• The higher BUP-to-NBUP ratio (0.7—19.19) than adults (0.165—1.4) has been observed in newborn patients studied. It might be due to immature hepatic function in newborns and compliance to BUP sublingual administration.

• ID 6 showed the two highest BUP/NBUP ratio among 12 newborns examined. It might be due to large percentage of BUP dose not being administrated (medical record errors) and/or analytical measurement errors.

• Large individual variability has been observed in sublingual bioavailability . Higher values of bioavailability observed might be due to underestimated CLsub and/or overestimated CLiv.

• Sublingual bioavailability was estimated from 13.5% to 56.6% in the 9 neonates treated exclusively with BUP during the first postnatal month. The bioavailability of these 9 neonates except ID 28 was increased gradually by 13% at least during this period. Linear increase trend is observed when plotting sublingual bioavailability versus PMA for each of the 9 neonates studied.

• Sublingual bioavailability of ID 13, 22, and 27 showed less linear increase with PMA, compared with that of all other patients. It might be due to enzyme induction of CYP 3A, CYP 2C, and UGTs caused by phenobarbital.

• Growth factors such as age, body weight can be important covariates to BUP exposure levels in newborns, given the fact of the significant changes of body fat content and enzyme levels of CYP 3A, 2C8, 2C9, 2C18 and 2C19.

• Dose adjustment is needed for BUP therapy in newborns based on lower sublingual bioavailability estimated in newborn patients, compared to 50% of BUP sublingual bioavailability observed in adults, and drug-drug interaction induced by phenobarbital. A larger, double-blind, properly powered clinical trial on BUP in young infants will enhance and validate this model and simulation.

CONCLUSIONS

• Nonlinear mixed-effect modeling was employed to characterize the pharmacokinetics of BUP and NBUP based on data from a pilot clinical trial in 12 newborn patients with NAS (Pediatrics, 2008, 122, e601-607).

• Population PK analysis was performed using non-linear mixed-effects modeling with the NONMEM software, Version VI, Level 1.1.  

• One compartment model with first order absorption, metabolism, and elimination was developed in describing the PK of BUP and NBUP.

MODEL DEVELOPMENT

1. To characterize PK and sublingual bioavailability of buprenorphine (BUP) in target population during newborn period.

2. To evaluate the developmental changes in newborns in order to assist dosing optimization in ongoing clinical studies.

OBJECTIVES

Background: About 55% to 94% of infants born to opioid dependent mothers have neonatal abstinence syndrome (NAS). Buprenorphine (BUP) is used clinically as an analgesic and a detoxification agent and a maintenance treatment for opioid dependence. No data, however, has been reported about the use of sublingual administration of BUP below the age of 4 year, especially for term infants with NAS. Objectives: Characterize pharmacokinetics (PK) of BUP in newborn patients; Evaluate the developmental changes in newborns in order to assist dosing optimization in ongoing clinical studies. Methods: In silico prediction of PK behavior and physiological development in newborn patients were evaluated using SIMCYP. Intravenous clearance was predicted through physiologically based simulation method in SIMCYP. Based on sublingual clearance obtained from a one compartmental model developed previously using NONMEM, individual changes of sublingual bioavailability were evaluated with physiological development in the first one and half month during the newborn period. Results: Intrinsic clearance of BUP in newborns were incorporated into enzyme kinetic data obtained from literature. Change of sublingual bioavailability for newborns was evaluated with bioavailability-postmenstrual age profiles. Sublingual bioavailability of BUP was estimated as 8.9--56.6% in newborn patients studied during the first one and half postnatal month. Conclusion: Developmental considerations for the PK of BUP in newborns are important for the characterization of the dose-exposure relationship. We have evaluated this from “bottom-up” and “top-down” approaches with SIMCYP and NONMEM respectively and found these approaches to be complementary and valuable for clinical trial design and routine clinical care. Presumably they would facilitate rational decision making in pediatric drug development as well.

ABSTRACT

Prediction of Sublingual Bioavailability of Buprenorphine in Newborns with Neonatal Abstinence Prediction of Sublingual Bioavailability of Buprenorphine in Newborns with Neonatal Abstinence Syndrome— a case study on physiological and developmental changes using NONMEM and SIMCYPSyndrome— a case study on physiological and developmental changes using NONMEM and SIMCYP

Di Wu1, Walter K Kraft2, Michelle E Ehrlich3, Jeffrey S Barrett1,4 1Division of Clinical Pharmacology & Therapeutics, The Children’s Hospital of Philadelphia, Philadelphia, PA 19104; 2Department of Pharmacology and Experimental Therapeutics, Thomas

Jefferson University, Philadelphia, PA 19107; 3Department of Pediatric and Neurology, Thomas Jefferson University, Philadelphia, PA 19107; 4Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104

• BUP is a semi-synthetic opioid derived from thebaine and used in clinics as an analgesic and as a detoxification and maintenance treatment for opioid dependence.

• BUP is administered via intravenous and sublingual routes, since BUP has very low oral bioavailability due to extensive first-pass metabolism.

• BUP has large volume of distribution and is extensively metabolized to norbuprenorphine (NBUP) by N-dealkylation mediated mainly by CYP3A4 (65%) and CYP2C8 (30%).

• BUP is metabolized in other metabolic pathways to a minor extent by CYP 2C9, 2C18, and 2C19 and to a major extent by CYP 3A.

• Both BUP and NBUP undergo glucuronidation by UGT1A3, 1A8, and 2B7.• NAS occurs in 55 to 94% of infants who are born to opioid-dependent mothers. • No data has been reported about the use of sublingual administration below the

age of 4 year, especially for term infants with NAS. Therefore, there is a great need to determine the PK characteristics to optimize pediatric therapy.

INTRODUCTION

MODEL DESCRIPTION• The population pharmacokinetic model is based on data from the 12 newborn patients with NAS. Three newborn

patients (ID13, 22, & 27) were co-treated with phenobarbital during the certain period of treatment. • Simulation of intravenous clearance of BUP was conducted in the 12 newborn patients with NAS. • In vitro enzyme kinetic data was incorporated into SIMCYP to simulate intravenous clearance in newborn patients.

* molecular weight, pKa, log P * dose* unbound fraction in plasma (fu) * dose interval* average renal clearance in adult * blood to plasma ratio* Vmax and Km values derived using human liver microsomes /recombinant CYPs* unbound fraction in human liver microsomes /recombinant CYPs (fuMic)

• Sublingual bioavailability was estimated by comparing intravenous clearance (CLiv) obtained from SIMCYP simulation to sublingual clearance (CLsub) generated using non-linear mixed effect modeling with NONMEM.

• Sublingual bioavailability—PMA profiles were drawn from the time they were born till BUP treatment completed for each of the 12 newborn patients.

SIMULATION with SIMCYP

• The final model fits the observed imatinib data well

• There appears to be no pharmacokinetic differences in the two populations evaluated

• Unlike in adults, the model does not predict an increase in CL/Fbetween single dose to steady state

• There appears to be considerable variablility in F, possibly due to enterohepatic recirculation

• A number of the subjects in this trial received transfusions on a routine basis, which possibly contributes to the observed variability in V/F (Imatinib distributes ~ 8 to 30% into RBCs)

• The metabolite model may have limited value given the uncertainyin structural parameters

• None of the covariates tested (weight, sex, age, albumin, study population) were found to have a significant effect on the observed variability

• The model will form the basis of additional covariate identification in pediatric patients and used to support design of future trials particularly with a view to managing toxicities

DISCUSSION AND CONCLUSIONS

BACKGROUND

OBJECTIVES

DESIGN / METHODS

• Peng B, et al., Clinical Pharmacokinetics of Imatinib, Clinical Pharmacokinetics, 2005, 44 (9) : 879-892

• Champagne MA, et al., Imatinib mesylate (STI571) for the treatment of children with Philadelphia chromosome-positive leukemia: results from a Children’s Oncology Group phase I study, Blood, 2004, 104(9):2655-2660

• Schmidli H, et al., Population Pharmacokinetics of Imatinib mesylate in pateints with chronic-phase chronic myeloid leukemia: results of a phase III study, British Journal of Clinical Pharmacology, 2005, 60(1):35-44

Acknowlegements: Children’s Oncology Group

REFERENCES

• Imatinib mesylate (Gleevec®, Novartis Pharmaceuticals, Hanover, NJ) is a bcr-abl tyrosine kinase inhibitor approved for use in the management of chronic myeloid leukemia (CML) in children

• It is completely absorbed on oral administration and extensivelybound to plasma proteins with an elimination half-life of ~ 18 hours

• It is metabolized predominantly in the liver by CYP3A4 to its N-demethylated piperazine derivative (CGP74588, fraction metabolized ~ 20%). CGP74588 has similar potency, selectivity and plasma protein binding profile as the parent

• Population pharmacokinetic (PPK) model developed from data in adults with CML have described a small effect of weight (doubling body weight increases CL by 12%), hemoglobin (doubling hemoglobin increases CL by 86%) and white blood cell count (halving the white blood cell count increases CL by 8%) on the clearance of imatinib

• PPK of imatinib in the pediatric population remains to be described

1. To develop a PPK model of imatinib and CGP74588 to describe exposure in children

2. To identify covariates that are predictors of variability

1521No. of Subjects

3-303-22Age (yr)

46 (19 – 95.6)45 (14.6 – 96.4)Weight (Kg)

-17/4Sex (M/F)

Predose, 1-3 hours, 6-9 hours and 24 hours and at 24 hours on day 8

Predose, 0.5, 1, 1.5, 2, 4, 8 and 24 hours on day1 and at predose, 0.5, 1, 1.5, 2, 4, 8, 24 and 48 hours on day 8

PK Sampling

Target oral dose of 440 mg/m2Escalating oral doses of 260, 340, 440 and 570 mg/m2

Dosage

Phase II / refractory or relapsed solid tumors

Phase I / Ph+ leukemiaPhase/Population

ADVL0122POG 9973

Divya Menon1, John Mondick1, Bhuvana Jayaraman1, Patrick Thompson2, Susan Blaney2, Peter Adamson1 and Jeffrey Barrett1

1Laboratory for Applied PK/PD, Division of Clinical Pharmacology and Therapeutics, The Children's Hospital of Philadelphia, Philadelphia, PA. 2 Baylor College of Medicine, Houston, TX.

A Population Pharmacokinetic Model of Imatinib Mesylate and its Metabolite in Children

Table 1: Study summary and demographics

• Data from two studies (Table 1) were used to develop the model

RESULTS (continued)

RESULTS

Figure 2: Top panel: Population (PRED) predicted, individual (IPRED) plasma concentrations and the goodness of fit plots for imatinib Bottom panel: Population (PRED) predicted, individual (IPRED) plasma concentrations and the goodness of fit plots for CGP74588

3000 8000 13000

0

4000

8000

12000

16000

PRED

DV

3000 8000 13000

0

4000

8000

12000

16000

IPRED

DV

5.5 7.0 8.5

PRED

-8

-6

-4

-2

0

2

WR

ES

0 100 200

TIME

-8

-6

-4

-2

0

2

WR

ES

0 1000 2000

0

500

1000

1500

2000

2500

PRED

DV

0 1000 2000

0

500

1000

1500

2000

2500

IPRED

DV

3 5 7

PRED

-5

-3

-1

1

3

WR

ES

0 100 200

TIME

-5

-3

-1

1

3

WR

ES

34.4 (43.7)0.47 (9.2)

0.56 (11.7)

0.43 (13.8)

Residual error

Imatinib (μg/L)

(POG 9973)

(ADVL0122)

CGP75488

-6.63 (23.1)V/F (L)

33.4 (30.1) 6.72 (14.7)

CGP74588

CL/F

51.3 (89.3)1 (FIXED)F1

74 (80.1)1.88 (9.46)D1 (hr)

68 (52.6)162 (15.1)V/F (L)

33 (28.4)7.42 (13.1)

Imatinib

CL/F (L/hr)

% Intersubject variability (%CV)Mean value (%CV)Parameter (units)

DESIGN / METHODS (continued)

Table 2: Parameter estimates of the final model

Model building

Model Evaluation

10 20 30 40 50

01

00

20

03

00

First Quartile AUC (mg.h/L)

Co

un

t

p = 0.44

30 40 50 60 70

05

01

00

15

02

00

25

0

Median AUC (mg.h/L)

Co

un

t

p = 0.7

40 60 80 100 120 140

05

01

00

15

02

00

25

03

00

Third Quartile AUC (mg.h/L)

Co

un

t

p = 0.87

Figure 3: Predictive Check for Area under the curve (AUC) on day 1 (single dose)Left: Population first quartile AUC Middle: Median AUC Right: Population third quartile AUC The observed values are indicated by a blue vertical line, along with the predictive check p-value (p-value is the proportion of the simulated AUCs above the observed value)

100 200 300 400

05

01

00

15

0

First Quartile AUC (mg.h/L)

Co

un

t

p = 0.67

200 300 400 500 600 700

05

01

00

15

02

00

25

0

Median AUC (mg.h/L)

Co

un

t

p = 0.94

400 600 800 1000 1200

05

01

00

15

02

00

25

03

00

Third Quartile AUC (mg.h/L)

Co

un

t

p = 0.98

Figure 4: Predictive Check for Area under the curve (AUC) on day 8 (steady state)Left: Population first quartile AUC Middle: Median AUC Right: Population third quartile AUC The observed values are indicated by a blue vertical line, along with the predictive check p-value (p-value is the proportion of the simulated AUCs above the observed value)

Model Evaluation

• One thousand Monte Carlo simulation replicates of the observed data were generated using the final PPK model. Distributions of the simulated AUC were compared to the observed values using histograms

Model buiding

• The model was developed using NONMEM (version 5, level 1.1, Globomax, Hanover, MD)

• First order, combined first and zero order, zero order, stepwiseabsorption and the weibull function were evaluated to describe the absorption process

• Exponential error model to describe intersubject variability

• Additive error model with an intersubject variability component to describe residual variability ie., lnY=lnF+EPS(n)*EXP (ETA(n)

• FOCE with - interaction

• Model selection was based on goodness of fit plots, precision of the parameter estimates and changes in the intersubject and residual variability

0 5 10 15 20 25

Time (hr)

0

2000

4000

6000

8000

10000

Obs

erve

d co

ncne

ntra

tions

(g

/L)

140 160 180 200 220

Time (hr)

0

4000

8000

12000

16000

Obs

erve

d co

ncen

trat

ion

(g

/L)

Figure 1: Observed concentration-time profiles of imatinibLeft: Day1 Right: Day 8

RESULTS (continued) Table 1 Summary of demographic characteristics for neonates (N = 12)

at baseline

Covariate Statistic Summary Gestational Age (weeks) Mean (SD)

Median (min, max) 39.28 (1.06) 39.31 (37, 41)

Postnatal Age (days)

Mean (SD) Median (min, max)

22 (11) 17 (11, 47)

Postmenstrual Age (PMA) (weeks)

Mean (SD) Median (min, max)

41.9 (1.22) 42.15 (39.6, 43.55)

Weight (kg) Mean (SD) Median (min, max)

3.002 (0.347) 3.060 (2.39, 3.459)

Gender Male Female

Number (%) Number (%)

10 (83.3%) 2 (16.7%)

Race White Hispanic / Latino

Number (%) Number (%)

10 (83.3%) 2 (16.7%)

mm CCLNCCLCM

dt

dCV

CCLCMCCLCRAKadt

dCV

AKadt

dA

3

2

Time (hr)

Con

c (n

g/m

L)

0 200 400 600 800 1000

0.0

0.2

0.4

0.6

0.8 Observed

Predicted Individual

0 200 400 600 800 1000 1200

Time (hours)

-2

-1

0

1

2

3

We

igh

ted

Re

sid

ua

ls

6

6

6 6

6

6

6

6

9

9

9

9

9

9

9

9

9

9

9

1212

12

12

12

12

1213

1313

13

13

13

13

13

1313

13

13

13

15

15

15

15

15

15

15

15

15

16

16

16

16

16

16

16

16

16

16

1616

16

18

18

18

18

18

18

1818

19

1919

19

191919

19

19

19

19

1919

19

19

19

19

19

19

22

22

22

2222

22

22

22

22 22

22

22

22 22

22

22

22

22

22

22

22

22

22 22

25

25

25

25 25

27

27

27

2727

27

27

27

27

28

28

28

BUP

0.0 0.1 0.2 0.3 0.4 0.5 0.6

Predicted Drug Concentration (ng/mL)

-2

-1

0

1

2

3

We

igh

ted

Re

sid

ua

ls

6

6

66

6

6

6

6

9

9

9

9

9

9

9

9

9

9

9

1212

12

12

12

12

12 13

1313

13

13

13

13

13

1313

13

13

13

15

15

15

15

15

15

15

15

15

16

16

16

16

16

16

16

16

16

16

1616

16

18

18

18

18

18

18

1818

19

1919

19

191919

19

19

19

19

1919

19

19

19

19

19

19

22

22

22

22 22

22

22

22

22 22

22

22

22 22

22

22

22

22

22

22

22

22

2222

25

25

25

2525

27

27

27

2727

27

27

27

27

28

28

28

RESULTS

We cordially thank Dr. Masoud Jamei and Mr. Steve Marciniak in Simcyp Limited for their technical assistance to operation and installation of SIMCYP software in our BUP project.

ACKNOWLEDGMENT

Time (hr)

Con

c (n

g/m

L)

0 200 400 600 800 1000

0.0

0.1

0.2

0.3

0.4

0.5 Observed

Predicted Individual

NBUP

0 200 400 600 800 1000

Time (hours)

-0.3

-0.2

-0.1

0.0

0.1

0.2

Re

sid

ua

ls

66

6

6

6

9

9

9

9

9

12

12

12

12

12

13

13

13

13

13

13

1313

13 13

13

13

15

15

15

15

15

15

15

15

1616

1616

16 16

16

16

16

1616

16

16

16

16

18

18

18

18

18

18

18

19

191922 22

22

22

22

22

27

27

27

0.0 0.1 0.2 0.3

Predicted Drug Concentration (ng/mL)

-0.3

-0.2

-0.1

0.0

0.1

0.2

Re

sid

ua

ls

66

6

6

6

9

9

9

9

9

12

12

12

12

12

13

13

13

13

13

13

1313

13 13

13

13

15

15

15

15

15

15

15

15

1616

1616

16 16

16

16

16

1616

16

16

16

16

18

18

18

18

18

18

18

19

191922 22

22

22

22

22

27

27

27

Table 2 Treatment information

Subject Treatment starting date

from birth (days)

Treatment duration (days)

Co-medication with phenobarbital

AN06 1 14 No AN09 4 15 No AN12 4 17 No AN13 3 30 Yes

(started at 143.29 hr & stopped at 287.29 hr)

AN15 1 15 No AN16 3 16 No AN18 4 11 No AN19 2 24 No AN22 2 47 Yes

(started at 222.62 hr & stopped at 381.5 hr)

AN25 1 26 No

AN27 2 39 Yes (started at 121.09 hr

& stopped at 409.09 hr ) AN28 2 15 No

PMA: Postmenstrual Age

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 6

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 9

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 12

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 15

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 16

PMA (week)

Sub

ling

ual B

ioa

vaila

bilit

y (%

)

38 40 42 44 46

020

4060

8010

0

ID = 18

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 19

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 25

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 28

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 13

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 22

PMA (week)

Sub

ling

ua

l Bio

ava

ilabi

lity

(%)

38 40 42 44 46

020

40

60

80

10

0

ID = 27

CLCM

CLNCLCR

V3

(NBUP)Depot

ka

Sublingual

V2

(BUP)

BUP/NBP Ratio: Concentration Ratio of BUP/NBUP at the Same Time PointPMA: Postmenstrual Age

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 6

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 9

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 12

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 15

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 16

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 18

PMA (week)

BU

P/N

BU

P R

atio

37 39 41 43

05

1015

20

ID = 19