When broken into smaller subgroups by disease process, the admission rates remain approximately the same. Six percent of aminophylline-treated asthmatics and 7% of aminophylline-treated patients with chronic obstructive pulmonary disease were admitted, but 19% of asthmatics who were receiving placebo and 26% of patients with chronic obstructive pulmonary disease who were receiving placebo were admitted (P = 0.07 and P = 0.08, respectively). Although statistical significance at the P = 0.05 level of these differences is lost because of smaller sample sizes, the clinical significance of these threefold reductions in admission rates is still compelling. That the reduction in hospital admissions is fairly constant despite grouping into a primary disease process strengthens the observation in the aggregate group. Furthermore, the observed difference in hospital admission rates was unaffected by adjusting for effects of baseline characteristics such as sex, duration of disease, frequency of attacks, duration off medications, and chronic use of corticosteroids. The randomized design of the trial, the similarity of the experimental and control groups at baseline, and the adjustment for baseline differences all strengthen the validity of the study results. We reduced bias in determining hospital admission as opposed to discharge from the emergency department by the double-blind nature of the study and also by removing the investigators from the decision-making process. We are unable to explain why treatment with aminophylline seems to have a measurable clinical effect in the absence of an objective improvement in pulmonary function as measured by spirometry. However, a study of the efficacy of corticosteroids in the treatment of asthma in an emergency department setting also found a decrease in admission rate despite no significant improvement in spirometric values (26). Because spirometric measurements improved by an average of 15% to 30% between time 0 and time 60 minutes for virtually all of our patients, it seems unlikely that lack of effort explains the failure to detect any effects of aminophylline treatment on pulmonary function. It is possible that further salutary effects of aminophylline on pulmonary function might have occurred later than 120 minutes after aminophylline was started, but we believe that the theophylline concentrations achieved in individual patients were likely to be near steady-state level by this time and that the curves shown in Figure 1 had begun to level off.

Table 3. Baseline Characteristics in Patients Who Were Admitted to the Hospital or Discharged Home from the Emergency Department Variable Discharged (n = 61) Median frequency of attack, nlmo Mean duration of attack, h Previous intubation, % "Severe" attack as assessed by the physician, %

Aminophylline Admitted (n = 4)

1 61 17 9

4 20 0 25

P Value*

Discharged (n = 54)

>0.2 0.0005 >0.2 >0.2

1 74 11 15

Placebo Admitted (n = 14) 1 74 23 29

P Value* > 0.2 > 0.2 >0.2 > 0.2

* Student /-test for categorical variables; Fisher exact test for proportions.

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

245

All the favorable effects of theophylline may not be due to bronchodilation. Although there is a small effect of theophylline on phosphodiesterase inhibition, it is believed to be insignificant at serum concentrations considered to be therapeutic (27). However, multiple molecular forms of phosphodiesterase have now been identified, and these differ in substrate specificity, kinetic characteristics, intracellular localization, and response to various allosteric effectors. The effects of theophylline on these particular forms of phosphodiesterase and the primary tissue focus of its effects have not been completely elucidated (28). The role of theophylline as an adenosine antagonist is also in question, because enprophylline, another xanthine that has little adenosine-blocking activity, retains comparable bronchodilating effects (29, 30). Other postulated mechanisms of action include increased circulating catecholamine levels (2), stimulation of the respiratory center (31), improved mucociliary clearance (32, 33), improved diaphragmatic contractility (34), decreased diaphragmatic fatigability (34), inotropic effects on both the right and left ventricles (35), diuretic effects (36), inhibition of prostaglandins (37), alterations in calcium ion fluxes (37), and anti-inflammatory effects, such as inhibition of anaphylactic mediator release (37), and the suppression of responses to mast cell mediators (38-40). Many of theophylline's effects do not correlate with its bronchodilator effects. Doses sufficient to raise the serum theophylline level to the range of 72 fjumoVL (13 mg/L) increased the provocative dose of methacholine, histamine, and acetylcholine required to induce bronchospasm but did not correlate with bronchodilator effects (38-40). In addition, the favorable effects of theophylline on ventricular function in patients with chronic obstructive pulmonary disease do not correlate with blood levels (33). Finally, the ability of theophylline to increase contractility and decrease fatigability of the diaphragm is not directly related to its bronchodilator effects (34). It may be that a combination of aminophylline's many actions contributes to the drug's favorable effects. If the putative anti-inflammatory or other effects of theophylline are responsible for the change in admission rate, it is possible that we do not have an adequate objective measure of these effects at this time. Studies have shown that subjective responses to longterm treatment with theophylline are often at odds with objective measures of bronchodilation (41); however, in our patients treated in an emergency department setting, there was no difference in patient satisfaction or in the investigator's assessment of response to therapy between the aminophylline and placebo groups. Because asthma and chronic obstructive pulmonary disease are not homogeneous diseases but often have multiple causes, it is possible that patients destined to be admitted to the hospital differ from those destined to be discharged. Although we were unable to show differences in baseline characteristics, we did not attempt to classify the patients into subgroups by precipitants such as infectious, inflammatory, or allergic exacerbations. Further investigations into the actions of theophylline as well as other objective methods of evaluating the response to treatment of patients with bronchospastic disease are needed. 246

Because theophylline has a narrow therapeutic index (4) and because several investigators have suggested that theophylline, when used in conjunction with betaagonists, contributes to the increased morbidity and mortality seen in the asthmatics who are receiving such therapy (42-45), it is important that evidence of some utility of this drug be demonstrated before it is routinely used. We believe that the threefold reduction in the hospital admission rate of patients with acute exacerbations of bronchospastic disease may be justification for the use of a theophylline preparation in the emergency department setting, particularly in patients at high risk for admission to the hospital or in settings where hospital-bed availability is limited. Further larger-scale trials of theophylline are needed to confirm the decrease in admission rates observed in our study. If these decreases are confirmed, considerable cost reductions from the use of theophylline are possible. Requests for Reprints: Keith Wrenn, MD, University of Rochester Medical Center, Emergency Department, Box 655, 601 Elmwood Avenue, Rochester, NY 14642. Current Author Addresses: Drs. Wrenn and Slovis: Strong Memorial Hospital, 601 Elmwood Avenue, Box 655, Rochester, NY 14642. Ms. Murphy: Department of Medicine, Emory University School of Medicine, Grady Memorial Hospital, Atlanta, GA 30335. Dr. Greenberg: Office of the Dean, Emory University School of Public Health, 1599 Clifton Road NE, Atlanta, GA 30329.

References 1. Littenberg B. Aminophylline treatment in severe, acute asthma. A meta-analysis. JAMA. 1988;259:1678-84. 2. Summer WR. Status asthmaticus. Chest. 1985;87(1 Suppl):87S-94S. 3. Rebuck AS, Chapman KR. Asthma: 2. Trends in pharmacologic therapy. Can Med Assoc J. 1987;136:483-8. 4. Rossing TH. Methylxanthines in 1989. Ann Intern Med. 1989;! 10: 502-4. 5. Fanta CH, Rossing TH, McFadden ER Jr. Treatment of acute asthma. Is combination therapy with sympathomimetics and methylxanthines indicated? Am J Med. 1986;80:5-10. 6. Benatar SR. Fatal asthma. N Engl J Med. 1986;314:423-9. 7. Tager I, Speizer FE. Role of infection in chronic bronchitis. N Engl J Med. 1975;292:563-70. 8. Rossing TH, Fanta CH, Goldstein DH, Snapper JR, McFadden ER Jr. Emergency therapy of asthma: comparison of the acute effects of parenteral and inhaled sympathomimetics and infused aminophylline. Am Rev Respir Dis. 1980;122:365-71. 9. McWilliams BC Jr, Menendez R, Kelly HW, Howick J. Comparison of the effects of intravenously administered aminophylline and inhaled isoproterenol on normal airways. Am Rev Respir Dis. 1986; 133:744-8. 10. Passamonte PM, Martinez AJ. Effect of inhaled atropine on metaproterenol in patients with chronic airways obstruction and therapeutic serum theophylline levels. Chest. 1984;85:610-5. 11. Siegel D, Sheppard D, Gelb A, Weinberg PF. Aminophylline increases the toxicity but not the efficacy of an inhaled beta-adrenergic agonist in the treatment of acute exacerbations of asthma. Am Rev Respir Dis. 1985;132:283-6. 12. Marciano D, Auclair M, Pariente R, Aubier M. A randomized, controlled trial of theophylline in patients with severe chronic obstructive pulmonary disease. N Engl J Med. 1989;320:1521-5. 13. Billing B, Dahlquist R, Hornblad Y, Leideman T, Skareke L, Ripe E. Theophylline in maintenance treatment of chronic asthma: concentration-dependent additional effect to beta 2-agonist therapy. Eur J Respir Dis. 1987;70:35-43. 14. Wolfe JD, Tashkin DP, Calvarese B, Simmons M. Bronchodilator effects of terbutaline and aminophylline and in combination in asthmatic patients. N Engl J Med. 1978;298:363-7. 15. Nassif EG, Weinberger M, Thompson R, Huntley W. The value of maintenance theophylline in steroid-dependent asthma. N Engl J Med. 1981;304:71-5. 16. Hill NS. The use of theophylline in "irreversible" chronic obstructive pulmonary disease. An update. Arch Intern Med 1988; 148:257984. 17. Mitenko PA, Ogilvie RI. Rational intravenous dose of theophylline. N Engl J Med. 1973;289:600-3. 18. Turner-Warwick M. Study of theophylline plasma levels after oral

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

19.

20.

21. 22. 23. 24. 25. 26. 27. 28. 29.

30. 31.

administration of new theophylline compounds. Br Med J. 1957;2: 67-9. Vozeh S, Kewitz G, Perruchoud A, Tschan M, Kopp C, Heitz M, et al. Theophylline serum concentration and therapeutic eflFect in severe acute bronchial obstruction: The optimal use of intravenously administered aminophylline. Am Rev Respir Dis. 1982; 125: 181-4. Klein JJ, Lefkowitz MS, Spector SL, Cherniack RM. Relationship between serum theophylline levels and pulmonary function before and after inhaled beta-agonist in "stable" asthmatics. Am Rev Resp Dis. 1983;127:413-6. Jenne JW. Theophylline as a bronchodilator in COPD and its combination with inhaled beta-adrenergic drugs. Chest (1 Suppl) 1987; 92:7S-14S. Ahrens RC, Milavetz G, Joad J. The eflFect of theophylline and beta 2 agonists on airway reactivity. Chest (1 Suppl) 1987;92:15S-21S. Glantz SA. Primer of Biostatistics. 2d edition. New York: McGrawHill, Inc.; 1987:88-91. Fletcher RH, Fletcher SW, Wagner EH. Clinical Epidemiology: The Essentials. 2d edition. Baltimore, Maryland: Williams & Wilkins; 1988:172-87. Weiss KB. Seasonal trends in US asthma hospitalizations and mortality. JAMA. 1990;263:2323-8. Littenberg B, Gluck EH. A controlled trial of methylprednisolone in the emergency treatment of acute asthma. N Engl J Med. 1986;314: 150-2. Poison JB, Kr/anowski JJ, Goldman AL, Szentivanyi A. Inhibition of human pulmonary phosphodiesterase activity by therapeutic levels of theophylline. Clin Exp Pharmacol Physiol. 1978;5:535-9. Weishaar RE. Multiple molecular forms of phosphodiesterase: an overview. J Cyclic Nucleotide Protein Phosphor Res. 1986-87;11: 463-72. Pauwels R, Van Renterghem DV, Van Der Straeten M, Johannesson N, Persson CG. The eflFect of theophylline and enprofylline on allergen-induced bronchoconstriction. J Allergy Clin Immunol. 1985;76:583-90. Lunell E, Andersson KE, Persson CG, Svedmyr N. Intravenous enprofylline in asthma patients. Eur J Respir Dis. 1984;65:28-34. Dowell AR, Heyman A, Sieker HO, Tripathy K. EflFect of aminophylline on respiratory center sensitivity in Cheyne-Stokes respiration and in pulmonary emphysema. N Engl J Med. 1965;273:144753.

32. Welsh MJ, Widdicombe JH, Nadel JA. Fluid transport across the canine tracheal epithelium. J Appl Physiol. 1980;49:905-9. 33. Iravani J, Melville GN. Wirkung von pharmaka und milieuanderangen auf die flimmertatigkeit der atemwege [Effects of drugs and environmental factors on ciliary movement.]. Respiration. 1975;32: 157-64. 34. Aubier M, De Troyer A, Sampson M, Macklem PT, Roussos C. Aminophylline improves diaphragmatic contractility. N Engl J Med. 1981;305:249-52. 35. Matthay RA, Berger HJ, Loke J, Gottschal A, Zaret BL. EflFects of aminophylline upon right and left ventricular performance in chronic obstructive pulmonary disease. Am J Med. 1978;65:903-10. 36. Barker AF. Strategies in managing asthma. West J Med. 1989; 150: 303-8. 37. Robertson C, Levison H. Bronchodilators in asthma. Chest (1 Suppl) 1985;87:64S-7S. 38. McWilliams BC, Menendez R, Kelly HW, Howick J. EflFects of theophylline on inhaled methacholine and histamine in asthmatic children. Am Rev Respir Dis. 1984;130:193-7. 39. Cartier A, Lemire I, L'Archev^que J, Ghezzo H, Martin RR, Malo JL. Theophylline partially inhibits bronchoconstriction caused by inhaled histamine in subjects with asthma. J Allergy Clin Immunol. 1986;77:570-5. 40. Koeter GH, Meurs H, Jonkman JH, de Vries K. Protective eflFect of choline theophyllinate on histamine, acetylcholine, and propranololinduced airflow obstruction. Respiration. 1984;45:139-45. 41. Guyatt GH, Townsend M, Norgradi S, Pugsley SO, Keller JL, Newhouse MT. Acute response to bronchodilator. An imperfect guide for bronchodilator therapy in chronic airflow limitation. Arch Intern Med. 1988;148:1949-52. 42. Smith SR, Kendall MJ. Potentiation of the adverse effects of intravenous terbutaline by oral theophylline. Br J Clin Pharmacol. 1986; 21:451-3. 43. Brady HR, Ryan F, Cunningham J, Tormey W, Ryan MP, O'Neill S. Hypophosphatemia complicating bronchodilator therapy for severe acute asthma. Arch Intern Med. 1989;149:2367-8. 44. Wilson JD, Sutherland DC, Thomas AC. Has the change to betaagonists combined with oral theophylline increased cases of fatal asthma? Lancet. 1981;1:1235-7. 45. Coleman JJ, Vollmer WM, Barker AF, Schultz GE, Buist AS. Cardiac arrhythmias during the combined use of beta-adrenergic agonist drugs and theophylline. Chest. 1986;90:45-51.

15 August 1991 • Annals of Internal Medicine • Volume 115 • Number 4

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19732/ by a University of California San Diego User on 03/30/2017

247