Skip navigation

Asthma & allergy

Immunotherapy for allergic conditions is one of those topics about which people seem to be wholly unmoved, or passionate believers. There may be circumstances where it is important - for insect sting reactions, perhaps - or in hay fever caused by airborne allergens.

For bronchial asthma it is a controversial topic. "What is the evidence that there is any benefit?" is the obvious question. What we decide to do with that evidence is the one that follows hot in its tracks. Bandolier certainly can't help with the second of these, but it can point to a meta-analysis of randomised studies which is likely to be of interest to those involved in trying to answer it.


The meta-analysis [1] was done by a group in Melbourne. They identified 20 randomised, placebo-controlled double-blind trials of immunotherapy for asthma between 1966 to 1990; English language reports only were used.

Data extracted

They looked at age of patients, allergen preparations used, adverse effects, and primary outcome measures like symptoms, medication, lung function and bronchial hyperreactivity (BHR). Data for house dust mites and other allergens (animal epithelium, mould or pollen) were presented separately. The results are given in terms of odds ratios, but the meta-analysis contains no information that would allow the calculation of numbers-needed-to-treat.


The results for symptom control, for reduction in medication and for BHR were all positive, as shown in the table with 95% confidence intervals. The effect of mite immunotherapy on lung function just achieved statistical significance, but was of little clinical significance.

There were many reports of local reactions at the injection site with allergens, with systemic reactions occurring in 32% (95%CI 20 - 44%). There were four reports of anaphylaxis. With placebo injections, systemic adverse effects occurred in 18% (95%CI 7 - 29%) of patients. There were no reports of anaphylaxis following placebo.

So what is the answer?

The authors calculated that 33 negative unpublished trials would be needed to reverse the conclusions of their meta-analysis. Their conclusion was that there is a solid basis to consider this treatment effective.

They also spend some time examining evidence about serious adverse effects like anaphylaxis. They report several surveys which examine this rare but life threatening event:
  • 31 of 14, 639 mite or pollen injections - a rate of 1 in 500.
  • 19 deaths in the USA in the five years till 1991 due to allergen immunotherapy, 16 of which were in asthmatics - with no denominator but with unstable asthma or overdose of allergen as a major precipitating factor as the main contributors.
Their conclusion, that allergen immunosuppression may be a useful adjunct to therapy in allergic asthma was followed by many caveats:-
  • an identifiable allergen
  • full assessment of risks and contraindications
  • initial consultant assessment
  • close consultant supervision
  • specific and effective allergen extract
  • structured follow-up
  • flexible dosing regimen
  • facilities and staff to allow for treatment of adverse effects
  • more trials - large, multicentre, randomised, with different outcomes
All in all, not the most enthusiastic of recommendations. But this is just an extreme case of weighing up the benefits and harms of an intervention. It would be nice to have a simple answer, but they are not always available, and Bandolier certainly doesn't have a view here.


  1. MJ Abramson, RM Puy, JM Weiner. Is allergen immunotherapy effective in asthma: a meta-analysis of randomized controlled trials. American Journal of Respiratory Critical Care Medicine 1995 151; 969-74.

previous or next story in this issue