Background

Hereditary angioedema (HAE) is an inherited disorder characterized by unpredictable attacks of localized swelling in the skin and/or mucosa [1, 2]. HAE is most frequently caused by mutations in SERPING1, resulting in reduced production (type I HAE) or dysfunction (type II HAE) of the C1-inhibitor (C1-INH) protein. This leads to vasodilation, increased capillary permeability and swelling, mediated by bradykinin (which is generated by the contact activation system) [1, 3,4,5]. HAE attacks are recurrent and frequently affect the face, extremities, upper airway and abdomen [2, 4]. Even mild attacks may cause transient discomfort and disfigurement, whereas abdominal attacks can be so painful that they may mimic an acute abdomen (leading to inappropriate surgery), and swelling of the larynx can be fatal [1, 2]. Because of the unpredictability and potential severity of HAE attacks, it is recommended that all patients are evaluated for the need for long-term prophylactic treatments [6]. Several prophylactic therapies are available (Table 1), which either replace deficient C1-INH or inhibit kallikrein – a component of the contact system that catalyses the production of bradykinin. An additional and historical option to targeted therapies for long-term prophylaxis is attenuated androgen (AA) treatment. AAs, such as danazol, stanozolol and oxandrolone, have not been studied in large, randomized, placebo-controlled trials, and available data are from limited numbers of patients [11, 19,20,21]. Treatment effects can be highly variable, and although some studies support the efficacy of androgens [11, 19], others show suboptimal outcomes [22, 23]. AAs are associated with side effects in approximately 80% of patients in some studies [6, 19, 20, 24]. As outlined in Table 1, these side effects may take a variety of forms including, but not limited to, weight gain, hypertension, proatherogenic lipid profile changes, headaches, cramps, mood disturbances (such as depression and anxiety), acne, and polycythaemia [11, 21, 24,25,26,27]. Although the safety profiles of AAs are derived from studies in small numbers of patients, with the potential for the majority of treated patients to be affected, the use of AAs and patient monitoring must be carefully evaluated. Further, AAs may not be appropriate in female patients because of potential virilisation and menstrual irregularities, and are contraindicated during pregnancy because of possible virilisation of female fetuses [6, 21, 24,25,26, 28]. In children and adolescents, AA use is not appropriate because of potential effects on bone development [6, 29], and the potential risk of early puberty [30, 31]. AAs are contraindicated in several conditions such as cardiovascular diseases or cancer, and also with a large number of drugs [18]. As effective and better tolerated targeted options for long-term prophylaxis are approved or developed [32,33,34,35,36,37], the use of AAs is decreasing and AA discontinuation is becoming an increasingly used option or necessity because of side effects, contraindications and/or patient/physician preference. Although other treatment options may be preferred to androgens, the higher cost of the former may be a hurdle to their use in some countries and regions [36]. However, a US study in 2015 indicated that the proportion of physicians who specified a preference for long-term prophylaxis with danazol decreased from 56 to 23% between 2010 and 2013 [38, 39].

Table 1 Prophylactic treatments for HAE
Full size table

AA discontinuation can result in destabilization of control of HAE attacks on one side, and a withdrawal syndrome on the other, with mood disturbances, anxiety, depression, insomnia, fatigue, hypersomnia and a flu-like syndrome, although some of these symptoms have only been studied in populations receiving high doses of androgens [40,41,42,43,44,45]. Studies of AA withdrawal in HAE have not been extensively conducted. A survey of 12 physicians treating HAE has shown that physicians had patients who had experienced complications and/or side effects of AA discontinuation including fatigue and mood disturbances [45]. Surveyed physicians were also concerned with the potential for changes to attack rates. While potential strategies for AA withdrawal in HAE—tapering, overlapping with other therapies prior to tapering or stopping, and immediate switching—were suggested based on this physician survey and the broader literature on the use of endocrine treatments [45], these strategies have not yet been systematically compared in terms of patient outcomes and further work is needed to understand the impact of different strategies. Through this case series, we describe AA discontinuation in patients with HAE caused by C1-INH deficiency. We examine the challenges associated with AA discontinuation, present patient outcomes, and describe how treatment strategies need to be modified following AA discontinuation in order to further understanding of this topic.

Case series

Methods

An advisory board of leading European experts in HAE was convened to discuss AA discontinuation in patients with HAE, a topic of current interest because of the expanding landscape of targeted prophylactics for HAE. The experts agreed that a case series could highlight challenges of AA discontinuation to healthcare professionals who treat HAE, and raise considerations for how to manage the transition to alternative treatments. This review is a retrospective case series of patients with confirmed type I HAE who have discontinued or attempted to discontinue prophylactic treatment with AAs. Descriptive statistics only are provided. Written informed consent for publication has been provided by all patients, except for one patient who was deceased and for whom consent has been provided by next of kin. All patient data are anonymized, and direct identifiers are not included [46].

Patient characteristics

The cases of 10 patients with confirmed type I HAE who either discontinued or are discontinuing AAs are presented. More comprehensive case details are provided as an Additional file 1. Three patients were female; the age range was between 31 and 76 years (median = 51 years). Patient characteristics and details of AA doses are shown in Table 2. The most commonly used AA was danazol (n = 8). Prior to AA discontinuation, all patients underwent AA dose modifications or a change of AA type (Table 2 and supplementary information). Time on AAs prior to discontinuation ranged from 1.5 to 36 years (median = 16.5 years).

Table 2 Patient characteristics, AA treatment, reasons for discontinuation and discontinuation strategy
Full size table

Reasons for AA discontinuation and methods of AA discontinuation

The most common reason for AA discontinuation was the occurrence of side effects (n = 5; Table 2). Side effects included headaches, hypertension and weight gain, among others. Insufficient control of HAE attacks affected the decision to discontinue AAs in 3 patients, and 1 patient was assessed as no longer requiring prophylaxis. Contraindications were responsible for discontinuation in a further 2 patients, while an unplanned pregnancy, participation in a clinical trial and loss of access to medication were other reasons for treatment switches (2 patients experienced side effects and insufficient control of HAE attacks; 1 patient experienced side effects and had no ongoing need for prophylaxis). In 7 patients, AAs were discontinued immediately with no gradual dose reductions. Of the remaining 3 patients, 2 decreased danazol gradually while a targeted therapy (lanadelumab or pdC1-INH) was introduced and 1 discontinued gradually.

Control of HAE attacks after AA discontinuation

Outcomes of AA discontinuation are summarized in Fig. 1 and described for individual cases in Table 3.

Fig. 1
figure 1

Initial treatments after attenuated androgen discontinuation, and patient outcomes. Patients were provided with a range of treatments, including prophylactic and on-demand options. In several patients, HAE attacks were not adequately controlled and further treatments were introduced. *On-demand; †prophylaxis; IV, intravenous.

Full size image
Table 3 Outcomes of discontinuation
Full size table

Side effects of AA discontinuation

Seven patients had no side effects of discontinuation other than changes to HAE attack frequency and/or severity. Three patients experienced the following: anxiety and depression (likely due to developing cancer as well as changes to attack frequency; n = 1), weight gain (n = 1) and fatigue (n = 1).

Patient outcomes after AA discontinuation

Of the 7 patients who remained off AA treatment, time since discontinuation ranged from 7 to 84 months (median = 48 months); the majority of patients went on to receive a different form of prophylaxis (n = 6), while 1 patient received only on-demand therapy. At data collection, these patients were experiencing no or very few HAE attacks, and quality of life (QoL) had largely improved. Of the 3 patients who restarted AAs, HAE attacks continued for 2 patients whose QoL was either moderately affected or described as ‘poor’. The third patient who reintroduced AAs has experienced no attacks, but remains anxious about introducing injectable treatments.

Discussion and conclusions

Although AAs have been the historical option for long-term prophylaxis for HAE [6], there is now a shift away from these drugs [38, 39], which may continue as further targeted therapies are developed and approved, such as IV or SC pdC1-INH, lanadelumab and berotralstat [32,33,34,35, 37, 47, 48]. Our case series illustrates the heterogeneity of AA discontinuation strategies and the risk of increased HAE attack frequency and severity, alongside the potential positive outcomes for patients with HAE caused by C1-INH deficiency if appropriate management is instigated after AA withdrawal.

The most appropriate protocols for managing discontinuation of long-term AA prophylaxis have not yet been identified [40,41,42,43]; while tapering, overlapping and immediate withdrawal strategies have been suggested, these have not been systematically compared in terms of patient outcomes [45]. It is likely that a whole range of factors beyond attack control, such as patient requirements/preferences and the availability of other options, will play a role in finalizing discontinuation strategies.

The most common challenge when patients discontinued AAs was an increase in HAE attack frequency and/or severity; a challenge previously highlighted by the survey of physicians treating HAE [45]. It is important for patients and clinicians to be aware of the risk of increased attack frequency and severity, and prior to discontinuing androgens, patients should be made aware of the potential need to manage severe breakthrough attacks, and should have both access to therapies and confidence in administering these therapies promptly. HAE attack frequency and severity should be monitored closely, using either the Angioedema Activity Score [49], or patient reports or diaries, to ensure that the provided therapies continue to manage attacks. Disease registries can be particularly useful in the monitoring of disease evolution, especially if patients can autonomously enter their data in real time. Patients may require training or retraining in administering therapies because, in some cases, patients may be switching from oral AAs to injectable therapies or patients may not have had to manage a breakthrough attack for several years. Patient training must be clearly communicated, with continued support to ensure that care plans align with any changes in HAE attack frequency or severity, and patient needs [6]. Healthcare professionals should be ready to modify the treatment strategy at any stage if HAE attacks are not controlled, and follow-up appointments can be systematically included in treatment plans to ensure that any changes in attacks are communicated promptly.

Beyond the physical risk of HAE attacks, fatigue, anxiety and depression have been reported in patients with HAE discontinuing AAs [45]. Although AA discontinuation in patients with HAE has not been systematically studied, in our case series 7/10 patients reported no side effects other than changes to HAE attack frequency or severity.

The post-AA period can also herald a psychological burden, particularly in patients experiencing minimal HAE attacks while being treated with AAs, and patients require additional support to manage these anxieties. For some patients, attachment to oral AAs can be high, even in the presence of side effects, and these patients emphasize the need for education and support during and after discontinuation. One patient in our case series reintroduced AAs due to anxieties over the use of injectable on-demand and prophylactic therapies and a psychological attachment to the AAs that had controlled his attacks for 36 years. The reasons for patients not completing discontinuation or returning to AAs are also highly dependent on available resources and therapy types. As exemplified by another of the cases here, the return to AA treatment after a severe laryngeal attack can be based on limited availability of other options.

Although our case series was limited by size, the seven patients who discontinued AAs and resumed treatment with a different option experienced improvements in HAE attack control. One patient who continued treatment with on-demand therapy only has experienced only one HAE attack in a 7-year period. This emphasizes the need to regularly assess patients for prophylactic requirements. Although the management of HAE attacks is crucial, the potential impact of continuous treatment must also be considered.

In our study of real-world cases, the limited numbers of patients combined with heterogeneity of clinical circumstances and variable long-term AA regimens do not permit us to draw firm conclusions on the most appropriate strategies for AA withdrawal. It is clear that patients must be monitored closely for increases in HAE attack frequency and severity, but with careful planning and monitoring, and appropriate resources and support, discontinuation can be well managed. While approaches to AA discontinuation in HAE have been suggested [45], to develop the required understanding of and provide standardized guidance for AA discontinuation in HAE, systematic studies in higher numbers of patients are required.

Such studies should be extensive and involve national or international networks of HAE experts, such as the global network of Angioedema Centers of Reference and Excellence (ACARE) [50]. Indeed, the ACARE network recently initiated the SHAERPA (Stopping Androgen Treatment in Patients with HAE—Characterization of Reasons and Protocols and Development of Advice for Patients and Physicians) project with the aim of developing consensus guidance on how to discontinue AA treatment based on patient data. The SHAERPA project will provide a platform for the systematic studies required to support future clinicians when transitioning patients from AAs to targeted therapies.

Recommendations on how to discontinue AAs should include details not only on how to manage discontinuation and changes to HAE attacks but also on how to support patient monitoring and education in order to help clinicians when transitioning patients from AAs to targeted therapies.

In conclusion, discontinuation of AAs is already, and will continue to be, a major topic in HAE management because of side effects, contraindications for AAs, and the availability of better tolerated drugs. While small, our case series highlights the heterogeneity of managing AA withdrawal and the possible destabilization of HAE control, and how replacement therapies are needed to support AA withdrawal for the majority of patients. The ongoing SHAERPA study followed by data‑driven recommendations will support the management of AA discontinuation to improve QoL for HAE patients.