The 5-Hydroxytryptamine3 (5-HT3) receptor antagonists (ondansetron, granisetron, dolasetron and tropisetron) are powerful, well tolerated antiemetic drugs for which very few hypersensitivity reactions have been described (1–6).
In October and November 2003, a 42-year-old man presented two episodes of isolated and generalized urticaria a few hours after chemotherapy for pancreatic adenocarcinoma, consisting of gemcitabine, oxaliplatine, ondansetron, alizapride chlorhydrate and methylprednisolone. The first episode appeared 12 h after the first course and the second 30 min after the second ondansetron injection (other drugs had not yet been given). For the third course of chemotherapy, ondansetron was replaced by metopimazine and a high dose of methylprednisolone. Hypersensitivity reaction did not occur following this replacement, although nausea and vomiting were severe. Before these episodes, the patient had never been given ondansetron. The patient had no other medical history, in particular other drug allergy, apart from an intermittent allergic rhinitis (because of plane tree and birch) for about 8 years. Since Levine demonstrated for penicillins that IgE-dependant reactions could occur up to the 72th hour after drug administration (7), an IgE-dependant mechanism is always suspected when a typical clinical pictures of drug hypersensitivity occurs within the first 24 h after drug intake.
To confirm the presence of IgE antibodies specific for ondansetron, skin test was performed with this drug and two other 5-HT3 receptor antagonists. Prick tests to ondansetron (pure concentration: 2 mg/ml), dolasetron (20 mg/ml) and granisetron (1 mg/ml) were all negative. Intradermal tests (IDT) to ondansetron proved positive at 0.2 mg/ml, with a wheal at 30 min of 2.5 times that of the initial papule. The IDTs tested negative to dolasetron and granisetron at 1/10 and at full concentrations. As these drugs are injected in short infusion, we choose to test up to pure concentration. Seven normal controls were tested and proved negative for IDTs to ondansetron at 0.2 mg/ml (wheals under 1.1 times that of the initial wheal); one emerged as positive at 2 mg/ml (wheal of 1.5 times). Because the patient required 5-HT3 receptor antagonists, we challenged him the same day with granisetron. He received intravenously at 30-min intervals and under strict hospital surveillance increasing doses of granisetron, starting with 1 mg up to the therapeutic dose of 100 mg. He consequently failed to experience any hypersensitivity reaction and could thereafter complete his chemotherapy trial.
The 5HT3-induced anaphylaxis has been described several times (1–6) and considered as a class effect (6) until a recent report (5) where the authors described an IgE-mediated allergy to granisetron and safe use of ondansetron. Both ondansetron and tropisetron share an indole heterocycle, however ondansetron and granisetron do not and this may possibly explain the absence of cross-reactivity between the latter two drugs. Considering the positive skin tests in our patient, the negative skin tests in seven controls and negative histamine release test in one case report (5), ondansetron does not appear to be a nonspecific histamine-releaser. Our case report underlines the importance of a thorough drug allergy work up.