I read with great interest the article by Armentia et al.1 recently published in Allergologia et Immunopathologia on food anaphylaxis in antiphospholipid syndrome (APS) and thrombosis. The authors suggested that seed lipoproteins might have a potential role in the APS and related thrombosis. I think that there are certain issues, which need clarification by the authors.
The first issue in need of explanation is about the randomisation of the patients. The authors mentioned that 52 patients with anaphylaxis were selected randomly from among patients with severe anaphylaxis. The number of patients with severe anaphylaxis (grade III–IV by Muller) is not clear. If the authors made this selection from among 21,879 patients, we could say that approximately, 28% of 21,879 patients may have thrombosis. This implies that development of thrombosis in patients with anaphylaxis is very high in these group patients. However, there is no clear evidence in the literature showing increased development of thrombosis in patients with anaphylaxis. Therefore the authors should give us the number of patients with severe anaphylaxis and clear information about randomisation. It seems to me that the selection of patients was made according to whether the patients have thrombosis or not. This could have led to a selection bias. The authors reported four cases with food anaphylaxis followed by thrombosis before.2 Were these patients included in this study?
Secondly, in this study, of 52 patients with anaphylaxis, 15 (28%) had ACA IgG positivity with medium or high titre. The authors described medium or high titre as 20–80 GPL medium, >80 GPL high positive, respectively. In Table 1A, the number of patients having positive ACA IgG with medium or high titre were only seven and one in Table 1B. As a result, unlike that mentioned by the authors in the text and Tables, only eight patients had ACA IgG antibody positivity with medium or high titre (15%).
Thirdly, it is seen in the Tables that vascular involvement exists in patients with anaphylaxis and the patients with APS. It is likely that some of these patients may use aspirin for the secondary prophylaxis for thrombosis. As is well known, aspirin may enhance wheat anaphylaxis or skin prick test results.3–5 Although this effect of aspirin may be seen in exercise-induced anaphylaxis, aspirin can be also an augmentation factor in wheat allergy without exercise.3–5 For this reason, it would be necessary to know the number of the patients using aspirin in both groups. In explaining their results, the authors should have mentioned this as a confounding factor.
Fourthly, there is no clear explanation as to what criteria were used to select the patients with APS in the text. The authors re-tested ACA IgG two months after initial testing. According to classification criteria for APS published in 2006,6 antibodies should be re-evaluated after three months to make a clear decision in patients with antibody positive. Why did the authors evaluate ACA IgG antibodies after two months? And did they use second antibody titres in their analysis.
In conclusion, we need well-designed studies investigating whether or not anticardiolipin antibody positivity and thrombosis are secondary to food allergies or a coincidence.
