Latex Anaphylaxis after Tourniquet Release during Total Knee Arthroplasty

A 63-yr-old woman with a history of asthma was scheduled for total knee arthroplasty. The preoperative interview revealed no allergy to medications. However, the patient described an incident of severe generalized edema after ingesting citrus fruit. Previous operations included appendectomy with general anesthesia, knee arthroscopy, and bunionectomy with epidural anesthesia. In each case, the intervention had gone smoothly. Medications included alprazolam and cetirizine hydrochloride for allergic rhinitis.

This article is accompanied by an Editorial View. Please see: Moss J: Allergic to anesthetics. Anesthesiology 2003; 99:521-3.

Following placement of a femoral catheter for continuous postoperative analgesia, general anesthesia was induced using midazolam (0.05 mg/kg), propofol (3.3 mg/kg), and sufentanil (0.4 μg/kg) and maintained with sevoflurane (1.1 minimal alveolar concentration) and sufentanil (0.4 μg · kg⁻¹· h⁻¹). Antibiotic prophylaxis consisting of cefamandole was administered a few minutes after orotracheal intubation. A latex-free tourniquet was placed around a cloth band applied around the proximal portion of the thigh and inflated.

After 1 h 18 min of uneventful anesthetic course, the tourniquet was deflated to check for hemostasis. Four minutes thereafter, the patient had signs of cardiovascular collapse with systolic arterial pressure of 40 mmHg, tachycardia (heart rate, 130 beats/min), severe bronchospasm with a Paco²of 21 mmHg, and an oxygen saturation measured by pulse oximetry of 88% despite manual ventilation with 100% fractional inspired oxygen tension. Redness of the face and neck was noted. Persistence of low systolic blood pressure (50 mmHg) and bradycardia (heart rate, 37 beats/min) despite infusion of 500 ml crystalloids (lactated Ringer's solution; Baxter Viaflex, France) prompted intravenous injection of 2 mg epinephrine. These measures gradually improved the patient's hemodynamic conditions. Three and 5 min after the injection of epinephrine, systolic blood pressure increased to 75 and 80 mmHg, respectively, and heart rate decreased from 135 to 110 beats/min with a return to normal tidal volumes, end-tidal CO², and oxygen saturation measured by pulse oximetry. A continuous infusion of epinephrine (0.25 μg · kg⁻¹· min⁻¹) was started 5 min after the onset of shock, and 60 mg intravenous methylprednisolone was administered. The surgeon ensured that the field of operation was completely sealed off in a latex-free environment. At 30 min, blood samples for the initial atopic appraisal were obtained. This workup included determination of plasma concentrations of tryptase and histamine as well as immunoglobulins (immunoglobulin E) specific for latex. In the immediate postoperative period, the patient was extubated without problem. The continuous infusion of epinephrine was stopped within 6 h. The patient was discharged 4 days after the operation.

The anaphylactic nature of the shock was confirmed by analysis of the perioperative samples, which showed an increase in plasma concentrations of tryptase (immunoenzymatic dosage, UniCAP Tryptase System; Pharmacia, Paris, France) and histamine (RIA Histamine®; Immunotech, Paris, France). The tryptase concentration was 26.4 μg/l (normal value, < 12 μg/l), and the histamine concentration was 68.2 nm (normal value, < 9 nm). Latex-specific immunoglobulin E (immunoenzymatic dosage, CAP Fluorescence System®; Pharmacia) was 2.39 kU/l (normal value, < 0.35 kU/l; class 2: moderate).

At a follow-up visit 6 weeks later, the patient reported an episode of edema of the lips after eating an avocado several years earlier. Interview found no occupational exposure to latex and no history of anaphylactoid reaction to domestic products containing latex.

The patient did not react to a prick test with 0.4% phenol in saline (negative control) but had a wheal of 6 mm with a prick test with 9% codeine phosphate (positive control). Prick tests with latex were performed using two different commercial fresh natural rubber latex extracts (Stallergene®[Stallergenes Laboratories, Antony, France] and Allerbio®[AB Laboratories, Varennes-en-Argonne, France]). Both extracts yielded a positive reaction. After 15 min, Stallergene® yielded a wheal of 8 mm, and Allerbio® yielded a wheal of 5 mm. A second test for latex-specific immunoglobulin E was performed, showing a high specific immunoglobulin E concentration of 12.4 kU/l (class 3). To reinforce the suspected diagnosis, an in vitro  study of basophil degranulation was also performed. This method uses flow cytometry to quantify the amount of CD63 induced by various antigens. 6The value of basophil degranulation was 16% for latex with 1 μl Stallergene® solution and 17% for latex with 10 μl Stallergene® solution. This value was 26% for avocado with 10 μl Stallergene® solution. No degranulation was induced by a kiwi antigen. No reaction was observed during antigenic stimulation with dust mites or grass seeds. Intradermal tests were performed with the other anesthetics used for the patient, but results of these tests were negative.

Despite that the release of metabolically active products and mediators from the acutely ischemic leg and acute loss of volume may have contributed to hypotension, the diagnosis of anaphylactic shock reaction to latex on removal of the tourniquet was quite likely. It was based on substantial clinical and biologic evidence, including positive results of skin tests. A female predominance of anaphylactic shock during anesthesia has been reported. 7The patient's history of fruit allergy and atopic disease should be emphasized. 8,9When the limb was revascularized, systemic dispersal of the putative antigen was immediately followed by a severe allergic reaction (grade III) in this patient. She may have developed sensitivity to latex during previous surgical or occupational exposure. This has already been demonstrated for children 10and adults. 11An anaphylactic reaction has been estimated to occur in one third of all patients who have a known allergy to latex. 12The originality of the current case lies in the onset by cardiovascular collapse after tourniquet release.

Very similar events have been described on release of a tourniquet after applying rifamycin to an operative wound in patients who had previously been sensitized to rifamycin. 13,14In these observations, shock occurred ≈ 10 min after tourniquet release. Although the delay was shorter in the current case, the clinical picture was comparable. In the cases involving rifamycin, contact between the allergen and the operative site was extensive. In our case, surgical contact primarily consisted of contact with surgical instruments. There was brief contact with the surgeon's gloves during use of the trial prosthesis and application of cement. Consequently, the embolization of very few particles of latex after removal of the tourniquet was apparently sufficient to trigger the anaphylactic reaction. The same type of mechanism has been hypothesized in cases of obstetric patients who received oxytocin infusions subsequent to endouterine contact with latex gloves. 7,15–17The allergen was thought to have been abruptly forced into the circulation by the induced uterine contractions leading to the immunoglobulin E–dependent reaction.

The current case report illustrates the usefulness of a thorough preanesthesia interview with an exhaustive list of all foods documented to have cross-allergies with latex (including avocado, banana, kiwi, pineapple, passion fruit, etc 6,18,19). Despite the absence of glove-related skin symptoms in our patient, the cross-reactivity between latex and avocado is obvious. 20,21Recent studies have reported hevein-like protein domains responsible for major cross-reacting allergen reactions with avocado. 22,23Including a list of such food allergens in the preanesthesia questionnaire is recommended. 1Findings of this interview would permit efficient scheduling of an allergy workup adapted to the anticipated surgical procedure, possibly reducing the number of perioperative anaphylactic incidents. This report also provides support for the policy of immediately clearing an operative wound of latex in cases of anaphylactic shock to an agent that has not been clearly identified.