Incidence and predictors of cetuximab hypersensitivity reactions in a North Carolina academic medical center

Abstract

Previous research has indicated a high incidence of cetuximab hypersensitivity reactions in the southern US. This study documents the incidence of hypersensitivity reactions in North Carolina, and explores whether factors such as patient demographics, allergy history, premedications, and cancer type are potential predictors for cetuximab reactions. This retrospective chart review consisted of 72 consecutively treated patients from Duke University's Morris Oncology treatment center between September 2005 and August 2007. Data regarding stage of malignancy, premedications, and hypersensitivity reactions were collected from electronic databases. The number and type of reactions were characterized. Patients with and without hypersensitivity reactions were compared using bivariate statistics and multivariate logistic regression. Of the 72 patients, 21 (29%) experienced hypersensitivity reactions. The majority of reactions (62%) were grades III or IV. Of the 21 patients having a reaction, 9 (43%) were sent to the emergency room and 5 (24%) had an overnight hospitalization. Three hospitalized patients were admitted to the intensive care unit. Both male gender (p = 0.039) and head/neck cancer (p = 0.014) were related to an increased likelihood of hypersensitivity reaction in bivariate analyses. In multivariate analyses, controlling for demographics, allergy history, premedications, and cancer type/stage, only type of cancer was predictive of hypersensitivity reaction (colon vs head/neck OR = 0.177; 95% CI 0.036-0.858). This study confirms a high rate of cetuximab hypersensitivity reactions in a southern region of the US. Patients with head/neck cancers were significantly more likely to have hypersensitivity reactions than patients with colon cancers.