Anaphylaxis is a severe, potentially life-threatening, acute hypersensitivity reaction. The agents most commonly involved in perioperative anaphylaxis are neuromuscular blocking drugs, latex, and antibiotics.1 Allergic reactions to ethylene oxide (EO), a highly reactive gas widely used for sterilization of medical devices, are rare and have been observed in patients during hemodialysis and in those with myelomeningocele.2–6 We report the case of severe anaphylaxis to EO in a 13-year-old boy with myelomeningocele during general anesthesia.
Written consent for publication was obtained from the patient.
A 13-year-old boy had congenital myelomeningocele accompanied by hydrocephalus due to Chiari II malformation, paraplegia, neurogenic vesical dysfunction, and progressive neuromuscular scoliosis. Hydrocephalus was treated with a ventriculoperitoneal shunt after birth. At 5 years of age, contact allergy to adhesive tape was observed and latex sensitivity was assumed. He previously had undergone several uneventful shunt revisions and orthopedic interventions under general anesthesia and had no history of atopy or drug hypersensitivity.
Surgery with a 2-stage approach was planned to treat progressive scoliosis: first, an anterior release of the vertebral column through the thorax with subsequent halo traction, followed by a posterior instrumented fusion. The complete procedure (Fig. 1) was scheduled free of all latex-containing materials. Anesthesia was induced using thiopental, fentanyl, remifentanil, and atracurium and maintained with sevoflurane, fentanyl, and atracurium. For antibiotic prophylaxis, cefazolin was infused preoperatively. Thirty minutes into surgery, hydroxyethyl starch was administered. Twenty minutes later, his exhaled carbon dioxide concentration suddenly decreased with a simultaneous decrease in mean arterial blood pressure from 70 to 40 mm Hg. His heart rate increased from 80 to 140 bpm. Apart from low carbon dioxide levels in the capnogram, ventilation and oxygenation were without problems. Anaphylaxis was suspected, and the thorax was emergently closed. After surgical drapes were removed, the entire skin was cyanotic and extensive urticaria was present. Treatment included epinephrine, clemastine, methylprednisolone, and Ringer’s lactate solution with slow but complete recovery. The mental status of our patient before surgery was slightly reduced within the limits of myelomeningocele patients. There were neither changes in mental status nor neurological changes after surgery. Serum tryptase levels were 14.1 µg·L−1 at 1 hour and 103 µg·L−1 at 4 hours; later baseline values were normal.
Skin prick, intradermal tests, and immunoglobulin (IgE) determinations were performed with latex and all drugs used during the procedure. Only the latex skin test was positive. IgE was positive to latex, disinfection mix (Pax6, Phadia, now ThermoFisher, Uppsala, Sweden), and its component EO (Table 1). All other IgE determinations (succinylcholine, rocuronium, thiopental, morphine, and chlorhexidine) were negative. Skin tests were done with propylene oxide as a surrogate for EO and tried with EO by running albumin through an EO-sterilized infusion tube 4 times as previously proposed.7,8 A patch test with propylene oxide was positive, a skin prick test with the EO albumin solution was equivocal, and the intradermal test was positive. Because the surgery was completely latex free, EO was the most likely cause of the allergic reaction. The same procedures for latex-free surgery were successfully used in patients monosensitized to latex. As a consequence, in addition to latex free, an EO-free environment was planned for surgery. A team of specialists was formed including members of the Allergy, Anesthesia, Orthopedic Surgery departments, the nursing team, as well as the Sterilization, Procurement, and Law departments. First, all sterile material not only for surgery and anesthesia but also in the intensive care unit and on the wards and its respective sterilization method was identified. Forty of 66 items (60%) were identified to be EO sterilized. Some could be replaced by non–EO-sterilized material reducing this number to 27 items. Infusion solutions and blood products were found to be EO free. Most material used for surgery could be replaced, whereas most material needed by anesthesia and intensive care was sterilized with EO and could not be replaced. An attempt to find an analytic method to measure EO levels in the materials was unsuccessful; the detection limit of 0.1 mg per examined medical device was not sensitive enough for our purposes.
Monbaliu et al.9 reported that 34% of materials could be purchased directly and nonsterile from the suppliers and then gamma or steam sterilized. However, this approach may raise liability issues in case of an infectious complication. Our procurement department ordered material either sterilized by an alternative method, for example, gamma sterilized or unsterilized. However, unsterile products do not fulfill CE (European Union Standards) or United States Food and Drug Administration requirements, and without this prerequisite most companies declined liability. Therefore, only a small proportion of EO-sterilized materials could be ordered unsterile.
To reduce residual EO, a prolonged low-temperature hydrogen peroxide gas plasma sterilization schedule to resterilize all EO-containing material was implemented. Plasma sterilization has been proposed as a new sterilization technology for thermolabile materials, without leaving any toxic residues or generating harmful waste.10 All packages were opened and resealed under sterile conditions. Before plasma sterilization all covers and caps had to be removed, because hydrogen peroxide needs to reach all cavities. Sterilization of filters bears the risk that they would not function properly; however, it was possible to provide them all in a functional state.
In parallel, off-label treatment with the humanized monoclonal anti-IgE antibody omalizumab (Xolair®, Novartis Consumer Health Schweiz AG, Bern, Switzerland) was started.11 This procedure was approved by the ethical committee and voluntarily paid for by the patient’s health insurance company. Injections were started 2 months before another surgical attempt was made (Fig. 1). Every 2 weeks 300 mg omalizumab was injected, although the initial total IgE levels were far beyond the upper levels used to calculate the omalizumab dose for asthma treatment.
Surgery was performed 7 months after the incident (Fig. 1). All materials to be used in the peri- and postoperative period were latex free, and EO was reduced with the described strategies. To this date, the patient had received 4 doses of omalizumab and pretreatment with clemastine and ranitidine was given; however, no corticosteroids were administered because of bone surgery. Despite all preventive measures, the patient again suffered from an erythematous rash and a decrease in systolic blood pressure first slowly from 100 to 85 mm Hg within 2 hours followed by a decrease from 85 to 45 mm Hg 6 hours into surgery. The patient was treated accordingly and recovered completely. This 6-hour delay before the reaction was interpreted as a partial effect of premedication and a putative slow buildup of the eliciting allergen level. Despite this incident, the implemented measures were retained and in the following weeks the patient underwent 3 additional surgeries, including ventriculoperitoneal shunt revision, and the second stage posterior spine fusion (lasting 14 hours) without any allergic symptoms. For the shunt, a device not sterilized by EO had to be obtained.
EO is a widely used sterilizing agent. Due to its chemical reactivity, it is an irritant, a hapten, and is considered to be potentially carcinogenic.12 EO can react with various protein moieties because it is capable of alkylating amino, carboxy, hydroxy, and sulfhydryl groups, which makes it a strong sensitizing hapten. Mechanism of type I hypersensitivity reactions was shown by the capability of EO to react with human serum albumin, and IgE antibodies were directed to the complex formed by EO and human serum albumin. Poothullil et al.13 reported the first case of anaphylaxis from EO in 1975 in a patient during hemodialysis. In hemodialysis, EO as the cause of hypersensitivity has been well documented.2 Several cases of EO causing perioperative anaphylaxis in myelomeningocele patients have been reported.3–6 It was assumed that EO might have been underestimated as some reports suggested latex, while EO was not excluded as an allergen.3 Pittman et al.14 found a prevalence of 23% of EO-specific IgE among children with myelomeningocele. EO levels were highest among children who had multiple surgeries and those who had a ventriculoperitoneal shunt, especially if it had been revised several times. These shunts may also be sterilized by EO possibly resulting in IgE sensitization, and a prolonged exposure could potentially result in local inflammatory complications.
Our latex-sensitized patient suffered from anaphylactic shock due to EO during latex-free surgery. In the allergy-related literature, EO is considered a relevant perioperative allergen. Allergists experienced in diagnosing perioperative anaphylaxis to hidden allergens, such as latex, EO, chlorhexidine as well as antibiotics and muscle relaxants, routinely exclude these allergens, especially in high-risk situations, such as in myelomeningocele patients who have had multiple surgeries and those who had a ventriculoperitoneal shunt, especially if it had been revised several times.
Because scoliosis correction was necessary, preventive measures had to be established. There were no standard procedures describing preventive strategies in detail. Published strategies included most often rinsing with sterile saline,3,4,9,12,15 prolonged storage time, and aeration to reduce residual levels of EO. For rinsing, there are concerns about sterility due to possible contamination. Furthermore, because the materials cannot be again packed sterile, this processing has to be done shortly before surgery which is too time consuming.
To provide an environment nearly free of EO is a complex and time-consuming task because most materials used perioperatively are EO sterilized and there is no sensitive analytical method. We had considered not only materials used for anesthesia or surgery, but also materials used afterwards in intensive care and in an emergency. An interdisciplinary team had to identify all materials needed for the specific surgery and assure whether EO-sterilized materials could be replaced by non–EO-sterilized materials. Because more than 40% of the material was only available EO sterilized, we used plasma sterilization to reduce EO, which is suitable for heat-sensitive medical devices.10 We assumed that the radicals built during plasma sterilization would react with EO and water further reducing EO. The advantage of this method is that medical devices can still be packed and stored under sterile conditions. Unlike EO sterilization, no problematic residuals remain after sterilization because hydrogen peroxide dissociates into oxygen and water. The drawback of plasma sterilization is that it cannot be used for material containing cellulose and sterilizing long, narrow-lumened tubes, like endoscopes, causes difficulties.10,16 Additional rinsing with physiological saline before plasma sterilization was also considered and tested, but when the material was not completely dry the plasma sterilization process was automatically interrupted. Furthermore, there was the risk that some filters would no longer properly function. Finally, all required EO-sterilized materials were resterilized for safety reasons by the prolonged plasma sterilization schedule we described.
Because EO-free conditions could not be guaranteed, further preventive measures had to be considered. Due to concern of the life-threatening anaphylaxis, treatment with omalizumab was initiated. The off-label use of omalizumab in this case was based on a report when omalizumab therapy was successfully administered for severe peanut allergy with recurrent anaphylaxis11 and experience in patients reacting to immune therapy for hymenoptera anaphylaxis. Later reports have shown that in severe IgE-mediated reactions omalizumab may have a protective effect.17,18 We still believe, however, that the best option is allergen avoidance, if at all possible. Since EO is present in such low concentrations, this additional preventive measure proved to be useful. Treatment with omalizumab was monitored by measurement of total, EO-, and latex-specific IgE levels and skin test reactivity to latex.19 However, 6 weeks after the start of omalizumab treatment, IgE levels neither decreased nor was allergen skin test reactivity suppressed. Technical issues such as measurement of IgE, which might have not only included free IgE, but also IgG–anti-IgE complexes, could have been the reason for high IgE levels. In contrast to our findings, suppression of a skin test response 98 days after administration of omalizumab was reported.19 A reason for the different results could be the prolonged administration of omalizumab.
In anticipation of emergencies, a nearly EO-free surgical environment was established. Expiration dates and the size of some materials had to be continuously kept up-to-date. For instance, tracheal tubes had to be adjusted to the growth of the patient.
In conclusion, creating an EO-free environment is complex and time consuming, needs interdisciplinary cooperation, and has a considerable economic impact. To reduce residual EO, plasma sterilization was used but its effect could not be validated. Because a completely EO-free environment could not be guaranteed, an off-label treatment with omalizumab was initiated and the patient was premedicated with a H1 and a H2 histamine receptor antagonist. With all these measures, subsequent surgeries were uneventful.
1. Moneret-Vautrin DA, Mertes PM. Anaphylaxis to general anesthetics. Chem Immunol Allergy. 2010;95:180–9
2. Ebo DG, Bosmans JL, Couttenye MM, Stevens WJ. Haemodialysis-associated anaphylactic and anaphylactoid reactions. Allergy. 2006;61:211–20
3. Bache S, Petersen JT, Garvey LH. Anaphylaxis to ethylene oxide - a rare and overlooked phenomenon? Acta Anaesthesiol Scand. 2011;55:1279–82
4. Moneret-Vautrin DA, Laxenaire MC, Bavoux F. Allergic shock to latex and ethylene oxide during surgery for spinal bifida. Anesthesiology. 1990;73:556–8
5. Goeters C, Theissen JL, Kästner H, Brunner W. An anaphylactic reaction under anesthesia due to a combined latex and ethylene oxide allergy. Anasthesiol Intensivmed Notfallmed Schmerzther. 1993;28:326–9
6. Demontoux V, Moreau E, Peyron I, Talon V, Khouri N, Morin JP, Bézie Y. Surgery in a patient allergic to ethylene oxide. Pharm World Sci. 2008;30:208–9
7. Birnie AJ, English JS. Ethylene oxide allergy may have been confirmed by patch testing to a similar epoxy compound. Contact Dermatitis. 2006;55:126
8. Dolovich J, Bell B. Allergy to a product(s) of ethylene oxide gas: demonstration of IgE and IgG antibodies and hapten specificity. J Allergy Clin Immunol. 1978;62:30–2
9. Monbaliu D, Van Breussegem A, Onsia A, Vandermeersch E, Segers C, Meert W, Kochuyt AM, Pirenne J, Claes K. Ethylene oxide allergy in patients on hemodialysis waiting for kidney transplantation: logistical nightmare or challenge? A case report. Transplant Proc. 2010;42:4375–7
10. Diab-Elschahawi M, Blacky A, Bachhofner N, Koller W. Lumen claims of the STERRAD 100NX sterilizer: testing performance limits when processing equipment containing long, narrow lumens. Am J Infect Control. 2011;39:770–4
11. Riffelmann F. Anti IgE therapy (off label use) in severe peanut allergy with recurrent anaphylaxis. Pneumologie. 2008;62:337–9
12. Bousquet J, Michel FB. Allergy to formaldehyde and ethylene-oxide. Clin Rev Allergy. 1991;9:357–70
13. Poothullil J, Shimizu A, Day RP, Dolovich J. Anaphylaxis from the product(s) of ethylene oxide gas. Ann Intern Med. 1975;82:58–60
14. Pittman T, Kiburz J, Steinhardt G, Krock J, Gabriel K. Ethylene oxide allergy in children with spina bifida. J Allergy Clin Immunol. 1995;96:486–8
15. Kaku N, Tsumura H, Kataoka M, Taira H, Torisu T. Influence of aeration, storage, and rinsing conditions on residual ethylene oxide in freeze-dried bone allograft. J Orthop Sci. 2002;7:238–42
16. Rutala WA, Weber DJ. Clinical effectiveness of low-temperature sterilization technologies. Infect Control Hosp Epidemiol. 1998;19:798–804
17. Galera C, Soohun N, Zankar N, Caimmi S, Gallen C, Demoly P. Severe anaphylaxis to bee venom immunotherapy: efficacy of pretreatment and concurrent treatment with omalizumab. J Investig Allergol Clin Immunol. 2009;19:225–9
18. Bell MC, Jackson DJ. Prevention of anaphylaxis related to mast cell activation syndrome with omalizumab. Ann Allergy Asthma Immunol. 2012;108:383–4
19. Corren J, Shapiro G, Reimann J, Deniz Y, Wong D, Adelman D, Togias A. Allergen skin tests and free IgE levels during reduction and cessation of omalizumab therapy. J Allergy Clin Immunol. 2008;121:506–11