SlideShare a Scribd company logo
1 of 2
Download to read offline
Computational Evaluation of Controlled Release Oral Delivery for Some Atypical Antipsychotics
GM Eichenbaum, VA Nguyen and WW van Osdol
ALZA Corporation, Mountain View, CA 94043
bvanosdo@alzus.jnj.com
Abstract Summary
We use computational ADME tools, physical
chemical measurements in vitro, and the results of site-
specific gastrointestinal dosing in the rat and dog to
predict the effect of controlled release oral delivery on
the plasma concentrations and duration of action of a
series of -carboline atypical antipsychotics.
Introduction
Historically, controlled release drug delivery
technologies have been applied to rescue or improve
marketed drugs and late stage clinical candidates.
However, controlled release (CR) drug delivery could
also offer an approach for reducing attrition of
compounds advancing from the lead optimization stage of
drug discovery. Drug discovery has generally not
considered CR drug delivery approaches at the lead
optimization stage, focusing instead on synthesizing
compounds that are suitable for immediate release (IR)
oral delivery. Compounds identified during lead
optimization that subsequently prove to be unsuitable for
IR oral delivery rarely advance, even if they show
promise in terms of their potency, selectivity and efficacy.
But, liabilities that make compounds unsuitable for IR
oral delivery may be overcome by CR oral delivery.
These liabilities include poor solubility, low permeability
and extensive metabolism in the upper GI tract, short
plasma half-life, and a narrow therapeutic index. If
recognized and evaluated during lead optimization, CR
drug delivery could enable successful advancement of
such compounds.
Optimization of a lead series for CR delivery requires
an assessment of how solubility, dissolution, absorption
and metabolism affect the pharmacokinetic (PK) profile
of the compound. Here, we apply computational
modeling to evaluate the impact of CR drug delivery on
the PK profile using in vitro and in vivo pre-clinical
measurements as inputs.
Experimental and Computational Methods
Absorption of a compound in the lower as well as the
upper GI tract is a key requirement for applying oral CR
drug delivery to sustain circulating plasma concentrations.
To evaluate the feasibility of applying controlled release
oral delivery to extend the duration of action of the -
carbolines, we applied the following strategy.
From a series of -carboline analogues, we selected
three lead compounds (A, B and C) and measured their
passive permeability in a parallel artificial membrane
permeability assay (PAMPA). Next, we measured their
solubilities versus pH to ensure that solubility would not
limit their absorption in vivo. We then performed intra-
duodenal and intra-colonic dosing studies in a
catheterized rat model for the three compounds and
risperidone to assess the relative upper and lower GI
absorption of each compound. Lastly, for the compound
(B) with the highest colonic absorption in the rat model
and for risperidone, we conducted site-specific dosing
studies in a catheterized dog model. Our goal was to
identify at least one member of the series with the
potential to achieve QD dosing through CR oral delivery.
Risperidone was used as a positive comparator because it
has substantial human colonic absorption and is related
structurally to the carbolines.
Our principal computational tools were GastroPlusTM
and its supporting program QMPRPlusTM
(SimulationsPlus, Inc., Lancaster, CA). GastroPlus
implements an advanced compartmental absorption and
transit (ACAT) model of the GI tract (Agoram et al.,
2001), based on the original model of Yu and Amidon
(1996, 1999), that couples to compartmental PK models
via rate equations governing the release, dissolution,
transit, absorption, distribution, metabolism and
elimination of a compound. The outputs of the model are
bioavailability, and liver and plasma concentrations as
functions of time.
To predict plasma profiles, values for the molecular
weight, Log P, pKas, solubility vs pH, aqueous diffusion
coefficient, and GI permeability must be provided. With
the exception of solubility as a function of pH, we
utilized QMPRPlus and ACD pKa DB (ACD Labs,
Toronto, Canada) to estimate these values in silico. In
addition, a pharmacokinetic model must be specified:
minimally, a clearance and volume of distribution. PK
analyses of the iv bolus data were conducted via the
PKPlus module within GastroPlus. Oral bioavailability
(BA) was calculated by comparing dose-normalized area
under the curves (AUCs) following site-specific and iv
bolus dosing.
Results and Discussion
The plasma concentration profiles for compound B
observed after duodenal and colonic dosing in rats are
shown in Figure 1. Absorption is rapid in both cases, with
tmax reached within an hour. The profiles are almost
identical, with a colonic-duodenal relative BA of 91%
(absolute BAs are 81% and 74%, respectively). The
predicted plasma profiles match the observed data well
between one and four hours. The predicted rate of
absorption may exceed the observed rate, but there are
insufficient data at early times to resolve this point.
Beyond four hours, the simulation predicts a slightly more
rapid decline than observed. On the whole, however, the
results of the simulation are quite good, with root mean
square (RMS) relative errors of 0.25 and 0.20,
respectively.
Fig. 1 Observed and predicted plasma concentrations (±
SEM, N=4) following site-specific dosing of compound B
in rats (0.1 mg/kg)
The plasma profiles observed after duodenal and
colonic dosing in dogs are shown in Figure 2. Absorption
is rapid, and the terminal phase is more rapid than in the
mouse. The profiles are quite similar, except at 0.25 hrs,
at which time there is large variability in the duodenal
data, which may be inaccurate. The similarity of the
profiles at all other times suggests so. The simulation of
duodenal dosing predicts the amplitudes well, but the
absorption rate appears to be too low, which shifts the
profile to later times. The predicted colonic profile
captures the absorption phase but the clearance appears to
be somewhat too large. The RMS relative errors are 0.34
and 0.47, respectively.
Fig. 2 Observed and predicted plasma concentrations (±
SEM, N=4) following site-specific dosing of compound B
in dogs (1.8 mg/kg)
Predictions for the CR delivery of compound B are
shown in Figure 3, along with a minimum effective
plasma concentration (dotted line) for D2 antagonism,
derived from studies on protection of dogs from
apomorphine-induced emesis. The release profile, shown
inset, is typical of an OROSTM
push-pull design, scaled to
match the transit time of the dog GI tract. Our
calculations suggest that CR delivery can achieve quasi-
steady plasma concentrations greater than the threshold
for efficacy for longer durations than IR delivery. CR
delivery may accomplish this at slightly lower doses than
IR delivery, although more time is required to reach
effective plasma levels initially.
Fig. 3 Predicted plasma concentrations following CR oral
delivery of compound B. The dotted line indicates a
minimum effective plasma concentration (5 ng/ml). The
release profile is inset.
Conclusions
We have used a combination of in vitro and in vivo
data and in silico estimates to simulate numerically the IR
and CR oral delivery of three -carboline analogues of
risperidone. We are able to reproduce the observed
plasma profiles of the compounds following duodenal and
colonic site-specific dosing in the rat and dog, and predict
the plasma profiles that would be observed in the dog
following CR delivery. These latter predictions, when
combined with observations from a model for efficacy,
suggest that CR delivery achieves quasi-steady plasma
concentrations and extended duration of action relative to
IR delivery.
Pending the availability of human PK data through
allometric scaling or studies in vivo, these kinds of
calculations can be extended to aid in the assessment of
lead series of compounds in the latter stages of drug
discovery.
References
(1) Agoram, B, Woltosz, WS and Bolger, MB. Adv. Drug
Delivery Rev. 50, S41-S67 (2001).
(2) Yu, LX, Crison, JR and Amidon, GL. Int. J. Pharm.
140, 111-118 (1996).
(3) Yu, LX and Amidon, GL. Int. J. Pharm. 186, 119-125
(1999).
JNJ-16558711
0
1
2
3
4
5
6
0 2 4 6 8
Time (hours)
PlasmaConc.
(ng/ml)
Duod Obs
Colon Obs
Duod Pred
Colon Pred
JNJ-16558711
0
1
2
3
4
5
6
0 2 4 6 8
Time (hours)
PlasmaConc.
(ng/ml)
Duod Obs
Colon Obs
Duod Pred
Colon Pred
0
16
32
48
64
80
0 2 4 6 8 10 12 14 16
Time (hours)
PlasmaConc.(ng/ml)
Duod Obs
Colon Obs
OROS 1 mg/kg
OROS 3 mg/kg
Min Effective Cp0
20
40
60
80
100
0 2 4 6 8 10
Time (hrs)
%Released
0
28
56
84
0 2 4 6 8
Time (hours)
PlasmaConc.
(ng/ml)
Duod Obs
Colon Obs
Duod Pred
Colon Pred

More Related Content

What's hot

IJBB-51-3-188-200
IJBB-51-3-188-200IJBB-51-3-188-200
IJBB-51-3-188-200
sankar basu
 
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MSDesign of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
saheli halder
 

What's hot (8)

SIMONA CAVALU_Raman and surface-enhanced Raman spectroscopy of tempyo spin la...
SIMONA CAVALU_Raman and surface-enhanced Raman spectroscopy of tempyo spin la...SIMONA CAVALU_Raman and surface-enhanced Raman spectroscopy of tempyo spin la...
SIMONA CAVALU_Raman and surface-enhanced Raman spectroscopy of tempyo spin la...
 
DM Garby_Vitamin B6 AACC 2012
DM Garby_Vitamin B6 AACC 2012DM Garby_Vitamin B6 AACC 2012
DM Garby_Vitamin B6 AACC 2012
 
IRJET- Kinetic Study of the Reaction of 5-Chlorosalicyaldehyde with M- To...
IRJET-  	  Kinetic Study of the Reaction of 5-Chlorosalicyaldehyde with M- To...IRJET-  	  Kinetic Study of the Reaction of 5-Chlorosalicyaldehyde with M- To...
IRJET- Kinetic Study of the Reaction of 5-Chlorosalicyaldehyde with M- To...
 
Renal physiology
Renal physiologyRenal physiology
Renal physiology
 
Nalbuphine hydrochloride
Nalbuphine hydrochlorideNalbuphine hydrochloride
Nalbuphine hydrochloride
 
Guided inquiry analysis the use of ft nmr of curcumin
Guided inquiry analysis the use of ft nmr of curcuminGuided inquiry analysis the use of ft nmr of curcumin
Guided inquiry analysis the use of ft nmr of curcumin
 
IJBB-51-3-188-200
IJBB-51-3-188-200IJBB-51-3-188-200
IJBB-51-3-188-200
 
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MSDesign of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
Design of ultrasensitive DNA-based fluorescent pH sensitive nanodevices_MS
 

Similar to CRS 2005 - ADME-PK anti-psychotics

Extrapolation of preclinical data to clinical data.pptx
Extrapolation of preclinical data to clinical data.pptxExtrapolation of preclinical data to clinical data.pptx
Extrapolation of preclinical data to clinical data.pptx
VincyDinakaran
 
Seahorse Poster DOS (JJW Edits 6-15-15)
Seahorse Poster DOS (JJW Edits 6-15-15)Seahorse Poster DOS (JJW Edits 6-15-15)
Seahorse Poster DOS (JJW Edits 6-15-15)
Zach Swanson
 
The aim of the use of meloxicam in patients with mild or mo
The aim of  the use of meloxicam in patients with mild or moThe aim of  the use of meloxicam in patients with mild or mo
The aim of the use of meloxicam in patients with mild or mo
MUSHTAQ AHMED
 
LPDT_A_938857-2014
LPDT_A_938857-2014LPDT_A_938857-2014
LPDT_A_938857-2014
Mansi Shah
 
Evaluation of the chronic kidney disease epidemiology 2010
Evaluation of the chronic kidney disease epidemiology 2010Evaluation of the chronic kidney disease epidemiology 2010
Evaluation of the chronic kidney disease epidemiology 2010
eduardo de avila
 

Similar to CRS 2005 - ADME-PK anti-psychotics (20)

Bioavailability and bioequivalane studies
Bioavailability and bioequivalane studiesBioavailability and bioequivalane studies
Bioavailability and bioequivalane studies
 
GI simulation final.pptx
GI simulation final.pptxGI simulation final.pptx
GI simulation final.pptx
 
Qsar and drug design ppt
Qsar and drug design pptQsar and drug design ppt
Qsar and drug design ppt
 
Martinez use of in silico models to support canine drug development
Martinez use of in silico models to support canine drug developmentMartinez use of in silico models to support canine drug development
Martinez use of in silico models to support canine drug development
 
Extrapolation of preclinical data to clinical data.pptx
Extrapolation of preclinical data to clinical data.pptxExtrapolation of preclinical data to clinical data.pptx
Extrapolation of preclinical data to clinical data.pptx
 
In-Vitro-In Vivo (IVIVC).pdf
In-Vitro-In Vivo (IVIVC).pdfIn-Vitro-In Vivo (IVIVC).pdf
In-Vitro-In Vivo (IVIVC).pdf
 
IN-VITRO-IN VIVO CORRELATION (IVIVC).pptx
IN-VITRO-IN VIVO CORRELATION (IVIVC).pptxIN-VITRO-IN VIVO CORRELATION (IVIVC).pptx
IN-VITRO-IN VIVO CORRELATION (IVIVC).pptx
 
Computational modelling of drug disposition active transport
Computational modelling of  drug disposition active transportComputational modelling of  drug disposition active transport
Computational modelling of drug disposition active transport
 
Theoretical background on GastroPlus Simulation Software
Theoretical background on GastroPlus Simulation SoftwareTheoretical background on GastroPlus Simulation Software
Theoretical background on GastroPlus Simulation Software
 
Seahorse Poster DOS (JJW Edits 6-15-15)
Seahorse Poster DOS (JJW Edits 6-15-15)Seahorse Poster DOS (JJW Edits 6-15-15)
Seahorse Poster DOS (JJW Edits 6-15-15)
 
Fed vs Fasted state KKR
Fed vs Fasted state KKRFed vs Fasted state KKR
Fed vs Fasted state KKR
 
Active transport
Active transportActive transport
Active transport
 
Madhu k s
Madhu k s Madhu k s
Madhu k s
 
Computational modelling of drug disposition
Computational modelling of drug disposition Computational modelling of drug disposition
Computational modelling of drug disposition
 
The aim of the use of meloxicam in patients with mild or mo
The aim of  the use of meloxicam in patients with mild or moThe aim of  the use of meloxicam in patients with mild or mo
The aim of the use of meloxicam in patients with mild or mo
 
Gastric absorption simulation
Gastric absorption simulation Gastric absorption simulation
Gastric absorption simulation
 
Computational modeling in drug disposition
Computational modeling in drug dispositionComputational modeling in drug disposition
Computational modeling in drug disposition
 
LPDT_A_938857-2014
LPDT_A_938857-2014LPDT_A_938857-2014
LPDT_A_938857-2014
 
Computational modeling in drug disposition
Computational modeling in drug dispositionComputational modeling in drug disposition
Computational modeling in drug disposition
 
Evaluation of the chronic kidney disease epidemiology 2010
Evaluation of the chronic kidney disease epidemiology 2010Evaluation of the chronic kidney disease epidemiology 2010
Evaluation of the chronic kidney disease epidemiology 2010
 

CRS 2005 - ADME-PK anti-psychotics

  • 1. Computational Evaluation of Controlled Release Oral Delivery for Some Atypical Antipsychotics GM Eichenbaum, VA Nguyen and WW van Osdol ALZA Corporation, Mountain View, CA 94043 bvanosdo@alzus.jnj.com Abstract Summary We use computational ADME tools, physical chemical measurements in vitro, and the results of site- specific gastrointestinal dosing in the rat and dog to predict the effect of controlled release oral delivery on the plasma concentrations and duration of action of a series of -carboline atypical antipsychotics. Introduction Historically, controlled release drug delivery technologies have been applied to rescue or improve marketed drugs and late stage clinical candidates. However, controlled release (CR) drug delivery could also offer an approach for reducing attrition of compounds advancing from the lead optimization stage of drug discovery. Drug discovery has generally not considered CR drug delivery approaches at the lead optimization stage, focusing instead on synthesizing compounds that are suitable for immediate release (IR) oral delivery. Compounds identified during lead optimization that subsequently prove to be unsuitable for IR oral delivery rarely advance, even if they show promise in terms of their potency, selectivity and efficacy. But, liabilities that make compounds unsuitable for IR oral delivery may be overcome by CR oral delivery. These liabilities include poor solubility, low permeability and extensive metabolism in the upper GI tract, short plasma half-life, and a narrow therapeutic index. If recognized and evaluated during lead optimization, CR drug delivery could enable successful advancement of such compounds. Optimization of a lead series for CR delivery requires an assessment of how solubility, dissolution, absorption and metabolism affect the pharmacokinetic (PK) profile of the compound. Here, we apply computational modeling to evaluate the impact of CR drug delivery on the PK profile using in vitro and in vivo pre-clinical measurements as inputs. Experimental and Computational Methods Absorption of a compound in the lower as well as the upper GI tract is a key requirement for applying oral CR drug delivery to sustain circulating plasma concentrations. To evaluate the feasibility of applying controlled release oral delivery to extend the duration of action of the - carbolines, we applied the following strategy. From a series of -carboline analogues, we selected three lead compounds (A, B and C) and measured their passive permeability in a parallel artificial membrane permeability assay (PAMPA). Next, we measured their solubilities versus pH to ensure that solubility would not limit their absorption in vivo. We then performed intra- duodenal and intra-colonic dosing studies in a catheterized rat model for the three compounds and risperidone to assess the relative upper and lower GI absorption of each compound. Lastly, for the compound (B) with the highest colonic absorption in the rat model and for risperidone, we conducted site-specific dosing studies in a catheterized dog model. Our goal was to identify at least one member of the series with the potential to achieve QD dosing through CR oral delivery. Risperidone was used as a positive comparator because it has substantial human colonic absorption and is related structurally to the carbolines. Our principal computational tools were GastroPlusTM and its supporting program QMPRPlusTM (SimulationsPlus, Inc., Lancaster, CA). GastroPlus implements an advanced compartmental absorption and transit (ACAT) model of the GI tract (Agoram et al., 2001), based on the original model of Yu and Amidon (1996, 1999), that couples to compartmental PK models via rate equations governing the release, dissolution, transit, absorption, distribution, metabolism and elimination of a compound. The outputs of the model are bioavailability, and liver and plasma concentrations as functions of time. To predict plasma profiles, values for the molecular weight, Log P, pKas, solubility vs pH, aqueous diffusion coefficient, and GI permeability must be provided. With the exception of solubility as a function of pH, we utilized QMPRPlus and ACD pKa DB (ACD Labs, Toronto, Canada) to estimate these values in silico. In addition, a pharmacokinetic model must be specified: minimally, a clearance and volume of distribution. PK analyses of the iv bolus data were conducted via the PKPlus module within GastroPlus. Oral bioavailability (BA) was calculated by comparing dose-normalized area under the curves (AUCs) following site-specific and iv bolus dosing. Results and Discussion The plasma concentration profiles for compound B observed after duodenal and colonic dosing in rats are shown in Figure 1. Absorption is rapid in both cases, with tmax reached within an hour. The profiles are almost identical, with a colonic-duodenal relative BA of 91% (absolute BAs are 81% and 74%, respectively). The predicted plasma profiles match the observed data well between one and four hours. The predicted rate of absorption may exceed the observed rate, but there are
  • 2. insufficient data at early times to resolve this point. Beyond four hours, the simulation predicts a slightly more rapid decline than observed. On the whole, however, the results of the simulation are quite good, with root mean square (RMS) relative errors of 0.25 and 0.20, respectively. Fig. 1 Observed and predicted plasma concentrations (± SEM, N=4) following site-specific dosing of compound B in rats (0.1 mg/kg) The plasma profiles observed after duodenal and colonic dosing in dogs are shown in Figure 2. Absorption is rapid, and the terminal phase is more rapid than in the mouse. The profiles are quite similar, except at 0.25 hrs, at which time there is large variability in the duodenal data, which may be inaccurate. The similarity of the profiles at all other times suggests so. The simulation of duodenal dosing predicts the amplitudes well, but the absorption rate appears to be too low, which shifts the profile to later times. The predicted colonic profile captures the absorption phase but the clearance appears to be somewhat too large. The RMS relative errors are 0.34 and 0.47, respectively. Fig. 2 Observed and predicted plasma concentrations (± SEM, N=4) following site-specific dosing of compound B in dogs (1.8 mg/kg) Predictions for the CR delivery of compound B are shown in Figure 3, along with a minimum effective plasma concentration (dotted line) for D2 antagonism, derived from studies on protection of dogs from apomorphine-induced emesis. The release profile, shown inset, is typical of an OROSTM push-pull design, scaled to match the transit time of the dog GI tract. Our calculations suggest that CR delivery can achieve quasi- steady plasma concentrations greater than the threshold for efficacy for longer durations than IR delivery. CR delivery may accomplish this at slightly lower doses than IR delivery, although more time is required to reach effective plasma levels initially. Fig. 3 Predicted plasma concentrations following CR oral delivery of compound B. The dotted line indicates a minimum effective plasma concentration (5 ng/ml). The release profile is inset. Conclusions We have used a combination of in vitro and in vivo data and in silico estimates to simulate numerically the IR and CR oral delivery of three -carboline analogues of risperidone. We are able to reproduce the observed plasma profiles of the compounds following duodenal and colonic site-specific dosing in the rat and dog, and predict the plasma profiles that would be observed in the dog following CR delivery. These latter predictions, when combined with observations from a model for efficacy, suggest that CR delivery achieves quasi-steady plasma concentrations and extended duration of action relative to IR delivery. Pending the availability of human PK data through allometric scaling or studies in vivo, these kinds of calculations can be extended to aid in the assessment of lead series of compounds in the latter stages of drug discovery. References (1) Agoram, B, Woltosz, WS and Bolger, MB. Adv. Drug Delivery Rev. 50, S41-S67 (2001). (2) Yu, LX, Crison, JR and Amidon, GL. Int. J. Pharm. 140, 111-118 (1996). (3) Yu, LX and Amidon, GL. Int. J. Pharm. 186, 119-125 (1999). JNJ-16558711 0 1 2 3 4 5 6 0 2 4 6 8 Time (hours) PlasmaConc. (ng/ml) Duod Obs Colon Obs Duod Pred Colon Pred JNJ-16558711 0 1 2 3 4 5 6 0 2 4 6 8 Time (hours) PlasmaConc. (ng/ml) Duod Obs Colon Obs Duod Pred Colon Pred 0 16 32 48 64 80 0 2 4 6 8 10 12 14 16 Time (hours) PlasmaConc.(ng/ml) Duod Obs Colon Obs OROS 1 mg/kg OROS 3 mg/kg Min Effective Cp0 20 40 60 80 100 0 2 4 6 8 10 Time (hrs) %Released 0 28 56 84 0 2 4 6 8 Time (hours) PlasmaConc. (ng/ml) Duod Obs Colon Obs Duod Pred Colon Pred