DDMoRe Model Repository

Overview
Files
History
Model Definition
Estimation Steps

Short description:

Semi-mechanistic Pharmacokinetic/Pharmacodynamic model of abexinostat-induced thrombocytopenia in both lymphoma and solid tumour patients, based on Friberg et al (2002). The model includes a different platelet baseline counts for each patient population (lymphoma or solid tumour) and a disease progression component for patients with lymphoma.

Format:	PharmML (0.6.1)
Related Publication:	Pharmacokinetic/Pharmacodynamic modeling of abexinostat-induced thrombocytopenia across different patient populations: application for the determination of the maximum tolerated doses in both lymphoma and solid tumour patients. Chalret du Rieu Q, Fouliard S, White-Koning M, Kloos I, Chatelut E, Chenel M Investigational new drugs, 10/2014, Volume 32, Issue 5, pages: 985-994 Affiliation: Clinical Pharmacokinetics Department, Institut de Recherches Internationales Servier, Suresnes, France. Abstract: BACKGROUND: In the clinical development of oncology drugs, the recommended dose is usually determined using a 3?+?3 dose-escalation study design. However, this phase I design does not always adequately describe dose-toxicity relationships. METHODS: 125 patients, with either solid tumours or lymphoma, were included in the study and 1217 platelet counts were available over three treatment cycles. The data was used to build a population pharmacokinetic/pharmacodynamic (PKPD) model using a sequential modeling approach. Model-derived Recommended Doses (MDRD) of abexinostat (a Histone Deacetylase Inhibitor) were determined from simulations of different administration schedules, and the higher bound for the probability of reaching these MDRD with a 3?+?3 design were obtained. RESULTS: The PKPD model developed adequately described platelet kinetics in both patient populations with the inclusion of two platelet baseline counts and a disease progression component for patients with lymphoma. Simulation results demonstrated that abexinostat administration during the first 4 days of each week in a 3-week cycle led to a higher MDRD compared to the other administration schedules tested, with a maximum probability of 40 % of reaching these MDRDs using a 3?+?3 design. CONCLUSIONS: The PKPD model was able to predict thrombocytopenia following abexinostat administration in both patient populations. A model-based approach to determine the recommended dose in phase I trials is preferable due to the imprecision of the 3?+?3 design.
Contributors:	Vincent Croixmarie

Validation Status:	Annotations have not been checked.
Certification Comment:	This model is not certified.

Mandatory file
- Chalret_2014_Abexinostat.xml

Additional Files

Model owner: Vincent Croixmarie
Submitted: Feb 12, 2016 5:24:28 PM
Last Modified: Feb 12, 2016 5:24:28 PM

Revisions

Version: 11
- Submitted on: Feb 12, 2016 5:24:28 PM
- Submitted by: Vincent Croixmarie
- With comment: Edited model metadata online.

Independent variable T

Function Definitions

combinedError1 (additive, proportional, f) = (additive + (proportional \times f))

Structural Model sm

Variable definitions

K67 = IKA

K70 = \frac{ICL}{IV2}

K78 = \frac{IQ3}{IV2}

K87 = \frac{IQ3}{IV3}

K79 = \frac{IQ4}{IV2}

K97 = \frac{IQ4}{IV4}

V7 = \frac{IV2}{IBIO}

S7 = V7

BAS0 = ((POP_BAS0 \times exp (ETA_BAS0)) \times (1 - PATH))

BAS1 = ((POP_BAS1 \times exp (ETA_BAS0)) \times PATH)

IMAT = (POP_IMAT \times (1 + ETA_IMAT))

K = \frac{4}{MTT0}

CP = \frac{CENT}{S7}

BASE = {\begin{matrix} (BAS0 - \frac{(IMAT \times \frac{T}{24})}{(POP_IT50 + \frac{T}{24})}) & if & (PATH = 0) \\ BAS1 & otherwise \end{matrix}

BASI = {\begin{matrix} BAS0 & if & (PATH = 0) \\ BAS1 & otherwise \end{matrix}

MTT = {\begin{matrix} 0.001 & if & (MTT0 \leq 0) \\ MTT0 & otherwise \end{matrix}

FBM = {\frac{BASE}{CIRC}}^{GAMMA}

FBK = {\frac{BASE}{CIRC}}^{DELTA}

E = \frac{(POP_IMAX \times CP)}{(IC50 + CP)}

\frac{dPROL}{dT} = ((((K \times PROL) \times FBM) \times (1 - E)) - ((K \times FBK) \times PROL))

\frac{dTRA1}{dT} = (((K \times FBK) \times PROL) - ((K \times FBK) \times TRA1))

\frac{dTRA2}{dT} = (((K \times FBK) \times TRA1) - ((K \times FBK) \times TRA2))

\frac{dTRA3}{dT} = (((K \times FBK) \times TRA2) - ((K \times FBK) \times TRA3))

\frac{dCIRC}{dT} = (((K \times FBK) \times TRA3) - (K \times CIRC))

\frac{dDEP}{dT} = (- K67 \times DEP)

\frac{dCENT}{dT} = ((((K67 \times DEP) + (K87 \times PERIF1)) + (K97 \times PERIF2)) - (CENT \times ((K78 + K79) + K70)))

\frac{dPERIF1}{dT} = ((K78 \times CENT) - (K87 \times PERIF1))

\frac{dPERIF2}{dT} = ((K79 \times CENT) - (K97 \times PERIF2))

Initial conditions

PROL = BASI

TRA1 = BASI

TRA2 = BASI

TRA3 = BASI

CIRC = BASI

DEP = 0

CENT = 0

PERIF1 = 0

PERIF2 = 0

Variability Model

Level	Type
DV	residualError
ID CYCL, parent level: ID	parameterVariability

Level

Type

residualError

CYCL, parent level: ID

parameterVariability

Covariate Model

Continuous covariate IBIO

Continuous covariate ICL

Continuous covariate IKA

Continuous covariate IQ3

Continuous covariate IQ4

Continuous covariate IV2

Continuous covariate IV3

Continuous covariate IV4

Continuous covariate PATH

Parameter Model

Parameters

POP_IC50

;

POP_IMAX

;

POP_MTT

;

POP_GAMMA

;

POP_DELTA

;

POP_BAS0

;

POP_BAS1

;

POP_IT50

;

POP_IMAT

;

Y_ADD

;

Y_PROP

;

OMEGA_IC50

;

OMEGA_IC50_OCC

;

OMEGA_MTT

;

OMEGA_GAMMA

;

OMEGA_DELTA

;

OMEGA_DELTA_OCC

;

OMEGA_BAS0

;

OMEGA_IMAT

;

ETA_IC50 \sim N (0.0, OMEGA_IC50)

— ID

ETA_MTT \sim N (0.0, OMEGA_MTT)

— ID

ETA_GAMMA \sim N (0.0, OMEGA_GAMMA)

— ID

ETA_DELTA \sim N (0.0, OMEGA_DELTA)

— ID

ETA_BAS0 \sim N (0.0, OMEGA_BAS0)

— ID

ETA_IMAT \sim N (0.0, OMEGA_IMAT)

— ID

ETA_IC50_OCC \sim N (0.0, OMEGA_IC50_OCC)

— CYCL

ETA_DELTA_OCC \sim N (0.0, OMEGA_DELTA_OCC)

— CYCL

EPS_Y \sim N (0.0, 1.0)

— DV

log (IC50) = (log (POP_IC50) + (ETA_IC50 + ETA_IC50_OCC))

log (DELTA) = (log (POP_DELTA) + (ETA_DELTA + ETA_DELTA_OCC))

log (MTT0) = (log (POP_MTT) + ETA_MTT)

log (GAMMA) = (log (POP_GAMMA) + ETA_GAMMA)

Observation Model

Observation Y
Continuous / Residual Data

Parameters

Y = (CIRC + (combinedError1 (Y_ADD, Y_PROP, CIRC) \times EPS_Y))

Estimation Steps

Estimation Step estimStep_1

Estimation parameters

Initial estimates for non-fixed parameters

$POP_IC50 = 0.0743$
$POP_IMAX = 0.881$
$POP_MTT = 95$
$POP_GAMMA = 0.498$
$POP_DELTA = 0.177$
$POP_BAS0 = 195$
$POP_BAS1 = 273$
$POP_IT50 = 68.2$
$POP_IMAT = 101$
$Y_ADD = 13.5$
$Y_PROP = 0.19$
$OMEGA_IC50 = 0.663$
$OMEGA_IC50_OCC = 0.304$
$OMEGA_MTT = 0.087$
$OMEGA_GAMMA = 0.097$
$OMEGA_DELTA = 0.228$
$OMEGA_DELTA_OCC = 0.224$
$OMEGA_BAS0 = 0.39$
$OMEGA_IMAT = 0.842$

Estimation operations

1) Estimate the population parameters

Algorithm FOCE

Step Dependencies

estimStep_1

DDMODEL00000078: Chalret_2014_Abexinostat

Revisions

Function Definitions

Structural Model sm

Variability Model

Covariate Model

Parameter Model

Observation Model

Observation YContinuous / Residual Data

Estimation Steps

Estimation Step estimStep_1

Estimation parameters

Estimation operations

Step Dependencies

Observation Y
Continuous / Residual Data