DISCUSSION
To our knowledge, few studies have focused on describing clinical features of ACEi-AE attacks in a cohort of patients with a diagnosis defined and confirmed by follow-up. Our results can be summarized as follows: (1) overdiagnosis of ACEi-AE is frequent at baseline, with at least a third of patients later reclassified as having histaminergic AE; (2) in our probable ACEi-AE population, attacks occurred mostly in elderly men, with a highly variable delay from ACEi introduction, and preferential involvement of oral cavity and throat; (3) classical histaminergic features (such as history of allergy and atopic conditions, attacks with preferential evening onset and spontaneous resolution < 24h) seemed non-rarely reported in ACEi-AE patients; (4) ACEi-AE attacks appear refractory to histamine-mediated AE treatments (antihistamine, corticosteroids and adrenalin) and sensitive to bradykinin-mediated AE treatments (tranexamic acid, icatibant and C1-INH).
Diagnosis of ACEi-AE is usually suspected in patients presenting with an AE attack, no urticaria and exposure to ACEi. However, idiopathic histaminergic AE, a frequent condition, can also occur in patients treated by ACEi; and there is no definitive way to discriminate between these 2 mechanisms at referral. This is a major issue, as it yields important therapeutic consequences (treatment modalities of AE attacks, unnecessary eviction of ACEi in case of misdiagnosed histaminergic AE…). Follow-up can help better differentiate these 2 mechanisms, as occurrence of urticaria and/or AE attacks after 6 months of ACEi discontinuation pleads for histaminergic AE. Using follow-up data, our study suggested that at least a third of patients referred for a suspected ACEi-AE were actually misdiagnosed idiopathic histaminergic AE. This highlights the relevance of performing systematic follow-up in these patients.
In order to avoid mistakenly-included histaminergic AE, we used a strict definition of ACEi-AE, using long-term follow-up. In this population, most of our results supported previous data from other studies with different designs (Table 2 ) (4,8,11–23): male predominance (although this is not consistently reported), elderly population, highly variable delay between ACEi introduction and first attack, preferential involvement of oral cavity (especially lips and tongue) and throat.
Interestingly, our data suggest that patients with probable ACEi-AE can present with classical histaminergic features. History of allergy and atopic conditions were reported in at least 20% of patients, although this stays within range of the prevalence of atopy in the general population (24). AE attacks with preferential evening onset and spontaneous resolution <24h occurred in about 25% of our patients. Previous studies have yielded heterogeneous results regarding the duration of untreated ACEi-AE attacks, which can be explained by the difficulty to accurately delineate the beginning and end of an attack (notably onset of symptom relief vs complete resolution) and the possible inclusion of idiopathic histaminergic AE(4,17,19–21). Nonetheless, spontaneous resolution <24h have previously been reported (4,23). Only one work has previously studied the time of day of ACEi-AE onset, suggesting that attacks frequently occurred outside regular work hours (11). ACEi-AE consists in bradykinin-mediated attacks, which are traditionally described as occurring at any time of day and lasting at least 24h. However, this classical semiology description mainly stems from hereditary AE patients; and these characteristics may not all apply to other types of bradykinin-mediated attacks (25). Overall, these data suggest that clinical features alone seem insufficient to discriminate between ACEi-AE and histaminergic AE at referral; and highlight an unmet need for diagnostic biomarkers.
Interestingly, our study confirmed that icatibant, a bradykinin receptor antagonist, seems effective in ACEi-AE attacks. Data regarding the efficiency of icatibant in this indication are conflicting, with one randomized clinical trial favoring treatment (20) and two others observing no difference from placebo (18) In line with other real-life data (15–17), our results plead for using icatibant in ACEi-AE patients, as now recommended in some expert guidelines (9).
Our study draws strength from its originality, as it uses an innovative and strict definition of ACEi-AE and focused heavily on the follow-up of a suspected ACEi-AE population to ascertain the diagnosis. It also has some limitations that mainly stem from its retrospective design, missing data and small sample size. Excluding patients with no follow-up data may have biased the study population; however, since it is a common practice in our department to perform at least one systematic follow-up visit to patients referred for ACEi-AE, we believe this had only a limited effect.