Transcript MOA Promotional Speaker Slide Kit

Dosage Determination from Preclinical
to Proof-of-Concept Trials,
(Including Toxicology)
Charlie Taylor, PhD
CpTaylor Consulting
Chelsea, MI, USA
Choosing Doses for POC:
• Preclinical and early studies that enable dose selection
• Reasons for drug failure in clinical phase 2-3
• Need to choose both low (ineffective) and high (maximum
tolerated) doses within dose range
• Biomarkers (one endpoint: animal → human translation)
• PK/PD modeling – EC50 as a target for efficacy or AEs
• Toxicology/toxicokinetics – daily AUC(0-24) as a limit
• Putting it together – visualizing multiple datasets
• Human population PK modeling – determine which doses
best fit the constraints
2
Sequence of Studies Needed
Prior to Clinical Proof-of-Concept
*
*
*
• Preclinical in vitro studies: action at drug target (pharmacology)
*
*
*
• Animal toxicology & toxicokinetic studies to identify safety issues
• Preclinical in vivo pain models: indicate treatment of disease
• Safety pharmacology: animal studies for adverse effects
• Preclinical (& human liver microsome) metabolism studies
• Clinical Phase I studies of pharmacokinetics and tolerance in healthy
human volunteers
• (Optional) Biomarker studies with both animal models and humans to
establish proof of pharmacology in vivo (apart from efficacy)
*
Requires in vivo unbound plasma drug concentration or daily
drug exposure to help choose human efficacy DOSES
3
Drug Development Failures – UK ’64-’85
Efficacy
29%
PharmKin
39%
Other
Com
mer
c.
Adv.
Events
Toxicol. 10%
11%
Prentis et al. (1988) Brit J Clin Pharmacol 25:387-396
4
Determine Initial Phase 2 Dose Range ??
• Clinical doses MUST encompass both low end (lack
of efficacy) and high end (maximum tolerance)
• Data from animal efficacy, animal safety, biomarker
and human tolerance ALL must be considered
• The peak unbound plasma drug concentration
(animal studies), daily exposure (AUC0-24 - tox) and
human multiple-dose PK each need consideration
• How to consider all these factors??
5
One Approach: Biomarkers – Surrogate
HUMAN Endpoints for Efficacy
• Defines drug action in vivo
• Examples: Imaging, Adverse Event or Mechanism
• e.g. PET to measure receptor occupancy in CNS
• e.g. Nystagmus, dizziness, balance platform
• e.g. Experimental pain model w/ volunteers
• e.g. Electrographic response (EEG, retinogram, TMS)
Biomarkers Allow no-go decision prior to proof of
efficacy, for example:
• Poor oral drug absorption or lack of CNS penetration
• Lack of receptor occupancy at highest safe dose
6
Hypothetical Human Biomarker:
• Criterion: 75% drug receptor occupancy in human brain @ high dose
• This criterion met at animal effective dose (animal PET study)
• Drug displacement of PET ligand in human brain:
18F-x-drugamine
given IV in tracer amount
• If greatest human volunteer dose of experimental drug reaches < 30%
occupancy, NO-GO
• If greatest human volunteer dose > 75% occupancy, GO (further
development)
• Caveat: Criteria must be selected based on results with a prior known
compound – Otherwise, risk of poor validation
7
Toxicology Findings (non-pharmacology) are
Based on Daily Drug Exposure (AUC0-24)
• Repeated-dose animal tox studies determine lowest
toxic dose and greatest no-effect (daily) dose
• Toxicokinetics determine drug exposure (AUC0-24)
in mg•hr/mL at greatest no-effect dose
• e.g. Drug X has 8 hr half-life; Cmax and AUC are
determined from plasma drug samples taken 0, 1, 4, 7,
12, and 24 hr after single oral dose at steady-state
• Similar human pharmacokinetic data and PK modeling
determine human drug exposure (AUC0-24) @ doses
• Analysis is adjusted for different drug binding of
plasma proteins between species
8
Calculation of Animal Drug Exposure Toxicokinetic AUC(0-24)
• Samples of drug in plasma of animal tox species
— Begin sampling after reaching repeated dose steady state
— Orange symbols are mean from n = 8
Dose =
50 mg/kg/day
Free Plasma Drug Conc. (mg/mL)
• Mathematical fit to curves of oral absorption & elimination
• Measure area under curve for 0-24 hr = Drug Exposure
14
Cmax = 10.8 mg/mL
12
10
AUC(0-24) =
136 mg•hr/mL
8
6
4
2
0
0
4
8
12
16
Hr After Dose
20
24
9
Measured Drug Exposure in Rat Tox Studies
Toxicokinetic Parameters in Multidose Oral Toxicity Studies
Species
Durat.
Dose
(mg/kg)
Plasma Concentration
(mg/mL)
Male
Rat
Gavage
500
1250
2500
2 Week
Diet
500
1250
2500
Female
AUC(0-24)
(mg·hr/mL)
Male
Female
13.5
27.6
47.4
9.92
25.2
40.7
120.0
332.0
626.0
102.0
334.0
602.0
11.5
25.9
50.7
10.7
19.0
32.4
199.0
491.0
921.0
181.0
336.0
606.0
10
Tumor
Decr fetal wt
Rhinitis
Skin Sores
Leukopenia
Hypoactivity
Death
Animal Toxicology: Human Exposure Limits
Are Set by Daily Drug Exposure (AUC0-24)
100000
DOG
MOUSE
AUC(0-24)
RAT
10000
1000
100
No-Effect Dose Limit:
200 mg•hr/mL – determines maximum permissible human exposure
11
Free Plasma Drug Conc. (nM)
(mg/mL)
PK Modeling of Drug in Human Plasma
(daily dosing of 50 mg oral)
Human Cmax Limit
based on Animal
Toxicology
(Max no-effect
dose AUC0-24)
100
10
AUC0-24
1
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (hours)
12
CYP2D6 Heterogeneity – Ca2+ Channel Blocker
Smith & Jones (1999) Curr Opin Drug Discov Devl 2:33-41
13
Q: How to Predict Human Efficacious
and Adverse Drug Doses Based on
Animal Efficacy, Animal AEs and Human
Pharmacokinetic Data??
A: Compare plasma drug Cmax
obtained in animal pharmacology
tests using a Napiergram to human
Pharmacokinetic Cmax data
14
“Napiergram”
• Named for John Napier of Merchistoun (aka
Marvelous Merchiston, Scotland)
• Inventor of Napier’s bones (slide rule),
popularization of logarithms and the decimal point
• Also: used a pet black rooster to tell fortunes and
devine truths
• Napiergram: graphic comparison
of log10 unbound plasma drug
concentrations associated with
pharmacology and with safety
concerns
John Napier (1550-1617)
15
From Dose:Response experiments:
Obtain ED5, ED50, ED95
16
Transform Pharmacology from ED50 to EC50
17
liver tox
decr blood p
ressure
ataxia (rotor
od)
in vivo functi
on C
in vivo functi
on B
in vivo functi
on A
Cmax = 2,500 nM
or 0.5 mg/mL
(unbound)
in vitro functi
on
receptor bin
d
ing
Napiergram: Many Pharmacology Datasets –
Animal Cmax for doses with 5% 50% & 95% effect
Cmin = 125 nM
or 0.026 mg/mL
(unbound)
Free Fraction (nM)
1000000
100000
10000
1000
100
10
1
18
Free Plasma Drug Conc. (nM)
(mg/mL)
PK Modeling of Drug in Human Plasma
(daily dosing of 50 mg oral)
100
Cmax (hi dose)
10
Cmax (mid dose)
Animal Adverse
Limit
(EC50 for ataxia)
Actual Human
PK – mid dose
Cmax (low dose)
1
EC20 for
Efficacy in
Animal Model
0 10 20 30 40 50 60 70 80 90 100 110 120
Time (hours)
19
Phase 2 Dose Selection (final chapter)
• Requires Deliberation from team of experts:
• Animal tox, Pharmacokinetics, PK/PD modeling, Clinical
research, Preclinical pharmacology, (Biomarkers)
• Who pays the clinical trial bills? Clinical Research
— Despite planning, dosage and regimen often are readjusted
during Phase 2 (toleration, efficacy or new safety findings)
— Dosages MUST continue to include both low (ineffective) and
maximal tolerated dosages to provide basis for FDA approval
— Dose toleration may vary between healthy volunteers and
patients with serious disease
20
SUMMARY: Preclinical Studies to Determine
Phase 2 Dose Selection
• In vitro and in vivo animal pharmacology
– target Cmax for therapy and adverse effects
• Animal toxicology & toxicokinetic studies
– determines maximal human drug exposure (AUC0-24)
• Phase 1 Clinical trials
- determines human pharmacokinetics & drug exposure
• Napiergram – allows consideration of Cmax from
multiple animal datasets & compare to human PK
• Phase 2 dose adjustment is common!
21
22
Example “Drug Killer” Problems
• Poor Oral Absorption (F < 25%)
• Poor Aqueous Solubility
• Poor Elimination Kinetics (t1/2 < 4 hr or t1/2 > 36 hr)
• Nonlinear Elimination Kinetics
(e.g. blocked clearance at high doses)
• Extensive metabolism to active or toxic compound
• Excessive plasma protein binding (> 99%)
PK
• Metabolism by variable CyP450 (CYP2D6, CYP2C19)
• Cardiac Q-T interval prolongation (hERG channel block)
• Genotoxic compound (Ames positive)
Tox
• Hepatic toxicity
23