Treatment patterns and healthcare resource utilization among patients with hereditary angioedema in the United States

Hereditary angioedema with C1 inhibitor deficiency (C1INH-HAE; hereafter “HAE”) is a rare disorder of genetic origin caused by C1 inhibitor deficiency or decreased functional activity [1]. Recent epidemiologic studies support HAE prevalence ranges from 1 in 60,000 to 1 in 100,0000 people, though misdiagnosis and delayed diagnosis of HAE are common [2, 3]. The characteristic features of HAE include intermittent and generally unpredictable attacks characterized by edematous swelling which can be very painful (abdominal attacks), disfiguring (peripheral swelling), or even fatal (laryngeal attacks). The disease can impart a considerable personal burden, significantly reducing quality of life (QoL), both during and between attacks; further, attacks lead to absenteeism from activities during attacks for both patients and their caregivers and decreased educational and career advancement [4‐9].

General treatment goals for HAE include minimizing morbidity and preventing mortality, as well as maximizing QoL for patients living with this chronic, burdensome disease [1]. These goals can be accomplished by effective “on-demand” treatment of attacks, as well as routine prophylaxis of attacks, if warranted. Prior to 2008, medications for the acute treatment of HAE in the United States (US) were limited to supportive treatment. There are now four US Food and Drug Administration (FDA)-approved on demand treatments for HAE attacks: intravenous (IV) plasma-derived C1-INH [C1-INH(IV) Berinert®, CSL Behring, Marburg, Germany]; subcutaneous (SC) icatibant (Firazyr®, Shire, Lexington, MA); SC ecallantide (Kalbitor®, Shire), and IV recombinant human C1-INH (Ruconest®, Pharming Healthcare, Inc., Berkeley Heights, NJ).

Prophylaxis options remain more limited, and historically included oral attenuated androgens and antifibrinolytic agents (eg, tranexamic acid). Long-term use of oral androgens, while convenient and inexpensive, are associated with a number of health risks, toxicities, and adverse side effects. Thus, they are generally not preferred for long-term prophylaxis, and are particularly unsuitable for young patients and women, especially during pregnancy or breast-feeding [1, 10‐12]. Antifibrinolytics are not recommended because of their lower efficacy relative to other prophylactic options [1, 13]. Newer disease-specific options for HAE prophylaxis include plasma-derived C1-INH(IV) (Cinryze®, Shire ViroPharma, Lexington, MA), FDA-approved in 2008, a SC formulation of C1-INH (C1-INH[SC]; HAEGARDA®, CSL Behring, Marburg, Germany) approved by the FDA in June 2017, and a SC monoclonal antibody (lanadelumab; TAKHZYRO™, Shire, Lexington, MA) FDA-approved in August, 2018. Routine prophylaxis with C1-INH(IV) reduces the median and mean attack frequency by half [14], while C1-INH(SC) at the approved dose of 60 IU/kg was shown to reduce the median (mean) attack frequency by 95% (84%) [15].

The introduction of HAE-specific therapies for both on-demand and prophylaxis treatment represented major advancements in HAE disease management, and self-administration has been embraced as a safe and feasible practice. Despite these new treatments which can effectively alleviate the disease burden for many patients, IV medication use may pose certain challenges, including difficulty gaining and/or maintaining venous access, or logistical issues pertaining to proper infusion procedure; some patients are simply reluctant to self-administer IV medication because of physical or psychological barriers [16, 17]. The extent to which these factors influence outcomes are currently poorly understood.

Real-world data are particularly useful for evaluating medication usage patterns and can help to quantify certain relevant outcomes. While long-term studies, observational cohorts, and registry data have been published for individual HAE products [18‐22], there are limited real-world data on general usage patterns encompassing all available disease-specific HAE medications in the US population, including concomitant medication usage patterns. Most notably, no studies have specifically evaluated outcomes associated with the use of IV versus non-IV medications.

The purpose of this retrospective study was to evaluate real-world treatment patterns of HAE-specific medications in the US and to assess their impact on healthcare resource utilization (HCRU).

This study of 631 unique patients is the largest study of its kind to evaluate real-world treatment patterns and outcomes in a US cohort of patients with HAE. During the study time window (2006–2014), C1-INH(IV), either as on-demand or prophylactic treatment, was the most frequently used medication for management of patients with HAE in the US. Slightly more than two-thirds of patients had claims for C1-INH(IV) during the 9-year study period. Our data suggest that C1-INH(IV) was used for routine HAE prophylaxis by at least 25% of patients, with the caveat that due to the nature of the prescription data analyzed, intended use (prophylactic vs on-demand) could only be assumed. As a surrogate means of assigning intention of use, we devised parameters for defining prophylactic use based on C1-INH(IV) refill activity. In more than half of episodes that were identified as prophylactic episodes, claims for ongoing prescriptions of HAE on-demand medications were observed. This is likely to reflect the implementation of current US HAE guidelines stating patients on a prophylactic treatment regimen must also have access to effective on-demand treatment for acute attacks [1]. These observations may also corroborate findings from a recent HAE patient survey which included 47 users of C1-INH(IV) for routine HAE prophylaxis who reported breakthrough attacks with a frequency of at least once per month [16].

The use of CVADs such as subcutaneous ports is strongly discouraged by the US Hereditary Angioedema Association Medical Advisory Board unless deemed as a last resort [1]. It is well-known that CVADs are associated with a number of medical risks including infections and thrombotic complications [26‐29]. Up until recently, only anecdotal reports were available describing complications in HAE patients with ports, including thrombosis and systemic fungal infection [30]. Based on our study we estimated the prevalence of port use in the US HAE population during the study period who used regular C1-INH(IV) infusions to be around 5% and that more than half of those patients with a CVAD experienced at least one major complication leading to CVAD replacement or repair. A similarly high risk of CVAD complications was reported in a recent patient survey which included HAE patients with ports and weekly intravenous C1-INH infusions [16]. Our data also revealed higher levels of healthcare utilization (eg, hospitalizations, ED visits) in the subgroup of patients who were port users. However, an important limitation of interpreting these findings is that the data could not determine causality between CVAD use and reasons for the higher rates of healthcare visits. It is certainly possible that patients with CVAD had more severe disease and/or had other underlying risk factors that predisposed them to more frequent hospitalization and/or emergency care, and issues other than port use likely contributed to this finding.

There are several additional limitations that need to be considered when interpreting the findings presented herein. Most notably, claims data do not provide a definitive means of identifying intent of medication usage. Therefore, while medication claims were interpreted as usage and a surrogate marker for breakthrough attacks for the purposes of this analysis, it is understood that prescription claims may not always accurately indicate clinical medication usage or HAE attack patterns. For analysis purposes, claims for on-demand medications that occurred during periods of prophylaxis were assumed to indicate breakthrough attacks, despite the possibility that some of these claims may have been for medications to keep on hand in case of an attack, thus exaggerating the interpreted prevalence of breakthrough attacks. Conversely, it is likely that many patients using C1-INH(IV) for routine prophylaxis may have used the same product to treat breakthrough attacks; this “rescue” usage could not have been differentiated from the prophylaxis refill activity of the same product and as a result, the number of breakthrough attacks may have been under-estimated in such cases. It is also possible that some patients requiring large doses of C1-INH(IV) as frequent on-demand treatment could have been incorrectly categorized as using regular prophylaxis based on the surrogate threshold of ≥1500 IU/week for ≥13 weeks. In addition, surrogate definitions based on sustained refill patterns exceeding a defined threshold were used to categorize prophylactic use of C1-INH(IV). Finally, since the claims database used in this study included only commercially-insured patients in the US, individuals that are uninsured or covered under government health plans, including the elderly, are under-represented and thus generalizability of the findings to the overall population of the US, as well as to other countries, is limited.

This analysis of a large real-world claims database suggests that, despite the introduction of multiple new HAE-specific medications in the US since 2008, a subset of patients with HAE requiring long-term prophylaxis continue to experience considerable disease and treatment burden. Specific treatment burdens suggested by these data include sub-optimal attack prevention efficacy and the need for central venous access in some US patients, along with corresponding higher consumption of hospital and emergency healthcare services.