Diverse Trial Designs, Populations, and Outcomes: A Systematic Literature Review of Trials for the Treatment of Hereditary Angioedema Attacks

Abstract

Introduction

Hereditary angioedema (HAE) is a rare genetic disorder affecting approximately 1 in 50,000 people. As of July 2025, five on-demand treatments for HAE attacks have been approved based on phase 2/3 and phase 3 randomized clinical trials (RCTs). This systematic literature review (SLR) aimed to evaluate clinical trial designs, populations, and outcomes to better understand factors influencing observed treatment effects.

Methods

Following health technology assessment methodology and standards, a comprehensive SLR was conducted through October 2024 using MEDLINE, MEDLINE In-Process, Embase, and the Cochrane Library. Eligible studies were phase 2–3 RCTs evaluating ecallantide, icatibant, plasma-derived C1 esterase inhibitor (C1-INH), recombinant human C1-INH (rhC1-INH), or sebetralstat, and reporting efficacy (e.g., time to onset of symptom relief [TOSR] or time to complete resolution [TTCR]). Assessments included trial design, attack characteristics, and patient-reported outcome (PRO) measures.

Results

Of the 12 RCTs evaluated, 11 were center-based and 1 was home-based with e-diary collection. Center-based trials required presentation within 5–8 h after attack onset, whereas the home-based trial instructed immediate treatment, resulting in faster treatment times (median, 215–630 min vs. 35–51 min, respectively). Baseline attack severity was a key differentiator between trials, with some trials requiring a minimum severity of ≥ 30 mm or ≥ 50 mm on the visual analog scale. Censoring criteria differed in timing and conditions (e.g., rescue medication, redosing), with placebo arm censoring rates exceeding 50% in some trials. Eight trials used TOSR as the primary endpoint, measured by six distinct PRO instruments with variable definitions of improvement and success.

Conclusion

Substantial heterogeneity in trial design, attack eligibility criteria, redosing and rescue protocols, censoring rules, and endpoint definitions limit cross-trial comparability of on-demand HAE therapies. Future comparative efficacy research should prioritize harmonization of endpoint definitions, map across PRO instruments, and adjust for treatment effect modifiers.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12325-026-03656-7.

Key Summary Points

Five on-demand hereditary angioedema (HAE) therapies are approved by the US Food and Drug Administration and have been evaluated across 12 randomized clinical trials (RCTs)

Differences in RCT design and enrollment criteria, including study setting, timing of intervention, and attack severity requirements, limit cross-trial comparability

Variability in baseline patient and attack characteristics, such as prophylaxis use, attack severity, and attack location, may influence observed treatment effects

Inconsistencies in endpoint definitions, including the use of different patient-reported outcome instruments and success criteria, introduce challenges in interpreting efficacy outcomes

Comparative analyses without harmonizing endpoint definitions and adjusting for treatment effect modifiers may yield misleading conclusions

Introduction

Hereditary angioedema (HAE) is a rare genetic disorder affecting approximately 1 in 50,000 people [1]. It is characterized by recurrent, unpredictable attacks of subcutaneous or submucosal swelling of the face, extremities, upper airways, and gastrointestinal tract, often resulting in considerable morbidity and negatively impacting patient quality of life [1–5]. The burden of disease is compounded by diagnostic delays, ranging from 4 to 26 years, and inconsistent access to treatment [6, 7]. Effective management of HAE attacks is critical to reducing disease morbidity and improving quality of life [1].

Although long-term prophylaxis is appropriate for some patients, current guidelines emphasize the universal need for access to effective on-demand therapy, regardless of prophylactic regimen, attack location, or severity [1, 4, 8]. In the USA, five on-demand therapies are approved by the US Food and Drug Administration (FDA) to treat HAE attacks: ecallantide, icatibant, plasma-derived C1 esterase inhibitor (pdC1-INH), recombinant human C1 esterase inhibitor (rhC1-INH), and sebetralstat [9–13].

Each of these therapies has demonstrated efficacy and safety in randomized clinical trials (RCTs) [14–23]. Accordingly, the 2021 US Hereditary Angioedema Association Guidelines characterize FDA-approved HAE therapies as very effective and generally safe [1]. Notably, sebetralstat was approved after the publication of these guidelines and, thus, is not included in their recommendations. Recent international guidelines recommend the use of ecallantide, icatibant, pdC1-INH, rhC1-INH, or sebetralstat as first-line therapy for acute treatment of angioedema attacks in pediatric patients 12 years and older with HAE [8]. However, Fijen et al. reported substantial heterogeneity in outcome measures and patient-reported outcome (PRO) instruments across acute HAE trials, complicating interpretation of efficacy findings between studies [24]. Consequently, the selection of an on-demand therapy remains a shared decision-making process between clinicians and patients. Additional evidence is needed for clinicians and patients to critically evaluate the clinical evidence base underlying these therapies.

This systematic literature review (SLR) aims to address this evidence gap by analyzing the design characteristics, populations, and endpoints of RCTs evaluating on-demand therapies for HAE. Specifically, the review examines trial design, attack eligibility criteria, baseline attack characteristics, and PRO measures to identify elements that may impact the interpretation and comparability of treatment effects. Understanding these trial design elements is essential for contextualizing reported treatment effects and identifying any potential sources of bias that may confound comparative effectiveness research.

Methods

Search Strategy and Study Selections

This SLR was conducted in accordance with the principles and standards for health technology assessment, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines, and guidelines from the Cochrane Collaboration [25–28]. Trial selection was performed in two stages: a primary screening based on titles and abstracts and a secondary full-text screening. In both stages, selection was completed by two independent reviewers, and discrepancies were resolved by a third independent reviewer.

The comprehensive literature search, conducted in accordance with the Peer Review of Electronic Search Strategies checklist [29], included the following: (1) electronic database searches in MEDLINE and MEDLINE In-Process, Embase, and the Cochrane Library (from inception to October 2024); (2) hand searches of bibliographies of published systematic reviews and meta-analyses as well as ClinicalTrials.gov, accessdata.fda.gov, journal sites, and corporate websites with reports containing relevant and complete data; and (3) gray literature searches of three international scientific conferences—American Academy of Allergy, Asthma & Immunology, European Academy of Allergy and Clinical Immunology, and American College of Allergy, Asthma, and Immunology—from 2022 to 2024.

Relevant publications were identified using prespecified selection criteria based on the population, intervention, comparison, outcome, study design, and time (PICOS-T) framework (Table 1). Included studies enrolled patients with acute HAE attacks; evaluated ecallantide, icatibant, pdC1-INH, rhC1-INH, or sebetralstat treatments; and reported efficacy (time to onset of symptom relief [TOSR], time to reduced severity [TTRS], time to minimal symptoms [TTMS], time to complete resolution [TTCR], based on any measure) outcomes. Eligible English-language publications included phase 2 and phase 3 RCTs as well as SLRs, meta-analyses, or network meta-analyses of relevant RCTs. Search strategies are presented in Tables S1–S3.

Table 1.

PICOS-T criteria

Category	Inclusion criteria	Exclusion criteria
Population	Patients (with HAE [acute attacks])	Healthy volunteers
Interventions	Ecallantide Icatibant rhC1-INH pdC1-INH Sebetralstat	Interventions other than those listed in the inclusion criteria Use of any intervention of interest in a prophylactic setting
Comparators	Placebo BSC Any of the interventions listed above	Any nonpharmacologic interventions Comparators other than those listed in the inclusion criteria
Outcomes	Efficacy outcome(s) as measured in the study (e.g., TOSR, TTRS, TTMS, TTCR based on any measure)	Trials not reporting outcomes listed in the inclusion criteria
Study design	Phase 2 or 3 RCTs SLRs, meta-analyses or NMAs of relevant RCTs	Any other study designs, including the following:  Observational studies (e.g., prospective studies, retrospective studies, case control studies, cross-sectional studies)  Case studies/case reports  Economic evaluations  Editorials, notes, comments, or letters  Narrative or non-SLRs  Studies not in humans (e.g., animal studies)  Pharmacokinetic or pharmacodynamic studies  Phase 1 RCTs  Open-label extension studies
Time limit	Database inception to date of search Conference abstracts from 2022 onward	Not applicable for full texts Conference abstracts published before 2022
Language	Title and abstract must be published in English
Countries	No restriction on geography or country will be applied

BSC best supportive care, HAE hereditary angioedema, NMA network meta-analysis, pdC1-INH plasma-derived C1 esterase inhibitor, PICOS-T population, intervention, comparison, outcome, study design, and timing, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, SLR systematic literature review, TOSR time to onset of symptom relief, TTCR time to complete resolution, TTMS time to minimal symptoms, TTRS time to reduced severity

Data Extraction

All data were extracted from eligible publications by two independent reviewers and verified against the original sources by a third independent reviewer. The third reviewer also resolved any discrepancies between the initial two reviewers.

If multiple publications were identified for the same trials, all reports were linked and extracted together.

Statistical Analyses

Data were tabulated from primary publications and summarized narratively according to the recommendations from the Cochrane Collaboration [28]. No formal statistical assessments of the data were performed.

Ethics Approval

This article is based on previously conducted studies and does not involve any new studies with human participants or animals performed by any of the authors.

Results

Study Characteristics

A total of 2037 records were identified via database search, of which 220 duplicate entries were removed (Fig. 1). Screening of titles and abstracts resulted in the exclusion of 1708 records, and 2 records were not retrieved. The remaining 107 records were assessed for eligibility. A total of 52 records were excluded for the following reasons: study design (18 records) [30–47]; population (11 records) [48–58]; intervention (10 records) [59–68]; publication type (8 records) [69–76]; lack of published results (2 records) [77, 78]; duplication [79], outcome [80], or comparator [81] (1 record each). A total of 35 records were identified via other methods. One record was not retrieved and the remaining 34 were assessed for eligibility; 1 was excluded due to publication type [82] and 1 due to a lack of published results [83]. In total, 12 RCTs were reported in 87 publications (including 3 data on file sources) [14–23, 79, 84–156] (Fig. 1).

Fig. 1.

PRISMA flow diagram. ^aIncludes searches within Cochrane Central Register of Controlled Trials and Cochrane Database of Systematic Reviews. ^bIncludes records identified from publicly accessible websites (n = 16), accessdata.fda.gov (n = 14), and EMA repository (n = 2). ^cIncludes records identified from ClinicalTrials.gov. EMA European Medicines Agency, PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses, RCT randomized clinical trial

Trial Design and Attack Criteria

Of the 12 included trials, 11 were center-based and 1 (KONFIDENT trial; sebetralstat) was home-based with e-diary data collection (Table 2). Among the 11 trials reporting study design, 9 used parallel and 2 used crossover assignment. After attack onset, 9 center-based trials required presentation to study center within 5 to 8 h, whereas the home-based KONFIDENT trial instructed immediate treatment. The EDEMA-1 trial evaluating ecallantide was excluded from synthesis due to insufficient reporting of study design, attack eligibility criteria, and baseline characteristics.

Table 2.

Trial design, attack eligibility criteria, and censoring^a of redosing and rescue therapy^b in RCTs of on-demand HAE therapies^c

Treatment	Trial	Years	Study design			Attack eligibility criteria		Redosing		Rescue medication
			Phase	Design	Site	Time from attack onset	Severity requirement	Allowed	Censored	Allowed	Censored
pdC1-INH	IMPACT-1	2005–2007	2/3	Parallel	Center	< 5 h	Moderate to severed	X	X	X	X
Ecallantide	EDEMA-1	NR	2	NRe	Center	NRe	NRe	NRe	NRe	NRe	NRe
	EDEMA-3	2005–2007	3	Parallel	Center	< 8 h	TOS > 0	X	X	X	X
	EDEMA-4	2007–2008	3	Parallel	Center	< 8 h	MSCS > 0	X	X	X	X
Icatibant	FAST-1	2004–2006	3	Parallel	Center	< 6 h	VAS ≥ 30 mm	–	–	X	–
	FAST-2	2005–2006	3	Parallel	Center	< 6 h	VAS ≥ 30 mm	–	–	X	–
	FAST-3	2009–2010	3	Parallel	Center	< 6 h	VAS ≥ 30 mm	–	–	X	–
rhC1-INH	C1 1304	2004–2007	3	Parallel	Center	< 5 h	VAS ≥ 50 mm	–	–	X	–
	C1 1205	2005–2008	2/3	Parallel	Center	< 5 h	VAS ≥ 50 mm	–	–	X	–
	C1 1310	2011–2012	2/3	Parallel	Center	< 5 h	VAS ≥ 50 mm	X	X	X	X
Sebetralstat	KVD900-201	2019–2020	2	Crossover	Center	NRe	PGI-S < severe	–	NRe	X	NRe
	KONFIDENT	2022–2023	3	Crossover	Home	Treated immediately	No criteria	X	–	X	X

HAE hereditary angioedema, MSCS mean symptom complex severity, NR not reported, not applicable, missing, or could not be found, pdC1-INH plasma-derived C1 esterase inhibitor, PGI-S Patient Global Impression of Severity, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, TOS treatment outcome score, VAS visual analog scale

^aCensored patients are those who received rescue therapy or were redosed (if allowed) or received other disallowed medication prior to achieving the primary endpoint. In the sebetralstat KONFIDENT trial, censoring occurred if patients received rescue treatment with conventional therapy or failed to meet the primary endpoint within 12 h

^bRescue was defined as any additional medication given other than the study treatment that may have affected observed treatment response

^cpdC1-INH (IMPACT-1); rhC1-INH (C1 1304, C1 1205, C1 1310); ecallantide (EDEMA-3, EDEMA-4); icatibant (FAST-1, FAST-2, FAST-3); sebetralstat (KVD900-201, KONFIDENT)

^dAs determined by clinician at the time of evaluation

^eData were not reported, missing, or could not be found

Only the icatibant (FAST-1, FAST-2, and FAST-3) and rhC1-INH (C1 1304, C1 1205, C1 1310) trials required a minimum visual analog scale (VAS) score ≥ 30 mm or ≥ 50 mm, respectively, without signs of worsening. The pdC1-INH IMPACT-1 trial enrolled patients with moderate to severe attacks, as determined by the clinician at the time of evaluation [147]. In the ecallantide trials (EDEMA-3 and EDEMA-4), severity assessments were based on PRO instruments that did not allow normal (“0”) severity ratings at enrollment (i.e., mean symptom complex severity [MSCS] score and treatment outcome score [TOS] greater than 0 were required to calculate percent change). No minimum attack severity thresholds were required in the KONFIDENT trial.

Attack and Treatment Characteristics

Attack locations were reported using heterogeneous categories and classifications across trials (Table 3). The proportion of abdominal attacks ranged from 23.1% to 79.5%, and peripheral attacks ranged from 23.1% to 70.8%. Multiple locations per attack were reported in the ecallantide EDEMA-3 and EDEMA-4 and sebetralstat KONFIDENT trials, whereas other trials reported either the primary attack location or most severe location.

Table 3.

Baseline attack and treatment characteristics in RCTs of on-demand HAE therapies

Trial	Treatment	Patient/attack count	Attack location			Baseline attack severity					Treatment regimen and timing
			Abdominal, n (%)	Peripheral, n (%)	Above the neck, n (%)	Baseline VAS, mean (SD)	Mild, n (%)	Moderate, n (%)	Severe, n (%)	Very severe, n (%)	LTP use, n (%)	Time to treatment,l median (IQR), min
IMPACT-1	pdC1-INH 10 IU/kg	39	31 (79.5)	NR	8 (20.5)a	NR	NR	32 (82.1)	7 (17.9)	NR	NR	NR
	pdC1-INH 20 IU/kg	43	34 (79.1)	NR	9 (20.9)a	NR	NR	27 (62.8)	16 (37.2)	NR	NR	NR
	Placebo	42	33 (78.6)	NR	8 (19.0)a	NR	NR	26 (61.9)	16 (38.1)	NR	NR	NR
EDEMA-3	Ecallantide 30 mg	36	20 (55.6)b	25 (69.4)b,c	4 (11.1)b	NR	NR	NR	NR	NR	NR	NR
	Placebo	36	21 (58.3)b	21 (58.3)b,c	9 (25.0)b	NR	NR	NR	NR	NR	NR	NR
EDEMA-4	Ecallantide 30 mg	48	18 (37.5)b	34 (70.8)b,d	22 (45.8)b,e	NR	NR	NR	NR	NR	NR	NR
	Placebo	48	27 (56.3)b	21 (43.8)b,d	22 (45.8)b,e	NR	NR	NR	NR	NR	NR	NR
FAST-1	Icatibant 30 mg	27	13 (48.1)	14 (51.9)d	NR	69.3 (NR)	NR	NR	NR	NR	NR	456 (NR)
	Placebo	29	16 (55.2)	13 (44.8)d	NR	67.7 (NR)	NR	NR	NR	NR	NR	600 (NR)
FAST-2	Icatibant 30 mg	36	12 (33.3)	24 (66.7)d	NR	63.7 (NR)	NR	NR	NR	NR	NR	630 (NR)
	Tranexamic acid	38	15 (39.5)	23 (60.5)d	NR	61.5 (NR)	NR	NR	NR	NR	NR	414 (NR)
FAST-3	Icatibant 30 mg	43	17 (39.5)	26 (60.5)d	NR	NR	NR	NR	NR	NR	NR	390 (NR)
	Placebo	45	19 (42.2)	26 (57.8)d	NR	NR	NR	NR	NR	NR	NR	330 (NR)
C1 1304	rhC1-INH 100 U/kg	16	7 (43.8)	9 (56.3)	2 (12.5)f	76.3 (17.4)	NR	NR	NR	NR	8 (50.0)	215 (186–268)
	Placebo	16	6 (37.5)	8 (50.0)	4 (25.0)f	78.5 (12.1)	NR	NR	NR	NR	13 (81.0)	235 (195–310)
C1 1205	rhC1-INH 100 U/kg	13	5 (38.5)	6 (46.2)	2 (15.4)f	82.5 (12.5)	NR	NR	NR	NR	NR	317 (245–389)
	rhC1-INH 50 U/kg	12g	5 (41.7)	6 (50.0)	0f	77.6 (14.4)	NR	NR	NR	NR	NR	347 (299–788)
	Placebo	13	3 (23.1)	3 (23.1)	6 (46.2)f	82.3 (17.2)	NR	NR	NR	NR	NR	315 (285–370)
C1 1310	rhC1-INH 50 U/kg	44	16 (36.4)	19 (43.2)	8 (18.2)h	73.5 (14.3)	NR	NR	NR	NR	22 (50.0)	235 (180–280)
	Placebo	31	12 (38.7)	14 (45.2)	5 (16.1)h	77.3 (12.6)	NR	NR	NR	NR	15 (48.0)	261 (212–312)
KVD900-201	Sequence 1	58	16 (27.6)	40 (69.0)	NR	NR	26 (44.8)	28 (48.3)	1 (1.7)	1 (1.7)	NR	NR
	Sequence 2	55	14 (25.5)	37 (67.3)	NR	NR	31 (56.4)	23 (41.8)	0	1 (1.8)	NR	NR
KONFIDENT	Sebetralstat 300 mg	87	35 (40.2)b	56 (64.4)b,i	11 (12.6)b,j	NR	36 (41.4)k	35 (40.2)k	12 (13.8)k	2 (2.3)k	19 (21.8)	35 (6–130)
	Sebetralstat 600 mg	93	42 (45.2)b	54 (58.1)b,i	13 (14.0)b,j	NR	41 (44.1)k	34 (36.6)k	16 (17.2)k	2 (2.2)k	21 (22.6)	41 (5–142)
	Placebo	84	37 (44.0)b	43 (51.2)b,i	13 (15.5)b,j	NR	36 (42.9)k	33 (39.3)k	10 (11.9)k	3 (3.6)k	18 (21.4)	51 (6–166)

HAE hereditary angioedema, IQR interquartile range, ITT intention-to-treat, LTP long-term prophylaxis, NR not reported, not applicable, missing, or could not be found, pdC1-INH plasma-derived C1 esterase inhibitor, PGI-S Patient Global Impression of Severity, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, VAS visual analog scale

^aOne patient in the IMPACT-1 study was originally randomized with a facial attack, which was later reassessed as a laryngeal attack. The treatment group of this patient was not specified

^bAn attack may be reported in more than one location

^cThis includes attacks in the genital or buttocks, external head or neck, or cutaneous locations

^dReported as cutaneous attack locations

^eThis includes attacks classified as occurring in the oropharyngeal head/neck and nonoropharyngeal head and neck locations

^fThis includes attacks classified as occurring in the orofacial-pharyngeal-laryngeal location

^gThe sample size of the ITT population of the rhC1-INH 50 U/kg arm in C1 1205 was 13, but only 12 patients were evaluated

^hThis includes attacks classified as occurring in the facial or oropharyngeal-laryngeal locations

ⁱThis includes attacks in the arms, hands, legs, feet, or torso

^jThis includes attacks classified as occurring in the head, face, neck, larynx, or throat

^kThe severity of attack was determined using the PGI-S (0 = none; 1 = mild; 2 = moderate; 3 = severe; 4 = very severe)

^lNot indicative of patient behavior or study drug mechanism of action

Baseline attack severity was not reported in the FAST-3, EDEMA-3, and EDEMA-4 trials. All attacks in the pdC1-INH IMPACT-1 trial were reported as moderate (68.5%) or severe (31.5%) by clinicians, whereas most attacks in the sebetralstat KONFIDENT trial were reported by patients as mild (42.8%) or moderate (38.6%) (Table 3). In the rhC1-INH C1 1304, C1 1205, and C1 1310 trials, baseline mean VAS scores for any of four anatomical regions (abdominal, genitourinary, orofacial-pharyngeal-laryngeal, or peripheral) ranged from 73.5 to 82.5. In the icatibant FAST-1 and FAST-2 trials, baseline mean VAS scores for three specific symptoms (cutaneous swelling, cutaneous pain, and abdominal pain) ranged from 61.5 to 69.3.

Only the C1 1304, C1 1310, and KONFIDENT trials reported the proportion of patients receiving long-term prophylaxis. In the rhC1-INH C1 1304 and C1 1310 trials, 50% of patients assigned to the rhC1-INH group received long-term prophylaxis. In the sebetralstat KONFIDENT trial, the use of long-term prophylaxis ranged from 21.8% to 22.6% among patients assigned to receive sebetralstat (Table 3).

The median time to treat was 35–51 min in the sebetralstat KONFIDENT trial and 215–630 min in the remaining trials (Table 3). The KONFIDENT trial was home-based, and patients were instructed to treat immediately, which led to substantially faster treatment times. The remaining trials were center-based, which required travel time and an evaluation in the center prior to treatment.

Redosing

Patients treated for HAE attacks may require additional doses of medication. A redose was defined as the administration of a second round of intervention with the intention of receiving the investigational medicinal product. Across studies, there were four different redosing protocols based on randomization design and redosing agent (Table 4):

1. No redose allowed regardless of intervention given (icatibant FAST-1, FAST-2, and FAST-3 trials; rhC1-INH C1 1304 and C1 1205 trials; and sebetralstat KVD900-201 trial)

2. Redose with active intervention regardless of the treatment group patients were assigned to at the time of randomization (rhC1-INH C1 1310 trial; ecallantide EDEMA-3 and EDEMA-4 trials)

3. Redose with same intervention at time of randomization (sebetralstat KONFIDENT trial)

4. Crossover redose with pdC1-INH 20 U/kg for patients originally assigned to placebo or placebo for patients originally assigned to pdC1-INH 20 U/kg; patients originally assigned pdC1-INH 10 U/kg were redosed with pdC1-INH 10 U/kg (pdC1-INH IMPACT-1 trial)

Table 4.

Redosing, rescue therapy, and censoring in RCTs of on-demand HAE therapies

Trial	Intervention	Patient/attack count	Redose given	Redose, n (%)	Rescuea, n (%)	Censoredb, n (%)
IMPACT-1	pdC1-INH 10 IU/kg	39	pdC1-INH 10 IU/kg	13 (33.3)	NR	NR
	pdC1-INH 20 IU/kg	43	Placebo	8 (18.6)	NR	13 (30.2)
	Placebo	42	pdC1-INH 20 IU/kg	24 (57.1)	NR	23 (54.8)
EDEMA-3	Ecallantide 30 mg	36	Ecallantide 30 mg	2 (5.6)	5 (13.9)	2 (5.6)
	Placebo	36	Ecallantide 30 mg	1 (2.8)	13 (36.1)	1 (2.8)
EDEMA-4	Ecallantide 30 mg	48	Ecallantide 30 mg	15 (31.2)	16 (33.3)	1 (2.1)
	Placebo	48	Ecallantide 30 mg	22 (45.8)	24 (50.0)	6 (12.5)
FAST-1	Icatibant 30 mg	27	NR	NR	6 (22.2)	1 (3.7)
	Placebo	29	NR	NR	15 (51.7)	1 (3.4)
FAST-2	Icatibant 30 mg	36	NR	NR	7 (19.4)	0
	Tranexamic acid	38	NR	NR	12 (31.6)	2 (5.3)
FAST-3	Icatibant 30 mg	43	NR	NR	3 (7.0)	0
	Placebo	45	NR	NR	18 (40.0)	3 (6.7)
C1 1304	rhC1-INH 100 U/kg	16	NR	NR	1 (5.9)	0
	Placebo	16	NR	NR	9 (56.2)	1 (5.9)
C1 1205	rhC1-INH 100 U/kg	13	NR	NR	0	0
	rhC1-INH 50 U/kg	12c	NR	NR	0	0
	Placebo	13	NR	NR	1 (7.7)	0
C1 1310	rhC1-INH 50 U/kg	44	rhC1-INH 50 U/kg	5 (11.4)	4 (9.1)	8 (18.2)
	Placebo	31	rhC1-INH 50 U/kg	13 (41.9)	2 (6.5)	14 (45.2)
KVD900-201	Sequence 1	58	NR	NR	NR	NR
	Sequence 2	55	NR	NR	NR	NR
KONFIDENT	Sebetralstat 300 mg	87	Sebetralstat 300 mg	34 (39.1)	12 (13.8)	21 (24.1)
	Sebetralstat 600 mg	93	Sebetralstat 600 mg	37 (39.8)	8 (8.6)	22 (23.7)
	Placebo	84	Placebo	47 (56.0)	21 (25.0)	43 (51.2)

HAE hereditary angioedema, ITT intention-to-treat, NR not reported, not applicable, missing, or could not be found, pdC1-INH plasma-derived C1 esterase inhibitor, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor

^aRescue was defined as any additional medication given other than the investigational medicinal product that may have affected observed treatment response

^bCensored patients are those who received rescue therapy, a redose (if allowed), or other disallowed medication prior to achieving the primary endpoint. In KONFIDENT, censoring occurred if patients received rescue therapy with conventional therapy or failed to meet the primary endpoint within 12 h

^cThe sample size of the ITT population of the rhC1-INH 50 U/kg arm in C1 1205 was 13, but only 12 patients were evaluated

Redosing rates were reported in five trials (pdC1-INH IMPACT-1, ecallantide EDEMA-3 and EDEMA-4, rhC1-INH C1 1310, and sebetralstat KONFIDENT trials); redosing rates were not reported in six trials (icatibant FAST-1, FAST-2, and FAST-3 trials; rhC1-INH C1 1304 and C1 1205 trials; and sebetralstat KVD900-201 trial). Results were reported heterogeneously, at various time points across studies and in different populations based on attack location or achievement of symptom relief. Redosing with study medication or placebo was captured at the end of the analysis window of each trial in the overall study populations (Table 4), except for the ecallantide EDEMA-3 trial, which only allowed redosing with open-label ecallantide for severe upper airway complications. Time points for the analysis window varied across trials based on when patients became eligible for redosing (e.g., 4 h post intervention) to the time of data cutoff (e.g., 24 h post intervention). Across studies, redosing rates ranged from 2.8% to 57.1% in placebo arms and 11.4% to 39.8% in treatments arms. In the active intervention arms, the sebetralstat KONFIDENT trial had the highest redosing rates (39.1–39.8%). In the placebo arms, the pdC1-INH IMPACT-1 and the KONFIDENT trials had the highest redosing rates (57.1% and 56.0%, respectively). Among the active treatment arms, the rhC1-INH C1 1310 trial had the lowest redosing rates both in the active treatment arm (11.4%) and in the placebo arm (41.9%); the ecallantide EDEMA-3 trial was excluded because it only allowed redosing for severe upper airway complications.

Rescue Therapy

Rescue therapy was defined as any administered medication other than the study intervention that may have affected observed treatment response. Rescue therapy was defined differently across trials that reported this information:

1. In the pdC1-INH IMPACT-1 trial, rescue therapy included rescue study medication, narcotic analgesics, non-narcotic analgesics, antiemetics, open-label C1 esterase inhibitor (C1-INH), androgens, and fresh frozen plasma.

2. In the ecallantide EDEMA-3 trial, rescue therapy included pain medications, antinausea medications, or C1-INH; in EDEMA-4, rescue therapy medications were not reported; in both trials, open-label study medication was permitted for attacks involving severe upper airway compromise.

3. In the icatibant FAST-1 and FAST-2 trials, rescue therapy (C1-INH, opioid and non-opioid analgesics, antiemetics, prokinetic agents, antihistamines, and epinephrine) was permitted for relief of any symptom but withheld as long as possible. In the FAST-3 trial, rescue medication was defined as any medications deemed by the investigator to be immediately necessary to alleviate acute symptoms, including morphine or equivalent low-dose narcotic, antiemetics, C1-INH, fresh frozen plasma, epinephrine, and nonsteroidal anti-inflammatory drugs.

4. In the rhC1-INH C1 1304, C1 1205, and C1 1310 trials, any medications that may have interfered with the assessment of rhC1-INH were listed as rescue medications, including bradykinin receptor antagonists, kallikrein inhibitors, and pdC1-INH.

5. In the sebetralstat KVD900-201 trial, rescue medication included pdC1-INH, rhC1-INH, or icatibant; in the sebetralstat KONFIDENT trial, patients with non-laryngeal attacks were required to receive a second dose of the randomized intervention before becoming eligible for rescue therapy with conventional medication (rhC1-INH, pdC1-INH, icatibant, ecallantide). In patients with laryngeal attacks, conventional on-demand treatments may have been administered at any time after the first dose of randomized intervention.

Across studies, rates of rescue therapy ranged from 6.5% to 56.2% in placebo arms and 5.9% to 33.3% in active treatments arms. In the active intervention arm, the ecallantide EDEMA-4 trial had the highest rescue therapy use (33.3%), and the rhC1-INH C1 1304 trial had the lowest rescue therapy use (5.9%). In the placebo arms, the rhC1-INH C1 1304 trial had the highest rescue therapy use (56%), and the rhC1-INH C1 1310 trial had the lowest rescue therapy use (6%) (Table 4). The sebetralstat KONFIDENT trial was the only trial that required patients to receive a second dose of the randomized intervention before permitting use of rescue therapy; a third dose was needed in 13.8% and 8.6% of attacks treated with 300-mg and 600-mg doses, respectively.

Censoring

Censoring criteria, methodology, and rates varied across studies and were classified into three censoring protocol categories (Table 2):

1. Censored rescue and redosing (pdC1-INH IMPACT-1, rhC1-INH C1 1310 trial, ecallantide EDEMA-3 and EDEMA-4 trials)

2. Censored only rescue (sebetralstat KONFIDENT trial)

3. Did not censor rescue or redosing (icatibant FAST-1, FAST-2, and FAST-3 trials; rhC1-INH C1 1205 and C1 1304 trials)

Censoring details were available for most trials except for the sebetralstat KVD900-201 trial. The censoring was categorized as “right-censored,” in which the patient is assumed to have not met the outcome of interest by the end of the analysis window. The one exception was the sebetralstat KONFIDENT trial, in which attacks without two or more postbaseline readings were censored at time 0. In the rhC1-INH C1 1310 trial, the TOSR was censored at the last time the Treatment Effect Questionnaire (TEQ) was assessed prior to the receipt of the disallowed concomitant medication or rescue medication. However, in the sebetralstat KONFIDENT trial, attacks treated with on-demand conventional therapy were right-censored at the end of the analysis period. Censoring rates were assessed over various analysis windows, defined from the time patients become eligible for rescue or redosing through the end of the analysis period. Among trials that included time censoring as part of the primary endpoint, the pdC1-INH IMPACT-1 and ecallantide EDEMA-3 and EDEMA-4 trials right-censored at 4 h, whereas the sebetralstat KONFIDENT trial right-censored at 12 h. In the ecallantide EDEMA-3 and EDEMA-4 trials and the rhC1-INH C1 1310 trial, patients receiving intervention or placebo could be redosed with open-label active intervention, resulting in censoring. Notably, the sebetralstat KONFIDENT trial was the only trial that allowed redosing but did not censor attacks that required redosing (Table 2).

In the active intervention arms, the pdC1-INH IMPACT-1 and sebetralstat KONFIDENT trials had higher censoring rates at 30.2% and 23.7% to 24.1%, respectively, whereas the ecallantide EDEMA-4 and icatibant FAST-1 trials had the lowest censoring rates at 2.1% and 3.7%, respectively. In the placebo arms, the pdC1-INH IMPACT-1, sebetralstat KONFIDENT, and rhC1-INH C1 1310 trials had high censoring rates at 54.8%, 51.2%, and 45.2%, respectively, whereas the icatibant FAST-1, rhC1-INH C1 304, and icatibant FAST-3 trials had the lowest censoring rates at 3.4%, 5.9%, and 6.7%. Table 4 provides censoring rates captured for the primary endpoint at the end of the analysis window.

Endpoints and PRO Instruments

Across clinical trials of on-demand HAE therapies, there was considerable variability in trial endpoints and PRO instruments. The criteria defining success for these outcomes also differed across studies (Table 5). Endpoints included TOSR, TTCR, TTMS, and TTRS. Eight of the 12 studies reported TOSR as the primary endpoint, with six different PRO instruments used to measure patient response (Table 5). These PRO instruments included a standard patient questionnaire, the Patient Global Impression of Change (PGI-C), the TEQ, and VAS (three variations), with varying criteria for success (Table 5).

Table 5.

Endpoints and PRO instruments in RCTs of on-demand HAE therapies

Treatment	Trial	Endpoint details
		Endpoint	PRO	Question structure	Response options	Success criteria
pdC1-INH	IMPACT-1	TOSR	NR	Taking into account all of the symptoms you experienced with this HAE attack, are you confident that it is starting to improve?	Yes, no	Answer “yes”
Icatibant	FAST-1 FAST-2	TOSR	VASa	The subject should draw a vertical line at the point along the scale that represents the current status of the measured symptom	0 mm, no symptom; 100 mm, worst possible symptom	30% reduction in score
Icatibant	FAST-3	TOSR	VAS-3b	The subject should draw a vertical line at the point along the scale that represents the current status of the measured symptom	Mean scores for skin swelling, skin pain, and abdominal pain, VAS-3	50% reduction in score
rhC1-INH	C1 1310	TOSR	TEQ	Q1: To what extent has the overall severity of your [attack location] HAE attack changed since you received the infusion? Q2: Overall, has the intensity of your [relevant attack location] HAE attack symptoms begun to decrease noticeably since you received the infusion?	Q1: Much worse, worse, a little worse, no change, a little better, better, much better Q2: Yes, no	Q1: Answer “a little better" or more Q2: Answer “yes” to Q2
rhC1-INH	C1 1205 C1 1304	TOSR	VAS	How severe are the angioedema symptoms now for this location?	0 mm, “no symptoms at all”; 100 mm, “extremely disabling”	≥ 20-mm reduction from baseline
Sebetralstat	KONFIDENT	TOSR	PGI-C	How would you describe your overall HAE symptoms right now, compared to how you were when you took the trial medication?	Much worse, worse, a little worse, no change, a little better, better, much better	Rating of at least “a little better” within 12 h
Sebetralstat	KONFIDENT	TTRS	PGI-S	What is your overall HAE attack severity right now?	None, mild, moderate, severe, very severe	1-point decrease from baseline within 12 h
rhC1-INH	C1 1310	TTMS	TEQ	Q3: At this moment, are your HAE attack symptoms minimal [barely noticeable]?	Yes, no	Answer “yes”
rhC1-INH	C1 1310 C1 1205 C1 1304	TTMS	VAS	How severe are the angioedema symptoms now for this location?	0 mm, “no symptoms at all”; 100 mm, “extremely disabling”	Score < 20 mm
pdC1-INH	IMPACT-1	TTCR	NR	Have all symptoms of the HAE attack resolved completely?	Yes, no	Answer “yes”
Icatibant	FAST-1 FAST-2 FAST-3	TTCR	VAS	The subject should draw a vertical line at the point along the scale that represents the current status of the measured symptom	0 mm, no symptom; 100 mm, worst possible symptom	Achieve score < 10 mm at all symptoms/locations
Sebetralstat	KONFIDENT KVD900-201	TTCR	PGI-S	What is your overall HAE attack severity right now?	None, mild, moderate, severe, very severe	Rating of "none “ within 24 h
Ecallantide	EDEMA-3 EDEMA-4	NR	MSCS	Describe the severity of the following [symptom complex] at this time	Normal (0), mild (1), moderate (2), and severe (3) averaged over select locations	Change from baseline MSCS ranging from + 2 to − 3 at 4 h
Ecallantide	EDEMA-3 EDEMA-4	NR	TOS	What [symptom complex(es)] are you experiencing? Describe the severity and level of improvement of [symptom complex] at this time	Severity = normal (0) to severe (3); significant improvement/worsening (100 to − 100)	Change from baseline TOS at 4 h

HAE hereditary angioedema, MSCS mean symptom complex severity, NR not reported, not applicable, missing, or could not be found, pdC1-INH plasma-derived C1 esterase inhibitor, PGI-C Patient Global Impression of Change, PGI-S Patient Global Impression of Severity, PRO patient-reported outcome, Q question, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, TEQ Treatment Effect Questionnaire, TOS treatment outcome score, TOSR time to onset of symptom relief, TTCR time to complete resolution, TTMS time to minimal symptoms, TTRS time to reduced severity

^aMost severe symptom was chosen by assessing the most severe “cutaneous swelling” or “pain (skin)”; if these were equal, pain was used. For abdominal attacks, pain was used

^bComposite of skin pain, skin swelling, and abdominal pain

The PRO instruments used in these trials also collected information about attacks, with differing levels of specificity (Fig. 2). The pdC1-INH IMPACT-1 and sebetralstat KONFIDENT and KVD900-201 trials used PRO instruments with global dimensions (PGI-C, Patient Global Impression of Severity [PGI-S]); the rhC1-INH trials (C1 1310, C1 1205, and C1 1304) used PRO dimensions that were specific to attack locations but not symptoms (TEQ, VAS); and ecallantide EDEMA-3 and EDEMA-4, and icatibant FAST-1, FAST-2, and FAST-3 trials used PRO instruments with both location- and symptom-specific dimensions (VAS-3, MSCS, TOS).

Fig. 2.

Specificity in attack information measured using patient-reported outcomes in RCTs of on-demand HAE therapies. pdC1-INH (IMPACT-1); rhC1-INH (C1 1304, C1 1205, C1 1310); ecallantide (EDEMA-3, EDEMA-4); icatibant (FAST-1, FAST-2, FAST-3); sebetralstat (KVD900-201, KONFIDENT). HAE hereditary angioedema, MSCS mean symptom complex severity, NR not reported, not applicable, missing, or could not be found, pdC1-INH plasma-derived C1 esterase inhibitor, PGI-C Patient Global Impression of Change, PGI-S Patient Global Impression of Severity, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, TEQ Treatment Effect Questionnaire, TOS treatment outcome score, VAS visual analog scale

Criteria for successful response to treatment varied across both studies and endpoints (Table 5, Fig. 3). The rhC1-INH trials (C1 1310, C1 1205, C1 1304) were the only trials that defined TOSR using a composite criterion requiring both sustained response (defined as any improvement from baseline at consecutive time points, i.e., a response of “a little better,” “better,” or “much better”) and sustained response without worsening (defined as maintaining or escalating improvement at consecutive time points, i.e., a response of “a little better,” “better,” or “much better” at the first time point and an equal or better response at the next time point) (Table 5). The pdC1-INH IMPACT-1 (TOSR), icatibant FAST-1, FAST-2, FAST-3 (TOSR, TTCR), and sebetralstat KONFIDENT (TOSR, TTRS) trials required only a sustained response.

Fig. 3.

Endpoint response requirements in RCTs of on-demand HAE therapies. pdC1-INH (IMPACT-1); rhC1-INH (C1 1304, C1 1205, C1 1310); icatibant (FAST-1, FAST-2, FAST-3); sebetralstat (KVD900-201, KONFIDENT). ^aDefined as any improvement from baseline at consecutive time points, i.e., a response of “a little better,” “better,” or “much better.” ^bDefined as maintaining or escalating improvement at consecutive time points, i.e., a response of “a little better,” “better,” or “much better” at the first time point and an equal or better response at the next time point. HAE hereditary angioedema, pdC1-INH plasma-derived C1 esterase inhibitor, RCT randomized clinical trial, rhC1-INH recombinant human C1 esterase inhibitor, TOSR time to onset of symptom relief, TTCR time to complete resolution, TTMS time to minimal symptoms, TTRS time to reduced severity

Discussion

This SLR identified 12 RCTs and encompassed all 5 FDA-approved on-demand therapies for acute HAE attacks. Although these studies provide robust evidence supporting the efficacy of these therapies, substantial methodological heterogeneity limits the ability to compare results and complicates clinical decision-making. Importantly, this review is descriptive in nature. While it identifies systematic differences in trial design, censoring, rescue and redosing protocols, and PROs, it does not formally quantify the independent effect of these features on treatment outcomes. Interpretations of potential bias mechanisms and their effects on treatment outcomes in the absence of a meta-analysis should be considered hypothesis-generating rather than empirical evidence.

Differences in trial design may meaningfully affect measured outcomes. For example, the sebetralstat KONFIDENT trial implemented a home-based design with instructions for immediate treatment initiation at attack onset, resulting in significantly shorter median time to treatment compared with the other trials, which were center-based [23]. In center-based studies, treatment could not be initiated until patients were evaluated at the study center, requiring time for decision-making and travel to the center, thereby increasing time to treatment initiation. Therefore, time-to-treatment differences reflect protocol logistics rather than route of administration (oral vs. intravenous).

The thresholds for attack severity at enrollment differed across trials; earlier trials often required higher baseline symptom intensity (e.g., VAS ≥ 50 mm), consistent with the urgent need for effective rescue therapies in a setting with limited treatment options [17, 19, 22]. More recent trials enrolled broader patient populations, aligning with improved access to therapy and evolving treatment guidelines supporting treatment of all attacks, regardless of severity level, and have included patients with less severe attacks relative to legacy studies [14, 16, 19, 22, 23, 157].

Attack location reporting was inconsistent across studies. HAE attacks often involve multiple anatomic sites, with symptoms that evolve over time [4, 158]. The trials included in this SLR typically classified attacks using simplified categories (e.g., abdominal, peripheral) that may not fully capture the nuance and complexity of the disease. Treatment efficacy in relation to attack location was not captured consistently across trials, which may limit interpretation of location-specific efficacy data in some studies.

Variability in redosing and rescue medication protocols was evident across studies. Notably, earlier trials commonly redosed patients receiving placebo with active treatment, reflecting ethical considerations in the management of high-severity attacks [17–19]. In contrast, later trials, such as the sebetralstat KONFIDENT trial, redosed patients with the same blinded intervention initially assigned [23]. These differences complicate assumptions of equivalent placebo groups across studies and may influence observed treatment effects. For example, placebo-group outcomes could plausibly differ depending on whether placebo-assigned patients were redosed with active treatment or remained on placebo during redosing.

Censoring in the included trials occurred due to redosing, rescue medication use, or loss to follow-up; however, the timing and rules for censoring differed substantially. Within individual trials, censoring functions as an informative measure—higher censoring rates in the placebo group relative to the active treatment group may reflect greater use of rescue therapy or redosing among placebo-treated patients, providing indirect evidence of treatment benefit. This raises the possibility of a form of selection or healthy user bias whereby censoring rules may preferentially retain patients from the placebo group with more favorable trajectories, potentially influencing observed placebo arm estimates [159]. However, because this review was descriptive and did not include formal comparative statistical analyses, the potential impact of censoring on observed treatment effects cannot be directly quantified. Furthermore, censoring changes which patients are included in the analysis and when their data are no longer included, potentially influencing summary statistics used to describe treatment effects (i.e., mean, median, mode). In the sebetralstat KONFIDENT trial, censoring patients with fewer than two postbaseline assessments at hour 0 versus hour 12 yielded a median TOSR of 1.79 h versus 2.00 h, respectively, for the sebetralstat 600-mg arm [13, 23], highlighting how censoring rules can shape endpoint interpretation.

Heterogeneity in outcome measures further limits comparability of trials without further steps to define equivalencies between different outcomes and instruments [24]. Although 8 of the 12 trials used TOSR as the primary endpoint, 6 different PRO instruments were used to assess this endpoint, each with varying definitions of improvement and success [14, 16, 19, 20, 22, 23]. Only the rhC1-INH trials (C1 1310, C1 1205, and C1 1304) required both sustained response and sustained response without worsening to meet the TOSR endpoint, a strict and difficult criterion that was not applied in other trials [19, 22]. Consequently, it can be difficult to parse how much the observed differences in efficacy outcomes may be attributable to endpoint definitions and measurement tools as much as the pharmacologic interventions themselves.

Together these findings expand upon a report by Fijen et al., which systematically summarized efficacy outcome measures (e.g., MSCS, TEQ, TOSR) used in acute HAE trials and demonstrated substantial heterogeneity in endpoint definitions and PRO instruments across studies [24]. Our review extends beyond endpoint selection by examining endpoint criteria for success (Fig. 3) as well as differences in trial design, including treatment setting, time-to-treatment study designs, attack eligibility criteria, redosing and rescue approaches, and censoring methodology.

In 2024, the Acute Treatment Outcomes in Hereditary Angioedema (AURORA) initiative identified nine outcomes that achieved consensus as critically important [160]. These included time to first onset of symptom relief, time to complete resolution of all symptoms, rebound angioedema symptoms within 24 h after complete resolution of previous angioedema symptoms, and the need for medical intervention(s) during the entire attack. Following a protocol amendment at the direction of the steering committee to ensure a concise and clinically applicable core outcome set, a consensus-based core outcome set was subsequently established. This refined set comprised five outcomes for consideration in future acute HAE clinical studies: change in overall symptom severity at a prespecified time point between 15 min and 4 h after treatment, time to end of progression of all symptoms, need for rescue medication during the attack, impairment of daily activities, and treatment satisfaction [160].

These recommendations represent an important step toward improving consistency and interpretability in future trials. While diverse methods were used across numerous trials to capture measures of symptom severity, symptom resolution, and rescue medication use, impairment of daily activities and treatment satisfaction were not routinely incorporated in these earlier HAE trials. Importantly, the AURORA initiative defined the core outcome set, but consensus has not yet been reached regarding specific PRO instruments that measure these outcomes. Therefore, this review supports mapping between legacy and contemporary PRO instruments, and reanalysis of legacy clinical trial data using the recently established AURORA core outcome set.

This review is subject to limitations inherent to the SLR methodology. Data from many of the RCTs are inconsistently reported, with limited access to unpublished data, thereby introducing challenges during the synthesis of the evidence. To minimize this impact, efforts were made to review all publicly available data, not limited to only published articles. Despite these efforts, the major limitation of this review was the inability to directly compare the results of the trials. In lieu of direct comparative data from head-to-head trials, clinicians and payers may rely on indirect treatment comparisons (ITCs) as a basis for their clinical decision-making [161]. In a 2025 publication by Li et al., an ITC focusing on the comparative efficacy of rhC1-INH and sebetralstat failed to account for population differences, distinct redosing, rescue, and censoring protocols, and endpoints that were evaluated using different PRO measures and definitions of success [162]. Without proper adjustment for these potential treatment effect modifiers and prognostic variables, ITCs such as this can be misleading. Finally, this SLR was not registered in PROSPERO.

The present review did not formally evaluate the quantitative impact of these methodological differences on comparative efficacy estimates; rather, it identified sources of heterogeneity that should be considered when comparing data. Future research should evaluate which methodological differences function as true treatment effect modifiers and prognostic variables and further explore the harmonization of PRO instruments and endpoint definitions to support more reliable comparative analyses across existing HAE trials. The adoption of harmonized outcome definitions and core outcome sets may improve consistency and comparability across future HAE trials.

Clinical Practice Implications

Findings from this SLR indicate that apparent differences in efficacy outcomes across on-demand HAE therapies may be driven by heterogeneity in trial design. Clinicians should interpret efficacy results within the context of individual trial protocols, including differences in treatment timing, baseline severity requirements, endpoint definitions, and censoring protocols related to redosing or rescue therapy, which may not reflect real-world use. Recognition of these methodological differences may support more informed interpretations of treatment outcomes in clinical practice.

Conclusion

This SLR provides a comprehensive summary of clinical trials evaluating all FDA-approved on-demand HAE therapies. These findings highlight the variations in trial design, baseline characteristics, and PRO endpoint definitions, which may influence observed treatment effects. Rigorous alignment of endpoints and adjustment for clinically relevant effect modifiers will be essential for future comparative effectiveness research. Methodological innovations, such as mapping studies that translate between PRO instruments, reanalysis of trials using harmonized endpoint criteria, and statistical adjustments for population differences, may help bridge these gaps and support more reliable comparative efficacy research. This review also supports the need to map legacy endpoints to the recently established AURORA initiative core outcome set to improve interpretability across future HAE trials [160].

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Nihal Narsipur, Amanda Harrington, Paranjoy Saharia, and Sakshi Jindal made substantial contributions to the conception or design of the work and contributed to data acquisition and analysis. Nihal Narsipur, Amanda Harrington, Paranjoy Saharia, Sakshi Jindal, Raffi Tachdjian, Nami Park, Joseph R. Harper, Anurag Relan, and John Anderson contributed to data interpretation, provided critical feedback during manuscript development, and approved the final manuscript for publication.

Funding

Pharming Healthcare, Inc., sponsored this study. The journal’s rapid service fee and open-access fee were also funded by Pharming Healthcare, Inc.

Data Availability

Most data used within this manuscript are publicly available; other data are available upon reasonable request. Researchers who have a methodologically sound research proposal should send inquiries or requests to info@pharming.com. Data requestors may be required to sign a data access agreement.

Declarations

Conflict of Interest

Raffi Tachdjian has received research or consulting support from Astria Therapeutics, BioCryst Pharmaceuticals, Inc, CSL Behring, Ionis Pharmaceuticals, Intellia, KalVista Pharmaceuticals, Pharming Healthcare, Inc., Pharvaris, and Takeda Pharmaceutical Company Ltd; and provided speaker presentations for ARS Pharmaceuticals, AstraZeneca, BioCryst Pharmaceuticals, CSL Behring, Grifols, GSK, Pharming Healthcare, Inc., Sanofi-Regeneron, and Takeda Pharmaceuticals Company Ltd; John Anderson is a speaker bureau member for CSL Behring, Pharming Healthcare, Inc, BioCryst Pharmaceuticals, Inc, Takeda Pharmaceutical Company Ltd, AstraZeneca, and GSK; has received consulting fees from CSL Behring, Pharming Healthcare, Inc., BioCryst Pharmaceuticals, Inc, Pharvaris, Ionis Pharmaceuticals, Takeda Pharmaceutical Company Ltd, and Novartis AG; and is a clinical trial investigator for BioCryst Pharmaceuticals, Inc, CSL Behring, Pharvaris, KalVista Pharmaceuticals, Astria Therapeutics, Takeda Pharmaceutical Company Ltd, Novartis AG, and CellDex Therapeutics; Paranjoy Saharia and Sakshi Jindal received consulting fees from Pharming Healthcare, Inc.; Nami Park, Joseph R. Harper, Anurag Relan, Amanda Harrington, and Nihal Narsipur are employees of Pharming Healthcare, Inc.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.