Death from allergic reactions to venoms from stinging insects (bees, hornets, yellowjackets, wasps, and fire ants) has been known since antiquity[1]. At least 40 fatalities from stings are reported each year in the United States. Those who are susceptible have been sensitized against specific venom antigens injected by a previous sting. Although not all those who demonstrate immunoglobulin E (IgE) antibody against venom have had a previous reaction[2], studies have shown that 40% to 74% of adults who have had at least one previous systemic allergic reaction will have a systemic allergic reaction with their next sting [3-5]. In a study [6] reported in this issue of the Annals, van der Linden and his colleagues from the Netherlands found that only 28% of their patients who had had an allergic reaction to a stinging insect before experienced reactions with challenge stings. Their patients volunteered to be stung as a qualification for venom therapy. By carefully observing these patients, the authors have extended the sparse literature on the physiology of human anaphylaxis. They show that compensatory mechanisms for restoration of blood pressure operate when allergic shock occurs. They also add to our knowledge of the variable responses of sensitized patients who undergo allergen challenge.

Although it is known that the risk for an allergic reaction to a sting in venom-sensitized persons can be reduced to less than 3% by immunizing them against the appropriate venoms [3, 7, 8], a predictor for reactivity has been sought. Some of our European colleagues, including van der Linden and associates, have advocated administering challenge stings as a prerequisite to venom immunotherapy. In the United States, others have abandoned the procedure [9]. Whether untreated, venom-allergic individuals are at risk for reactions of increasing severity was questioned in a recent prospective study in children in which the outcome of "natural" (accidental) stings was documented [10]. This study found that children whose systemic allergic reactions had been only cutaneous were not at risk for progressing to more severe reactions and that the risk for any systemic reaction to a future sting was only about 10%. Van der Linden reports a similar stereotype in the responses of his patients to stings administered in a controlled setting.

Most U.S. allergists believe that in order to recommend venom immunotherapy in an adult patient with a history of a previous systemic sting reaction of any severity, it is sufficient to show that the patient has either skin-test or in-vitro evidence for the presence of venom-specific IgE antibodies. This position is predicated on the observations that the risks associated with treatment are no greater than those associated with other allergen immunotherapy [11]; that there is more than an even chance for a systemic reaction with the next sting in unimmunized patients; and the benefit to the majority at risk outweighs the risks and inconveniences of immunotherapy to the minority who might not react.

Our group has avoided deliberate stings in unimmunized patients since observing, as part of a controlled clinical trial of venom immunotherapy, the near-catastrophic consequences of several challenge stings in placebo-treated, venom-allergic patients even when conditions for urgent treatment seemed ideal [12]. Although in this country the challenge sting has continued to be a useful research tool, enabling us to define measures of successful therapy [13, 14] and showing that immunotherapy may be discontinued in certain patients with relative safety after a specific, finite period of immunization [15], challenge stings are not part of the usual diagnostic evaluation of insect allergic patients.

That approaches differ on the opposite sides of the Atlantic speaks to the need for improved understanding of human anaphylaxis.