Lecture 12: Drug development and clinical trials

Question 1

Long and costly drug development

Answer 1

• 10 years from discovery to market
• $3 billion per drug at least (partly due to 1/10 working)
• patent protection very important

Question 2

biomarker/surrogate/outcome?

Answer 2

biomarker

• readily measuable marker of response
• eg. EEG response to anaesthetic induction agent

surrogate

- biomarker used for regulatory approval

• eg. reduction in HIV viral load

outcome

• how the patient functions/feels/survives

Question 3

Phase 0

Answer 3

Predictions for humans

• data from non-human animals
• probable mechanism of action
• likely effective concentration*
• major routes of elimination*
• oral absorption properties*

* = similar in humans as in rats

Question 4

Phase 1

Answer 4

Tolerability (different from tolerance)

• start with small doses and slowly increase
• stop when adverse effects noted

Question 5

Phase 2

Answer 5

Effectiveness

PHASE 2A

• Proof of concept, the yes/no decision point

PHASE 2B

• learn about dose response curve
• learn about effective doses
• learn about taget concentration

Question 6

Phase 3

Answer 6

Safety

• learn adverse effects in target population
• confirm effective doses
• Learn pharmacodynamics of surrogate/outcome
• Learn PK and PD covariates (age, sex, other drugs)

Question 7

Phase 4

Answer 7

Post-marketing

• confirm effective dose(s)
• confirm common adverse events and learn about uncommon adverse events (1/10,000)
• learn “use effectiveness” (does it work in real life)
• learn pharmacoeconomics (use effectiveness)

Question 8

Clinical Trial Design?

Answer 8

Assignment

Blinding

Comparison

Sequence

Question 9

Assignment?

Answer 9

First come, First served

Randomised - balanced or stratified (age, previous stroke etc)

Question 10

Blinding?

Answer 10

open

single (doctor knows)

doubble (neither investigtor or patient) -“doubble dummy”

tripple (error)

Question 11

Comparison?

Answer 11

As well as a control group it pays to investigate within the trial group on the new drug or procedure.

Active - dose, conc and biomarker control

Placebo - inactive

standard treatment - non-inferiority, add-on

Question 12

Sequence -Parallel

Answer 12

Parallel

• different treatments are assigned to different groups of subjects
• Good for finding out if the drug works but gives unclear answers to learning questions that ask about the shape of the dose response relationship of what dose is needed to achieve a particular effect.

EFFECTIVE, BUT NOT EFFICIENT

Question 13

Sequence - crossover

Answer 13

Crossover

• Two or more treatments is used in each subject. Allows individual dose response curves to be observed and the true shape of the dose response relationship can be determined.
• Also may have advantages in terms of statistical power. If it is assumed that within subject variability is small, fewer subjects need to be studied to detect a treatment effect.
• Can have carryover effect, period effect
• higher risk of dropout due to several treatments needed.

Question 14

Sequence - Titration

Answer 14

Forced

costant fixed doses given to all patients to learn about dose response relationship

Flexible

starting with lower dose and building up untill response is seen and then keeping it constant there

Question 15

Analysis perspectives?

Answer 15

Intention to treat - looks at use effectiveness and a pharmacoeconomic perspective. Is always biased as use effectiveness includes people who don’t actually take the drug.

As treated - method effectiveness (more accurate) and develops a science perspective so useful for making scientific decisons. Only looks at people who took the drug as instructed.