Transcript A Population Model, as made by Breugel!

Multiple Model
Dosage Design
Roger Jelliffe, M.D.
USC Lab of Applied Pharmacokinetics
7/17/2015
USC LAPK
1
Conventional Design of
Drug Dosage Regimens
•
•
•
•
•
Use Population Model
Select single target goal
Develop the dosage regimen to “hit target”
But will it? How precisely?
How to evaluate the expected precision?
7/17/2015
USC LAPK
2
7/17/2015
USC LAPK
3
Optimal Drug Dosage
• Consider the effects of the pop outliers.
• Do a clinical simulation to validate each
regimen.
• Standard Doses: no feedback. Optimize
them to hit desired targets.
• Cancer Rx, Veterinary Rx.
• Individualized Doses: Optimize them also.
• Targets are desired Conc, AUC, etc
• AIDS, Transplants, Cancer, Antibiotics,
Cardiovascular, etc.
7/17/2015
USC LAPK
4
What is the BEST Pop Model?
• The correct structural PK/PD Model.
• The collection of each subject’s
exactly known parameter values
for that model.
7/17/2015
USC LAPK
5
An NPML Population Joint Density,
as made by Mallet
7/17/2015
USC LAPK
6
“Multiple Model” Dosage Design
• Start with multiple models in pop model
• Best starting tool = NPAG joint density.
• Compute regimen having least weighted
squared error in target goal
achievement.
7/17/2015
USC LAPK
7
7/17/2015
8
Continuous IV Vanco. Predictions when regimen
7/17/2015
9
based on means is given to all subjects
Vanco, continuous IV. Predictions from
7/17/2015
10
MM regimen
MM Optimal Dosage Regimens:
•
•
•
•
•
•
Consider the quantitative effect of outliers.
A built-in simulated clinical trial each time.
Achieve target goals with max precision.
Get the best overall “standard dose”.
Best for Vet use, without feedback.
Best for Patient use, with or without
feedback - cancer, AIDS, inf. Disease,
CV disease.
• Best optimally coordinated combination
regimens in future.
7/17/2015
USC LAPK
11
Clinical Illustrations
• Planning the initial regimen
• Adjusting it based on TDM data
7/17/2015
USC LAPK
12