Sirs, In their letter,1 Casteele et al. agree with our findings that acute severe infusion reactions to infliximab (IFX) are associated with anti-IFX antibody formation, and that risk of reactions is particularly high at second infusion during episodic therapy.2
However, they question the prevalence of reactions necessitating discontinuation of IFX in our tertiary single centre cohort.2 Previously, reported prevalences have varied from 0% to about 6%.3–13 We observed reactions in 8% of patients. There may be several explanations for this apparent discrepancy.
First, different definitions of reactions; acute severe reactions have been defined as acute reactions leading to IFX discontinuation,7, 8, 14 acute reactions judged life-threatening,3 acute reactions leading to hospital admission3, 6 and acute reactions resulting in drop in blood pressure and dyspnoea.5, 9–11, 15 Due to the retrospective nature of our study, we only had limited data available, and we defined acute severe infusion reactions as reactions occurring during IFX infusion, which were judged severe by the treating physician, and resulted in immediate and permanent IFX cessation.
Secondly, treatment regimens may not be comparable between studies; we found a significantly higher proportion of reactions during episodic treatment when compared with continuous treatment (17% vs. 3%; P < 0.001). However, treatment regimen is not stated in the vast majority of studies, and the definition of episodic therapy may differ between studies.
Finally, the nature of data collection (prospective vs. retrospective) may differ, as may number of infusions per patient, which is usually not stated, and the definition of prevalence (events compared with number of patients treated with IFX or compared with number of infusions administered).
Despite all our patients receiving medical prophylaxis (steroid and antihistamine) prior to IFX, prevalence of reactions was still high indicating a negligible effect of this intervention, and contradicting the clinical practice described by Casteele et al.