Risk of anaphylactic reaction to aprotinin. Update article.

Objective: To analyse the risk of anaphylactic reaction with the administra tion of aprotinin, either by iv route or as a biological sealent applicatio n and to propose updated guidelines in accordance with current data of the literature. Data sources: Search in the Medline(R) data base of articles in French, Eng lish and German, published since 1960, using following key words: aprotinin , allergy, anaphylaxis. Study selection: All categories of articles on this topic have been selecte d. Data extraction: Articles have been analysed for history, incidence and mec hanisms of anaphylactic reactions, symptomatology, factors of risk, diagnos is and precautions of use. Data synthesis: Aprotinin is widely used for decreasing peroperative bleedi ng, especially in cardiac and orthopaedic surgery. This heterologue protein can cause anaphylactic reactions in 0.5 to 5.8% of patients, depending of the inclusion criteria. They are mediated by IgG and IgE antibodies. Aproti nin has also a direct, non specific, histaminoliberation effect. The clinic al presentation includes various degrees of severity, up to cardiac arrest. Documented factors of risk are a previous parotinin administration, 15 day s to 6 months before, and intolerance to beef meat, white of egg, cheese an d milk. The immediate biological diagnosis is obtained on assessing the deg ranulation of basophiles (histamine) and mastocytes (tryptase), as well as the concentration of anti-aprotinin antibodies (RAST IgE), with a test of i nhibition. The secondary assessment, six weeks later, includes prick-tests and intradermoreactions if the former are negative. The mean precaution con sists to search factors of risk at preanaesthetic assessment. The predictiv e value of systematic prick-tests has not yet been validated. Anti H-1 and anti H-2 premedication is inefficient. A test dose can trigger a severe rea ction. Conclusion: Considering a significant anaphylactic risk, aprotinin administ ration becomes only licit after a careful evaluation of the benefit-risk ra tio. (C) 2000 Editions scientifiques et medicales Elsevier SAS.