The Journal of Emergency Medicine, Vol. 50, No. 4, pp.

567–580, 2016
Ó 2016 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
0736-4679

http://dx.doi.org/10.1016/j.jemermed.2015.11.008

Original
Contributions

EFFICACY OF DIFFERENT MEDICAL THERAPIES FOR THE TREATMENT OF ACUTE

LARYNGEAL ATTACKS OF HEREDITARY ANGIOEDEMA DUE TO C1-ESTERASE
INHIBITOR DEFICIENCY

Konrad Bork, MD,* Jonathan A. Bernstein, MD,† Thomas Machnig, MD,‡ and Timothy J. Craig, MD§
*Department of Dermatology, University of Mainz, Mainz, Germany, †University of Cincinnati Medical Center and Bernstein Clinical Research
Center, Cincinnati, Ohio, ‡CSL Behring GmbH, Marburg, Germany, and §Penn State University College of Medicine, Hershey, Pennsylvania
Reprint Address: Konrad Bork, MD, Department of Dermatology, University of Mainz, Johannes Gutenberg-Universität/Hautklinik,
Langenbeckstr. 1, Mainz 55131, Germany

, Abstract—Background: Hereditary angioedema (HAE)
is a rare disease characterized by C1-esterase inhibitor
(C1-INH) deficiency, resulting in periodic attacks of acute
edema, which can be life-threatening if they occur in the
upper airway. No head-to-head comparisons of different
treatment options for acute HAE attacks are available.
Because immediate symptom relief is critical for potentially
life-threatening laryngeal attacks, it is important to deter-
mine the treatment option that provides optimal treatment
response. Objective: Review and compare data from clinical
studies that evaluated the efficacy and safety of treatments
for laryngeal HAE attacks. Methods: We conducted an indi-
rect comparison of clinical outcomes from prospective
studies for treatment of 881 acute laryngeal attacks with
plasma-derived C1-INH concentrate (pdC1-INH) at fixed
doses (500 or 1000 U) or a body weight-adjusted dose (20
U/kg), recombinant C1-INH concentrate at a fixed dose
(2100 U), or a body weight-adjusted dose (50 U/kg), icatibant
Konrad Bork is a consultant of CSL Behring, Shire, and Viro-
Pharma. He has not received compensation for his work on this
manuscript. Jonathan A. Bernstein and Timothy J. Craig are
speakers/consultants of CSL Behring, Dyax, Shire, and Viro-
Pharma, and have received financial support for research from
CSL Behring, ViroPharma, Pharming, Shire, and Dyax. They
have not received compensation for their work on this manu-
script. Thomas Machnig is an employee of CSL Behring.
No institutional review board approval was required for this
study.
(30 mg), or ecallantide (30 mg). Comparisons included time
to onset of symptom relief and need for re-dosing or emer-
gency procedures. Results: The median time to onset of
symptom relief ranged between 15 min and approximately
2 h, and was shortest with body weight-adjusted doses of
pdC1-INH. The proportion of laryngeal attacks with
re-dosing ranged between 0% and 72%. No re-dosing
was needed after treatment with a single body weight-
adjusted dose of pdC1-INH (48 attacks). Conclusions:
Available data suggest that among different HAE treat-
ments, body weight-adjusted pdC1-INH (20 U/kg) provides
the most reliable treatment response for treatment of laryn-
geal HAE attacks. Ó 2016 The Authors. Published by
Elsevier Inc. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/).
, Keywords—C1-INH; efficacy; HAE; laryngeal;
re-dosing
INTRODUCTION
Hereditary angioedema (HAE) due to C1-esterase inhib-
itor (C1-INH) deficiency (HAE–C1-INH, hereafter called
HAE) is a rare autosomal dominant disorder caused by
reduced expression of normal C1-INH (type I HAE) or
expression of less functional C1-INH (type II HAE)
(1,2). Patients with HAE experience intermittent
episodic swellings that may affect the skin or

RECEIVED: 18 August 2015;

ACCEPTED: 10 November 2015

567
568 K. Bork et al.

gastrointestinal tract and are potentially life-threatening include efficacy in terms of treatment response and
in case of laryngeal attacks. Clinical symptoms of laryn- need for re-dosing.
geal attacks include hoarseness, stridor, dyspnea, the
feeling of having a lump in the larynx, dysphagia, and METHODS
voice change (3). Although laryngeal attacks are rare
Search Methods and Data Collection
(approximately 1% of HAE attacks), at least 50% of
patients with HAE experience a laryngeal attack at least
Reports on the treatment of laryngeal attacks with licensed
once in their lifetime (3). Mortality of 14% to 33% due
HAE treatments (see Table 1) were identified by a system-
to untreated and unrecognized laryngeal attacks has
atic database search in PubMed and EMBASE in May
been reported, highlighting the importance of early diag-
2015. The full search strategies are provided in Appendix
nosis of HAE and providing patients with appropriate
1 (available online). In addition, we searched the Internet,
treatment for potentially life-threatening HAE attacks
specifically, publications listed in clinicaltrials.gov for
(4,5).
completed HAE studies (as of April 2015) and Web sites
International consensus guidelines recommend
of regulatory agencies. Titles and abstracts from electronic
plasma-derived C1-INH concentrate (pdC1-INH;
databases and publications associated with completed
BerinertÒ [CSL Behring, King of Prussia, PA] or
HAE studies in clinicaltrials.gov were examined for eligi-
CinryzeÒ [Shire, Lexington, MA], at fixed doses of
bility. We obtained the full text of all relevant records. Data
1000 U and body weight-adjusted doses of 20 U/kg)
extracted on treatment of laryngeal attacks in patients with
or recombinant C1-INH concentrate (rhC1-INH,
HAE type I or II with different treatments included efficacy
RuconestÒ [Salix Pharmaceuticals, Inc., Raleigh, NC]),
endpoints, the need for re-dosing, and the need for emer-
bradykinin B2 receptor antagonist icatibant (FirazyrÒ
gency procedures (e.g., intubation).
[Shire]), or kallikrein inhibitor ecallantide (KalbitorÒ
[Dyax Corp., Burlington, MA]) for the treatment of acute
HAE attacks (6–10). Assessment of Risk of Bias
To date, no head-to-head comparisons of the efficacy
and safety of different treatment options for acute HAE The risk of selective outcome reporting was assessed
attacks are available. Because fast response to treatment based on the process described by Dwan et al. for all rele-
is especially important for potentially life-threatening vant studies identified in the systematic literature search
laryngeal attacks, we reviewed and compared available (11). For each study, we extracted results for the
data from clinical studies on treatment of laryngeal outcomes used in this comparison of treatment options
attacks that evaluated the efficacy and safety of different for laryngeal attacks. Outcomes that were not fully
treatment options for the rare but potentially life- reported but may have been assessed based on the
threatening cases of laryngeal attacks. Comparisons reported methods were highlighted. The assessment of

Table 1. Treatment Options for Acute HAE Attacks

Recommended Dose & Route

Trade Name (Manufactured for) Active Substance Licensed for of Administration

BerinertÒ (CSL Behring)
CinryzeÒ (ViroPharma)
RuconestÒ (Pharming)
FirazyrÒ (Shire)
KalbitorÒ (Dyax)
Human pasteurized,
nanofiltered pdC1-INH
Human pasteurized,
nanofiltered pdC1-INH
rhC1-INH
Icatibant, bradykinin 2
receptor antagonist
Ecallantide, kallikrein inhibitor
Europe: acute treatment & 20 U/kg, intravenous
short-term prophylaxis
US: acute treatment only
Europe: acute treatment & 1000 U, intravenous
prophylaxis
US: prophylaxis only
Acute treatment (not for 50 U/kg,* intravenous
treatment of laryngeal ($84 kg: 4200 U)
attacks in the US)
Acute treatment 30 mg,† subcutaneous
Acute treatment (US only) 3
10 mg,‡ subcutaneous

HAE = hereditary angioedema; pdC1-INH = plasma-derived C1-esterase inhibitor concentrate; rhC1-INH = recombinant C1-esterase in-
hibitor concentrate; US = United States.
* In case of insufficient clinical response, an additional dose can be administered. Not more than two doses should be administered within
24 h.
† In case of insufficient relief or recurrence of symptoms, additional injections may be administered at intervals of at least 6 h. No more than
3 injections should be administered within 24 h.
‡ If attack persists, an additional dose may be administered within 24 h.
Comparison of Treatments for Acute HAE Attacks 569

selective outcome reporting bias for each study was based
on the published results.
Data Analysis
The following treatment options were compared: pdC1-
INH at fixed doses of 500 or 1000 U or at body weight-
adjusted doses of 20 U/kg, rhC1-INH fixed doses of
2100 U or at body weight-adjusted doses of 50 U/kg,
30 mg of icatibant, and 30 mg of ecallantide.
Laryngeal attacks are potentially life-threatening, and
effective treatments are available for more than 30 years
in some countries. Thus, it is generally considered uneth-
ical to include laryngeal attacks in placebo-controlled
studies of HAE. Except for the treatment of five laryngeal
attacks in a placebo-controlled study of icatibant, data on
the treatment of laryngeal attacks are available only from
open-label studies (12).
All studies used patient-reported outcomes to assess
onset of symptom relief, although time to onset of
symptom relief was not the primary endpoint in all
studies, and definitions varied for studies with different
drugs [see Table 2 (13–19)]. As secondary efficacy
endpoint, some studies assessed the time to complete
resolution, others the duration of attacks, the time to
minimal symptoms, or the time to almost complete
symptom relief. Comparisons of time to complete
resolution of laryngeal attacks with different
treatments need to be interpreted with caution because
assessments were even more diverse between different

Table 2. Patient-reported Time to Onset of Symptom Relief for Laryngeal Attacks–Definitions and Assessments

Question Asked/
Drug Endpoint Assessment Time Points Reference

pdC1-INH (BerinertÒ)* Onset of symptom relief ‘‘Taking into account all Up to 24 h after Craig et al. (13),
(first of 2 consecutive symptoms, are you administration (20 Bernstein et al. (14)
reports) confident that the time points; every
attack is starting to 15 min for the first
improve?’’ 2 h, every 30 min for
the next 2 h, and at 5,
6, 7, 8, 12, 16, 20,
and 24 h after
administration)
Time to first signs of Assessed Patients were issued a Data on file
symptom resolution consecutively at patient diary and
or end of symptom least every 6 months asked to record data
progression by patient interviews on symptom
during office visits or resolution of HAE
by telephone attacks
pdC1-INH (CinryzeÒ) Time to unequivocal Symptoms (defining Up to 4 h after Riedl et al. (15)
relief (first of 3 site): ‘‘absent but administration or
consecutive reports) present before’’ or substantial relief (16
‘‘present, symptoms time points; every
better’’ or ‘‘absent 15 min)
now and before’’?
rhC1-INH Time to first onset of VAS decrease Up to 48 h after Moldovan et al. (16)
relief of symptoms by $ 20 mm administration (11
(first of 2 consecutive compared with time points; baseline,
reports) baseline for any 15 and 30 min, and 1,
eligible location 2, 4, 8, 12, 16, 24,
(abdominal, and 48 h after
urogenital, orofacial- administration)
pharyngeal-
laryngeal, or
peripheral)
Icatibant Patient-reported time to Time at which initial Patients were asked to Cicardi et al. (17),
initial symptom improvement of record the date and Malbrán et al. (18)
improvement (first symptoms was time when they felt
report) noticed their symptoms
started to improve
Ecallantide Beginning of Overall response Up to 24 h after Sheffer et al. (19)
improvement (first assessment: ‘‘a little administration (5
report) better’’ or ‘‘a lot time points; 1, 2, 3, 4,
better or resolved’’ and 24 h after
administration)

HAE = hereditary angioedema; pdC1-INH = plasma-derived C1-esterase inhibitor concentrate; rhC1-INH = recombinant C1-esterase in-
hibitor concentrate; VAS = visual analog scale.
* Definition used in the I.M.P.A.C.T.2 study.
570
studies and treatments than assessments of onset of
symptom relief.
Due to the heterogeneity of open-label studies and due
to variations in the definitions of efficacy endpoints, a
meta-analysis (i.e., an adjusted indirect comparison based
on differences between active treatment and placebo) was
not possible. However, it is extremely unlikely that HAE
treatments for laryngeal attacks will ever be directly
compared in a head-to-head analysis using validated
‘‘common ground’’ criteria for the assessment of treatment
response (20). Therefore, we performed a descriptive
comparison of the available efficacy data on the treatment
of laryngeal attacks with different drugs. Furthermore, the
percentages of laryngeal attacks treated with additional
doses were compared descriptively; 95% confidence inter-
vals (CIs) were calculated from available data.
Differences in treatment response (in terms of time to
onset of symptom relief and time from start of attack to
complete resolution of symptoms) between treatment
with pdC1-INH at fixed doses of 500 or 1000 U and
body weight-adjusted doses of 20 U/kg were assessed
descriptively, and by a Wilcoxon test.
RESULTS
Overview of Studies
The search strategy identified 196 articles, eight of which
were identified as eligible for the comparison of efficacy
for the treatment of laryngeal attacks. In addition, four
relevant articles and reports were identified based on pub-
lications listed in clinicaltrials.gov for completed HAE
studies and on Web sites of regulatory agencies.
A tabular overview of the 12 different eligible studies
or analyses, which included a total of 881 treated laryn-
geal attacks, is provided in Table 3 (12,13,15–19,21–
25). Even though some studies were controlled studies,
laryngeal attacks were excluded from randomized
treatment and were treated open-label, except for three
attacks treated with icatibant during the placebo-
controlled part of study FAST-3 (two laryngeal attacks
were treated with placebo in this study) (12).
Risk of Bias
The risk of selective outcome reporting was considered
low for most of the included studies and analyses
[Table 4 (13,15–19,21–25)]. It should be noted that the
focus of publications was often on other attack
locations that were included in double-blind, randomized
parts of the studies. Potentially life-threatening laryngeal
attacks were generally only treated open-label and were
therefore not always fully assessed and reported.
K. Bork et al.
Efficacy of Different Treatments (Per-attack Analysis)
The median time to onset of symptom relief after treat-
ment of laryngeal attacks was shortest with body
weight-adjusted doses of pdC1-INH and ranged between
15 min and approximately 2 h for the different drugs
(Figure 1). Fast onset of symptom relief was generally
achieved with all treatment options and was reported
within 1 h after the start of treatment for 60% to 100%
of laryngeal attacks, and within 4 h after the start of treat-
ment for 77% to 100% of laryngeal attacks [Table 5
(12,13,15,16,19,21,22,24)].
The median time from start of attacks to complete res-
olution of laryngeal attacks ranged between 9 and 16 h,
and the median time from treatment to complete resolu-
tion generally ranged between 3 and 12 h. However,
comparisons between treatments should be interpreted
with caution due to differences in the definition of time
to complete resolution.
The time between onset of symptoms and start of treat-
ment was stipulated in the study protocol for several
studies (between 4 and 8 h, see Table 3), but the actual
times to treatment are only rarely published. The median
time to treatment for laryngeal attacks was 3.15 h in the
I.M.P.A.C.T.2 study with the pdC1-INH Berinert, and
53% of laryngeal attacks were treated within # 4 h in
studies with the pdC1-INH Cinryze (13,15).
Efficacy of Different Doses of pdC1-INH
Most patients in the pdC1-INH body weight-adjusted and
fixed-dose groups experienced onset of symptom relief
within 4 h after injection (Table 5). In a per-patient analysis
of laryngeal attacks treated with the pdC1-INH Berinert, the
median average time to onset of symptom relief (95% CI)
was 26.1 min (18.6–41.5) in the body weight-adjusted
dose group and 42.5 min (30.0–60.0) in the fixed-dose
group, with the difference being significant (p = 0.019),
and with no significant differences between fixed doses of
500 vs. 1000 U pdC1-INH. This difference in time to onset
of symptom relief between the body weight-adjusted and
the fixed-dose group was also observed in a per-attack
Kaplan-Meier analysis (Figure 2). The median average
time from start of attack to complete resolution (95% CI)
in a per-subject analysis of laryngeal attacks treated with
the pdC1-INH Berinert was 10.5 h (5.1–29.4) in the body
weight-adjusted dose group and 17.8 h (16.0–27.0) in the
fixed-dose group. Also, this difference was significant
(p = 0.037), with no significant difference between the
500 and 1000 U dose group. However, the aforementioned
differences in definitions of endpoints make the comparison
of the studies difficult; therefore, the results have to be
considered with caution.

Table 3. Overview of Included Studies
Efficacy Outcomes for
Treatment of Laryngeal
Drug; Study Study Design Dose Regimen Attacks
BerinertÒ (pdC1-INH,
CSL Behring,
Marburg, Germany)
I.M.P.A.C.T.2 (13) Open-label 20 U/kg Time from start of
treatment to onset of
symptom relief, time
to complete
resolution of HAE
symptoms
Bork & Barnstedt (21) Case collection 500 or 1000 U Time to first signs of
symptom resolution
or end of symptom
progression,
duration of laryngeal
edema from first
symptoms to
attainment of
normality
CinryzeÒ (pdC1-INH,
ViroPharma,
Brussels, Belgium)
LEVP 2006-1 (15) Open-label 1000 U Time to unequivocal
relief (3 consecutive
reports), clinical relief
(1 or more
assessment of
improvement
followed by
discharge),
beginning of relief
(first assessment of
improvement)
Pooled analysis (15)* Placebo-controlled†/ 1000 U None*
open-label
Comparison of Treatments for Acute HAE Attacks
Patients with Number of
Notes Laryngeal Attacks Laryngeal Attacks
241
No restrictions for time 16, HAE type I or II 48
between onset of an
attack and start of
treatment (median:
3.15 h), no
stipulations
regarding re-dosing
Second 500-U injection 42, HAE type I or II 193
was administered if
symptoms did not
resolve within 30 to
60 min; if a second
dose was necessary
in 3 consecutive
attacks, patients
received 1000 U for
all subsequent
attacks
267
Treatment within 4 h 37, HAE type I or II 84
from onset of attack
(not always
possible); re-dosing
after 60 min if there
was no substantial
relief
See LEVP 2006-1; 85, HAE type I or II 267
determination of
clinical effectiveness
was confined to
study LEVP 2006-1
because data on time
to attack resolution
were not
systematically
(Continued )
571
Table 3. Continued

572
Efficacy Outcomes for
Treatment of Laryngeal Patients with Number of
Drug; Study Study Design Dose Regimen Attacks Notes Laryngeal Attacks Laryngeal Attacks

collected in the other

studies
RuconestÒ (rhC1-INH, 65
conestat alfa)
1304-01-OLE (16) Open-label 2100 U Time to first onset of Treatment within 6 h 24, HAE type I or II 45‡
symptom relief; time (presentation within
to ‘‘minimal’’ 5 h of onset with an
symptoms overall VAS severity
score of at least
50 mm and no
spontaneous
regression after 1 h);
up to 2 additional
vials (2100 U each)
could be
administered within
4h
Pooled analysis (22)§ Open-label 50 U/kg or 2100 U See above (1304-01- See above (1304-01- 25, HAE type I or II 33
OLE) OLE)
FirazyrÒ (icatibant) 88
FAST-1 and FAST-2 Placebo-controlled†/ 30 mg Median time to initial Treatment within 6 h; 11, HAE type I or II 11
(17) open-label symptom rescue medication
improvement (C1-INH, antiemetic
(patient-reported) agents, or opiates)
withheld for as long
as possible (ideally 8
or 9 h after initial
treatment)
FAST-1 OLE (18) Open-label 30 mg Median time to initial Further icatibant 19, HAE type I or II 37
symptom injections permitted
improvement if symptoms
(patient-reported), worsened within 48 h
change in of the initial treatment
investigator- (maximum of 3 doses
assessed symptom at least 6 h apart).
score severity Rescue medication
for relief of any
symptom permitted.
FAST-2 OLE (23) Open-label 30 mg Median time to initial See above (FAST-1) 10, HAE type I or II 30
symptom
improvement
(patient-reported),

K. Bork et al.
change in
investigator-

(Continued )
Table 3. Continued

Comparison of Treatments for Acute HAE Attacks

Efficacy Outcomes for
Treatment of Laryngeal Patients with Number of
Drug; Study Study Design Dose Regimen Attacks Notes Laryngeal Attacks Laryngeal Attacks

assessed symptom
score severity,
patient-assessed
symptom scores for
difficulty swallowing
and voice change
FAST-3 incl. OLE (12) Placebo-controlled†/ 30 mg Median times to 50% Treatment within 6 h 21, HAE type I or II 21
open-label reduction in 5- (severe laryngeal
symptom composite attacks were always
VAS score, onset of treated open-label);
primary symptom no stipulations
relief, reduction in regarding re-dosing
laryngeal symptom reported
score, almost
complete symptom
relief, and initial
symptom
improvement (patient
and investigator), any
reduction in laryngeal
VAS score, and any
reduction in patient-
and investigator-
assessed laryngeal
symptom score
Pooled analysis Open-label 30 mg Time to first symptom See above 60, HAE type I or II 60
(FAST-1, -2, and improvement
-3, incl. OLE; (according to
interim) (24,25) patient); Clinical
Global Improvement
as assessed by the
investigator
KalbitorÒ (ecallantide) 220
Pooled analysis Placebo-controlled†/ 30 mg Mean Symptom Treatment within 8 h 98, HAE type I or II 220
(EDEMA2, open-label Complex Severity (4 h in EDEMA2) of
EDEMA3, (MSCS) score, onset of a moderate
EDEMA4, DX-88/ Treatment Outcome or severe attack (no
19) (19) Score (TOS), restrictions on time
beginning of or intensity in DX-88/
improvement, onset 19); Re-dosing:
of sustained additional open-label
improvement, onset dose within 4 h for
of significant severe upper airway
improvement compromise

573
(Continued )
574 K. Bork et al.

‡ Forty-five orofacial/pharyngeal/laryngeal attacks were treated among the first five attacks of each patient. The total number of laryngeal attacks included in the study was not reported

§ Based on an interim pooled analysis of studies 1205-01-OLE (50 U/kg) and 1304-01-OLE (2100 U). Of 53 attacks at the oro-facial/pharyngeal/laryngeal location, 33 pharyngeal/laryn-
Need for Re-dosing or Emergency Procedures
Laryngeal Attacks

HAE = hereditary angioedema; OLE = open-label extension; pdC1-INH = plasma-derived C1-esterase inhibitor concentrate; rhC1-INH = recombinant C1-esterase inhibitor concentrate;

* Pooled analysis of 4 studies including LEVP 2006-1. No efficacy outcomes in terms of onset of relief or resolution, but assessment of number of infusions needed for each attack.
Number of
Across different treatments and studies, the proportion of
laryngeal attacks with re-dosing ranged between 0% and
72% (Figure 3). No re-dosing was needed for laryngeal
attacks treated with body weight-adjusted pdC1-INH
(20 U/kg). Of the laryngeal attacks treated with fixed
Laryngeal Attacks

doses of pdC1-INH (500 or 1000 U), 0% to 62% of

Patients with

attacks required re-dosing. The highest proportion of

attacks with re-dosing was reported for laryngeal attacks

† Irrespective of the overall study design, laryngeal attacks were treated open-label in all studies except for five laryngeal attacks in the FAST-3 study.
The total number of laryngeal attacks treated with a specific drug includes each attack only once, even if it was included in more than one analysis. treated with rhC1-INH at a fixed dose of 2100 U (72%)
(16). With icatibant, the proportion of laryngeal attacks
with re-dosing in the open-label extension phases of
FAST-1 and FAST-2 ranged between 7% and 14%, and
with ecallantide, the proportion of laryngeal attacks
EDEMA4); additional

for incomplete or no
response or relapse
between 4 and 24 h

with re-dosing was 10%.

open-label dose

Three patients were intubated, one had received a fixed

(EDEMA3 and
Notes

dose of 1000 U pdC1-INH, one was treated with icatibant,

and one was treated with ecallantide (15,17,19). No
emergency procedures were reported for any of the
other laryngeal attacks treated.

DISCUSSION
Treatment of Laryngeal
Efficacy Outcomes for

Potentially life-threatening laryngeal edema represents a

Underlined outcomes were used in the comparison of efficacy of different HAE treatment options.
Attacks

continuous threat to patients with HAE. Laryngeal attacks

mostly occur spontaneously and without warning, and can
even be the first clinical symptom of HAE (26). Despite the
availability of targeted and effective therapies for the
emergency management of HAE attacks, patients with
HAE still die from untreated laryngeal attacks (26–28).
Due to the danger of asphyxiation, it is vital that a
Dose Regimen

laryngeal edema be treated effectively as early as

possible, prior to reaching its maximum development.
Emergency procedures such as intubation or
tracheotomy can be avoided with treatment that provides
fast onset of symptom relief (5). Treatments for laryngeal
attacks should therefore ensure rapid treatment response
with a single dose, as the need for re-dosing and associated
progression of the laryngeal edema may endanger patients
Study Design

with HAE experiencing an acute laryngeal attack. Because

a large part of the medical community lacks knowledge of
geal attacks were included in the analysis.

the rare disorder of HAE, it is recommended that patients

have an emergency supply of an effective HAE treatment
at home or with them when they travel (29).
As no head-to-head comparisons of the efficacy of
VAS = visual analog scale.

and may be more than 45.

different treatment options for acute HAE attacks are

available to date, we reviewed and compared available
Table 3. Continued

data from clinical studies on treatment of laryngeal

Drug; Study

attacks to assess potential differences in efficacy and

need for re-dosing. The available data suggest that body
weight-adjusted doses of 20 U/kg pdC1-INH provide
the shortest median time to onset of symptom relief
(15 min), followed by approximately 30 to 45 min with
 Comparison of Treatments for Acute HAE Attacks
Table 4. Reporting of Outcome Measures Used for Descriptive Comparisons and Assessment of Selective Outcome Reporting

Outcome

Time to Onset of Symptom Relief

Time to Time from Risk of Bias Due to
Median Proportion Proportion Complete Start of Attack Emergency Selective Outcome
Drug; Study/Publication (95% CI) #1h #4h Resolution to Resolution Re-dosing Procedures Reporting

BerinertÒ (pdC1-INH, CSL Behring, Marburg, Germany)

I.M.P.A.C.T.2 (13) U U U U U U U No risk (all outcomes
reported)
Bork & Barnstedt (21) U* U U – U U U No risk (all outcomes
reported)
CinryzeÒ (pdC1-INH, ViroPharma, Brussels, Belgium)
LEVP 2006-1 (15) U U U – – U† U† No risk (all assessed
outcomes reported)
Pooled analysis (15)‡ – – – – – U U No risk (all assessed
outcomes reported)
RuconestÒ (rhC1-INH, conestat alfa)
1304-01-OLE (16) B – B B – B – High risk (partial reporting,
times to onset of
symptom relief, and
complete resolution
reported for only 20 of at
least 45 laryngeal
attacks [no explanation
for missing data]; no
total number of laryngeal
attacks; proportion of
patients with relief within
4 h only reported by
attack number for first
five attacks; no number
of attacks reported for
re-dosing, just
percentage)
Pooled analysis (22)§ U – U U – U U Not applicable (report from
regulatory agency, no
original publication)
FirazyrÒ (icatibant)
FAST-1 OLE (18) B – – – – U U Low risk (partial reporting
of time to onset of
symptom relief [no
overall results, just
subgroups by attack
number, legend of figure
unclear/incomplete])

(Continued )

575
Table 4. Continued

576
Outcome

Time to Onset of Symptom Relief

Time to Time from Risk of Bias Due to
Median Proportion Proportion Complete Start of Attack Emergency Selective Outcome
Drug; Study/Publication (95% CI) #1h #4h Resolution to Resolution Re-dosing Procedures Reporting

FAST-2 OLE (23) B B U – – U B (Rescue Low risk (partial reporting

medication) of time to onset of
symptom relief [no
overall results, just
subgroups by attack
number, legend of figure
unclear/incomplete]; no
reporting on emergency
procedures although it
can be assumed that it
was assessed)
FAST-1 and B† –† – – – U U Low risk (only partial
FAST-2 (17) reporting for time to
onset of relief)
FAST-3 incl. U – – U – – B (Rescue Low risk (no reporting on
OLE (12) medication) re-dosing and
emergency procedures
although it can be
assumed that it was
assessed)
Pooled analysis U U – – – U B (Rescue Not applicable (reports
(24,25)k medication) from regulatory
agencies, no original
publication)
Pooled analysis U – U U – U U Low risk (it can be assumed
(19){ that data on proportion
of patients with onset of
symptom relief of # 1 h
are available; however,
not a defined outcome in
this study)

CI = confidence interval; OLE = open-label extension; pdC1-INH = plasma-derived C1-esterase inhibitor concentrate; rhC1-INH = recombinant C1-esterase inhibitor concentrate;
SD = standard deviation.
U indicates full reporting; B indicates partial reporting; - indicates no reporting.
* Data on file for median (95% CI) by dose (500 or 1000 U), publication reports mean (SD) for all attacks (irrespective of dose).
† Included in pooled analysis.
‡ Pooled analysis of 4 studies (LEVP 2005-1/A and B, LEVP 2006-1, and LEVP 2006-4).
§ Pooled analysis of interim data from studies 1205-01-OLE (50 U/kg) and 1304-01-OLE (2100 U).

K. Bork et al.
k Pooled analysis of interim data from studies FAST-1, -2, and -3, incl. OLEs.
{ Pooled analysis of 4 studies (EDEMA2, EDEMA3, EDEMA4, and DX88/19).
Comparison of Treatments for Acute HAE Attacks 577

Analysis (19)

CI = confidence interval; n = number of attacks; n.a. = data not available; OLE = open label extension; pdC1-INH = plasma-derived C1-esterase inhibitor concentrate; rhC1-
Ecallantide
pdC1-INH (Berinert) (13, 21) 15

n = 220

2.8–3.8
Pooled

30 mg,
20 U/kg (N=48)
30

3.1
n.a.

n.a.
n.a.
500 U (N=48)

88
1000 U (N=145)
pdC1-INH (Cinryze) (15) 40
45
1000 U (N=82)
rhC1-INH (17, 18)
115
50 U/kg or 2100 U (N=33)
subgroup: 2100 U (N=20)
FAST-3 (12)

2.2–24.3
Icatibant (12, 25) 120
30 mg,

36
n = 21

FAST-1+2+3 (N=60)
n.a.
n.a.

n.a.
n.a.

6.0
subgroup: FAST-3 OLE (N=21)
Ecallantide (19) 42
93
Pooled analysis (N=216)
Icatibant

0 20 40 60 80 100 120 140

Median time to onset of symptom relief [minutes]
FAST-1+2 (24)

± 95% confidence intervals

30 mg,
n = 61

Figure 1. Median time to onset of symptom relief for laryn-

62.3
n.a.

n.a.
n.a.

n.a.
n.a.
geal attacks (per-attack analysis). The whiskers show the
95% confidence intervals. N = Number of attacks;
OLE = open-label extension; pdC1-INH = plasma-derived
C1-esterase inhibitor concentrate; rhC1-INH = recombinant
C1-esterase inhibitor concentrate. aData on file. bTwo of 84
† Although more laryngeal attacks were treated in study 1304-01-OLE, time to complete resolution was only reported for 20 attacks.
1304-01-OLE (16)

attacks did not have time to beginning of unequivocal relief

reported. cNo confidence intervals available. dIncluding
2100 U†,
n = 20

12.1

data from 27 attacks treated in FAST-1, 12 attacks treated

n.a.
n.a.

n.a.
n.a.

n.a.

* Based on an interim pooled analysis of studies 1205-01-OLE (dosing with 50 U/kg) and 1304-01-OLE (dosing with 2100 U).

in FAST-2, and 21 attacks treated in FAST-3. eFour of 220 at-

tacks did not have beginning of improvement reported.
rhC1-INH

fixed doses of pdC1-INH or icatibant, and approximately

50 U/kg or 2100 U*,
Pooled Analysis(22)

1.5 h with ecallantide and 2 h with rhC1-INH. No

re-dosing was needed for laryngeal attacks treated with
4.0–15
n = 33

4.4
90.9
n.a.

n.a.
n.a.

body weight-adjusted doses of 20 U/kg pdC1-INH.

‡ Definitions for time to complete resolution: minimal symptoms or almost complete symptom relief.

Whereas in one study, a second dose was needed for

30% of laryngeal attacks treated with 500 U pdC1-INH,
a single dose was sufficient for attacks treated with
Table 5. Time to Onset of Symptom Relief for Laryngeal Attacks (Per-attack Analysis)

1000 U. In another study, using a different pdC1-INH,

LEVP 2006-1 (15)

62% of the attacks treated with 1000 U required re-

1000 U,
n = 84

dosing. Therefore, the available data suggest that a

59.5
77.4

n.a.
n.a.

n.a.
n.a.

body weight-based dosing regimen of 20 U/kg pdC1-

INH provides faster and more reliable treatment response
than fixed doses of 500 or 1000 U pdC1-INH.
More than 70% of laryngeal attacks treated with rhC1-
1000 U,
n = 145

9.0–9.0
Bork & Barnstedt (21)

INH at a fixed dose of 2100 U needed a second dose to

9.0
99.5

n.a.
n.a.
100

achieve symptom relief. With icatibant or ecallantide,

pdC1-INH

the percentage of laryngeal attacks with re-dosing (be-

tween 7% and 14%) was between that observed with
INH = recombinant C1-esterase inhibitor concentrate.
16.0–17.5
500 U,
n = 48

Time from start of attack to complete resolution (h)

16.0

n.a.
n.a.
100
100

Time from treatment to complete resolution‡ (h)

100

90

80
I.M.P.A.C.T.2 (13)

Percentage of attacks (%)

20 U/kg, n = 48

70
9.5–26.8

6.2–21.5

60
93.8

15.0

8.4
100

50
Onset of relief (% of attacks)

40
500 U Berinert (N=48; censored: 0)
30
1000 U Berinert (N=145; censored: 1)
20
Efficacy Endpoint

1000 U Cinryze (N=82; censored: 20)

10
20 U/kg Berinert (N=48; censored: 0)
Within 1 h
Within 4 h

0
0 15 30 45 60 75 90 105 120
95% CI

95% CI
Median

Median

Time to onset of symptom relief (minutes)

Figure 2. Time to onset of symptom relief for different doses

of pdC1-INH (per-attack analysis). N = Number of attacks.
578
pdC1-INH (Berinert) (13, 21)
0%
20 U/kg (N=48)
30%
500 U (N=69)
0%
1000 U (N=124)
pdC1-INH (Cinryze) (15)
Pooled analysis 1000 U (N=265)
rhC1-INH (16, 22) 20%
1205-01-OLE 50 U/kg (N=20)
1304-01-OLE 2100 U (N=45)
Icatibant (18, 23) 14%
FAST-1 OLE (N=37)
FAST-2 OLE (N=30)
7%
Ecallantide (19) 10%
Pooled analysis (N=220)
0 10 20 30
Percent of attacks with re-dosing
± 95% confidence intervals
Figure 3. Need for re-dosing after laryngeal attacks. N =
Number of attacks; OLE = open-label extension; pdC1-
INH = plasma-derived C1-esterase inhibitor concentrate;
rhC1-INH = recombinant C1-esterase inhibitor concentrate.
a
For two of 267 attacks, dosing data were not available. bAlth-
ough the total number of laryngeal attacks was not provided
in Moldovan et al. and may be higher, it was reported that 45
laryngeal attacks were treated among the first five attacks of
each patient (16). Re-dosing frequency was calculated based
on this number and deviates from the number of attacks
requiring re-dosing given in Moldovan et al. (72%) due to
rounding (16).
body weight-adjusted doses of pdC1-INH (no re-dosing)
and that observed with rhC1-INH. Emergency procedures
like intubation were reported for only three of 881 treated
laryngeal attacks.
Overall, the descriptive comparison of available data
from clinical studies on treatment of laryngeal HAE
attacks suggests that pdC1-INH at a single body
weight-adjusted dose of 20 U/kg provides the most reli-
able treatment response. If pdC1-INH is not available,
all other treatment options analyzed in this study are
also appropriate for the treatment of laryngeal attacks
and provide rapid and reliable relief. However, it should
be noted that differences in efficacy or need for
re-dosing in this descriptive comparison may be biased
by methodological differences between studies.
Limitations
Due to the lack of data from placebo control arms (consid-
ered unethical for potentially life-threatening laryngeal
attacks) and the heterogeneity of studies, a meta-
analysis (i.e., an adjusted indirect comparison) of different
treatments for laryngeal attacks based on the treatment
response relative to placebo was not possible. In addition
to the inherent problems of a descriptive across-study
comparison, other limitations of this comparison include
differences in study design and definitions of efficacy end-
points. For example, some study protocols stipulated
re-dosing at a specific time when symptom relief had
not been reported (e.g., after 1 h for Cinryze), whereas
other study protocols did not specify a time point for
K. Bork et al.
re-dosing (e.g., I.M.P.A.C.T.2 study with Berinert). Dif-
ferences in the percentage of attacks with reported time
to onset of symptom relief between fixed doses of 1000
U of the pdC1-INH products Cinryze and Berinert
62% (Figure 2) should be interpreted with caution because
they may be confounded by differences in study design
71% and timing of assessments (see Tables 2 and 3). Time to
onset of symptom relief and complete resolution may
also be biased by different times from start of an attack
to treatment, which are only rarely reported. It should
40 50 60 70 80 90 100
also be noted that the high rate of re-dosing and the rela-
tively long time to onset of symptom relief observed
with rhC1-INH may be biased by the fact that there was
no clear distinction between laryngeal and facial attacks
in the studies with rhC1-INH; all ‘‘orofacial-pharyngeal-
laryngeal’’ attacks were included in the analysis (16).
CONCLUSIONS
Due to the differences in study design and limitations of
descriptive cross-study comparisons, the results of this
study can only be hypothesis-generating. A prospective
study comparing different treatment options, using vali-
dated ‘‘common ground’’ criteria for efficacy assess-
ments, and thoroughly investigating differences that
may exist between treatments for laryngeal attacks
regarding efficacy would be ideal, but is extremely
unlikely to be conducted given the complexity of a
head-to-head trial in a clinical emergency condition
such as an acute laryngeal attack and given the small
size of the population of HAE patients. Therefore,
clinician-driven consensus recommendations can only
resort to the comparative value of the available data on
treatment of laryngeal attacks. Currently available clin-
ical evidence suggests that pdC1-INH at a body weight-
adjusted dose of 20 U/kg should be considered as the
preferred treatment choice for treatment of acute laryn-
geal HAE attacks to ensure optimal treatment response.
Acknowledgments—We thank Heinz-Otto Keinecke (Accovion
GmbH, Marburg, Germany) for statistical consultancy on behalf
of CSL Behring GmbH, and Christina Eichhorn and Eva Kest-
ner (Trilogy Writing & Consulting GmbH, Germany) for med-
ical writing services on behalf of CSL Behring GmbH.
REFERENCES
1. Longhurst H, Cicardi M. Hereditary angio-oedema. Lancet 2012;
379:474–81.
2. Cicardi M, Aberer W, Banerji A, et al. Classification, diagnosis, and
approach to treatment for angioedema: consensus report from the
Hereditary Angioedema International Working Group. Allergy
2014;69:602–16.
Comparison of Treatments for Acute HAE Attacks
3. Bork K, Meng G, Staubach P, Hardt J. Hereditary angioedema: new
findings concerning symptoms, affected organs, and course. Am J
Med 2006;119:267–74.
4. Moore GP, Hurley WT, Pace SA. Hereditary angioedema. Ann
Emerg Med 1988;17:1082–6.
5. Bork K, Hardt J, Witzke G. Fatal laryngeal attacks and mortality in
hereditary angioedema due to C1-INH deficiency. J Allergy Clin
Immunol 2012;130:692–7.
6. Bowen T, Cicardi M, Farkas H, et al. 2010 International consensus
algorithm for the diagnosis, therapy and management of hereditary
angioedema. Allergy Asthma Clin Immunol 2010;6:24.
7. Bowen T. Hereditary angioedema: beyond international consensus
– circa December 2010 – The Canadian Society of Allergy and Clin-
ical Immunology Dr. David McCourtie Lecture. Allergy Asthma
Clin Immunol 2011;7:1.
8. Cicardi M, Bork K, Caballero T, et al. Evidence-based recommen-
dations for the therapeutic management of angioedema owing to
hereditary C1 inhibitor deficiency: consensus report of an Interna-
tional Working Group. Allergy 2012;67:147–57.
9. Craig T, Aygören-Pürsün E, Bork K, et al. WAO guideline for the
management of hereditary angioedema. World Allergy Organ J
2012;5:182–99.
10. Zuraw BL, Banerji A, Bernstein JA, et al. US Hereditary Angioe-
dema Association Medical Advisory Board 2013 recommendations
for the management of hereditary angioedema due to C1 inhibitor
deficiency. J Allergy Clin Immunol Pract 2013;1:458–67.
11. Dwan K, Gamble C, Kolamunnage-Dona R, et al. Assessing the po-
tential for outcome reporting bias in a review: a tutorial. Trials 2010;
11:52.
12. Lumry WR, Li HH, Levy RJ, et al. Randomized placebo-controlled
trial of the bradykinin B2 receptor antagonist icatibant for the treat-
ment of acute attacks of hereditary angioedema: the FAST-3 trial.
Ann Allergy Asthma Immunol 2011;107:529–37.
13. Craig TJ, Bewtra AK, Bahna SL, et al. C1 esterase inhibitor concen-
trate in 1085 Hereditary Angioedema attacks - final results of the
I.M.P.A.C.T.2 study. Allergy 2011;66:1604–11.
14. Bernstein JA, Ritchie B, Levy RJ, et al. Hereditary angioe-
dema: validation of the end point time to onset of relief by cor-
relation with symptom intensity. Allergy Asthma Proc 2011;32:
36–42.
15. Riedl MA, Lumry WR, Li HH, et al. Nanofiltered C1 esterase inhib-
itor for treatment of laryngeal attacks in patients with hereditary
angioedema. Am J Rhinol Allergy 2013;27:517–21.
16. Moldovan D, Reshef A, Fabiani J, et al. Efficacy and safety of
recombinant human C1-inhibitor for the treatment of attacks of
579
hereditary angioedema: European open-label extension study.
Clin Exp Allergy 2012;42:929–35.
17. Cicardi M, Banerji A, Bracho F, et al. Icatibant, a new bradykinin-
receptor antagonist, in hereditary angioedema. N Engl J Med 2010;
363:532–41.
18. Malbrán A, Riedl M, Ritchie B, et al. Repeat treatment of acute he-
reditary angioedema attacks with open-label icatibant in the FAST-1
trial. Clin Exp Immunol 2014;177:544–53.
19. Sheffer AL, MacGinnitie AJ, Campion M, et al. Outcomes after
ecallantide treatment of laryngeal hereditary angioedema attacks.
Ann Allergy Asthma Immunol 2013;110:184–8.
20. Caballero T. Efficacy assessments in randomized controlled studies
of acute therapy for hereditary angioedema. J Clin Immunol 2012;
32:1204–12.
21. Bork K, Barnstedt SE. Treatment of 193 episodes of laryngeal
edema with C1 inhibitor concentrate in patients with hereditary an-
gioedema. Arch Intern Med 2001;161:714–8.
22. European Medicines Agency. CHMP assessment report: Rucon-
est; June 2010. Available at: http://www.ema.europa.eu/docs/en_
GB/document_library/EPAR_-_Public_assessment_report/human/
001223/WC500098546.pdf. Accessed July 31, 2015.
23. Baş M, Greve J, Hoffmann TK, et al. Repeat treatment with icati-
bant for multiple hereditary angioedema attacks: FAST-2 open-la-
bel study. Allergy 2013;68:1452–9.
24. European Medicines Agency. CHMP assessment report for Firazyr;
2008. Available at: http://www.ema.europa.eu/docs/en_GB/document_
library/EPAR_-_Public_assessment_report/human/000899/WC
500022970.pdf. Accessed July 31, 2015.
25. FDA Advisory Committee Briefing Package, Firazyr, June 2011.
Available at: http://www.fda.gov/downloads/AdvisoryCommittees/
CommitteesMeetingMaterials/Drugs/Pulmonary-AllergyDrugs
AdvisoryCommittee/UCM260022.pdf. Accessed July 31,2015.
26. Bork K, Siedlecki K, Bosch S, Schopf RE, Kreuz W. Asphyxiation
by laryngeal edema in patients with hereditary angioedema. Mayo
Clin Proc 2000;75:349–54.
27. Moellman JJ, Bernstein JA. Diagnosis and management of heredi-
tary angioedema: an emergency medicine perspective. J Emerg
Med 2012;43:391–400.
28. Bork K, Barnstedt SE. Laryngeal edema and death from asphyxia-
tion after tooth extraction in four patients with hereditary angioe-
dema. J Am Dent Assoc 2003;134:1088–94.
29. Craig TJ, Wasserman RL, Levy RJ, et al. Prospective study of rapid
relief provided by C1 esterase inhibitor in emergency treatment of
acute laryngeal attacks in hereditary angioedema. J Clin Immunol
2010;30:823–9.
580 K. Bork et al.

ARTICLE SUMMARY
1. Why is the topic important?
Acute attacks of hereditary angioedema (HAE) can be
life-threatening if they occur in the upper airway. Prompt
and effective therapies are available and can help avoid
emergency procedures such as intubation or tracheotomy
and decrease the rate of mortality among patients with this
condition.
2. What does this study attempt to show?
We reviewed and compared data from clinical studies
that evaluated the efficacy and safety of different treat-
ment options for acute laryngeal HAE attacks in terms
of time to onset of symptom relief, need for re-dosing,
and need for emergency procedures. Treatment options
included plasma-derived C1-esterase inhibitor concen-
trate (pdC1-INH), recombinant C1-esterase inhibitor
concentrate, icatibant, and ecallantide.
3. What are the key findings?
Of the different treatment options, pdC1-INH at body
weight-adjusted doses of 20 U/kg provides the most reli-
able response. However, our findings indicated that, if
pdC1-INH is not available, all other options are also
appropriate in providing rapid and reliable relief.
4. How is patient care impacted?
A prospective study comparing different treatment op-
tions would be ideal, but is extremely unlikely to be con-
ducted given the complexity of a head-to-head trial in a
clinical emergency condition such as an acute laryngeal
attack and the small size of the population of HAE pa-
tients. Because clinician-driven consensus recommenda-
tions can therefore only resort to the comparative value
of the available data, our comparison provides useful in-
formation on the emergency management of laryngeal at-
tacks of HAE. It aims at raising awareness of HAE, the
risk of asphyxiation during an (untreated) laryngeal
attack, and the availability of safe and effective therapies
among health care professionals and patients.
Comparison of Treatments for Acute HAE Attacks 580.e1

APPENDIX 1 #18 #17 AND (‘conference abstract’/it OR ‘confer-

ence paper’/it OR ‘review’/it)
EMBASE search criteria:
#19 #17 NOT #18
#20 #17 NOT #18 AND [English]/lim AND [humans]/
#1 ‘hereditary angioedema’/exp OR ‘hereditary angioe-
lim
dema’
#2 ‘larynx’/exp
PubMed search criteria:
#3 ‘laryngeal’
#4 ‘larynx’
‘‘hereditary angioedema’’[All Fields] AND (‘‘larynx’’
#5 #2 OR #3 OR #4
[MeSH Terms] OR ‘‘larynx’’[All Fields] OR ‘‘laryngea-
#6 #1 AND #5
l’’[All Fields]) NOT acquired[All Fields] AND ((‘‘ecal-
#7 ‘ecallantide’/exp
lantide’’[Supplementary Concept] OR ‘‘ecallantide’’[All
#8 ‘c1 inhibitor’/exp
Fields]) OR ‘‘C1 inhibitor’’[All Fields] OR (‘‘icatibant’’
#9 ‘icatibant’/exp
[Supplementary Concept] OR ‘‘icatibant’’[All Fields])
#10 ‘serping1 protein human’
OR (‘‘SERPING1 protein, human’’[Supplementary
#11 ‘cinryze’/exp
Concept] OR ‘‘SERPING1 protein, human’’[All Fields]
#12 ‘complement component c1s inhibitor’
OR ‘‘cinryze’’[All Fields]) OR berinert[All Fields]
#13 ‘berinert’
OR (‘‘conestat alpha’’[Supplementary Concept] OR
#14 ‘conestat alfa’
‘‘conestat alpha’’[All Fields] OR ‘‘ruconest’’[All
#15 ‘ruconest’
Fields])) AND english[Language] NOT review[Publica-
#16 #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13
tion Type]
OR #14 OR #15
#17 #6 AND #16