Defining ENDOtypes in Perioperative Hypersensitivity by Extensive Cellular and Molecular PHENotyping (ENDOPHEN)

Perioperative Acute Hypersensitivity (PAH) is a systemic reaction that occurs rapidly following injection of a drug during anesthesia.The HSA-PA reaction must occur within a maximum of one hour after the induction of anesthesia or a new product by the anesthetist. The main mechanism evoked is an immune response of immediate systemic hypersensitivity or anaphylaxis. Anaphylactic reactions are classically described as IgE-dependent and triggered by the injection of allergen which by bridging specific IgE present on the surface of mast cells, induces a massive release of histamine responsible for the observed symptoms. The diagnosis of this mechanism (IgE endotype) requires the determination of associated circulating mediators (histamine and mast cell tryptase) as well as skin tests performed during an allergologic evaluation. However, our previous work on patients with PAH (NASA study, ClinicalTrials.gov: NCT01637220) demonstrated that classical markers of IgE endotype are present in only 42% of patients. This finding has three consequences: * a diagnostic inaccuracy with deleterious consequences for the patient, * the existence of undocumented endotypes explaining the observed clinical manifestations, * a lack of formal identification of culprit drug, with uncertainty about the eviction recommendations leading to consequences for the safety of the patient. The investigators hypothesize that symptoms associated with PAH are caused by several distinct endotypes involving different cellular effectors and molecular mediators. These endotypes may be related to the immune system but independent of IgE, or independent of the immune system. To assess these endotypes, The investigators will be measuring the activation status of blood cells and a wide range of secreted mediators in blood drawn as soon as possible after PAH onset, and at steady state during a subsequent allergology visit. These data will be analyzed along with clinical data in multivariate analysis and clust