We thank the authors of the four letters to the editor  for their interest in our work regarding our article  and would like to respond to them as follows.

Valach et al.  suggest intramuscular epinephrine should be used as first-line treatment in perioperative anaphylaxis. We disagree with this suggestion as perioperative anaphylaxis is a life-threatening adverse reaction with an estimated mortality rate of 4 to 4.7%, a mortality rate higher than in food, Hymenoptera, or other drug-related anaphylaxis.  This perioperative mortality difference has multiple potential causes related to the amount of antigen presented when drugs are injected intravenously in the setting of the anesthetized patient with decreased sympathetic tone, vasodilation, and/or negative inotropic effects of inhaled anesthetic agents. In any acute shock state, rapid restoration of normal hemodynamics is critical to limit the potential of organ injury. In particular, recommendations for epinephrine use in perioperative anaphylactic shock are IV. Furthermore, during anesthesia, patients are fully monitored, vascular access is available, and anesthesiologists are trained to use vasopressors, including IV epinephrine, to rapidly restore baseline hemodynamics. IV administration allows epinephrine to be rapidly titrated to the therapeutic target of blood pressure restoration, an effect not possible by the intramuscular route. Some patients may require a higher dose of epinephrine, especially those undergoing regional anesthesia or taking other drugs that may interfere with sympathetic tone. The critical reason for choosing to use IV epinephrine rather than intramuscular epinephrine is that in severe anaphylaxis, muscle perfusion is severely impaired. Microdialysis studies in experimental models of anaphylaxis have shown that exchange between the muscle compartment and plasma is decreased, particularly for catecholamines.  This could lead to a decrease in the effectiveness of intramuscular epinephrine in the most severe forms of anaphylaxis, which are frequent in the perioperative period with refractory anaphylaxis. 

Nonetheless, intramuscular epinephrine remains important when IV access is not immediately available (e.g., in the prehospital setting), when the treating physician is not trained in the use of IV epinephrine, or when the patient cannot be properly monitored. Furthermore, the adverse effects of epinephrine usually occur when inappropriate dosing is used. We agree with the authors that ongoing education of anesthesiologists is essential to avoid misusing epinephrine. However, as stated in all the guidelines issued by the perioperative anaphylaxis groups, IV epinephrine remains the first-line treatment. 

Nelson et al.  suggest the term hypovolemia in anaphylaxis is misleading. The hypovolemia occurring during anaphylaxis is explained in part by reduced vascular tone, increased venous and splanchnic sequestration of blood volume.  However, profound hypovolemia can also occur during anaphylaxis due to massive capillary leak produced by the multiple mediators released (e.g., histamine, platelet activating factor, bradykinin, prostaglandins, and NO).  These mediators produce acute endothelial injury, barrier disruption, increased capillary permeability, and intravascular hypovolemia that occurs in the initial phase of anaphylactic shock. This hypovolemic response is similar to septic shock, but occurs much more rapidly with an acute volume shift into the interstitial space causing profound hypovolemia, as shown by clinical reports and hemoconcentration observed after anaphylactic shock. 

Thus, vascular filling (i.e., volume repletion) remains the cornerstone of treating severe anaphylaxis. As a reminder, in studying fatal cases of perioperative anaphylaxis, even when epinephrine is appropriately dosed and administered within minutes of the onset of the reaction, hypovolemia is frequent during initial resuscitation due to the increased capillary permeability, an important factor that contributes to refractory hypotension.  In patients without hemodynamic monitoring receiving a high volumes of fluid during initial anaphylaxis resuscitation, euvolemia is rarely achieved.  For these reasons, restoration of intravascular volume should remain a primary goal in perioperative anaphylaxis, with the administration of 20 to 30 ml.kg-1 of crystalloids during the initial phase of resuscitation. For patients who remain unstable after this initial resuscitation, additional hemodynamic monitoring, including echocardiography to evaluate patients where appropriate volume repletion may be important, to guide further fluid administration.

Balzani et al.  comment on pretreatment strategies for nonimmunologic reactions. We agree that antihistamines and corticosteroids are not able to prevent allergic reactions to radiologic contrast agents. Although the mechanism of allergy to contrast agents is not well-understood, some of these reactions can be severe and lead to shock or death despite antihistamine and corticosteroid pretreatment. Findlay et al. showed that antihistamines were able to reduce reactions due to nonspecific histamine release to hyperosmolar products.  This was of particular interest with older hyperosmolar radiocontrast agents, but with low osmolar nonionic agent use, such situations are less common.

During anesthesia, some drugs with histamine-releasing properties are still used (e.g., atracurium, morphine, or vancomycin).  Based on this consideration, with certain known agents such as vancomycin, a slow infusion of a dilute solution should be considered, and in these specific situations, premedication with antihistamines and corticosteroids may be useful. 

As mentioned by Balzani et al., these drugs may also be of interest in other pathologic conditions, such as mastocytosis, where nonspecific mast cell activation occurs. These patients frequently report clinical signs of mast cell activation, for a wide variety of stimuli. This pathology is rare, but in this situation, antihistamines and corticosteroids may be useful. All patients with a history of reactions to anesthetic drugs or radiologic contrast material should be evaluated, and as a reminder, premedication will not prevent anaphylaxis.

We also agree with Arora et al.  that the level of evidence is low with sugammadex administration in rocuronium-induced anaphylaxis. Our inclusion of sugammadex was based on previous reports as a treatment for reactions refractory to epinephrine, but we agree there are limited data supporting its use.  Sugammadex itself is associated with anaphylaxis, and the incidence was estimated to be 0.02% of exposure, exceeding even the risk of anaphylaxis to rocuronium.  For all these reasons, we agree that sugammadex should not be used as first-line therapy for rocuronium-induced anaphylaxis. Nevertheless, when all other treatments have failed, sugammadex may remain a therapy of last resort.