Background
Angioedema is a swelling of the deeper layers of the skin and mucous membranes subsequent to blood vessel dilation and increased vascular permeability induced by vasoactive mediators such as histamine and bradykinin. Angioedema can be life-threatening when involving the upper airways or very debilitating when involving the gastrointestinal tract. Most cases of angioedema are non-hereditary in nature [
1]. A significant share is mediated by mast-cell mediators such as histamine, but a substantial proportion are unrelated to mast-cell activation and mediated by bradykinin (Bk-AE). Bradykinin is a potent vasodilator, increases vascular permeability, and mediates pain [
2‐
5]. While mast cell -mediated angioedema can be successfully treated with antihistamines and glucocorticosteroids and with omalizumab as a prophylactic treatment [
6], Bk-AE requires interventions that target the synthesis or receptor activity of bradykinin. This distinction is essential in life-threatening situations such as swellings of the upper airways where the use of the wrong therapy translates into unnecessary suffering and the risk of suffocation.
Bk-AE results from a variety of circumstances, such as a slow-down in bradykinin degradation subsequent to the use of angiotensin converting enzyme inhibitor (ACEI-AE), or the uncontrolled generation of bradykinin, such as observed in hereditary angioedema (C1-INH-HAE) and recurrent angioedema due to acquired C1-Inhibitor deficiency (C1-INH-AAE) [
7] (Table
1). ACE is critical in the degradation of bradykinin, which is hypothesized to accumulate excessively in some patients taking ACEI for their hypertension. Uncontrolled generation of bradykinin, on the other hand, occurs when complement-1 inhibitor (C1-INH) is either deficient or non-functioning. The physiological role of C1-INH is to control the generation of bradykinin after activation of the contact system. C1-INH-HAE is inherited as an autosomal dominant mutation that results in low levels of C1-INH (type I) or non-functioning C1-INH (type II). An increasing number of patients are also reported to have Bk-AE of hereditary nature, despite normal levels of C1-INH [
8,
9]. Mutations of factor XII, and more recently of plasminogen and angiopoietin-1 were found in a small subgroup, but the majority lack clear defining criteria [
10,
11]. HAE with normal C1-INH is not covered by our study. C1-INH-AAE is acquired as a result of C1-INH depletion predominantly in association with lymphoproliferative disorders.
Table 1
Classification of angioedema (adapted from Craig et al. [
7])
C1-INH Deficiency / Defect | C1-INH Normal |
Inherited | Acquired | Inherited | Acquired |
HAE-1 & HAE-2 | C1-INH-AAE | HAE-3 | ACEI-AE |
ACEI-AE is a relatively new phenomenon that is largely determined by the use of ACEI in the population and appears to be twice as likely in patients of African ancestry [
12]. The prevalence of C1-INH-HAE should be determined by the incidence of spontaneous mutations, the mortality of the disease, and the average numbers of children of C1-INH-HAE patients. In other words, a stable prevalence would suggest equilibrium among the birth-rates and mortality rates of patients with spontaneously developed and familial C1-INH-HAE.
Prevalence and incidence rates of ACEI-AE and C1-INH-HAE are frequently found in reviews that cover these conditions. For example, the typical prevalence for C1-INH-HAE reported is 1 in 50,000 [
13,
14], although the evidence of this has not been systematically assessed. In the present study we report the results of a systematic review of the epidemiological literature on BK-AE and make an attempt of providing evidence-based estimates of its expected prevalence. These estimates are critical in raising awareness among physicians about the differential diagnosis and expected frequency of bradykinin-mediated angioedema [
15] and they may help to promote the use of the available targeted treatment approaches that are required for correct medical management of these patients.
Methods
Two separate search strategies were deployed to obtain publications on the epidemiology of (1) ACEI-AE and (2) C1-INH-HAE and C1-INH-AAE. The following databases were searched:
The search strategy for ACEI-AE in Medline and EMBASE was the following: exp. angioedema/ AND (angiotensin:.tw. OR ace.tw.) AND (prevalence.tw. OR incidence.tw. OR epidemiol:.tw.), with subsequent deduplication. The search strategy for C1-INH-HAE and C1-INH-AAE was as follows: (*Angio(o)edema/cl, ep, pc OR (hereditary adj angio(o)edema).tw. OR *angio(o)edema, hereditary/ OR (quincke adj angio(o)edema).tw. OR ((acquired adj angio(o)edema).tw.) AND (prevalen: or inciden: or epidemiol:).tw. with subsequent deduplication of references from both databases. The CRD database was searched using the text word combination “angioedema” and “angiotensin” for induced angioedema, and “hereditary angioedema” for C1-INH-HAE.
The titles and abstracts of all references retrieved through the searches were independently scanned by all four authors and had to meet the following criteria to be included: (1) For ACEI-AE the publication had to be a systematic review of randomized controlled trials of ACEI or an inception cohort of new ACEI users; (2) for C1-INH-HAE or C1-INH-AAE the publication had to contain an original epidemiologic data collection in patients with C1-INH-HAE (Type I or II) or C1-INH-AAE, using a survey or sampling strategy that allowed a comprehensive accounting of affected individuals within a defined national area. Discrepancies in the assessments of evaluators were resolved by consensus measures.
To determine the prevalence of ACEI-AE at the population level, a subsequent literature search of Medline and EMBASE was conducted to identify National population-based surveys from, the US, France and Germany that provided information about the proportion of the population being treated with anti-hypertensives and among those, the proportion of patients being prescribed ACEI. Countries representative of large populations and different proportions of ACEi use among antihypertensives were exemplarily selected. The incidence of ACEI-AE from systematic reviews of randomized controlled trials or inception cohort-based epidemiological was then combined with population-based estimates of the use of anti-hypertensives to derive population-based estimates of the incidence or annual prevalence of ACEI-AE. This research is exempt from IRB review.
Discussion
We summarized the studies reporting on the annual prevalence of ACEI-AE, and prevalence of C1-INH-HAE and C1-INH-AAE. The prevalence of ACEI-AE is determined by the country-specific use of antihypertensive; in the US, the prevalence was estimated to vary between 0.7 and 1.7 per 10,000 inhabitants when weighting the risk of ACEI-AE among Caucasians and African-Americans. In Germany it was estimated to vary between 1.0 and 2.6 per 10,000 while in France it was estimated to be lower and vary between 0.4 and 1.0 in 10,000. The prevalence of C1-INH-HAE was estimated to vary between 1.1 and 1.6 per 100,000. The prevalence of C1-INH-AAE was estimated to be 1.5 per 1 Million from the one available epidemiological investigation of C1-INH-AAE in Denmark.
Overall, the prevalence figures would be approximately 1.5 per 10,000 for ACEI-AEs, 1.5 per 100,000 for C1-INH-HAE and 1.5 per 1 Million for C1-INH-AAE. For every 100 patients with ACEI-AE there would be 10 patients with C1-INH-HAE and 1 patient with C1-INH-AAE. However, patients with C1-INH-HAE have on average 1 angioedema attack per month [
25]. For this reason, the number of annual angioedema attacks due to C1-INH-HAE and ACEI use should be similar, given that ACEI are typically withdrawn at the time of diagnosis and re-occurrence is less likely. For a country with 100 Million inhabitants there would be approximately 31,500 bradykinin mediated angioedema attacks each year, calculated by multiplying the prevalence standardized to 1 million, 150 for ACEI-AE, 15 per for C1-INH-HAE, and 1.5 for C1-INH-AAE, with 100 for each entity, and multiplying the resulting figures with an annual attack rate of approximately 1 per year for ACEI-AE and 10/year for C1-INH-HAE and C1-INH-AAE.
Treatments for histamine-mediated angioedema are inappropriate for bradykinin-mediated angioedema. These numbers highlight the significant need to recognize the nature of the underlying cause in each patient who represents to the emergency room with signs and symptoms of an angioedema attack.
In the absence of confirmatory investigations, many publications referenced the prevalence of C1-INH-HAE to be 1 in 10,000 to 1 in 50,000. Several studies have now reported the population-based prevalence of C1-INH-HAE. These studies were supported by intense efforts of patient societies and providers to register patients in order to provide access to adequate treatment and also to raise awareness and shorten the delay to a correct diagnosis. We have summarized these studies and determined that the prevalence of diagnosed cases of C1-INH-HAE is approximately 1.5 per 100,000 (or ~ 1 in 67,000).
A population-based estimate of ACEI-AEs has not been published before. The annual (1st-year) cumulative incidence of angioedema among ACEI users is fairly stable, as the meta-analyses of the randomized controlled trials and the database analyses suggest. However, the population-based prevalence is necessarily variable and depends on the actual use of ACEI in the population, how this use was measured, and the demographic mix, especially with respect to the representation of African-Americans. For this reason we were only able to provide approximate estimates with potentially large variation. Other antihypertensives addressing the renin-angiotensin system such as angiotensin-receptor blockers or direct renin inhibitors have not been conclusively linked to increased risk of AE. Any incremental risk of AE in users of such drugs would translate into a higher prevalence in the population. Also, because the ratio of patients with ACEI-AE vs. C1-INH-HAE is approximately 10:1, we calculated that the expected number of angioedema attacks is similar, due to the frequency of attacks among C1-INH-HAE patients.
We searched both the Medline and EMBASE database using appropriate keywords to select epidemiological investigations and systematic reviews. A cross-check of the CRD database confirmed that no systematic reviews of ACEI were missed. The C1-INH-HAE-associated literature is not as extensive as can be seen from the overall number of search hits, because the disease is rare and is covered by a small community of researchers. Because there were not so many publications, we did not attempt to quantitatively combine the estimates, but rather describe each individual study. This will enable the readers to come to their own conclusion regarding the manifested prevalence of C1-INH-HAE or ACEI-AE in their own setting.
It appears that the epidemiology of C1-INH-HAE is stable across the geographic regions covered in this review. This may be the result of a stable mix of spontaneous mutations and inherited patterns [
30,
31]. It is likely that the prevalence of 1 in 67,000 provides a reference ceiling to the estimation of diagnosed patients with C1-INH-HAE in a specific country, given the current pattern of detection in countries with modern healthcare systems. This number might, therefore, be helpful to policymakers in estimating the number of patients expected in their constituencies. Moreover, as investigations of family members of affected patients have shown [
32], there still persists a percentage of undiagnosed family members. Further efforts in raising disease awareness, educating providers and improving access to appropriate therapies may shorten the current lagtime of 10 [
33] to 16 [
25] years between onset of first symptoms and date of diagnosis.
There are several potential clinical implications of our results: ACEI-AE may currently be underdiagnosed in some populations. Up until 2017, the German national database on suspected adverse drug reactions listed about 2000 angioedema incidents in connection with ACEI use while prescription rates of ACEI-AE suggest a considerably larger number [
34]. This points to a gap in the diagnosis of angioedema among ACEI users and highlights the need for expanding the knowledge on clinical findings, course and potential therapy of acute ACEI-AE in the medical community. Vice versa, there are factors that promote potential misdiagnosis of C1INH deficiency cases as ACEI-AE cases: ACEI use is quite common in many populations, At the same time, ACEI are a common precipitating factor of acute angioedema in patients with C1-INH deficiency. In undiagnosed C1-INH deficiency cases, an angioedema episode in the presence of ACEI therapy would most likely be attributed to the ACEI. This might lead to a, probably minor, effect on prevalence numbers of C1-INH-HAE and C1-INH-AAE. The recognition of an ACEI-AE should prompt the exclusion of C1-INH deficiency as in diagnosed C1-INH deficiency cases investigation of family members is warranted. To characterize clinical differences in these different entities more deeply is an unmet need and will be the goal of future efforts.
Limitations of the study include its inherent restriction to US and European data, while the epidemiology of Bk-AE might be different in other parts of the world. Also, the prevalence of ACEI-AE may be higher as estimated in our analysis as in some of the studies patients with previous angioedema were excluded [
12,
18]. A few trials did not report incidence of angioedema, as adverse effects were only reported if the incidence was > 1% [
12]. Another limitation is the low number of prevalence studies in Europe, for example but not restricted to Germany and France that a lack systematic nationwide data collection. Additionally, the present analysis is focused on ACEI-AE and C1-INH-HAE. Further efforts to assess the epidemiology of additional types of HAE with normal C1-INH, for example HAE due to a FXII mutation (FXII-HAE), are needed. It has yet to be determined if angioedema attacks in patients with HAE due to a novel plasminogen mutation (HAE-PLG), HAE with a novel angiopoietin-1 gene mutation (ANGPT-1-HAE) or HAE of unknown origin (U-HAE) are bradykinin mediated, in which case future analysis of Bk-AE epidemiology should include these novel types. A further limitation of the present analysis might be the fact that the studies indentified essentially covered the last two decades. As awareness of Bk-AE increased considerably over time, there is a chance that older prevalence data may underestimate the real prevalence. Continued efforts in assessing the prevalence of Bk-AE are therefore warranted.