Critically Appraised Topic: Does the addition of theophylline improve outcomes

for asthma patients using inhaled steroids

Appraised by: Eric Terman, MD

November 24, 1997

Clinical Bottom Lines:

1. Although there were decreases in the use of B agonists as well as FVC and FEV1 it is difficult to know whether it had a substantial impact overall.

2. This might be a reasonable regimen to try, in order to avoid high dose inhaled steroids, but there is no evidence that it will prevent hospitalization or death.

3. The data presented make it impossible to calculate a number needed to treat.

4. You can use theophylline in attempts to get better control over asthmatic patients, but this study really does not support your decision.

The Methods: Randomized trial of Budesonide 400 ug bid plus low dose theophylline vs. budesonide 800 ug bid plus placebo. This appears to have been randomized, but the treatment groups differed substantially. It is ambiguous whether or not the data was analyzed by an intention to treat model.

The Evidence:

The data for reduction of use of Albuterol is stated as significant, but no actual data nor frequencies are shown. It really does not matter if the frequency was decreased from two to one time, but it would matter if the reduction were from five to one.

Data is shown for an increase in FVC and FEV1, but these are not correlated well with symptoms and are therefore fairly meaningless.

Comments:

1. Although there was an attempt to randomize and blind patients, it does not appear that either were successful. It appears that the group randomized to 800 ug of budesonide plus placebo could have known that they were not getting theophylline by simple deduction, as could the investigators. It is a simple matter of how many puffs they took. In addition there was no control group. The groups also seemed to be somewhat different in terms of sex of the patients.

2. The study was small, which is not terrible except for the fact that there were other problems.

3. The time of comparison for the two groups for FVC was different, but this might have been a typographical error, nonetheless it makes analysis more difficult.

4. One of the more important outcomes, number of exacerbations, did not differ between the two groups.

5. There was an improvement in FVC and FEV1 as well as the number of times a patient used their inhaler, which is somewhat promising, but really can not replace hard outcomes.

6. The observed difference of serum cortisol levels is suggestive that theophylline could help spare steroid use, but the groups were different at the start. Although this difference was not significant, they claim that the high start then showed suppression. This is not clear from their data.

7. The actual numbers for peak flows are so similar it really makes little or no difference.

Evans DJ, Taylor DA, Zetterstrom O, Chung KF, O’Connor BJ, Barnes PJ. A comparison of low-dose inhaled budesonide plus theophylline and high-dose inhaled budesonide for moderate asthma. NEJM. 1997; 337(20):1412-1418.