Anaphylaxis caused by benzodiazepines is very rare. In a recent report by the 6th National Audit Project (NAP6) in the UK [14], of the 199 agents identified as causative for perioperative anaphylaxis, the most common agents were antibiotics (94/199), neuromuscular blocking agents (65/199), and chlorhexidine (18/199), with no reports for benzodiazepines. On the other hand, there have been several reports of anaphylaxis caused by remimazolam, a new anesthetic, in recent years; 11 cases have been reported to date. The frequency of anaphylaxis caused by remimazolam is estimated to be 0.18% based on the report by Kim et al. [12]. The characteristics of previously reported remimazolam anaphylaxis cases and the present case are presented in Table 1. All patients had circulatory collapse, and cardiopulmonary resuscitation was needed in three patients, including our patient [9, 12]. Respiratory manifestations were observed in 58% (7/12) of patients, but bronchospasm has been reported in only one patient other than ours [8]; laryngeal edema was reported in 2 patients [10, 11]. According to the report by NAP6, hypotension was observed in all of the patients with perioperative anaphylaxis reported nationwide, bronchospasm/high airway pressure was reported in 48% of patients, and a reduced/absent capnography trace was reported in 33% of patients [14]. Based on these data, circulatory collapse was consistently observed in all cases of remimazolam anaphylaxis; however, bronchospasm due to remimazolam anaphylaxis was less frequent than other manifestations of anaphylaxis. Onset of anaphylaxis occurred within 5 min after remimazolam administration in 10 of the 12 cases. All patients were successfully treated with adrenaline. No biphasic anaphylaxis was reported.

In the present case, anaphylaxis was characterized by severe bronchospasm/high airway pressure preceding circulatory collapse. If bronchospasm with remimazolam anaphylaxis develops, it may overlap the timing of tracheal intubation due to its rapid onset. Therefore, the diagnosis of remimazolam-induced anaphylaxis is sometimes challenging, as it can be confused with intubation complications. In addition, because remimazolam is a drug that can be administered continuously, a delay in responding to anaphylaxis can lead to more severe complications. Kim et al. reported a higher incidence of anaphylaxis in patients with a higher infusion rate [12].

Regarding circulatory management, The Australian and New Zealand College of Anaesthetists (ANZCA) and the Australian and New Zealand Anaesthetic Allergy Group (ANZAAG) recommend intravenous adrenaline under close monitoring for intraoperative anaphylaxis [15]. The dosages are 0.01–0.2 mg intravenously or 0.5 mg intramuscularly according to severity. If cardiac arrest occurs, 1 mg of adrenaline should be administered intravenously. In this case, we administered 0.5 mg of adrenaline intravenously, a dose lower than the recommendation. Although the patient was quickly resuscitated, we had difficulty regulating post-resuscitation hypertension; his systolic arterial blood pressure temporarily reached 300 mmHg and exceeded 250 mmHg for an extended period. In the other two previously reported cases of cardiac arrest due to remimazolam anaphylaxis [9, 12], both patients recovered spontaneous circulation with the administration of 1 mg of intravenous adrenaline, and post-resuscitation hypertension was greater than 200 mmHg. Therefore, in the case of cardiac arrest due to remimazolam anaphylaxis, a lower dose of intravenous adrenaline may be sufficient, and clinicians must be aware of the possibility of post-resuscitation hypertension.

It is essential to ascertain the allergy history of patients to prevent anaphylaxis. Anesthesiologists should fully review and evaluate past anesthesia records and allergy details [16]. If drugs suspected of inducing anaphylaxis are present, the best method of prevention is to avoid exposing patients to these drugs, including any cross-antigenic material [17]. Reports have suggested the presence of cross-reactivity between remimazolam and midazolam [9, 11]. In the present case, the patient had previously developed a severe allergic reaction thought to be caused by fresh-frozen plasma transfusion; however, midazolam had also been administered shortly before. Severe hypotension also occurred during the induction of anesthesia for cardiac surgery in which midazolam was administered. Considering the series of anaphylactic events, it is probable that the cause of the allergic reaction was midazolam, and the exposure to midazolam may have been the root cause of remimazolam anaphylaxis through the activation of an IgE-mediated pathway. However, we did not perform a skin prick test or intradermal test for midazolam and remimazolam; therefore, this assumption is not confirmed. On the other hand, the patient had a history of taking brotizolam, a benzodiazepine, but was not allergic to it. Regarding cross-reactivity among benzodiazepines, some reports showed cross-reactivity among benzodiazepines [18, 19], while others did not [20]. To date, cross-reactivity among benzodiazepines is poorly understood. The structures of remimazolam, midazolam and brotizolam are similar. However, remimazolam and midazolam are formed by a diazepine ring bonded to a benzene ring, whereas brotizolam is formed by a diazepine ring bonded to a thiophene ring; this structural difference may have caused a difference in cross-reactivity.

Some patients who experienced anaphylaxis induced by remimazolam had no history of exposure to remimazolam or midazolam [7] or negative skin test results for remimazolam [7, 10, 12]. Remimazolam preparations used worldwide, such as Anelem®, ByFavo®, and Aptimyda™, contain dextran 40 as an additive. Dextran occasionally causes an anaphylaxis-like, acute hypersensitivity reaction due to complement activation via immune complexes of non-IgE antibodies; this reaction is indistinguishable from IgE-mediated anaphylaxis in clinical findings [21]. It is possible that remimazolam anaphylaxis involves anaphylaxis via the IgE-mediated pathway and anaphylaxis-like reactions caused by dextran via the non-IgE-mediated pathway.

A limitation of our case report is that we did not perform either skin prick tests and intradermal tests for remimazolam and midazolam due to considerations of patient safety. Then we also did not perform intradermal tests for any of the suspected drugs including latex which is also a common cause of anaphylaxis. Therefore, the causative agent was not confirmed by allergic tests. However, the other suspected drugs yielded negative results on the prick tests, and those drugs could be used clinically during anesthetic management of the patient during subsequent surgery without causing an allergic reaction. Given the patient’s history, there is a very high likelihood that the cause of anaphylaxis is remimazolam. However, the gold standard for post-anaphylaxis management is to make an accurate diagnosis and to provide patients with detailed information about their anaphylaxis. Our review shows that the positive rate of skin tests for remimazolam is low (Table 1), and according to a recent report, even skin tests have a positive rate of only about 57% for diagnosing intraoperative anaphylaxis. Therefore, it is necessary to combine the skin test and various tests (e.g., specific antibody test and basophil activation test) to improve the diagnostic accuracy [22].