Extrapolation of Fenfluramine Pharmacokinetics from Adults to Children: Estimating PK Exposure in Patients with Dravet Syndrome

Li Zhang,1 Brooks Boyd,2 Christopher M. Rubino1,3

1Institute for Clinical PharmacoDynamics, Schenectady, NY; 2Zogenix, Inc., Emeryville, CA; 3Department of Pharmaceutical Sciences, University at Buffalo, State University of New York, Buffalo, NY

INTRODUCTION f Fenfluramine is being developed as adjunctive therapy

for the treatment of seizures associated with Dravet syndrome, a rare and severe epileptic encephalopathy1-3

f Fenfluramine was previously marketed as an appetite suppressant for the treatment of adult obesity but was withdrawn from worldwide markets when its use, often in combination with phentermine, was associated with cardiovascular adverse events, including cardiac valvulopathy and primary pulmonary hypertension4

f The doses of fenfluramine being considered for Dravet syndrome are 0.2-0.8 mg/kg/day, with a maximum of 30 mg/day. Labeled doses used in adults for treatment of obesity were 60-120 mg/day, although there are reports in the literature of higher doses being used5

– In adults, the risk of cardiac valvulopathy was 9-fold higher in patients taking 60 mg/day or more compared with patients taking <40 mg/day5

f This study was performed to estimate the exposure to fenfluramine in children at doses between 0.2 and 0.8 mg/kg/day with a maximum dose of 30 mg/day

METHODS f The most recent published pharmacokinetic (PK) data

for fenfluramine or dexfenfluramine in adults was used to provide an estimate of the clearance for fenfluramine following oral dosing (CLt/F)6,7

– 2 additional studies provided additional fenfluramine PK data in adults8,9

f 2 studies that reported blood levels of fenfluramine in children with autism were examined to determine steady-state blood levels during treatment with 1.5 mg/kg/day10,11

– Inadequate information regarding the timing of blood collection relative to time of dosing made quantitative comparison of these data to the adult data infeasible

Monte Carlo Simulation f Monte Carlo simulation was performed to estimate the

fenfluramine exposure likely to be observed in children with Dravet syndrome receiving fenfluramine at doses of 0.2-0.8 mg/kg/day with a maximum dose of 30 mg/day

f Key assumptions – ZX008 contains racemic fenfluramine with equal parts

d- and l-fenfluramine – More robust PK data are available for d-fenfluramine.

Therefore, these simulations were based upon expected d-fenfluramine clearance, which was then converted to total fenfluramine clearance based upon the following assumptions: • PK parameters of the 2 enantiomers are similar• d-fenfluramine exposure is predictive of racemic

fenfluramine exposures in children and adults – Oral bioavailability (F) is 69%7

– Clearance of fenfluramine (CLt) scales across the weight range from adults to children according to an allometric relationship12:

CL child = CL adult • [body weight child (kg)/80 kg] 0.75

– Maturational factors are not relevant for fenfluramine when extrapolating to children aged 2 years and above12-14

f A simulated database of 2000 pediatric patients was constructed f Age was simulated to approximate a uniform age distribution from

2-18 years f The distribution of male and female patients was about 1:1 f A randomly sampled distribution was used to identify age- and

gender-specific values from the Centers for Disease Control and Prevention growth charts15 of pediatric subjects

f Estimation of d-fenfluramine clearance – Based on the literature values in adults, a mean CLt value of 40 L/h for

an 80 kg adult was assumed – For each simulated child, the following equation was used to estimate CLt

CLt = 40 L/h • [body weight (kg)/80 kg] 0.75

– CLt was further adjusted based on an assumed coefficient of variation (CV%) of 33%, which was randomly and log-normally applied to the calculated CLt

f Estimation of steady-state fenfluramine exposure in children – Assuming a dose of 0.2-0.8 mg/kg/day, with a maximum dose of

30 mg/day, administered in divided doses, BID – For each simulated patient, AUC0-24h was estimated

AUC0-24h = [(F • Dose (mg/kg) • body weight (kg)]/CLt f Estimation of steady-state fenfluramine exposure in adults

– A distribution of expected AUC0-24h in adults receiving 60-120 mg/day was generated assuming a mean of 40 L/h with a CV% of 33%

RESULTS f A dataset of 2000 simulated children aged 2-18 years was

generated – 52% were male – Distribution of ages is shown in Figure 1

Figure 1: Distribution of Age in Years for the Simulated Pediatric Dataset

Co

unt 50

75

25

0

2 4 6 8 10 12 14 16 18Age (yr)

f The relationship between age and body weight for male and female patients is shown in Figure 2

Figure 4: Steady-state Exposure to Fenfluramine in Adults and Children

Fenfluramine Dose Regimen

Adult, 20–40 TID (60–120 mg/d) Pediatric, 0.2–0.8 mg/kg/d

Fenf

lura

min

e S

tea

dy-

sta

te A

UC

0-24

h (m

g*h

/L)

0

2

4

6

Figure 5: Comparison of the Distribution of Fenfluramine Steady-state AUC0-24h Estimates in Children to Literature Values for Adults8

Caccia 1979 Acute (n=3)Caccia 1979 Chronic (n=8)Caccia 1979 Chronic (n=12)Caccia 1979 Chronic (n=8)Pediatric Simulation, 0.8 mg/kg/d

Fenf

lura

min

e S

tea

dy-

sta

teA

UC

0-24

h (m

g*h

/L)

2

3

1

0

30

Adult AUC values are normalized to a dose of 20 mg TID.

60 90 120Weight (kg)

CONCLUSIONS f Prediction of steady-state fenfluramine exposure in a simulated population of 2000 children with age-appropriate weights was enabled by applying allometric scaling to adult CLt estimates for fenfluramine

f The expected fenfluramine exposure from the proposed ZX008 dosing regimen (0.2-0.8 mg/kg/day; maximum 30 mg/day) is predicted to be significantly lower than that of adults treated with the originally approved fenfluramine product at labeled doses of 60-120 mg/day

Figure 2: Body Weight vs Age for the Simulated Pediatric Dataset

120

90

60

30

0

We

ight

(kg

)

Age (yr)2

FemaleMale

4 6 8 10 12 14 16 18

f Figure 3 shows the relationship between CLt and body weight in the simulated population and comparison with adult values. Estimated distribution of clearances as a function of body weight is consistent with values from the literature

Figure 3: Relationship Between CLt and Body Weight in the Simulated Population and Comparison to Adult Population from the Literature6,7

Cheymol IVCheymol POGross EMGross PM

Fenf

lura

min

e C

Lt (

L/h)

60

90

30

0

30

Black dots are simulated values in each of 2000 pediatric subjects; red line is mean CLt at each WTKG.EM, extensive metabolizers; PM, poor metabolizers; WTKG, weight in kg.

60 90 120Weight (kg)

f Fenfluramine exposures – Figure 4 presents box-and-whisker plots that show the distributions

of fenfluramine steady-state AUC0-24h estimates in simulated children treated with 0.2-0.8 mg/kg/day with a maximum dose of 30 mg/day compared with adults receiving 60-120 mg/day• Even at the highest expected dose of fenfluramine in children,

the exposure in children is expected to be less than that seen in adults

• Only 0.4% of the AUC estimates in the pediatric group were above the estimated adult mean values

– Figure 5 shows the distribution of fenfluramine steady-state AUC0-24h in simulated children treated with 0.2-0.8 mg/kg/day with a maximum dose of 30 mg/day compared with literature values in adults treated with 60-120 mg/day• Only the outlier values from predictions in children overlap the

mean reported values in adults from the literature

Presented as part of the Zogenix Scientific Exhibit during the 70th Annual Meeting of the American Epilepsy Society, December 2-6, 2016, Houston, TX

REFERENCES1. Ceulemans, B. et al. Epilepsia. 2016; 57: e129-342. Dravet, C. Epilepsia. 2011; 52 Suppl 2: 3-93. Schoonjans, A.-N. et al. Eur J Neurol. 2016. [Epub ahead of print]4. Onakpoya, I.J. et al. Crit Rev Toxicol. 2016; 46: 477-895. Li, R. et al. Int J Obes Relat Metab Disord. 1999; 23: 926-86. Gross, A.S. et al. Br J Clin Pharmacol. 1996; 41: 311-77. Cheymol, G. et al. Br J Clin Pharmacol. 1995; 39: 684-7

8. Caccia, S. et al. Eur J Drug Metab Pharmacokinet. 1979; 4: 129-329. Caccia, S. et al. Eur J Clin Pharmacol. 1985; 29: 221-410. Beeghly, J.H. et al. J Autism Dev Disord. 1987; 17: 541-811. Stern, L.M. et al. J Child Psychol Psychiatry. 1990; 31: 569-8512. Anderson, B.J. and Holford, N.H. Annu Rev Pharmacol Toxicol. 2008; 48: 303-3213. Salem, F. et al. J Clin Pharmacol. 2013; 53: 857-6514. Johnson, T.N. et al. Clin Pharmacokinet. 2006; 45: 931-5615. Centers for Disease Control and Prevention, National Center for Health Statistics. http://www.cdc.gov/growthcharts/

percentile_data_files.htm. Accessed January 19, 2011.

ACKNOWLEDGMENTSThe authors received professional medical writing and editing assistance in the preparation of this poster that was provided by Edward Weselcouch, PhD (PharmaWrite, LLC, Princeton, NJ) and was funded by Zogenix, Inc. (Emeryville, CA).