14 CLINICAL STUDIES

The safety and efficacy of BERINERT in the treatment of acute abdominal or facial HAE attacks in subjects with hereditary angioedema were demonstrated in a placebo-controlled, double-blind, prospective, multinational, randomized, parallel-group, dose-finding, three-arm, clinical study, referred to as the randomized clinical trial (RCT). The RCT assessed the efficacy and safety of BERINERT in 124 adult and pediatric subjects with C1 esterase inhibitor deficiency who were experiencing an acute moderate to severe attack of abdominal or facial HAE. Subjects ranged in age from six to 72 years of age; 67.7% were female and 32.3% were male; and approximately 90% were Caucasian.

The study objectives were to evaluate whether BERINERT shortens the time to onset of relief of symptoms of an abdominal or facial HAE attack compared to placebo and to compare the efficacy of two different doses of BERINERT. The time to onset of relief of symptoms was determined by the subject's response to a standard question posed at appropriate time intervals for as long as 24 hours after start of treatment, taking into account all single HAE symptoms. In addition, the severity of individual HAE symptoms was assessed over time.

Subjects were randomized to receive a single 10 IU/kg body weight dose of BERINERT (39 subjects), a single 20 IU/kg dose of BERINERT (43 subjects), or a single dose of placebo (42 subjects) by slow intravenous infusion (recommended to be given at a rate of approximately 4 mL per minute) within 5 hours of an HAE attack. At least 70% of the subjects in each treatment group were required to be experiencing an abdominal HAE attack.

If a subject experienced no relief or insufficient relief of symptoms by 4 hours after infusion, investigators had the option to administer a second infusion of BERINERT (20 IU/kg for the placebo group, 10 IU/kg for the 10 IU/kg group), or placebo (for the 20 IU/kg group). This masked (blinded) "rescue study medication" was administered to subjects and they were then followed until complete resolution of symptoms was achieved. Adverse events were collected for up to 7 to 9 days following the initial administration of BERINERT or placebo.

In the rare case that a subject developed life-threatening laryngeal edema after inclusion into the study, immediate start of open-label treatment with a 20 IU/kg body weight dose of BERINERT was allowed.

All subjects who received confounding medication (rescue medication) before symptom relief were regarded as "non-responders". Therefore, time to onset of symptom relief was set at 24 hours if a subject received any rescue medication (ie, rescue study medication, narcotic analgesics, non-narcotic analgesics, anti-emetics, open-label C1 inhibitor, androgens at increased dose, or fresh frozen plasma) between 5 hours before administration of blinded study medication until time to onset of relief.

For the trial to be considered successful, the study protocol specified the following criteria for the differences between the BERINERT 20 IU/kg and the placebo group:

The time to onset of relief of symptoms of the HAE attack had to achieve a one-sided p-value of less than 0.0249 for the final analysis, and at least one of the following criteria had to demonstrate a trend in favor of BERINERT with a one-sided p-value of less than 0.1:
- The proportion of subjects with increased intensity of clinical HAE symptoms between 2 and 4 hours after start of treatment with study medication compared to baseline, or
- The number of vomiting episodes within 4 hours after start of study treatment.

Subjects treated with 20 IU/kg body weight of BERINERT experienced a significant reduction (p=0.0016; "Wilcoxon Rank Sum test") in time to onset of relief from symptoms of an HAE attack as compared to placebo (median of 48 minutes for BERINERT 20 IU/kg body weight, as compared to a median of >4 hours for placebo). The time to onset of relief from symptoms of an HAE attack for subjects in the 10 IU/kg dose of BERINERT was not statistically significantly different from that of subjects in the placebo group.

Figure 9 is a Kaplan-Meier curve showing the percentage of subjects reporting onset of relief of HAE attack symptoms as a function of time. Individual time points beyond 4 hours are not presented on the graph because the protocol permitted blinded rescue medication, analgesics, and/or anti-emetics to be administered starting 4 hours after randomized blinded study medication had been administered.

*: Included rescue study medication (as blinded C1 inhibitor or placebo given as rescue medication), open-label C1 inhibitor, narcotic and non-narcotic analgesics, anti-emetics, androgens at increased dose, or fresh frozen plasma.

Figure 9. Time to Onset of Symptom Relief with Imputation to >4 hours for subjects Who Received any Rescue Medication* or Non-narcotic Analgesics Before Start of Relief

In addition, the efficacy of BERINERT 20 IU/kg body weight could be confirmed by observing a reduction in the intensity of single HAE symptoms at an earlier time compared to placebo. For abdominal attacks Figure 10 shows the time to start of relief of the last symptom to improve that was already present at baseline. Pre-defined abdominal HAE symptoms included pain, nausea, vomiting, cramps and diarrhea. Figure 11 shows the respective time to start of relief of the first symptom to improve that was already present at baseline.

*: Included rescue study medication (as blinded C1 inhibitor or placebo given as rescue medication), open-label C1 inhibitor, narcotic and non-narcotic analgesics, anti-emetics, androgens at increased dose, or fresh frozen plasma.

Figure 10. Time to Start of Relief of the Last Symptom to Improve (Abdominal HAE Attacks) with Imputation to >4 hours for subjects Who Received any Rescue Medication* Before Start of Relief

*: Included rescue study medication (as blinded C1 inhibitor or placebo given as rescue medication), open-label C1 inhibitor, narcotic and non-narcotic analgesics, anti-emetics, androgens at increased dose, or fresh frozen plasma.

Figure 11. Time to Start of Relief of the First Symptom to Improve (Abdominal HAE Attacks) with Imputation to >4 hours for subjects Who Received Any Rescue Medication* Before Start of Relief

For facial attacks, single HAE symptoms were recorded. In addition, photos were taken at pre-determined time points and assessed by the members of an independent Data Safety Monitoring Board (DSMB), who were blinded as to treatment, center and other outcome measures. The change in the severity of the edema when compared to baseline was assessed on a scale with outcomes "no change", "better", "worse" and "resolved". Figure 12 shows the time to start of relief from serial facial photographs by DSMB assessment.

*: Includes facial HAE attacks in subjects with concomitant abdominal HAE attacks.

Figure 12. Time to Start of Relief from Serial Facial Photographs*

Table 5 compares additional endpoints, including changes in HAE symptoms and use of rescue medication in subjects receiving BERINERT at 20 IU/kg body weight and placebo.

Table 5. Changes in HAE Symptoms and Use of Rescue Medication in subjects Receiving BERINERT 20 IU/kg Body Weight vs. Placebo
Additional Endpoints	Number (%) of subjects BERINERT 20 IU/kg Body Weight Group (n=43)	Number (%) of subjects Placebo Group (n=42)
* p-value = 0.033 † p-value = 0.00008
Onset of symptom relief within 60 minutes after administration of study medication (post-hoc)	27 (62.8%)	11 (26.2%)
Onset of symptom relief within 4 hours after administration of study medication	30 (69.8%)	18 (42.9%)
Number of vomiting episodes within 4 hours after start of study treatment*	6 episodes	35 episodes
Worsened intensity of clinical HAE symptoms between 2 and 4 hours after administration of study medication compared to baseline†	0 (0%)	12 (28.6%)
Number (percent) of combined abdominal and facial HAE attack subjects receiving rescue study medication, analgesics, or anti-emetics at any time prior to initial relief of symptoms	13 (30.2%)	23 (54.8%)
At least one new HAE symptom not present at baseline and starting within 4 hours after administration of study medication	2 (4.6%)	6 (14.3%)

Both the proportion of subjects with increased intensity of clinical HAE symptoms between 2 and 4 hours after start of treatment compared to baseline, and the number of vomiting episodes within 4 hours after start of study treatment demonstrated trends in favor of BERINERT in comparison to placebo (p-values <0.1). Tables 6 through 9 present additional information regarding responses to treatment.

Table 6. Proportion of subjects Experiencing Start of Self-Reported Relief of Symptoms by 4 hours by HAE Attack Type
HAE Attack Type	BERINERT 20 IU/kg Body Weight (Abdominal subjects = 34) (Facial subjects = 9) (Other subjects = 0)	Placebo Group (Abdominal subjects = 33) (Facial subjects = 8) (Other subjects = 1)*
* Laryngeal edema initially classified as facial edema.
Abdominal	24 (70.6%)	15 (45.5%)
Facial	6 (66.7%)	3 (37.5%)

Table 7. Proportion of subjects Experiencing Reduction in Severity of at Least One Individual HAE Attack Symptom by 4 hours
HAE Attack Type	BERINERT 20 IU/kg Body Weight (Abdominal subjects = 34) (Facial subjects = 9)	Placebo Group (Abdominal subjects = 33) (Facial subjects = 8)
Abdominal	33 (97.1%)	29 (87.9%)
Facial	6 (66.7%)	4 (50%)

Table 8. Proportion of subjects with Facial HAE Attacks Demonstrating Improvement in Serial Facial Photographs by 4 hours*
Attack Type	BERINERT 20 IU/kg Body Weight (Subjects = 9)	Placebo Group (Subjects = 8)
* Based on masked (blinded) evaluation by data safety monitoring board.
Facial	7 (77.8%)	2 (25%)

Table 9. Proportion of subjects with Abdominal and Facial HAE Attacks Receiving Rescue Study Medication at any Time Prior to Complete Relief of Symptoms
HAE Attack Type	BERINERT 20 IU/kg Body Weight	Placebo Group
HAE Attack Type	(Abdominal subjects = 34) (Facial subjects = 9)	(Abdominal subjects = 33) (Facial subjects = 8)
Abdominal	7 (20.6%)	17 (51.5%)
Facial	1 (11.1%)	6 (75%)

No subjects treated with BERINERT at 20 IU/kg body weight reported worsening of symptoms at 4 hours after administration of study medication compared to baseline.

The study demonstrated that the BERINERT 20 IU/kg body weight dose was significantly more efficacious than the BERINERT 10 IU/kg body weight dose or placebo.

Open-Label Extension Study

BERINERT was evaluated in a prospective, open-label, uncontrolled, multicenter extension study conducted at 15 centers in the US and Canada in subjects who had participated in the RCT study for the treatment of acute abdominal or facial HAE attacks in subjects with hereditary angioedema.

The purpose of this extension study was to provide BERINERT to subjects who had participated in the RCT study and who experienced any type of subsequent HAE attack (ie, abdominal, facial, peripheral, or laryngeal).

The safety analysis of the open-label extension study included a total of 57 subjects (19 males and 38 females, age range: 10 to 53 years) with 1085 HAE attacks treated with 20 IU/kg body weight dose of BERINERT per attack, who were observed at the study site until onset of relief of HAE symptoms, and were followed up for adverse reactions for 7 to 9 days following treatment of each HAE attack [see Adverse Reactions (6.1)]. During the extension study, 51 subjects experienced 747 abdominal attacks, 21 subjects experienced 51 facial HAE attacks, 30 subjects experienced 235 peripheral HAE attacks, and 16 subjects experienced 48 laryngeal HAE attacks. Some study subjects may have experienced HAE attacks in more than one location.

An analysis of laryngeal HAE attacks showed that the median time to initial onset of symptom relief and median time to complete resolution in the per-attack analysis were 0.25 hours and 8.4 hours, respectively (Table 10), which were the shortest times among the various HAE attack locations.

Table 10. Time to Initial Onset of Symptom Relief and Time to Complete Resolution of HAE Symptoms for Laryngeal HAE Attacks
Statistic	Laryngeal (n=48)
CI = confidence interval
HAE = hereditary angioedema
N = number of attacks
* The maximum time to complete resolution of 61.8 hours was an imputed value. Subject 29301 had 2 laryngeal HAE attacks with missing times to complete resolution of HAE symptoms, which were imputed with the maximum time to complete resolution of HAE symptoms observed for an abdominal HAE attack in this subject.
Time to initial onset of symptom relief [hours]
Median (range)	0.25 (0.10 - 1.25)
95% CI for median	[0.23; 0.42]
Time to complete resolution of HAE symptoms [hours]
Median (range)	8.4 (0.6 - 61.8*)
95% CI for median	[6.2; 21.5]

There were no clinically relevant or consistent data suggesting that gender, age group, race/ethnic group, type of HAE, routine use of androgens, or presence of detectable anti-C1 Esterase Inhibitor antibodies had an effect on the time to initial or complete relief of symptoms following BERINERT.

The prospective open-label extension study demonstrated that, in comparison to untreated historical control data retrospectively collected at a study center in Germany over a 20 year period8, the BERINERT 20 IU/kg body weight dose appeared to be effective in ameliorating laryngeal HAE attacks by achieving complete resolution of HAE symptoms within 24 hours from attack onset in the majority of subjects. The treatment effects observed with BERINERT in the extension study are consistent with the findings from the placebo-controlled efficacy trial.