Hi,
I am working on a pop PK model to estimate PK parameters in pediatric and adult
patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per subject
whereas the adult study (n=50, median age 20 yrs) has 12 samples per subject. A
two-compartment model best describes the data for each data set. Although a
two-compartment model best describes the combined data, the individual
parameter estimates in pediatric population are different compared to those
obtained using with pediatric data alone. Note that the parameter estimates in
adults were not significantly altered with either combined or adult data alone.
Body weight is the only covariate included in the model with allometric
exponents fixed to 0.75 on CL and 1 on V1.
I would like to hear your thoughts on this and any suggestions on how to
proceed with modeling combined data from pediatric and adult studies.
Regards,
- Chandra
Sparse (pediatric) and rich (adult) data
Chandra,
Pediatric data alone may not be able to support (with 3 samples per patient) a two compartment model. So combined adult/pediatric model is more appropriate. You may also want to scale peripheral compartment parameters (Q as CL, V2 as V, K12and K21 as CL/V ~ 1/WT^0.25). Remaining dependence of CL on WT (if any is noticeable) for very young kids could be attributed to maturation and explained by AGE covariate
Leonid
--------------------------------------
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel: (301) 767 5566
Chandrasekhar Udata wrote:
> Hi,
>
> I am working on a pop PK model to estimate PK parameters in pediatric and adult patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per subject whereas the adult study (n=50, median age 20 yrs) has 12 samples per subject. A two-compartment model best describes the data for each data set. Although a two-compartment model best describes the combined data, the individual parameter estimates in pediatric population are different compared to those obtained using with pediatric data alone. Note that the parameter estimates in adults were not significantly altered with either combined or adult data alone. Body weight is the only covariate included in the model with allometric exponents fixed to 0.75 on CL and 1 on V1. I would like to hear your thoughts on this and any suggestions on how to proceed with modeling combined data from pediatric and adult studies. Regards,
>
> - Chandra
Chandra,
With such a small sample its hard to learn much about differences between adults and children. Your principled approach using allometric scaling is a reasonable way to bridge the gap in recognizing that adults and children are all the same species (see reference below).
"Children are just small adults"
I would not be too worried about individual parameter estimate in children being different. With only 3 samples per child and a 2 cmt model requiring at least 4 parameters you will always get different results if you use different assumptions.
Nick
Anderson BJ, Holford NH. Mechanism-Based Concepts of Size and Maturity in Pharmacokinetics. Annu Rev Pharmacol Toxicol. 2008;48:303-32.
Chandrasekhar Udata wrote:
> Hi,
>
> I am working on a pop PK model to estimate PK parameters in pediatric and adult patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per subject whereas the adult study (n=50, median age 20 yrs) has 12 samples per subject. A two-compartment model best describes the data for each data set. Although a two-compartment model best describes the combined data, the individual parameter estimates in pediatric population are different compared to those obtained using with pediatric data alone. Note that the parameter estimates in adults were not significantly altered with either combined or adult data alone. Body weight is the only covariate included in the model with allometric exponents fixed to 0.75 on CL and 1 on V1. I would like to hear your thoughts on this and any suggestions on how to proceed with modeling combined data from pediatric and adult studies. Regards,
>
> - Chandra
--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
[EMAIL PROTECTED] tel:+64(9)373-7599x86730 fax:+64(9)373-7090
www.health.auckland.ac.nz/pharmacology/staff/nholford
Chandra, Nick et al
It is worth noting that while three samples won't support a 4 parameter
model if all patients contribute these samples at exactly the same time
(i.e. the patients are exchangeable from a design perspective) this is not
necessarily the case if the design is optimized to learn about the PK.
We have designed and conducted a number of studies where the number of
samples is less than the number of parameters and achieved good results.
Some of the issues that you need to consider are:
1) Your design will probably lead to some shrinkage in the empirical Bayes
estimates which may be problematic if you intend to use the EBEs for
inferential purposes. However if you're after the population estimates only
(which is often the case) then this is not an issue.
2) Your design is unbalanced with respect to covariates. Adults are
providing much more information about the model and parameter values than
the children (even if the design in children was optimized) - which will
affect your ability to identify some covariate relationships with accuracy.
This can be assessed relatively easily using both optimal design and
simulation based investigations.
Regards
Steve
--
Professor Stephen Duffull
Chair of Clinical Pharmacy
School of Pharmacy
University of Otago
PO Box 913 Dunedin
New Zealand
E: [EMAIL PROTECTED]
P: +64 3 479 5044
F: +64 3 479 7034
Design software: www.winpopt.com
Quoted reply history
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of Chandrasekhar Udata
> Sent: Thursday, 29 May 2008 9:16 a.m.
> To: [email protected]
> Subject: Re: [NMusers] Sparse (pediatric) and rich (adult) data
>
> Thank you Nick and Leonid for your comments.
>
> Follow-up question:
> I do understand that 3 samples per subject may not support 4
> parameters model. However, historically, the compound showed
> bi-phasic characteristics (in adults) and I do like to use
> the same model in pediatrics. Also, the model (ADVAN3,
> TRANS4) did converge with no issues/errors (with pediatric
> data alone). Is there something I am missing? or is TRANS5
> (AOB, ALPHA, BETA) an alternative for such limited data?
>
> Regards,
> - Chandra
>
> >>> Nick Holford <[EMAIL PROTECTED]> 5/28/2008 1:38:07 PM >>>
>
> Chandra,
>
> With such a small sample its hard to learn much about
> differences between adults and children. Your principled
> approach using allometric scaling is a reasonable way to
> bridge the gap in recognizing that adults and children are
> all the same species (see reference below).
>
> "Children are just small adults"
>
> I would not be too worried about individual parameter
> estimate in children being different. With only 3 samples per
> child and a 2 cmt model requiring at least 4 parameters you
> will always get different results if you use different assumptions.
>
> Nick
>
> Anderson BJ, Holford NH. Mechanism-Based Concepts of Size and
> Maturity in Pharmacokinetics. Annu Rev Pharmacol Toxicol.
> 2008;48:303-32.
>
>
> Chandrasekhar Udata wrote:
> > Hi,
> >
> > I am working on a pop PK model to estimate PK parameters in
> pediatric
> > and adult patients. Pediatric study (n=20, age <6 yrs) has fewer
> > samples (3) per subject whereas the adult study (n=50,
> median age 20
> > yrs) has 12 samples per subject. A two-compartment model best
> > describes the data for each data set. Although a
> two-compartment model
> > best describes the combined data, the individual parameter
> estimates
> > in pediatric population are different compared to those obtained
> > using with pediatric data alone. Note that the parameter
> estimates in
> > adults were not significantly altered with either combined or adult
> > data alone. Body weight is the only covariate included in the model
> > with allometric exponents fixed to 0.75 on CL and 1 on V1.
> >
> > I would like to hear your thoughts on this and any
> suggestions on how
> > to proceed with modeling combined data from pediatric and
> adult studies.
> >
> > Regards,
> > - Chandra
> >
>
> --
> Nick Holford, Dept Pharmacology & Clinical Pharmacology
> University of Auckland, 85 Park Rd, Private Bag 92019,
> Auckland, New Zealand
> [EMAIL PROTECTED] tel:+64(9)373-7599x86730 fax:+64(9)373-7090
> www.health.auckland.ac.nz/pharmacology/staff/nholford
>
>
>
>
Steve,
I hope that you do not dispute that in this particular case you need to use adult data (50 full profiles) rather than discard them and use only kids data (3 sample per subject, 20 subjects)? While optimal design can be used to extract more information from the same number of samples, it is not a substitute for the real data. Even with optimal design of the pediatric study (with the same 20 subjects, 3 optimal sample points) I bet you would gain by using adult data as well.
Leonid
--------------------------------------
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web: www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel: (301) 767 5566
Stephen Duffull wrote:
> Chandra, Nick et al
>
> It is worth noting that while three samples won't support a 4 parameter
> model if all patients contribute these samples at exactly the same time
> (i.e. the patients are exchangeable from a design perspective) this is not
> necessarily the case if the design is optimized to learn about the PK.
>
> We have designed and conducted a number of studies where the number of
> samples is less than the number of parameters and achieved good results.
>
> Some of the issues that you need to consider are:
> 1) Your design will probably lead to some shrinkage in the empirical Bayes
> estimates which may be problematic if you intend to use the EBEs for
> inferential purposes. However if you're after the population estimates only
> (which is often the case) then this is not an issue.
> 2) Your design is unbalanced with respect to covariates. Adults are
> providing much more information about the model and parameter values than
> the children (even if the design in children was optimized) - which will
> affect your ability to identify some covariate relationships with accuracy.
> This can be assessed relatively easily using both optimal design and
> simulation based investigations.
>
> Regards
>
> Steve
> --
> Professor Stephen Duffull
> Chair of Clinical Pharmacy
> School of Pharmacy
> University of Otago
> PO Box 913 Dunedin
> New Zealand
> E: [EMAIL PROTECTED]
> P: +64 3 479 5044
> F: +64 3 479 7034
>
> Design software: www.winpopt.com
>
> > -----Original Message-----
> >
Quoted reply history
> > From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of Chandrasekhar Udata
> >
> > Sent: Thursday, 29 May 2008 9:16 a.m.
> > To: [email protected]
> > Subject: Re: [NMusers] Sparse (pediatric) and rich (adult) data
> >
> > Thank you Nick and Leonid for your comments. Follow-up question: I do understand that 3 samples per subject may not support 4 parameters model. However, historically, the compound showed bi-phasic characteristics (in adults) and I do like to use the same model in pediatrics. Also, the model (ADVAN3, TRANS4) did converge with no issues/errors (with pediatric data alone). Is there something I am missing? or is TRANS5 (AOB, ALPHA, BETA) an alternative for such limited data? Regards,
> >
> > - Chandra
> >
> > > > > Nick Holford <[EMAIL PROTECTED]> 5/28/2008 1:38:07 PM >>>
> >
> > Chandra,
> >
> > With such a small sample its hard to learn much about differences between adults and children. Your principled approach using allometric scaling is a reasonable way to bridge the gap in recognizing that adults and children are all the same species (see reference below).
> >
> > "Children are just small adults"
> >
> > I would not be too worried about individual parameter estimate in children being different. With only 3 samples per child and a 2 cmt model requiring at least 4 parameters you will always get different results if you use different assumptions.
> >
> > Nick
> >
> > Anderson BJ, Holford NH. Mechanism-Based Concepts of Size and Maturity in Pharmacokinetics. Annu Rev Pharmacol Toxicol. 2008;48:303-32.
> >
> > Chandrasekhar Udata wrote:
> >
> > > Hi,
> > >
> > > I am working on a pop PK model to estimate PK parameters in
> >
> > pediatric
> >
> > > and adult patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per subject whereas the adult study (n=50,
> >
> > median age 20
> >
> > > yrs) has 12 samples per subject. A two-compartment model best describes the data for each data set. Although a
> >
> > two-compartment model
> >
> > > best describes the combined data, the individual parameter
> >
> > estimates
> >
> > > in pediatric population are different compared to those obtained using with pediatric data alone. Note that the parameter
> >
> > estimates in
> >
> > > adults were not significantly altered with either combined or adult data alone. Body weight is the only covariate included in the model with allometric exponents fixed to 0.75 on CL and 1 on V1. I would like to hear your thoughts on this and any
> >
> > suggestions on how
> >
> > > to proceed with modeling combined data from pediatric and
> >
> > adult studies.
> >
> > > Regards,
> > >
> > > - Chandra
> >
> > --
> > Nick Holford, Dept Pharmacology & Clinical Pharmacology
> >
> > University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
> >
> > [EMAIL PROTECTED] tel:+64(9)373-7599x86730 fax:+64(9)373-7090
> > www.health.auckland.ac.nz/pharmacology/staff/nholford
Leonid
> I hope that you do not dispute that in this particular case
> you need to use adult data (50 full profiles) rather than
> discard them and use only kids data (3 sample per subject, 20
> subjects)?
I definitely do not dispute the need to have both adult and paediatric data
in the analysis (so I agree :-) ). I see two reasons for this (perhaps more
if I took more time). The first and most important reason is combining
adult and paediatric data together is a great (only) way to learn how
children differ pharmacokinetically from adults and how doses can be scaled
to achieve equivalent exposures. Secondly, especially in this case, it is
often helpful to combine data sets together to improve the informativeness
of the overall design. This latter point, however was the point of my
previous email. Some care must be taken to assess the accuracy of covariate
effects given the unbalanced nature of the design.
> While optimal design can be used to extract more
> information from the same number of samples, it is not a
> substitute for the real data. Even with optimal design of the
> pediatric study (with the same 20 subjects, 3 optimal sample
> points) I bet you would gain by using adult data as well.
You always gain by summing over data (unless the new data is negatively
informative which is unlikely in any PK situation). So I don't exactly
follow your point. The question to me is simply, what chance do I have of
identifying a model that allows me to draw appropriately accurate
conclusions. Optimal design is a way that allows investigators to improve
the informativeness of data. Obviously, no data = no information.
Steve
--
Professor Stephen Duffull
Chair of Clinical Pharmacy
School of Pharmacy
University of Otago
PO Box 913 Dunedin
New Zealand
E: [EMAIL PROTECTED]
P: +64 3 479 5044
F: +64 3 479 7034
Design software: www.winpopt.com
Leonid and Steve,
Thanks for your comments. While optimal design with/without combining with adults can help learn more about children in both cases it comes with assumptions. One of the assumptions is adults are big children.
Exactly how to implement that assumption is controversial but it will determine what is learned eg. should we assume allometric theory with fixed coefficients of 3/4 for clearance and 1 for volume? Or should we use up 2 degrees of freedom and estimate the coefficents? Depending on whether or not you use 2 parameters to describe how big adults are will influence the conclusions about other covariate relationships which may not be so well understood.
Nick
Leonid Gibiansky wrote:
> Steve,
>
> I hope that you do not dispute that in this particular case you need to use adult data (50 full profiles) rather than discard them and use only kids data (3 sample per subject, 20 subjects)? While optimal design can be used to extract more information from the same number of samples, it is not a substitute for the real data. Even with optimal design of the pediatric study (with the same 20 subjects, 3 optimal sample points) I bet you would gain by using adult data as well.
>
> Leonid
>
> --------------------------------------
> Leonid Gibiansky, Ph.D.
> President, QuantPharm LLC
> web: www.quantpharm.com
> e-mail: LGibiansky at quantpharm.com
> tel: (301) 767 5566
>
> Stephen Duffull wrote:
>
> > Chandra, Nick et al
> >
> > It is worth noting that while three samples won't support a 4 parameter
> > model if all patients contribute these samples at exactly the same time
> >
> > (i.e. the patients are exchangeable from a design perspective) this is not
> >
> > necessarily the case if the design is optimized to learn about the PK.
> >
> > We have designed and conducted a number of studies where the number of
> > samples is less than the number of parameters and achieved good results.
> >
> > Some of the issues that you need to consider are:
> >
> > 1) Your design will probably lead to some shrinkage in the empirical Bayes
> >
> > estimates which may be problematic if you intend to use the EBEs for
> >
> > inferential purposes. However if you're after the population estimates only
> >
> > (which is often the case) then this is not an issue.
> > 2) Your design is unbalanced with respect to covariates. Adults are
> >
> > providing much more information about the model and parameter values than
> >
> > the children (even if the design in children was optimized) - which will
> >
> > affect your ability to identify some covariate relationships with accuracy.
> >
> > This can be assessed relatively easily using both optimal design and
> > simulation based investigations.
> >
> > Regards
> >
> > Steve
> > --
> > Professor Stephen Duffull
> > Chair of Clinical Pharmacy
> > School of Pharmacy
> > University of Otago
> > PO Box 913 Dunedin
> > New Zealand
> > E: [EMAIL PROTECTED]
> > P: +64 3 479 5044
> > F: +64 3 479 7034
> >
> > Design software: www.winpopt.com
> >
> > > -----Original Message-----
> > >
Quoted reply history
> > > From: [EMAIL PROTECTED] [ mailto:[EMAIL PROTECTED] On Behalf Of Chandrasekhar Udata
> > >
> > > Sent: Thursday, 29 May 2008 9:16 a.m.
> > > To: [email protected]
> > > Subject: Re: [NMusers] Sparse (pediatric) and rich (adult) data
> > >
> > > Thank you Nick and Leonid for your comments. Follow-up question: I do understand that 3 samples per subject may not support 4 parameters model. However, historically, the compound showed bi-phasic characteristics (in adults) and I do like to use the same model in pediatrics. Also, the model (ADVAN3, TRANS4) did converge with no issues/errors (with pediatric data alone). Is there something I am missing? or is TRANS5 (AOB, ALPHA, BETA) an alternative for such limited data? Regards,
> > >
> > > - Chandra
> > >
> > > > > > Nick Holford <[EMAIL PROTECTED]> 5/28/2008 1:38:07 PM >>>
> > >
> > > Chandra,
> > >
> > > With such a small sample its hard to learn much about differences between adults and children. Your principled approach using allometric scaling is a reasonable way to bridge the gap in recognizing that adults and children are all the same species (see reference below).
> > >
> > > "Children are just small adults"
> > >
> > > I would not be too worried about individual parameter estimate in children being different. With only 3 samples per child and a 2 cmt model requiring at least 4 parameters you will always get different results if you use different assumptions.
> > >
> > > Nick
> > >
> > > Anderson BJ, Holford NH. Mechanism-Based Concepts of Size and Maturity in Pharmacokinetics. Annu Rev Pharmacol Toxicol. 2008;48:303-32.
> > >
> > > Chandrasekhar Udata wrote:
> > >
> > > > Hi,
> > > >
> > > > I am working on a pop PK model to estimate PK parameters in
> > >
> > > pediatric
> > >
> > > > and adult patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per subject whereas the adult study (n=50,
> > >
> > > median age 20
> > >
> > > > yrs) has 12 samples per subject. A two-compartment model best describes the data for each data set. Although a
> > >
> > > two-compartment model
> > >
> > > > best describes the combined data, the individual parameter
> > >
> > > estimates
> > >
> > > > in pediatric population are different compared to those obtained using with pediatric data alone. Note that the parameter
> > >
> > > estimates in
> > >
> > > > adults were not significantly altered with either combined or adult data alone. Body weight is the only covariate included in the model with allometric exponents fixed to 0.75 on CL and 1 on V1. I would like to hear your thoughts on this and any
> > >
> > > suggestions on how
> > >
> > > > to proceed with modeling combined data from pediatric and
> > >
> > > adult studies.
> > >
> > > > Regards,
> > > >
> > > > - Chandra
> > >
> > > --
> > > Nick Holford, Dept Pharmacology & Clinical Pharmacology
> > >
> > > University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
> > >
> > > [EMAIL PROTECTED] tel:+64(9)373-7599x86730 fax:+64(9)373-7090
> > > www.health.auckland.ac.nz/pharmacology/staff/nholford
--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
[EMAIL PROTECTED] tel:+64(9)373-7599x86730 fax:+64(9)373-7090
www.health.auckland.ac.nz/pharmacology/staff/nholford
Hi Chandra
If you want some ideas on how to proceed on modelling with combined adult
and paediatric data, you may want to refer to the following posters:
Cella et al., PAGE 16 (2007) Abstr 1203
[www.page-meeting.org/?abstract=1203]
Cella et al., ACOP 2008 Abstr 46
[ http://www.mosaicnj.org/acop/pdfs/46_DellaPasqua.pdf]
Regards
Massimo
_______________________________________
Massimo Cella
LACDR / Pharmacology, room 636
Gorleaus Laboratories
Einsteinweg 55 2333 CC Leiden
The Netherlands
Tel : ++31 71 527 6207
Email: [EMAIL PROTECTED]
_____
Quoted reply history
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Chandrasekhar Udata
Sent: Wednesday, May 28, 2008 5:47 PM
To: [email protected]
Subject: [NMusers] Sparse (pediatric) and rich (adult) data
Hi,
I am working on a pop PK model to estimate PK parameters in pediatric and
adult patients. Pediatric study (n=20, age <6 yrs) has fewer samples (3) per
subject whereas the adult study (n=50, median age 20 yrs) has 12 samples per
subject. A two-compartment model best describes the data for each data set.
Although a two-compartment model best describes the combined data, the
individual parameter estimates in pediatric population are different
compared to those obtained using with pediatric data alone. Note that the
parameter estimates in adults were not significantly altered with either
combined or adult data alone. Body weight is the only covariate included in
the model with allometric exponents fixed to 0.75 on CL and 1 on V1.
I would like to hear your thoughts on this and any suggestions on how to
proceed with modeling combined data from pediatric and adult studies.
Regards,
- Chandra
Steve - Thanks for making the point about the importance of experimental design. Often times when pooling adult and pediatric data, data are imbalanced, and pediatric PK designs are much less informative than the adult data. If, for a particular drug and disease state, pediatric patients really are just small adult patients, the design deficiency isn't much of a concern - but that's not always the case.
Although very useful for scaling body-size related differences in PK parameters from adults to peds, the allometric "small adult" assumption, doesn't always provide the complete story. There are other bits of information about pediatric PK (e.g. developmental changes, pediatric disease state effects) that we'd like to learn about directly from the pediatric data.
The analysis of the pooled data in this case (sparse, poorly- optimized pediatric data with more informative adult data) is similar to a Bayesian data analysis, with informative prior distributions for most/all model parameters. An alternative approach to analyzing the sparse pediatric data could be:
1. Assess the expected precision of PK parameters under the pediatric data alone, using a PFIM-type method. 2. Analyze the pediatric data, using a full Bayesian estimation method. Informative prior distributions based on adults would be selectively applied to those parameters with poor design support in the pediatric data alone, while other parameters which are of particular interest in the pediatric population could be estimated with diffuse prior distributions.
This approach allows the pediatric data alone to influence the estimation of a subset of parameters (hopefully, those components you'd like to learn about), while relying on prior adult information to anchor some of the more poorly supported components of the model.
Marc
Marc R. Gastonguay, Ph.D.
President & CEO, Metrum Research Group LLC [www.metrumrg.com]
Scientific Director, Metrum Institute [www.metruminstitute.org]
Direct: 860-670-0744 Main: 860-735-7043
Email: [EMAIL PROTECTED]
Quoted reply history
On May 28, 2008, at 9:49 PM, Stephen Duffull wrote:
> Leonid
>
> > I hope that you do not dispute that in this particular case
> > you need to use adult data (50 full profiles) rather than
> > discard them and use only kids data (3 sample per subject, 20
> > subjects)?
>
> I definitely do not dispute the need to have both adult and paediatric data in the analysis (so I agree :-) ). I see two reasons for this (perhaps more if I took more time). The first and most important reason is combining
>
> adult and paediatric data together is a great (only) way to learn how
>
> children differ pharmacokinetically from adults and how doses can be scaled to achieve equivalent exposures. Secondly, especially in this case, it is often helpful to combine data sets together to improve the informativeness
>
> of the overall design. This latter point, however was the point of my
>
> previous email. Some care must be taken to assess the accuracy of covariate
>
> effects given the unbalanced nature of the design.
>
> > While optimal design can be used to extract more
> > information from the same number of samples, it is not a
> > substitute for the real data. Even with optimal design of the
> > pediatric study (with the same 20 subjects, 3 optimal sample
> > points) I bet you would gain by using adult data as well.
>
> You always gain by summing over data (unless the new data is negatively informative which is unlikely in any PK situation). So I don't exactly follow your point. The question to me is simply, what chance do I have of
>
> identifying a model that allows me to draw appropriately accurate
>
> conclusions. Optimal design is a way that allows investigators to improve
>
> the informativeness of data. Obviously, no data = no information.
>
> Steve
> --
> Professor Stephen Duffull
> Chair of Clinical Pharmacy
> School of Pharmacy
> University of Otago
> PO Box 913 Dunedin
> New Zealand
> E: [EMAIL PROTECTED]
> P: +64 3 479 5044
> F: +64 3 479 7034
>
> Design software: www.winpopt.com
Marc
I agree that the use of a fully Bayesian analysis might be valueable here.
In option 2 when you suggest analysing the paediatric data - do you mean by
itself without the adult data or a combined analysis?
If the former then Aris and Leon (Manchester group) have done some work
comparing simultaneous analysis with sequential Bayesian analysis. I can't
recall where it was published - but it's worth a read.
Steve
--
Professor Stephen Duffull
Chair of Clinical Pharmacy
School of Pharmacy
University of Otago
PO Box 913 Dunedin
New Zealand
E: [EMAIL PROTECTED]
P: +64 3 479 5044
F: +64 3 479 7034
Design software: www.winpopt.com
Quoted reply history
> -----Original Message-----
> From: Gastonguay, Marc [mailto:[EMAIL PROTECTED]
> Sent: Saturday, 31 May 2008 2:20 a.m.
> To: Stephen Duffull
> Cc: 'Leonid Gibiansky'; 'Chandrasekhar Udata'; [email protected]
> Subject: Re: [NMusers] Sparse (pediatric) and rich (adult) data
>
> Steve - Thanks for making the point about the importance of
> experimental design. Often times when pooling adult and
> pediatric data, data are imbalanced, and pediatric PK designs
> are much less informative than the adult data. If, for a
> particular drug and disease state, pediatric patients really
> are just small adult patients, the design deficiency isn't
> much of a concern - but that's not always the case.
>
> Although very useful for scaling body-size related
> differences in PK parameters from adults to peds, the
> allometric "small adult" assumption, doesn't always provide
> the complete story. There are other bits of information about
> pediatric PK (e.g. developmental changes, pediatric disease
> state effects) that we'd like to learn about directly from
> the pediatric data.
>
> The analysis of the pooled data in this case (sparse,
> poorly-optimized pediatric data with more informative adult
> data) is similar to a Bayesian data analysis, with
> informative prior distributions for most/all model
> parameters. An alternative approach to analyzing the sparse
> pediatric data could be:
>
> 1. Assess the expected precision of PK parameters under the
> pediatric data alone, using a PFIM-type method.
> 2. Analyze the pediatric data, using a full Bayesian
> estimation method. Informative prior distributions based on
> adults would be selectively applied to those parameters with
> poor design support in the pediatric data alone, while other
> parameters which are of particular interest in the pediatric
> population could be estimated with diffuse prior distributions.
>
> This approach allows the pediatric data alone to influence
> the estimation of a subset of parameters (hopefully, those
> components you'd like to learn about), while relying on prior
> adult information to anchor some of the more poorly supported
> components of the model.
>
> Marc
>
> Marc R. Gastonguay, Ph.D.
> President & CEO, Metrum Research Group LLC [www.metrumrg.com]
> Scientific Director, Metrum Institute [www.metruminstitute.org]
> Direct: 860-670-0744 Main: 860-735-7043
> Email: [EMAIL PROTECTED]
>
>
>
>
> On May 28, 2008, at 9:49 PM, Stephen Duffull wrote:
>
>
> Leonid
>
>
>
> I hope that you do not dispute that in this
> particular case
>
>
> you need to use adult data (50 full profiles)
> rather than
>
>
> discard them and use only kids data (3 sample
> per subject, 20
>
>
> subjects)?
>
>
>
> I definitely do not dispute the need to have both adult
> and paediatric data
> in the analysis (so I agree :-) ). I see two reasons
> for this (perhaps more
> if I took more time). The first and most important
> reason is combining
> adult and paediatric data together is a great (only)
> way to learn how
> children differ pharmacokinetically from adults and how
> doses can be scaled
> to achieve equivalent exposures. Secondly, especially
> in this case, it is
> often helpful to combine data sets together to improve
> the informativeness
> of the overall design. This latter point, however was
> the point of my
> previous email. Some care must be taken to assess the
> accuracy of covariate
> effects given the unbalanced nature of the design.
>
>
>
> While optimal design can be used to extract more
>
>
> information from the same number of samples, it
> is not a
>
>
> substitute for the real data. Even with optimal
> design of the
>
>
> pediatric study (with the same 20 subjects, 3
> optimal sample
>
>
> points) I bet you would gain by using adult
> data as well.
>
>
>
> You always gain by summing over data (unless the new
> data is negatively
> informative which is unlikely in any PK situation). So
> I don't exactly
> follow your point. The question to me is simply, what
> chance do I have of
> identifying a model that allows me to draw
> appropriately accurate
> conclusions. Optimal design is a way that allows
> investigators to improve
> the informativeness of data. Obviously, no data = no
> information.
>
> Steve
> --
> Professor Stephen Duffull
> Chair of Clinical Pharmacy
> School of Pharmacy
> University of Otago
> PO Box 913 Dunedin
> New Zealand
> E: [EMAIL PROTECTED]
> P: +64 3 479 5044
> F: +64 3 479 7034
>
> Design software: www.winpopt.com
>
>
>
>
>
>